Synthesis of Organic Inorganic Lead Halide Perovskite Nanoplatelets: Towards High-Performance Perovskite Solar Cells and Optoelectronic Devices

Transcription

1 COMMUNICATION Synthesis of Organic Inorganic Lead Halide Perovskite Nanoplatelets: Towards High-Performance Perovskite Solar Cells and Optoelectronic Devices Son Tung Ha, Xinfeng Liu, Qing Zhang, David Giovanni, Tze Chien Sum, and Qihua Xiong * Recently, organic-based lead halide perovskites have received much attention for their high performance as light absorbers in thin-film solar cells. [1 8 ] They exhibit not only a high optical absorption coefficient, optimal bandgap, and long electron/ hole diffusion lengths, [9,1 ] which are advantageous for solar cells, but also good optical and electrical transport properties, making them suitable for other opto-electronic devices, such as, field-effect transistors, light-emitting diodes, and photodetectors. [11 14 ] Although these structures were first synthesized a long time ago, many intrinsic physical questions still remain unresolved for these types of material, such as, the nature of their excited states, the relative fraction of free and bound charge pairs, and the interplay between two species, as well as the question relating to the function of the halide atom towards the charge-transport behavior in the perovskite structure. [9 ] Up to now there have been several methods to prepare organic-based lead halide perovskite films, which are suitable forms to study the optical and electrical properties, as well as to fabricate opto-electronic devices. The simplest method is spincasting from a solution of perovskite compounds. However, this method usually results in inhomogeneous, polycrystalline structures with a large surface roughness. Thermal evaporation has been employed to obtain a better homogeneity and higher crystalline perovskite film through either co-evaporation [3 ] of lead halide and alkyl amino halide or single source (perovskite) evaporation. [15 ] The former technique requires the fine balancing of the two source evaporation rates which is difficult due to the much higher vapor pressure of the organic component compared to the inorganic counterpart, whereas the latter requires S. T. Ha, Dr. X. F. Liu, Dr. Q. Zhang, D. Giovanni, Prof. T. C. Sum, Prof. Q. H. Xiong Division of Physics and Applied Physics School of Physical and Mathematical Sciences Nanyang Technological University , Singapore Qihua@ntu.edu.sg Prof. T. C. Sum, Prof. Q. H. Xiong Singapore-Berkeley Research Initiative for Sustainable Energy 1 Create Way 13862, Singapore Prof. Q. H. Xiong Division of Microelectronics School of Electrical and Electronics Engineering Nanyang Technological University , Singapore DOI: 1.12/adom dedicated equipment. Another method to prepare lead halide perovskite films is to dip a pre-deposited lead halide film into a solution of alkyl amino halide to convert the lead halides to their corresponding perovskites. [16 ] This method works well with a layered perovskite compound (RNH 3 ) 2 PbI 4 in which R- is an alkyl group with two or more carbon atoms in the chain, which makes it easier for alkyl amino halide molecules to diffuse into the lead halide octahedron network through the van der Waals gap within the organic bilayer and, as a result, the reaction could be done in a few minutes time. However, in the case of CH 3 NH 3 PbX 3 perovskite, due to the lack of a van der Waals gap in the threedimensional structure of lead halide, the required dipping time would be much longer ( i.e., 1 3 hours). Such a long immersion time in an organic solvent could dissolve and/or disperse parts of the lead halide and the formed perovskite, reducing the quality of the perovskite film. [2,17 ] In conventional semiconductors, there have been extensive investigations on the synthesis in forms of nano-platelets (2D), [18 ] nano-wires (1D), [19,2 ] nanobelts (quasi-1d), [21,22 ] and their extraordinary optical and electrical properties compared to their bulk counterparts, leading to a number of exciting applications in nano-electronics and nanophotonics. To the best of our knowledge, nobody has reported a lead halide perovskite species in two-dimensional platelets. Herein, we report for the first time the synthesis of lead halide perovskite family nano-platelets with lateral dimensions from 5 3 µm and thicknesses from several atomic layers to several hundred nanometers. The CH 3 NH 3 PbI 3 platelets prepared by our method have an electron diffusion length of more than 2 nm, which is two times higher than the recently reported value for a film prepared by conventional solution spin-coating. We believe that this new synthesis method will push forward the fundamental study of the lead halide perovskites family as well as to explore their new applications in opto-electronics. The presented method involves two steps: First, the growth of lead halide nano-platelets on muscovite mica utilizing van der Waals epitaxy in a vapor transport chemical deposition system. [23 25 ] Next, the as-grown platelets are converted to perovskites by a gas solid hetero-phase reaction with methyl ammonium halide molecules. Figure 1 a shows the optical and scanning electron microscopy (SEM) images of lead halides grown on muscovite mica substrates. The difference in color corresponds to different thicknesses as shown in Figure 1 b for particular lead iodide platelets. The in-plane orientation of the platelets in the case of PbCl 2 and PbBr 2 (Figure 1 a: A, B) is evident of the van der Waals epitaxial growth on the muscovite mica substrate because of the three-fold symmetry of the mica 838 wileyonlinelibrary.com 214 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Optical Mater. 214, 2,

2 COMMUNICATION Figure 1. Morphological characterizations of lead halides nano-platelets as-grown on muscovite mica substrate: a) Optical (above) and SEM (below) images of lead halides: A,D: PbCl2; B,E: PbBr2; C,F: PbI2. b) Optical images of individual PbI2 nano-platelets with different colors corresponding to different thicknesses as measured by AFM. c) XRD pattern of the platelets, indexed in blue for lead halides and in red for muscovite mica. d) Raman spectra measured for individual lead halide platelets. Insets: Black curves: experimental data, purple curves: simulation data, green curves: peak fitting. surface lattices.[26,27] The platelets show a highly flat and smooth surface with a surface roughness of only ±1.5 nm as seen by SEM and atomic force microscopy (AFM) (see Figure 1a and Figure S1 in the Supporting Information). The as-grown lead halide platelets on mica were characterized by powder X-ray analysis (Figure 1c) in θ θ geometry, meaning that only planes parallel to the surface of the substrate contribute to the patterns. Multiple strong peaks indexed in red correspond to the basal planes of muscovite mica of the 2M1 poly-type [KAl2(Si3Al)-O1(OH)2, monoclinic, space group: C2/c],[28] whereas peaks indexed in blue correspond to PbCl2, PbBr2, and PbI2. It should be noted that lead halide platelets have a highly oriented growth direction along the a-axis in the case of PbCl2 and PbBr2 and along the c-axis for PbI2. Raman spectroscopy was used to further characterize the crystalline structure of individual platelets for each lead halide compound. All Raman spectra were taken under 633 nm excitation with a laser power of.5 mw through a 1 objective at room temperature. The Raman spectra of the as-grown lead halide platelets agree well with their corresponding single-crystal spectra as Adv. Optical Mater. 214, 2, reported in the literature.[29,3] The detailed assigned vibrational modes for each spectrum are shown in Table S1 in the Supporting Information. We then converted the as-grown lead halide platelets or nanowires into perovskites by reacting with gas-phase methyl ammonium halides. The experimental setup is demonstrated in Figure 2a below. The converting reaction was carried out in a quartz tube in vacuum with an inert carrier gas such as nitrogen or argon. The methyl ammonium halide source was synthesized by a solution method and re-crystallized in diethylether/ methanol following the procedure published elsewhere.[4] The source was placed in the center of the tube furnace where the set temperature (ca. 12 C) is normally achieved whereas the pre-grown lead halide platelets were placed downstream. The pressure was about 2 Torr. Figure 2b,c shows the crystal structures of lead halide and perovskite with methyl ammonium (CH3NH3+) as the cation. As can be seen, both crystal structures have a similar network unit of lead halide octahedrons with the lead atom located in the center surrounded by halide atoms. Whereas in lead halide each octahedron shares 214 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 839

3 COMMUNICATION Figure 2. Conversion of lead halide nano-platelets to perovskites by gas solid hetero-phase reaction with methyl ammonium halide (CH 3 NH 3 X, X = Cl, Br, I). a) Schematic of the synthesis setup using a home-built vapor-transport system. b) Structure of the lead halide in which the Pb atoms are at the center of the halide octahedrons. In the same layer, each octahedron shares 2 equatorial halide atoms with its neighbor whereas two octahedrons from two continuous layers share one axial halide atom. c) Structure of lead halide perovskite CH 3 NH 3 PbX 3 (X = Cl, Br, I) in which each lead halide octahedron shares one equatorial halide atom with its neighbors in the same layer and shares one axial halide atom with neighbors from the next layers. The methyl ammonium group CH 3 NH 3 + denoted as a red sphere is located within the center of eight lead halide octahedrons. The similarity of the lead halide and perovskite structures makes it possible to convert the lead halide solid structure into its perovskite by intercalating methyl ammonium halide molecules. d) Thickness of PbI 2 platelets before (images above data line) and after being converted to CH 3 NH 3 PbI 3 (images below data line). Note that the color of the PbI 2 platelets changed corresponding to the change in thickness (as measured by AFM). The thickness of the CH 3 NH 3 PbI 3 platelets was about 1.8 times higher compared to the corresponding PbI 2 platelets, which agrees well with the ratio of the c lattice constant between the two compounds. two equatorial halide atoms with its neighbors in the same layer and shares one axial halide atom with its neighbors from different layers forming a layered structure, the octahedrons in lead halide perovskite form a 3D network structure in which each octahedron shares only one halide atom with its neighbors either in the same layer or in a different layer. XRD analysis revealed the hexagonal structure of lead iodide having a lattice constant c =.695 nm with an orientation perpendicular to the substrate (Figure 1 c). The perovskite CH 3 NH 3 PbI 3 normally has a tetragonal structure at room temperature with a lattice constant c = nm. [31 ] The difference in lattice constant c is due to the insertion of a methyl ammonium group in the center of eight octahedrons and the relocation of the equatorial halide atoms resulting in a twisting of the lead halide octahedrons as illustrated in Figure 2 b and c. Interestingly, the thickness of PbI 2 and CH 3 NH 3 PbI 3 platelets (before and after conversion) correlated to each other by a factor of 1.81 (as shown in Figure 2 d), which agrees well with the lattice constant ratio of the two compounds along the c axis. Our observation is also in good agreement with previous work on a PbI 2 film, where the film thickness increased by a factor of 1.75 (from 2 nm to 35 nm) after converting to CH 3 NH 3 PbI 3. [32 ] This provides a good way to control the thickness of perovskite platelets by monitoring the thickness of the corresponding lead halide platelets. In order to confirm whether the conversion of the lead iodide platelets into their perovskite form had been successful, we investigated the crystalline structure by XRD and the optical properties of the platelets before and after conversion as shown in Figure 3. Figure 3 a shows the XRD pattern of as-grown platelets on muscovite mica substrate before and after conversion) in the θ θ geometry. It is clear that after conversion the identical peaks corresponding to 1, 2, 3, 4 of the 2-H lead iodide crystals (space group: P3m 1(164), JCPDS file No ) [33 ] disappeared (marked by the dashed-red circles in the XRD pattern of CH 3 NH 3 PbI 3 ) and several new peaks 84 wileyonlinelibrary.com 214 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Optical Mater. 214, 2,

4 COMMUNICATION Figure 3. Characterizations of lead iodide platelet after conversion to CH 3 NH 3 PbI 3 perovskite. a) XRD pattern of as-grown PbI 2 platelets on muscovite mica (below) and after conversion to CH 3 NH 3 PbI 3 platelets (above). After conversion, the identical peaks of PbI 2 (1, 2, 3, 4) disappeared (as shown by the red dashed circle in the XRD spectrum of CH 3 NH3PbI 3 ). Instead, several peaks of tetragonal CH 3 NH 3 PbI 3 (indexed in blue color) were detected. b) Raman spectra of the same PbI 2 platelet before (green) and after (red) conversion. The blue curve is the Raman spectrum of a bulk CH 3 NH 3 PbI 3 crystal that was synthesized by a solution method (see details in the Supporting Information, Figure S3) for comparison. c) Optical absorption and photoluminescence (at 77K) of PbI 2 platelet before (black curve) and after conversion to perovskite (red curve). d) PL lifetime of PbI 2 platelet before (black squares) and after (red dots) conversion. The PL lifetime of a CH 3 NH 3 PbI3 platelet is approximately 4 times larger than that of the corresponding PbI 2 platelet. of tetragonal-phase lead iodide perovskite were observed. [31 ] Because of the strong peaks of the mica substrate and the slightly twisted structure of the lead iodide octahedrons after conversion, we could not observe the peak corresponding to planes perpendicular to the c -axis as would be expected in the XRD pattern. However, the disappearance of the PbI 2 peaks confirmed the completed conversion. The complete conversion of the other halide perovskites, CH 3 NH 3 PbBr 3 and CH 3 NH 3 PbCl3, were also confirmed by XRD, as shown in Figure S2 in the Supporting Information. Raman spectroscopy was conducted before and after conversion (Figure 3 b). In the PbI 2 platelet, the peak at 73 cm 1 was assigned to the shearing motion between two iodide layers, E g, whereas the vibration at 97 cm 1 corresponded to the symmetric stretch A 1g. [29 ] On the other hand, the Raman spectrum of the CH 3 NH 3 PbI 3 platelets shows a low-frequency vibration located at 13 cm 1 and a broad band featured at around 215 cm 1. The other vibration peaks of the perovskites are quite similar to that of lead iodide probably due to the similarity in their crystal structures. Nevertheless, the perovskite platelets that were converted from PbI 2 platelets showed identical peaks to that of a reference perovskite crystal (see additional text and Figure S3 in the Supporting Information) implying that it has the same tetragonal structure as that of a solution-grown perovskite crystal. [31 ] The optical absorption and photoluminescence of lead iodide and its perovskite were also characterized in individual platelets having similar thicknesses (18 nm for PbI 2 and 175 nm for CH 3 NH 3 PbI 3 ) to minimize the effect of the thickness on the optical density as shown in Figure 3 c. It is well-known that lead iodide has an optical absorption at around 5 nm whereas that of CH 3 NH 3 PbI3 is 77 nm. [34,35 ] Moreover, the absorption coefficient of perovskite is also much higher than that of lead iodide. Our data shows similar observation for the two platelets with identical thickness. In addition, after conversion, the platelet showed Adv. Optical Mater. 214, 2, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 841

5 COMMUNICATION strong photoluminescence (PL) at room temperature whereas the PL of PbI 2 could be obtained only at low temperatures (<2 K). Figure 3 c also shows the PL of a platelet before and after conversion at 77 K, which is consistent with the optical absorption spectrum. One of the properties that makes CH 3 NH 3 PbI 3 perovskite suitable for solar cell applications is the long diffusion length of its charge carriers, which can be characterized by time-resolved photoluminescence spectroscopy. [9,1 ] The lifetime of the charge carriers in the perovskite is exceptionally long so that they can reach the electrodes of the cells before recombination and therefore reduce the loss in power conversion. In order to verify this property of perovskites, we carried out time-resolved photoluminescence of PbI 2 and CH 3 NH 3 PbI 3 platelets. The results in Figure 3 d show that after conversion, the perovskite platelet has a PL lifetime that is more than 4 times higher than that of PbI 2. In summary, it is confirmed that the lead iodide platelet was successfully converted to perovskite by thermally intercalating methyl ammonium iodide. This approach can be applied to other lead halide perovskites even with a mixed halide composition as shown below. Figure 4 a shows the optical absorption and photoluminescence of different lead halide perovskites synthesized in a similar manner as the CH 3 NH3PbI 3 platelets above. The optical absorption reveals that the bandgaps for CH 3 NH 3 PbCl 3, CH 3 NH 3 PbBr3, and CH 3 NH 3 PbI 3 are at 3.1 ev (4 nm), 2.34 ev (53 nm), and 1.61 ev (77 nm), respectively, which is in good agreement with previous reports. [36,37 ] All perovskite compounds show a strong band-edge photoluminescence at room temperature. Figure 4 b displays the optical characterizations of the mixed halide perovskites prepared by intercalating different methyl ammonium halides (CH 3 NH 3 X, with X = Cl, Br, I) into the PbI 2 platelets. The results show that CH 3 NH 3 PbI 3 and CH 3 NH 3 PbIx Cl3- x have a broad absorption covering the entire visible range (4 75 nm), whereas CH 3 NH 3 PbIx Br 3- x only has a strong absorption in the range of 4 55 nm. This may partially explain why tri-iodide and iodide chloride perovskites are more suitable for solar cell applications. [1,2 ] The mixed chloride iodide perovskite also shows a stronger absorption in the near-uv regime whereas the pure iodide perovskite has a larger absorption near the 5 6 nm region. This result also suggests that if we use a combination of the perovskites in the absorption layer of solar cells, it would be possible to obtain a higher photo-to-electric conversion efficiency thanks to the higher absorption in the whole range of the visible spectrum. By using our synthesis strategy, it is possible to further tune the composition of the lead halide perovskite to obtain an optimal material for solar cell applications, such as co-intercalating a mixture of methyl ammonium halides into lead halide. Our simple method has shown the advantages of a high crystallinity as demonstrated by the characterizations discussed previously. In order to prove that our perovskite platelets exhibit a higher crystalline quality compared to conventional solution-prepared films, we measured the electron diffusion length in our platelets using CH 3 NH 3 PbI 3 as a case study. We believe that the charge generation and transportation in the perovskite layer are well-correlated with the order and quality of its crystal network. Recently, two groups have reported that the diffusion length for a solution-processed CH 3 NH 3 PbI 3 film is about 1 nm for both the electron and hole. [9,1 ] We characterized the electron Figure 4. Optical absorption (dashed line) and room temperature PL (solid line) of converted lead halide perovskite platelets. a) Optical properties of different lead halide perovskites (CH 3 NH 3 PbX 3 ) showing a bandgap of 4 nm for X = Cl, 53 nm for X = Br, and 77 nm for X = I, which are in good agreement with previous reports. b) Mixed halide perovskite platelets prepared by conversion of lead iodide platelets with different methyl ammonium halides (CH 3 NH3 X ). diffusion length in our CH 3 NH 3 PbI 3 platelets using phenyl- C61-butyric acid methyl ester (PCBM) as a quenching layer. Figure 5 displays the experimental results for the estimation of the electron diffusion length in our CH 3 NH 3 PbI3 nano-platelets. Figure 5 a shows the steady-state PL spectrum of a CH 3 NH 3 PbI 3 platelet with a thickness of 7 ± 5 nm with and without a PCBM layer. The thickness of the perovskite platelet used in the experiment was characterized by AFM as shown in Figure 5 b. Using a homogeneous platelet with a small deviation of about 7% the uncertainties of the diffusion length estimation arising from a large variation in the perovskite film thickness [9,1 ] could be reduced. Figure 5 c shows the time-resolved PL decay transient of the perovskite platelet with (purple dots) and without (green dots) a PCBM layer. By fitting the decay dynamics, the PL lifetime of CH 3 NH 3 PbI 3 ( τ ) and CH 3 NH 3 PbI 3 /PCBM (τ PL ) were found to be 6.8 ±.4, and.278 ±.4 ns, respectively. We then plotted the dependence curve of the charge-carrier diffusion length on the PL lifetime 842 wileyonlinelibrary.com 214 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Optical Mater. 214, 2,

6 COMMUNICATION Figure 5. Determination of electron-diffusion length in CH 3 NH 3 PbI 3 platelets. a) Time-integrated PL spectra of as-synthesized CH 3 NH3PbI 3 platelet on mica (black curve) and after coating with a PCBM layer (red curve). Inset: Optical image of the measured platelet. b) Thickness measurement of the platelet using AFM. c) Time-resolved PL decay transient measured at 76 ± 1 nm for CH 3 NH 3 PbI 3 platelet (green dot) and CH 3 NH 3 PbI 3 platelet/ PCBM (purple dot) after excitation at 4 nm. d) A plot of excitation length versus PL lifetime quenching ratios based on Equation S2 (see Supporting Information). The diffusion length is scaled in multiples of CH 3 NH3PbI 3 platelet thickness (7 nm). quenching ratio (Figure 5 d) obtained from an analytical model that was reported elsewhere. [1 ] (Detailed calculations can be found in the Supporting Information.) Using the same conservative approach, the electron-diffusion length was estimated to be 21 ± 5 nm, which is longer than the minimal estimated values of at least 1 nm reported earlier. [1 ] This longer diffusion length can be attributed to the high crystal quality of the perovskite platelet prepared by the present method. In conclusion, we have reported a facile method to prepare organic-based lead halide perovskite nano-platelets with a high crystal quality and good optical properties. This synthesis approach will create a new platform to exploit the physical properties of organic-based lead halide perovskites. The synthesized perovskite platelets can be readily applied to numerous applications, such as, single-crystal perovskite solar cells, lasing devices, LEDs, as well as other opto-electronic devices. Furthermore, this synthesis approach can also be applied to prepare perovskite films in planar solar cell configurations, which we believe will further boost the efficiency limits of solar cells. Experimental Section Synthesis of Lead Halide Platelets : Either PbI 2, or PbBr 2, or PbCl 2 powder (99.999%, Aldrich) was used as a single source and put into a quartz tube mounted on a single-zone furnace (Lindberg/Blue M TF5535C-1). The freshly cleaved muscovite mica substrate (1 cm 3 cm) was pre-cleaned with acetone and placed in the downstream region inside the quartz tube. The quartz tube was fi rst evacuated to a base pressure of 2 mtorr, followed by a 3 sccm flow of high purity Ar gas premixed with 5% H 2 gas. The temperature and pressure inside the quartz tube were set and stabilized to desired values for each halide (38 C and 2 Torr for PbI 2 ; 35 C and 75 Torr for PbBr 2 ; and 51 C and 2 Torr for PbCl 2 ). In all cases, the synthesis was carried out within 2 minutes and the furnace was allowed to cool down naturally to ambient temperature. Conversion of Lead Halides to Perovskites : The conversions were done using a similar CVD system. Methyl ammonium halides synthesized by a solution method (detailed synthesis method can be found in the Supporting Information) were used as a source and placed in the center of the quartz tube while mica substrates having as-grown lead halide platelets or nanowires were placed around 5 6 cm away from the center in the downstream region. The quartz tube was fi rst evacuated to a base pressure of 2 mtorr, followed by a 3 sccm flow of high purity Ar or N 2 gas. The pressure was then stabilized to 5 Torr and the temperature was elevated to 12 C and kept there for 1 hour after which the furnace was allowed to cool down naturally to ambient temperature. Characterizations : The structure of the as-grown samples was characterized using an optical microscope (Olympus BX51), AFM (Veeco Dimension V) in the tapping mode, fi eld-emission scanning electron microscopy (FE-SEM, JEOL JSM-71F), and X-ray powder diffraction (XRD, Bruker D8 advanced diffractometer, Cu Kα radiation) in the θ θ geometry. Absorption spectra were measured by a commercial transmission/reflectance micro-spectrometer (Craic 2/2). The linearly Adv. Optical Mater. 214, 2, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 843

7 COMMUNICATION polarized white light from a Xe lamp was focused onto the sample normally from the bottom. The transmitted light was collected by a reflective objective (36, numerical aperture:.4) and spectrally analyzed by a monochromator. An aperture was used to acquire the transmission of light from an area of 15 µm 15 µm, which was chosen to ensure adequate transmission flux and multiple measurements over the whole pattern. Raman spectra were obtained on a triple-grating micro-raman spectrometer (Horiba-JY T64). The signal was collected through a 1 objective, dispersed with a 18 g/mm grating, and detected by a liquid nitrogen cooled charge-coupled device. PL spectra were obtained from the same micro-raman spectrometer, but with a single-grating setup to improve effi ciency. For low-temperature PL measurements the samples were put into a cryostat in advance. The signal was collected through a 5 objective with a long focal length. If not specifi ed, the laser power was kept under.5 mw to avoid possible damage and oxidation on the samples. TRPL Measurements : For time-resolved PL measurements, frequency doubled pulses (4 nm) from a Coherent Mira titanium:sapphire oscillator (12 fs, 76MHz at 8 nm) was used as the excitation source. The time-resolved PL spectra were obtained using a streak camera system (Optronis GmbH) confi gured with a fast synchroscan sweep unit (FSSU1-ST) which had an ultimate temporal resolution of around 2 ps including jitter (or ca. 6 ps after coupling with a monochromator) at the fastest scan speed of 15 ps mm 1. Typical operating scan speeds in this work were 1 ps mm 1. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. Acknowledgements Q.X. gratefully thanks the Singapore National Research Foundation for a fellowship grant (NRF-RF29 6), a Competitive Research Program (NRF-CRP ), the Ministry of Education for an AcRF tier2 grant (MOE212-T2 2 86) and Nanyang Technological University via a start-up grant support (M581161). T.C.S acknowledges the fi nancial support from the NTU start-up grant M48514, SPMS collaborative Research Award M48536, Ministry of Education AcRF Tier 2 grant MOE213-T and a Competitive Research Program NRF-CRP Q.X. and T.C.S also acknowledge the funding of this research program/project by the National Research Foundation (NRF), Prime Minister s Offi ce, Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE) program. Received: March 12, 214 Revised: April 13, 214 Published online: May 23, 214 [1] M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami, H. J. Snaith, Science 212, 338, 643. [2] J. Burschka, N. Pellet, S. J. Moon, R. H. Baker, P. Gao, M. K. Nazeeruddin, M. Grätzel, Nature 213, 499, 316. [3] M. Liu, M. B. Johnston, H. J. Snaith, Nature 213, 51, 395. [4] J. H. Heo, S. H. Im, J. H. Noh, T. N. Mandal, C. S. Lim, J. A. Chang, Y. H. Lee, H. J. Kim, A. Sarkar, M. K. Nazeeruddin, M. Grätzel, S. I. Seok, Nat. Photonics 213, 7, 486. [5] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, J. Am. Chem. Soc. 29, 131, 65. [6] J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal, S. I. Seok, Nano Lett. 213, 13, [7] L. Etgar, P. Gao, Z. S. Xue, Q. Peng, A. K. Chandiran, B. Liu, M. K. Nazeeruddin, M. Grätzel, J. Am. Chem. Soc. 212, 134, [8] B. Cai, Y. Xing, Z. Yang, W. H. Zhang, J. S. Qiu, Energy Environ. Sci. 213, 6, 148. [9] S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, H. J. Snaith, Science 213, 342, 341. [1] G. C. Xing, N. Mathews, S. Y. Sun, S. S. Lim, Y. M. Lam, M. Grätzel, S. Mhaisalkar, T. C. Sum, Science 213, 342, 344. [11] C. R. Kagan, D. B. Mitzi, C. D. Dimitrakopoulos, Science 1999, 286, 945. [12] K. Chondroudis, D. B. Mitzi, Chem. Mater. 1999, 11, 328. [13] T. Hattori, T. Taira, M. Era, T. Tsutsui, S. Saito, Chem. Phys. Lett 1996, 254, 13. [14] L. C. Schmidt, A. Pertegas, S. G. Carrero, O. Malinkiewicz, S. Agouram, G. M. Espallargas, H. J. Bolink, R. E. Galian, J. P. Prieto, J. Am. Chem. Soc. 214, 136, 85. [15] D. B. Mitzi, M. T. Prikas, K. Chondroudis, Chem. Mater. 1999, 11, 542. [16] K. Liang, D. B. Mitzi, M. T. Prikas, Chem. Mater. 1998, 1, 43. [17] D. B. Mitzi, Chem. Mater. 21, 13, [18] J. Zhang, Z. Peng, A. Soni, Y. Y. Zhao, Y. Xiong, B. Peng, J. B. Wang, M. S. Dresselhaus, Q. H. Xiong, Nano Lett. 211, 11, 247. [19] N. Chen, S. H. Chen, C. B. Ouyang, Y. W. Yu, T. F. Liu, Y. J. Li, H. B. Liu, Y. L. Li, NPG Asia Mater. 213, 5, 1. [2] P. Krogstrup, H. I. Jørgensen, M. Heiss, O. Demichel, J. V. Holm, M. Aagesen, J. Nygard, A. F. i Morral, Nat. Photonics 213, 7, 36. [21] D. H. Li, J. Zhang, Q. H. Xiong, ACS Nano 212, 6, [22] J. Zhang, D. H. Li, R. J. Chen, Q. H. Xiong, Nature 213, 493, 54. [23] M. I. B. Utama, M. de la Mata, C. Magen, J. Arbiol, Q. H. Xiong, Adv. Funct. Mater. 213, 23, [24] M. I. B. Utama, Z. P. Peng, R. Chen, B. Peng, X. L. Xu, Y. J. Dong, L. M. Wong, S. J. Wang, H. D. Sun, Q. H. Xiong, Nano Lett. 211, 11, 351. [25] M. Gillet, R. Delamare, E. Gillet, J. Cryst. Growth 25, 279, 93. [26] I. Matolınova, M. Gillet, E. Gillet, V. Matolın, Nanotechnology 29, 2, [27] Z. X. Wang, T. Kong, K. Zhang, H. L. Hu, X. P. Wang, J. G. Hou, J. Chen, Mater. Lett. 27, 6, 251. [28] M. I. B. Utama, F. J. Belarre, C. Magen, B. Peng, J. Arbiol, Q. H. Xiong, Nano Lett. 212, 12, [29] G. K. Kasi, N. R. Dollahon, T. S. Ahmadi, J. Phys. D: Appl. Phys. 27, 4, [3] G. A. Ozin, Can. J. Chem. 197, 48, [31] T. Baikie, Y. Fang, J. M. Kadro, M. Schreyer, F. X. Wei, S. G. Mhaisalkar, M. Grätzel, T. J. White, J. Mater. Chem. A 213, 1, [32] Q. Chen, H. Zhou, Z. Hong, S. Luo, H. S. Duan, H. H. Wang, Y. S. Liu, G. Li, Y. Yang, J. Am. Chem. Soc. 214, 136, 622. [33] Z. Zheng, A. Liu, S. Wang, Y. Wang, Z. S. Li, W. M. Lau, L. Z. Zhang, J. Mater. Chem. 25, 15, [34] M. R. Tubbs, Phys. Stat. Sol. 1972, 49, 11. [35] C. C. Stoumpos, C. D. Malliakas, M. G. Kanatzidis, Inorg. Chem. 213, 52, 919. [36] N. Kitazawa, Y. Watanabe, Y. Nakamura, J. Mater. Sci. 22, 37, [37] K. Tanaka, T. Takahashi, T. Ban, T. Kondo, K. Uchid, N. Miura, Sol. Stat. Comm. 23, 127, wileyonlinelibrary.com 214 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Optical Mater. 214, 2,

8 Copyright WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 214. Supporting Information for Advanced Materials., DOI: 1.12/adom Synthesis of Organic Inorganic Lead Halide Perovskite Nanoplatelets: Towards High-Performance Perovskite Solar Cells and Optoelectronic Devices Son Tung Ha, Xinfeng Liu, Qing Zhang, David Giovanni, Tze Chien Sum, and Qihua Xiong*

9 Copyright WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 213. Supporting Information Organic-inorganic lead halide perovskite nano-platelets: Towards high performance perovskite solar cells and opto-electronic devices Son Tung Ha, Xinfeng Liu, Qing Zhang, David Giovanni, Tze Chien Sum and Qihua Xiong * Materials and methods Synthesis of methyl ammonium halide: Methyl ammonium halides were synthesized in a 5 ml 3-neck round flask equipped with a thermometer, a condenser, and a dropping funnel. First, 11.29g (.12 mol) of 33 wt % in ethanol solution of methylamine (CH 3 NH 2, Aldrich) was charged into the flask. The solution was then emerged in an ice bath and stirred by a magnetic stirrer. Next, hydrohalic acid solution- HX (23.26 g of 55 wt % HI solution in water(aldrich); g of 48 wt % HBr solution in water(aldrich); or g of 37 wt% HCl solution in water (Aldrich)) was slowly dropped into the vessel through the dropping funnel. The temperature was controlled below 1 C throughout the reaction by controlling the dropping rate of HX solution and ice feeding to the ice bath. After completion of dropping, the reaction was further maintained at 2 C for 2 more hours. The solvents were then evaporated at 78 C until 1/3 of original volume remained. The precipitates were filtered and recrystallized in diethyl ether/methanol mixture (1:1 v/v) two times. The white crystals were dried at 6 C for 24 hours and stored in a desiccator. Solution synthesis of CH 3 NH 3 PbI 3 crystal for Raman comparison In order to compare Raman spectra of the perovskite synthesized by our CVD method and the bulk crystal grown by solution method, we synthesized CH 3 NH 3 PbI 3 following the procedure published elsewhere. [1] First, CH 3 NH 3 I (.1 mol), which was synthesized by the procedure

10 described above, and PbI 2 (.1 mol) were mixed in 8 ml -butyrolactone in a 5-mL round flask. The mixture was stirred at 6 C over night to obtain a transparent yellow solution. The solution was then transferred to a glass Petri dish and placed onto a hot-plate at 1 C. The black crystals were formed as -butyrolactone completely evaporated. XRD was used to characterize crystal structure of the perovskite as shown in Figure S2. The XRD pattern reveals tetragonal structure of CH 3 NH 3 PbI 3 which is agreed well with the previous report. [2] Calculation of electron diffusion length in CH 3 NH 3 PbI 3 platelet The optical transmittance and total reflectance spectra of CH 3 NH 3 PbI 3 platelet on mica substrate and blank mica substrate were measured using micro-spectrometer (Craic 2/2) with 36 objective and an aperture of 15 m 15 m. The absorption coefficient of CH 3 NH 3 PbI 3 platelet on mica substrate was calculated using equation below: [3] plate = d mica d total 1 d plate ln 1 R total T total 1 d mica ln 1 R mica T mica (S1) where plate, d plate are the absorption coefficient and thickness of CH 3 NH 3 PbI 3 platelet, respectively. R mica, T mica and d mica are the substrate reflectance, transmittance and thickness, respectively. R total, T total and d total are the reflectance, transmittance and thickness of mica/ch 3 NH 3 PbI 3 platelet, respectively. The electron diffusion length was calculated based on diffusion model which was described elsewhere [3] using equation below: N(t) = 2n L π exp ( kt) m = (exp ( π2 D (m + 1 ) 2 t) exp ( αl)π m ( 1)m αl L 2 2 ((αl) 2 +π 2 (m+ 1 2 )2 )(m ) (S2) where

11 N(t): total charge number within perovskite layer n : initial charge carrier density L: perovskite platelet thickness k: original charge carrier consumption rate without PCBM D: charge carrier diffusion coefficient : absorption coefficient (calculated from equation S1) 1 nm (a) 1 (b) 2 1 nm -1-2 um Thickness (nm) Position (um) 1 um um Position (um) um um Thickness (nm) 1 Figure S1. AFM data of PbBr 2 (a) and PbI 2 (b) showing the surface roughness of grown platelets. Table S1. Assignments for low frequency Raman peaks of lead halide platelets. [4,5] Lead halide Frequency (cm -1 ) Symmetry PbI PbBr E g A 1g B 2g B 2g PbCl A 1g + B 2g?? B 2g

12 a Mica peaks CH 3 NH 3 PbBr 3 peaks A 1g + B 2g + (B 3g?) Normalized intensity(a.u) (deg) b Normalized intensity(a.u) Mica peaks CH 3 NH 3 PbCl 3 peaks (deg) Figure S2. XRD pattern of CH 3 NH 3 PbBr 3 (a) and CH 3 NH 3 PbCl 3 (b) nanoplatelets prepared by converting PbX 2 (X = Br, Cl) nanoplatelets with its corresponding methyl ammonium halide (CH 3 NH 3 X). The XRD data shows that after conversion, the identical peak of PbBr 2 at 22.1 (2 peak) and that of PbCl 2 at 23.2 (2 peak) disappear which means the conversion has been completed. The appearance of perovskite peaks are in good agreement with literature. [6]

13 Intensity (a.u) Intensity (a.u) 4.5 mm (deg) CH 3 NH 3 PbI 3 PbI (deg) Figure S3. XRD pattern of solution grown CH 3 NH 3 PbI 3 crystal as a reference for Raman spectrum comparison in Fig.3b. Inset: zoom-in XRD pattern of CH 3 NH 3 PbI 3 and optical image of the grown single crystal CH 3 NH 3 PbI 3. References [1] J. H. Heo, S. H. Im, J. H. Noh, T. N. Mandal, C. S. Lim, J. A. Chang, Y. H. Lee, H. J. Kim, A. Sarkar, M. K. Nazeeruddin, M. Grätzel and S. I. Seok, Nature Photonics 213, 7, 486. [2] T. Baikie, Y. Fang, J. M. Kadro, M. Schreyer, F. X. Wei, S. G. Mhaisalkar, M. Grätzel and T. J. White, J. Mater. Chem. A 213, 1, [3] G. C. Xing, N. Mathews, S. Y. Sun, S. S. Lim, Y. M. Lam, M. Grätzel, S. Mhaisalkar, T. C. Sum, Science 213, 342, 344. [4] G. K. Kasi, N. R. Dollahon and T. S. Ahmadi, J. Phys. D: Appl. Phys. 27, 4, [5] G. A. Ozin, Canadian J. of Chem. 197, 48, [6] B. Cai, Y. Xing, Z. Yang, W. H. Zhang, J. Qiu, Energy Environ. Sci. 213, 6, 148.