Materials Chemistry A

Transcription

1 Journal of Materials Chemistry A COMMUNICATION View Article Online View Journal View Issue Cite this: J. Mater. Chem. A, 2015,3, 9081 Received 29th October 2014 Accepted 30th October 2014 DOI: /c4ta05819d Formation chemistry of perovskites with mixed iodide/chloride content and the implications on charge transport properties Tsz-Wai Ng, ab Chiu-Yee Chan, a Ming-Fai Lo,* ab Zhi Qiang Guan a and Chun-Sing Lee* ab Although Cl-doping is a common technique for achieving high photovoltaic (PV) performance, the Cl content is negligibly small and cannot easily be tuned. Therefore, we herein study the formation chemistry of Cl-doped perovskites by examining the chemical interactions between thermally evaporated MAI and PbCl 2 through X-ray photoemission spectroscopy (XPS). We show that PbCl 2 is not stable at the MAI/PbCl 2 contact surface. The Cl atom readily detaches from the PbCl 2, which subsequently initiates electron transition from Pb to MAI. Via thermal-evaporation, a perovskite with a high PbCl 2 content can be prepared and examined. We found that the presence of metallic Pb, associated with increased Cl content, can quench the photogenerated exciton in PV devices. By optimizing the ratio of MAI : PbCl 2, a perovskite solar cell with 6% efficiency was obtained. Introduction Organolead halide perovskites have recently emerged as a new revolutionary light absorber for photovoltaic (PV) devices. 1 3 Despite continuous breakthroughs in power conversion efficiencies, fundamental understanding of perovskite solar cells is limited. 3,4 Recent studies showed that the perovskite device performances are dominated by the intrinsic properties of the perovskite layer. 5 7 It is thus important to control and custom tune the properties of perovskites Methylammonium (MA is the abbreviation for CH 3 NH 3 ) lead iodide (MAPbI 3 ) is a typical perovskite for photon harvesting. Fully or partially replacing iodine in this model structure with other halide atoms (e.g. boride/chloride, and iodide/chloride) can open the band gap from 1.57 to 2.29 ev. 11,12 Perovskites with a mixed halide content of iodide/chloride (MAPbI 3 x Cl x ) are a Center of Super-Diamond and Advanced Films (COSDAF), Department of Physics and Materials Science, City University of Hong Kong, Hong Kong SAR, P. R. China. ming o@cityu.edu.hk; apcslee@cityu.edu.hk b City University of Hong Kong Shenzhen Research Institute, Shenzhen, P. R. China Electronic supplementary information (ESI) available. See DOI: /c4ta05819d reported to show good PV performance even in devices using a simple planar structure PV devices consisting of MAPbI 3 x Cl x show remarkably high open circuit voltages 8 and extraordinary long charge transport distances of over 1 mm. 11,17 These impressive ndings aroused recent interest in the Cldoped perovskites. 18,19 Since then, there have been attempts to increase the Clcontent by increasing the ratio of PbCl 2 /MAI in the precursor solution. 18 However, the Cl content is still negligibly small 19,20 and the ratio of Cl/I in the resulting perovskite lms is difficult to tune. Despite the large amount of recent publications on MAPbI 3 x Cl x based PV devices, there are relatively few studies on its chemical structure, mechanism of formation and charge transport. 11,21 23 Herein, we perform a surface analysis-based study of the formation chemistry of MAPbI 3 x Cl x by investigating the charge interaction between MAI and PbCl 2. The implications of the MAI : PbCl 2 ratio in the precursor solution on the charge transport properties of MAPbI 3 x Cl x are discussed. Experimental Photoemission studies All photoemission analyses were performed using a VG ESCA- LAB 220i-XL surface analysis system equipped with a monochromatic Al-Ka X-ray gun giving photons of ev for XPS investigation. 24 The base pressure of the system is 10 9 Torr. The evaporation source of MAI (from Luminescence Technology Corp.) and the PbCl 2 (from Sigma Aldrich) were used asreceived. A bilayer lm of MAI/PbCl 2 is in situ deposited on an ITO glass substrate by thermal evaporation; while mixed lms of MAI : PbCl 2 are prepared by co-evaporation from PbCl 2 and MAI. The formed lms are then transferred to a preparation chamber where they are annealed at 100 C for 1 hour without vacuum break. The prepared lms are nally transferred to the analysis chamber for XPS measurements. XPS core level spectra of Pb 4f (signal from PbCl 2 ) and I 3d (signal from MAI) show spin orbit splitting. Their peaks are tted according to a This journal is The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015,3,

2 View Article Online Journal of Materials Chemistry A Gaussian Lorentzian peak shape along with a Shirley background. The pairs of Pb (i.e. Pb 4f 7/2 and Pb 4f 5/2 ) and I (i.e. I3d 3/2 and I 3d 5/2 ) peaks are tted with a constrained intensity ratio of 4 : 3 and 2 : 3, respectively. Device fabrication The patterned and routinely cleaned indium tin-oxide (ITO) glass substrate is ultraviolet ozone treated for 15 min. Perovskite-based solar cells with a con guration of ITO/perovskite (MAI : PbCl 2 ratio 0.5 : 1, 1 : 1, 1.5 : 1, 2.3 : 1 and 4 : 1, 150 nm)/c 60 (40 nm)/bphen (8 nm)/al (80 nm) are fabricated. The MAI (from Lumtech.) and PbCl 2 (from Sigma Aldrich) are used as-received. The perovskite lms are prepared by co-evaporation of PbCl 2 and MAI onto ITO coated glass substrates, followed by annealing at 100 C for 1 hour. A er cooling to room temperature, the organic layers are deposited. The Al cathode (80 nm) is deposited by thermal evaporation through a shadow mask. All of the lm thicknesses are monitored using an INFICON crystal monitor with the thickness calibrated using an ellipsometer. The active device area is 0.1 cm 2. The devices are encapsulated in a glove box immediately a er fabrication without ambient exposure. More details of the device fabrication can be found in the ESI. 25 The performances of the devices are measured using an Oriel 150 W solar simulator with AM1.5G (AM: air mass, G: global) lters at 100 mw cm 2. Results and discussion Formation chemistry of the perovskites at the atomic level To examine the chemical interaction between MAI and PbCl 2, we performed a surface study of the MAI/PbCl 2 interface using XPS. Simple bilayer lms of MAI/PbCl 2 are thermally evaporated onto an ITO substrate for examination. Chemical changes and electronic interactions are monitored at each incremental lm deposition. Fig. 1a shows the XPS Cl 2p core level spectra of PbCl 2 on an ITO/MAI (15 nm) substrate with increasing thickness. It is intriguing that no Cl signal can be measured during the initial deposition (0.2 to 0.5 nm) of PbCl 2 on the MAI surface, taking into account the detection limit of the XPS measurement Communication (0.1 atm%). 9 The Cl signal is only observed a er the coverage of the PbCl 2 lm reaches over 1.0 nm thick. Fig. 1b shows the Cl content near the MAI/PbCl 2 surface. The Cl : Pb ratio gradually increases from 0 to 1.4 as the PbCl 2 spreads away from the MAI surface. We repeat the experiment, replacing the ITO/MAI (15 nm) substrate with a bare ITO substrate. The deposited PbCl 2 lm shows a stoichiometric Cl : Pb ratio of 2 even at a thickness as thin as 0.2 nm. Fig. S1 compares the Pb 4f and Cl 2p XPS core level spectra of PbCl 2 lms prepared on ITO/MAI and ITO substrates. While PbCl 2 deposited on bare ITO shows a strong Cl signal, the same lm prepared on the ITO/MAI substrate shows a negligible Cl signal (lower line, Fig. S1b ). This suggests that no Cl loss occurs during thermal evaporation of the PbCl 2 source. The absence of Cl in Fig. 1 is therefore attributed to interactions between PbCl 2 and MAI. This observation is consistent with the theoretical calculation by Colella et al. The structural mismatch due to the abrupt difference in the ionic radii of Cl and I ions hinders the incorporation of Cl ions in the network of MAPbI 3 perovskites. 18 Although Cl is energetically unfavorable to be formed at the MAI/PbCl 2 surface, it can sit as an inclusion somewhere >1.0 nm away from MAI/PbCl 2 surface as shown in Fig. 1b. We continue our discussion on the structural chemistry of perovskites by examining the core level XPS spectra of different elements at the MAI/PbCl 2 contact surface. Fig. 2a compares the Pb 4f core level spectra for 0.2 and 10.0 nm PbCl 2 deposited on an ITO/MAI (15.0 nm) substrate (bottom line). When 0.2 nm of PbCl 2 is deposited on the MAI surface, the Pb 4f 7/2 and the Pb 4f 5/2 peaks are observed at and ev, respectively. Fig. 1 (a) The XPS Cl 2p core level spectra with increasing PbCl 2 thickness on an ITO/MAI (15.0 nm) substrate. (b) The atomic ratio of Cl/ Pb from different thicknesses of PbCl 2 deposited on ITO/MAI and ITO surfaces. Fig. 2 High resolution XPS core level spectra comparing the in situ thermally prepared PbCl 2 on an ITO/MAI surface with the ITO/MAI surface. (a) Pb 4f spectra from the PbCl 2 and (b) I 3d, (c) C 1s, and (d) N 1s spectra from the MAI, are shown respectively J. Mater. Chem. A, 2015,3, This journal is The Royal Society of Chemistry 2015

3 View Article Online Communication Journal of Materials Chemistry A Their positions shi ed by 0.9 ev, towards the high binding energy (BE) side, when the PbCl 2 thickness was increased to 10 nm. Similar peak shi s in Fig. 2b are observed for the I 3d core level peaks, for which the signals come from the underlying MAI lm. The 15.0 nm MAI shows I 3d 7/2 and I 3d 5/2 peaks at and ev, respectively. These peaks progressively shi by 0.8 ev, to the high BE side, upon depositing 10.0 nm of PbCl 2.BothXPScorelevelpeakshi s for the Pb 4f (signal from PbCl 2 ) and I 3d (signal from MAI) suggest a strong interfacial charge transfer at the MAI/PbCl 2 contact surface. Neither the Pb 4f nor the I 3d spectra show any new peaks, indicating there is no new chemical bond formation. 26 On the other hand, it is interesting to note that both C 1s and N 1s show negligible shi s (Fig.2candd). Pristine MAI shows C 1s peaks with two C 1s peaks located at and ev, from C C andc N bondsrespectively. Although both of the two C 1s peaks show no energy shi, their intensity ratio changes obviously upon the PbCl 2 deposition. Some of the C N bonds change to C C bonds during the contact of PbCl 2 with MAI. 19 Fig. 3 summarizes the net energy shi of the core level signals from both MAI (red) and PbCl 2 (blue) during the MAI/ PbCl 2 interface formation. This gure depicts the chemical response of the different elements during the MAI/PbCl 2 contact formation. Pb and I are the only elements showing a vigorous shi and chemical response during the MAI/PbCl 2 contact formation. Fig. 4 schematically illustrates the structural formation of the perovskites and the electron transfer that occurs when PbCl 2 meets MAI. When PbCl 2 is in contact with MAI molecules, the Cl atoms will detach from the Pb atom due to a lattice mismatch with the I atom. 18 The Pb, with excess valence electrons, would then subsequently pass its electrons to the I in MAI. During the electron transfer process, the Pb atom retains its oxidation state of 2+ with no sign of further oxidization or reduction, as no new peaks emerged in the Pb 4f spectra (Fig. 2a). 26 Meanwhile, a er getting an additional Fig. 4 A schematic diagram illustrating the core level charge transfer and chemical interaction at the MAI/PbCl 2 contact surface. electron from Pb, the charge balance in the MAI is regained by the breaking of some of its C N bonds, and the formation of C C bonds (or C species). Although the interfacial surface studies of the MAI/PbCl 2 bilayer lm (Fig. 2) can disclose the detailed formation chemistry of the perovskite lm, it might not totally re ect the charge interactions in a mixed MAI : PbCl 2 lm a er annealing at 100 C. We therefore also examined the chemical structure of a mixed MAI : PbCl 2 (3 : 1) lm before and a er annealing at 100 C for 1 hour. The XPS results are shown in Fig. S2 in the ESI. The XPS spectra of the MAI : PbCl 2 mixed lms are similar to those for the MAI/PbCl 2 bilayer contact (Fig. 2). Most importantly, no Cl signal can be detected even when the MAI : PbCl 2 mixed lm is thick (15 nm). This indicates that the mixed lm and the bilayer share similar chemistry, and also that the 100 C annealing has no observable effect on the chemical bond structure. 22 Fig. 3 The energy shifts observed in the different elemental core level peaks. Solid symbols show the elements from MAI, while the open symbols show the elements from PbCl 2. Impact of increasing the PbCl 2 precursor content We found that the Cl content in the thermally prepared perovskite lm can be more easily tuned compared to that from solution processing. 27 Fig. 5a shows the Cl 2p signals from annealed MAI : PbCl 2 (i.e. perovskite) lms with different MAI : PbCl 2 ratios. Fig. 5b compares the Pb 4f core level spectra of the perovskites lms with MAI : PbCl 2 ratios of 1 :1,2 :1 and 3 : 1. In these perovskite samples, the signals due to the main pair of peaks for Pb 4f 7/2 and 4f 5/2 are found at and ev, respectively. This corresponds to Pb atoms with oxidation states of +2. In the lm prepared with a high PbCl 2 :MAI ratio of 1 : 1, an additional pair of Pb peaks emerges at (Pb 4f 7/2 ) and ev (Pb 4f 5/2 ). These peaks are attributed to a Pb atom with an oxidation state of 0, suggesting that a small amount of Pb 2+ is reduced to Pb metal. This observation matches well with the theoretical simulation reported by Miller et al. 28 This journal is The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015,3,

4 View Article Online Journal of Materials Chemistry A Communication Fig. 5 XPS core level spectra for the (a) Cl 2p and (b) Pb 4f of the perovskite films with MAI : PbCl 2 ratios of 3 : 1, 2 : 1 and 1 : 1, respectively. In order to study the effectofincreasingtheclcontentin perovskites, perovskite-based solar cells with a con guration of ITO/perovskite 150 nm (MAI : PbCl 2 ratios 0.5 : 1, 1 : 1, 1.5 : 1, 2.3 : 1, 4 : 1)/C nm/bphen 8 nm/al were fabricated. Fig. 6 shows an XRD pattern (top line) of a perovskite lm deposited on an ITO substrate. The pattern shows diffraction peaks at 14,28,and43 that correspond respectively to the (110), (220), and (330) plane of perovskite. 4 It is noted that neither MAI nor PbCl 2 signals can be detected in the perovskite lm. In addition, our XPS data show that no detectable contamination occurred during our fabrication process. The perovskite lm shows a characteristic photoluminescence (PL) emission at 780 nm (Fig. S3 ). It is noted that the major PL emission at 780 nm is decreased with increasing doping ratio. Also, a blue shi in the PL emission peak is found on increasing the PbCl 2 content. This is a common sign of undesirable impurities (i.e. excess Pb metal), which form a new radiation pathway for the lm. 29 All these characterization data show that the thermally evaporated perovskite lm has the same crystal structure, PL emission and absorption as reported by other groups. 4,13 Fig. 6b compares the conductivities of the perovskite devices with different MAI : PbCl 2 ratios. When the MAI : PbCl 2 ratio changes from 1 :1 to 4 :1, the dark current of the formed devices increases by two orders of magnitude. When the PbCl 2 content is further increased (i.e. a MAI : PbCl 2 ratio of 0.5 : 1), the device shows a large leakage current with no diode behavior. The corresponding J V characteristics under illumination are shown in Fig. 6c. It is noted that the device with a MAI : PbCl 2 ratio of 2.3 : 1 shows the best performance with a PCE up to 6.1%. The detailed photoresponses and the external quantum efficiency (EQE) of the formed devices are summarized in Table 1 and Fig. S4, respectively.fig.6dshowstherelationshipbetweenv oc and the dark current (obtained at 1 V for the Fig. 6b). When the MAI : PbCl 2 ratio changes from 4 : 1 to 1 : 1, the device s V oc is found to decrease with an increasing dark current. This observation is consistent with the theoretical calculation using the diode equation, which shows that the dark current is inversely proportional to the V oc. Although a perovskite lm with a higher PbCl 2 content can show better conductivity, an excessive PbCl 2 content can degrade the PV performance. The formation of metallic Pb could quench the photocurrent in the device by providing charge recombination sites, leading to a high dark current (Fig. 6b) and poor PV device performance (Fig. 6c). It is noted that the morphology may affect the device performance. We thus also examined the surface morphologies of the perovskite lms with MAI : PbCl 2 ratios of 2.3 : 1 and 1 : 1 using atomic force microscopy (AFM) and scanning electron microscopy (SEM). The two perovskite lms with MAI : PbCl 2 ratios of 2.3 : 1 and 1 : 1 show a similar morphology with a root-mean-square (R.M.S.) roughness of 1 nm (Fig. S5 and S6 ). No further morphology differences were observed when C 60 /BCP was deposited (Fig. S6 ). Both the AFM and SEM results suggested that the doping process has no signi cant impact on the lm morphology. The notable large leakage current in Fig. 6d is not attributed to any morphological issue (e.g. valley or protrusion), but to the increased PbCl 2 content. By optimizing the MAI : PbCl 2 ratio, an OPV device with an energy conversion efficiency of 6% was obtained. Table 1 Photovoltaic responses of the ITO/perovskite/C 60 /Bphen/Al devices with different MAI : PbCl 2 ratios MAI : PbCl 2 ratio J a sc (ma cm 2 ) V a oc (V) FF a PCE a (%) Fig. 6 (a) XRD patterns for comparing the perovskite, the PbCl 2, the MAI and the ITO substrate. The current density voltage (J V) characteristics of the perovskite devices with MAI : PbCl 2 ratios of 4 : 1, 3 : 1, 2 : 1 and 1 : 1 in (b) dark and (c) illumination conditions. (d) The relationship between V oc and the dark current density for the corresponding devices. 1 : : : : a J sc : short circuit current density; V oc : open circuit voltage; FF: ll factor; PCE: power conversion efficiency J. Mater. Chem. A, 2015,3, This journal is The Royal Society of Chemistry 2015

5 View Article Online Communication Conclusion The formation chemistry of a MAPbI 3 x Cl x perovskite from its precursor materials, MAI and PbCl 2, was examined using XPS. Due to the differences in ionic radii of Cl and I ions, the Cl leaves the Pb atom when PbCl 2 meets with the MAI molecules. The Pb transfers its excess electrons to the MAI during the MAI/ PbCl 2 contact formation. As a result, a negligible amount of Cl could be observed close to the MAI/PbCl 2 contact surface. By controlling the thermal co-evaporation ratio of the MAI and PbCl 2, the PV effect of perovskites with a high PbCl 2 content was also examined. The excess Pb ions, accompanied with an increasing Cl content, would lead to partial reduction of the Pb ions to Pb metal. The metallic Pb is considered as the cause of the reduced PV performance in some perovskite lms prepared with a high content of PbCl 2. Acknowledgements This work was supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project no. T23-713/11); and the National Natural Science Foundation of China (no ). References 1 I. Chung, B. Lee, J. Q. He, R. P. H. Chang and M. G. Kanatzidis, Nature, 2012, 485, M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami and H. J. Snaith, Science, 2012, 338, J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Gratzel, Nature, 2013, 499, P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon and H. J. Snaith, Nat. Commun., 2013, 4, E. Edri, S. Kirmayer, S. Mukhopadhyay, K. Gartsman, G. Hodes and D. Cahen, Nat. Commun., 2014, 5, V. Gonzalez-Pedro, E. J. Juarez-Perez, W. S. Arsyad, E. M. Barea, F. Fabregat-Santiago, I. Mora-Sero and J. Bisquert, Nano Lett., 2014, 14, H. S. Kim, I. Mora-Sero, V. Gonzalez-Pedro, F. Fabregat- Santiago, E. J. Juarez-Perez, N. G. Park and J. Bisquert, Nat. Commun., 2013, 4, E. Edri, S. Kirmayer, D. Cahen and G. Hodes, J. Phys. Chem. Lett., 2013, 4, E. Edri, S. Kirmayer, M. Kulbak, G. Hodes and D. Cahen, J. Phys. Chem. Lett., 2014, 5, 429. Journal of Materials Chemistry A 10 G. E. Eperon, S. D. Stranks, C. Menelaou, M. B. Johnston, L. M. Herz and H. J. Snaith, Energy Environ. Sci., 2014, 7, B. Suarez, V. Gonzalez-Pedro, T. S. Ripolles, R. S. Sanchez, L. Otero and I. Mora-Sero, J. Phys. Chem. Lett., 2014, 5, J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal and S. I. Seok, Nano Lett., 2013, 13, M. Z. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, Q. Chen, H. P. Zhou, Z. R. Hong, S. Luo, H. S. Duan, H. H. Wang, Y. S. Liu, G. Li and Y. Yang, J. Am. Chem. Soc., 2014, 136, G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and H. J. Snaith, Adv. Funct. Mater., 2014, 24, J. Y. Jeng, Y. F. Chiang, M. H. Lee, S. R. Peng, T. F. Guo, P. Chen and T. C. Wen, Adv. Mater., 2013, 25, S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza and H. J. Snaith, Science, 2013, 342, S. Colella, E. Mosconi, P. Fedeli, A. Listorti, F. Gazza, F. Orlandi, P. Ferro, T. Besagni, A. Rizzo, G. Calestani, G. Gigli, F. De Angelis and R. Mosca, Chem. Mater., 2013, 25, H. Yu, F. Wang, F. Xie, W. Li, J. Chen and N. Zhao, Adv. Funct. Mater., 2014, DOI: /adfm E. Edri, S. Kirmayer, A. Henning, S. Mukhopadhyay, K. Gartsman, Y. Rosenwaks, G. Hodes and D. Cahen, Nano Lett., 2014, 14, D. Bi, L. Yang, G. Boschloo, A. Hagfeldt and E. M. Johansson, J. Phys. Chem. Lett., 2013, 4, C. Bi, Y. Shao, Y. Yuan, Z. Xiao, C. Wang, Y. Gao and J. Huang, J. Mater. Chem. A, 2014, DOI: /c4ta04007d. 23 Z. Xiao, C. Bi, Y. Shao, Q. Dong, Q. Wang, Y. Yuan, C. Wang, Y. Gao and J. Huang, Energy Environ. Sci., 2014, 7, T. W. Ng, M. F. Lo, M. K. Fung, W. J. Zhang and C. S. Lee, Adv. Mater., 2014, 26, H. W. Mo, M. F. Lo, Q. D. Yang, T. W. Ng and C. S. Lee, Adv. Funct. Mater., 2014, 24, P. Schulz, E. Edri, S. Kirmayer, G. Hodes, D. Cahen and A. Kahn, Energy Environ. Sci., 2014, 7, L. E. Polander, P. Pahner, M. Schwarze, M. Saalfrank, C. Koerner and K. Leo, APL Mater., 2014, 2, E. M. Miller, Y. Zhao, C. C. Mercado, S. K. Saha, J. M. Luther, K. Zhu, V. Stevanović, C. L. Perkins and J. van de Lagemaat, Phys. Chem. Chem. Phys., 2014, 16, H. Tetsuka, H. Takashima, K. Ikegami, H. Nanjo, T. Ebina and F. Mizukami, Chem. Mater., 2008, 21, 21. This journal is The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015,3,

6 Electronic Supplementary Material (ESI) for Journal of Materials Chemistry A. This journal is The Royal Society of Chemistry 2014 Supporting Document: Figure S1. High resolution XPS (a) Pb 4f and (b) Cl 2p core level spectra comparing the thermal evaporated 5 nm PbCl 2 formed on ITO/MAI (15.0 nm) and bare ITO substrate respectively. While deposition of PbCl 2 on ITO is stoichiometric with Cl:Pb ratio of 2.0, the same deposition of PbCl 2 on ITO/MAI(15.0 nm) shows no measurable Cl content.

7 Figure S2. High resolution XPS (a) Pb 4f, (b) Cl 2p, (c) C 1s, (d) N 1s and (e) I 3d XPS core level spectra comparing the MAI:PbCl 2, before and after 100 o C for 1 hour. No special chemical difference is observed in the testing MAI:PbCl 2 sample before and after 100 o C annealing.

8 (a) Intensity (a.u.) Survey C KVV N KLL O KLL Perovskite I 3d In3d O1s Pb4d N1s C1s Pb4f I 4f In4s Pb5d Binding Energy (ev) (b) PL intensity (a.u.) 0.8 4: : : : PbCl wavelength (nm) (c) 1.0 Absorption (a.u.) Wavelength (nm) Figure S3 (a) XPS survey scan, (b) photoluminescence spectra and (c) absorption spectrum of perovskite film.

9 EQE (%) : : :1: : Wavelength (nm) Absorption (a.u.) Wavelength (nm) Figure S4 EQE spectra of perovskite solar cells with different MAI:PbCl 2 ratios. Inset shows the corresponding absorption of perovskite films.

10 (a) (b) 2.3:1 1:1 (c) m R.M.S. 1.2 nm m (d) m R.M.S nm m Figure S5 AFM of perovskite film with MAI:PbCl 2 ratios of (a) 2.3:1 and (b) 1:1. The related cross-section morphologies are shown in (c) and (d) respectively.

11 (a) Metal Organic Perovskite ITO 100nm (b) (c) Perovskite 2.3:1 Perovskite 1:1 1 m 1 m (d) Perovskite 2.3:1/ C 60 /BcP (e) Perovskite 1:1 C 60 /BcP 1 m 1 m Figure S6 SEM images of (a) cross-section of the perovskite devices. Perovskite film with MAI:PbCl 2 ratios of (b) 2.3:1 and (c) 1:1. Film morphologies with C 60 /BCP film deposited on perovskites with MAI:PbCl2 ratios of (d) 2.3:1 and (e) 1:1.

12 Experimental details: Thermal evaporator consist of a vacuum chamber of base vacuum better than 10-6 Torr equipped with a vacuum pump (i.e. CTI Cryo pump) and electrically heated quartz crucibles for holding the source materials. A substrate holder is set about 10 inches about the crucibles. During sample preparation, an ITO coated glass substrate is put on the substrate holder with the ITO side facing the crucibles. CH 3 NH 3 I (from Lumtech.) and PbCl 2 (from Sigma Aldrich) are used as-received and load into two different crucibles. Evaporation temperatures for CH 3 NH 3 I and PbCl 2 are ~120 and ~ 325 C respectively. A quartz crystal monitor is positioned close to the substrate and used for monitoring the thickness of the deposited film. Doping ratio in the film is precisely controlled with two additional crystal monitors. Detailed operation procedure is as follows, 1. PbCl 2 or MAI are deposited on the Si substrates for 30 minutes at a fixed evaporation rate and the final deposited thickness can be recorded on the crystal monitor. 2. The actual deposited films thickness were characterized using ellipsometer. 3. The tooling factors were justified by comparing the system-estimated thickness and the actual measured thickness. It is noted that this factor highly depends on the mean-free-path of evaporated source and the source-to-substrate distance, this would vary system-by-system. 4. With all these deposition parameters, doping ratio of co-evaporation can be controlled. 5. Finally the doping ratio of two sources is recorded and verified with XPS.