The Influence of Solvent Coordination on Hybrid Organic-Inorganic

7. Figure 2: a) Environmental SEM image of MAPbI3 crystals formed in a 1 M PC solution comprising stoichiometrically equivalent MAI and PbI2 precursor...
0 downloads 7 Views 3MB Size
Download PDF
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Influence of Solvent Coordination on Hybrid Organic−Inorganic Perovskite Formation J. Clay Hamill, Jr.,† Jeffrey Schwartz,‡ and Yueh-Lin Loo*,†,§ †

Department of Chemical and Biological Engineering, Princeton University, Princeton, New Jersey 08544, United States Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States § Andlinger Center for Energy and the Environment, Princeton University, Princeton, New Jersey 08544, United States ‡

S Supporting Information *

ABSTRACT: Solution-processed hybrid organic−inorganic perovskites (HOIPs) from organoammonium halide and lead halide precursors form efficacious active layers for photovoltaics, light-emitting diodes, and flexible electronics. Though solvent−solute coordination plays a critical role in HOIP crystallization, the influence of solvent choice on such interactions is poorly understood. We demonstrate Gutmann’s donor number, DN, as a parameter that indicates the coordinating ability of the processing solvent with the Pb2+ center of the lead halide precursor. Low DN solvents interact weakly with the Pb2+ center, favoring instead complexation between Pb2+ and iodide and subsequent crystallization of perovskite. High DN solvents coordinate more strongly with the Pb2+ center, which in turn inhibits iodide coordination and stalls perovskite crystallization. Varying the concentration of high-DN additives in precursor solutions tunes the strength of lead− solvent interactions, allowing finer control over the crystallization and the resulting morphology of HOIP active layers. bond unsaturation23 have been used to describe solvent− precursor interactions in HOIP solutions. The dielectric constant of the processing solvent is thought to correlate with its “coordinating ability” with the lead halide salt,17 which influences the formation iodoplumbate intermediates (PbIn2−n, where n = 2−6) and lead−solvent complexes17 in precursor solutions comprising methylammonium iodide (MAI) and PbI2. The existence of these species subsequently determines the defect density and morphology of the fully formed active layers.15 On the other hand, Saidaminov et al. showed that methylammonium lead iodide (MAPbI3) single crystals can be grown from γ-butyrolactone (GBL) but not from N,Ndimethylformamide (DMF)24 and later speculated this difference to arise from different coordinating abilities of the solvents due to differences in polarity.20 Hansen’s solubility parameters, on the other hand, have been used as a starting point to screen solvents but fail to account for ionic interactions and molecular complexation.22 Still other studies state that the effects of other parameters are secondary to Lewis acid−base interactions.23,25 The result of these inconsistencies is an unclear picture of how solvent choice affects the formation of intermediates or how it can be leveraged to control HOIP crystallization. The complicated phase space of solution chemistry in HOIP

R

ecent years have seen significant interest surrounding the use of hybrid organic−inorganic perovskites (HOIPs) as absorbers in photovoltaic (PV) devices,1,2 emitters in light-emitting diodes (LEDs),3,4 and active layers in flexible electronics.5−7 The rapid climb in efficiencies of PV devices comprising HOIP active layers from 3.8%8 in 2009 to 22.1% today,9 the low cost of precursors,10,11 and the ease of processing from solution are all promising indicators for efficacious harnessing of solar energy at scale.12,13 Unfortunately, many challenges exist that still hamper the deployment of HOIP-based technologies. 2 Important among these challenges is poor control of solution-mediated crystallization of HOIP materials.2,12 The formation of solution-based intermediates has been shown to influence both crystallization processes for HOIPs and the performance of devices comprising the resulting HOIP active layers.14−18 While it is generally accepted that solvent choice dictates the formation of such intermediates,17,19 the mechanism by which solvent− precursor coordination affects crystallization is not fully understood. In this study, we show how solvent interactions dictate intermediate formation in precursor solutions and elucidate how these interactions influence perovskite crystallization. Precursor−solvent interactions have been shown to influence the formation of HOIPs,17,20,21 but it is unclear what solvent properties dictate these interactions and how these interactions influence the formation of intermediates. Thus, far, dielectric constant,17,20 Hansen’s solubility parameters,22 and Mayer © XXXX American Chemical Society

Received: October 26, 2017 Accepted: November 29, 2017 Published: November 29, 2017 92

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97

Letter

Cite This: ACS Energy Lett. 2018, 3, 92−97

Letter

ACS Energy Letters

solvents’ ability to form stable precursor solutions. In general, and consistent with what has been reported in the literature,22,26,27 solvents with dielectric constants > 30 appear to solubilize the precursors, but our data highlights a number of exceptions to this rule. For example, TMS and ACN, with dielectric constants exceeding 30, do not support solubilization of the precursors; the precursors instead readily crystallize to form MAPbI3. EC and PC, both of which both have higher dielectric constants relative to the other solvents tested, also do not support the formation of stable solutions of the precursors. Finally, nitromethane (NME), with a dielectric constant of 35.9, does not dissolve the precursors at all. We find DN to exhibit a stronger correlation with the ability of the solvent to solubilize the precursors; solvents having DN > 18 kcal/mol are effective at solubilizing the precursors, whereas solvents having lower DN allow for crystallization of the precursors from 1 M solutions. In fact, precursor solutions in solvents with DN > 18.0 kcal/mol are stable even when heated to temperatures approaching the respective boiling points of the solvents. The observation that solubility is more strongly correlated with DN than with dielectric constant indicates that acid−base interactions between the solvent and precursor, rather than electrostatics, dominate the dissolution process. Because the Pb2+ center of the lead salt precursor exhibits an affinity for basic ligands, which can be iodide anions, solvent molecules, or a combination of the two,17,19 the solubility of the lead salt increases when its affinity for the solvent ligands increases. By employing a solvent with DN < 18 kcal/mol, we can induce controlled precipitation of large single crystals. When MAI and PbI2 are at stoichiometric equivalence in PC (DN = 15.1 kcal/mol) at 1 M total concentration, for example, the mixture turns black within 3 min of stirring at room temperature, which we attribute to perovskite formation and precipitation. To confirm, we conducted environmental scanning electron microscopy (E-SEM) on an aliquot of this mixture. Figure 2a shows faceted crystals; powder X-ray diffraction (XRD) of these crystals (XRD trace shown in Figure S1) confirmed that they are MAPbI3 in its tetragonal form. We can grow crystals exceeding 2 mm in all dimensions (Figure 2b) by heating a 0.7 M precursor solution in PC to 70

precursor solutions, including electrostatic interactions, ionic interactions, and sterics, necessitates a focused study of each type of interaction and its corresponding effect on the formation of HOIPs relevant for device applications. In this study, we focused on elucidating the role of Lewis basicity of the processing solvent, quantified by Gutmann’s donor number, DN, on HOIP formation. We find that DN is a stronger predictor of a solvent’s ability to solvate HOIP precursors than dielectric constant because the former describes the strength of interactions between Lewis-basic solvents and the borderline soft Lewis-acidic Pb2+ center of iodoplumbate complexes in solution. High DN solvents compete with I− for coordination sites around Pb2+ and consequently suppress the formation of iodoplumbates; such solvents form stable precursor solutions for thin-film processing. Low-DN solvents coordinate less strongly with Pb2+ and thus favor iodoplumbate formation and subsequent single-crystal growth. This understanding allows the judicious selection of processing solvents or solvent additives tailored for the desired application. Figure 1 summarizes the dielectric constant and DN of several processing solvents, categorized by whether these solvents

Figure 1. Dielectric constant (red triangles) and Gutmann’s donor number (DN, blue circles) of solvents used for HOIP processing. The unshaded region on the left shows solvents that form stable solutions at total MAI and PbI2 concentrations exceeding 1 M. The gray region on the right shows solvents that do not support the formation of stable solutions. Instead, HOIP crystals form spontaneously and they precipitate from solution.

support the solubilization versus rapid crystallization of 1:1 MAI and PbI2 at 1 M. Solvents that solubilize the precursors at total concentrations = 1 M are shown on the left in the unshaded region; dimethylpropyleneurea (DMPU), dimethyl sulfoxide (DMSO), dimethylacetamide (DMAC), N-methyl-2pyrrolidone (NMP), and DMF fall in this category. Solvents that do not form stable precursor solutions at this concentration and instead promote crystallization and precipitation of MAPbI3 (indicated by the color change of the mixtures from yellow to black, characteristic of MAPbI3) are shown in the gray region on the right of Figure 1; these include ethylene carbonate (EC), propylene carbonate (PC), tetramethylsulfone (TMS), and acetonitrile (ACN). In GBL, the precursors are soluble at room temperature at 1 M, but MAPbI3 single-crystal growth can be induced by heating the precursor solution and maintaining the solution temperature at 120 °C for several minutes.24 Such behavior is not observed in solvents with higher DN than that of GBL. Figure 1 reveals a weak correlation between the dielectric constants of solvents and the

Figure 2. (a) Environmental SEM image of MAPbI3 crystals formed in a 1 M PC solution comprising stoichiometrically equivalent MAI and PbI2 precursors. (b) MAPbI3 crystal grown from a 0.7 M PC solution after heating to 70 °C for 3 h. (c) PEA2PbI4 platelets that formed in a droplet of a 1 M PC solution comprising PEAI and PbI2 at stoichiometric equivalence. (d) FA0.83MA0.17PbI3 crystal grown from a 0.7 M PC solution after heating to 72 °C for 12 h. 93

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97

Letter

ACS Energy Letters °C for 3 h and removing the resulting crystal before the solution cools, similar to the inverse-temperature crystallization method employed by Saidaminov et al.24 While Nayak et al. found that temperature-induced solvent degradation is responsible for crystal growth in some solvent systems,28 we did not observe a decrease in the onset temperature of crystallization when the degradation products of PC were intentionally added to the solutions used for single-crystal growth. We conclude that crystallization in PC does not occur due to solvent degradation at elevated temperature but instead occurs because of temperature-dependent changes in the coordinating ability of the solvent. The method of growing HOIP single crystals by choosing solvents with low DN is generalizable beyond systems with MA+ cations; we can routinely form high-quality single crystals when MA+ is replaced with phenylethylammonium (PEA+) or butylammonium (BA+) cations. Consistent with the PEA2PbI4 crystal structure, we obtain 50−200 μm “flakes” of single crystals. PEA2PbI4 crystals precipitate from solution, as observed in Figure 2c; their corresponding XRD pattern in Figure S2 confirms the formation of two-dimensional perovskites with the appropriate layer spacing. Precipitation of perovskite crystals also occurs spontaneously in TMS, ACN, and EC, all of which exhibit DN < 18 kcal/mol. To further demonstrate the generalizability of this approach, we have additionally grown large single crystals (>5 mm in all dimensions) of FA0.83MA0.17PbI3 (Figure 2d) from a 0.7 M precursor solution in PC at 72 °C. Its corresponding XRD pattern is provided in Figure S3 and agrees with the reported pattern for the expected cubic phase of FA0.85MA0.15PbI3,29 with reflections at slightly higher 2θ°, corresponding to a slightly smaller unit cell dimension, presumably due to the higher MA:FA ratio in our crystals. To explore how Lewis basicity of the solvent impacts solvent−precursor interactions, we conducted Benesi−Hildebrand (BH) analyses14,16,30 to determine the relative strength of interactions between iodide and PbI2 in the solvent of interest. Starting with 0.2 mM PbI2 solutions in PC (DN = 15.1 kcal/ mol) and DMF (DN = 26.6 kcal/mol), we measured the absorbance at increasing concentrations of MAI to determine how PbI2 coordinates with iodide to form PbIn2−n complexes in solution. While PbI53− and PbI64− formation is requisite to perovskite formation,17,31 they are not observed in the absorbance data due to the intentionally low concentrations of precursors used during our BH study. Instead, we monitored PbI3− iodation to PbI42− to probe the relative strength of I− interaction with the Pb2+ center of PbI3−. Figure 3a shows one such set of absorbances with increasing concentrations of iodide. At low [MAI] in PC, we observe a peak absorbance at 368 nm; this absorbance is characteristic of PbI3−.32 Increasing the concentration of MAI results in a decrease in absorbance at this wavelength and a concurrent increase in absorbance at 410 nm, consistent with the formation of PbI42−.32 Figure 3b plots the absorbances at 368 and 410 nm as a function of [MAI]. BH analysis tracking the evolution of these absorbance peaks quantifies the strength of I− interaction with PbI3−, from which we can infer how strongly the solvent coordinates with the Pb2+ center of the complex. Figure 3c shows the result of our BH analysis of 0.2 mM PbI2 in PC with increasing [MAI]. Our calculations yielded a PbI42− formation constant (Kf) of 94 ± 2 M−1. We conducted an equivalent analysis for 0.2 mM PbI2 in DMF (higher DN than PC; analysis in Figure S4a−c) and obtained Kf = 9.8 ± 0.3 M−1. The larger Kf in PC compared to

Figure 3. (a) Stacked UV−vis absorbance traces for PbI2 in PC with increasing [MAI]:[PbI2]. As [MAI]:[PbI2] increases, the characteristic absorbance of PbI3− at λ = 368 decreases while the absorbance of PbI42− increases at 368 and 410 nm as a function of [MAI]. The spectra are offset along the y-axis for clarity. (b) Peak absorbance of PbI3− (λmax = 368 nm; red triangles) and PbI42− (λmax = 410 nm; blue circles) in PC as a function of [MAI]:[PbI2]. (c) BH analysis of the UV−vis data presented in (a) and (b). The equilibrium constant for the formation of PbI42− in PC is 94 ± 2 M−1.

that in DMF indicates a stronger propensity to form PbI42− in PC. This comparison implies weaker coordination between Pb2+ and PC than that between Pb2+ and DMF, the difference of which we attribute to the basicity of the solvents. The stronger basicity solvent (DMF; higher DN) more readily coordinates with the borderline soft Lewis-acidic Pb2+ center of PbI3−. Depending on the relative basicity of the solvent and I−, competitive ligand exchange19,33 thus takes place around the Pb2+ center. A more Lewis-basic solvent will competitively inhibit the formation of iodoplumbate complexes that is requisite for perovskite formation, whereas a less Lewis-basic solvent will allow I− to complex with Pb2+ because they do not 94

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97

Letter

ACS Energy Letters

additional control over active layer film formation. Unexpectedly, DMPU−Pb2+ coordination is so strong that a yellow paste results when a stoichiometric equivalence of MAI and PbI2 is dissolved in DMPU at concentrations appropriate for for spincoating (between 1.5 and 3 M from our group’s experience). To take advantage of the strong coordination we expect from DMPU while still maintaining solubility, we instead used DMPU as a solvent additive as opposed to a processing solvent. The addition of DMPU increases the DN of the solvent mixture;33 the addition of fractional amounts of DMPU thus provides a “tuning knob” to access the desired basicity of the solvent mixture. Figure 5 shows scanning electron micrographs of films spincoated from 2.5 M DMF solutions containing stoichiometri-

compete as strongly for coordination sites, allowing for the formation of iodoplumbate complexes and subsequent crystallization of MAPbI3. This finding that solvents with higher DN coordinate more strongly to the Pb2+ center extends to solid-state complexes (i.e., “solvates”), as Cao et al. found the decomposition temperature of solid-state complexes of PbI2 with various Lewis basic solvents (e.g., NMP, DMSO, and HMPA) to increase with increasing DN of the solvent.34 Because the described interactions occur between Pb2+ and the processing solvent and do not involve the countercation of the iodide source, these results are generalizable to systems beyond ones with MA+ as the organic cation, including those containing FA+, as demonstrated in Figure 2, and likely extendable to state-of-the-art mixed-cation systems. With a DN of 34.0 kcal/mol, we expect DMPU to coordinate strongly with Pb2+; iodoplumbate formation should be suppressed in this solvent. Figure 4 shows the absorbance of

Figure 5. SEM images showing the morphology of MAPbI3 films cast on PEDOT:PSS-coated glass substrates from precursor solutions in DMF with increasing volume fractions of DMPU. a) No DMPU; (b) 0.5 vol % DMPU; (c) 10 vol % DMPU; and (d) magnified image of the sample whose wide-view image is shown in (c).

Figure 4. Absorbance of 0.05 mM PbI2 with increasing [MAI] in DMPU. Spectral features corresponding to PbI3− at 370 nm and PbI42− at 425 nm are evident only at [MAI]:[PbI2] equal to or exceeding 1500. For [MAI]:[PbI2] less than 1500, the absorbances of PbI3− and PbI42− are negligible. The spectra are stacked along the y-axis for clarity.

cally equivalent MAI and PbI2 precursors with increasing amounts of DMPU on PEDOT:PSS-coated glass substrates. Spin-coating the precursor solution without DMPU results in a thin film that exhibits needle-like morphology and incomplete surface coverage that is typical of what had previously been reported.21 With the addition of 0.5 vol % DMPU, the morphology of the resulting film changes and surface coverage is dramatically improved. At a DMPU content of 10 vol %, we observe a transition from needle-like morphology to grains with 100% surface coverage. The morphology of these films is similar to those attained from DMF-processed films when toluene washing (i.e., antisolvent treatment38,39) is performed during spin-coating. Consistent with our picture of complexation, the addition of DMPU substantially retards perovskite crystallization. Films processed from DMPU-containing solvent mixtures remain translucent and yellow after 30 s of spincoating and turn black only after 3 min of thermal annealing at 70 °C. Comparatively, films spin-coated from a solution employing only DMF as the solvent turn black after 1500, the absorbances of PbI3− at 370 nm and PbI42− at 425 nm rise concurrently. The λmax of the absorbance bands in DMPU are red-shifted compared to that in PC due to the lower polarity of DMPU relative to PC.35 For comparison, we see a significant rise of PbI42− absorbance at [MAI]:[PbI2] = 10 and 250 in PC and DMF, respectively. We attribute the lack of iodoplumbate absorbance in precursor solutions in DMPU to its strong coordination with Pb2+, which makes it difficult for I− to compete for the same coordination sites. We sought to take advantage of the strong coordinating ability of DMPU to control crystallization of MAPbI3 thin films formed via one-step deposition.36,37 We hypothesized that crystallization would not proceed until DMPU−Pb2+ coordination is dissociated during thermal annealing, thus retarding crystallization and rendering 95

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97

Letter

ACS Energy Letters with the incorporation of low Pvap solvent additives related to slow crystallization.41 Our results instead suggest Lewis basicity of the solvent additives to be responsible, with a more Lewisbasic solvent additive complexing more strongly with the Pb2+ center and delaying the onset of perovskite crystallization. To rule out the possibility that our observation of delayed crystallization is due to the low vapor pressure of DMPU (Pvap = 0.039 mm Hg), we also spin-coated DMF solutions comprising 2.5 M MAI and PbI2 precursors with PC (Pvap = 0.023 mm Hg) instead of DMPU as the additive. While PC has a lower Pvap than DMPU, it does not coordinate as strongly with Pb2+ as DMPU. Films spin-coated from precursor solutions containing 10 vol % PC do not exhibit delayed crystallization; the films instead turn black after just 4 s of spincoating. We therefore attribute delayed crystallization in the films cast from DMF + DMPU solutions to the strong interactions between DMPU and Pb2+. To fabricate functional solar cells, we found we still need to incorporate toluene washing38,39 during spin-coating to extract residual processing solvent. The incorporation of DMPU, however, substantially extends the time available for which toluene washing can be applied compared to that available for films cast from DMF-only solutions. Figure 6 summarizes the

decreased density of recombination centers, can be beneficial for the scalable fabrication of large-area PV devices comprising HOIP active layers.41 We demonstrate that solvent coordination with Pb2+ is correlated with Lewis basicity, which in turn influences solidstate perovskite formation. In solvents with low DN, iodide coordination with Pb2+ dominates and the reaction proceeds to allow the formation of single crystals, as seen with the facile growth of bulk MAPbI3, BA2PbI4, and PEA2PbI4. At the other end of the spectrum, solvents with high DN complex strongly with Pb2+ and competitively inhibit I− coordination with Pb2+ that is requisite for perovskite formation. This framework formed the basis with which we selected DMPU as a high-DN solvent additive to alter thin-film morphology and to afford greater flexibility in postdeposition processing. Our study has shed light on the nature of solvent−solute interactions in these complex solutions and demonstrates the importance of judicious selection of solvent to tailor solid-state structural development in HOIPs.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsenergylett.7b01057. Materials and methods, powder XRD traces of MAPbI3 and PEA2PbI4 crystals, Benesi−Hildebrand analysis for the PbI2−DMF system with increasing [MAI], and solar cell performance versus antisolvent delay time (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Address: Andlinger Center for Energy and the Environment, Princeton University, Princeton, NJ 08544. ORCID

Jeffrey Schwartz: 0000-0001-9873-4499 Yueh-Lin Loo: 0000-0002-4284-0847

Figure 6. Comparison of the power conversion efficiency (PCE) of solar cells whose active layers were cast from precursor solutions comprising only DMF as the processing solvent (red triangles) and those whose active layers were cast from precursor solutions comprising 10 vol % DMPU (blue circles) as a function of delay time to antisolvent treatment. The former devices show a complete decay of PCE if antisolvent treatment is not carried out within 3 s of precursor solution deposition, while the latter devices exhibit respectable efficiencies over a wider processing window.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge financial support from the National Science Foundation through Award Numbers CMMI-1537011 and DMR-1420541. The latter originates from NSF’s MRSEC program that supports the Princeton Center for Complex Materials. J.C.H. is supported by the Department of Defense (DoD) by a National Defense Science and Engineering Graduate Fellowship (NDSEG).

performance of PV devices cast from DMF-only solutions; the efficiencies of these solar cells are highly sensitive to the time at which toluene washing is applied. In comparison, the same graph shows the efficiencies of solar cells whose active layers are cast from solutions with DMPU. These efficiencies are stable across a wider processing window during which toluene washing is applied. We attribute the ca. 2% improvement in efficiency in devices comprising films cast from 10% DMPU precursor solutions to a decreased density of charge recombination centers stemming from a decrease in PbI42− in the precursor solution per Stewart et al.15 Because DMPU coordinates strongly with Pb2+, its addition impedes PbI42− formation in precursor solutions. The easing of time constraints for postdeposition processing by selecting high-DN solvent additives, coupled with an increased efficiency due to a

■

REFERENCES

(1) McGehee, M. D. Perovskite Solar Cells: Continuing to Soar. Nat. Mater. 2014, 13, 845−846. (2) Berry, J.; Buonassisi, T.; Egger, D. A.; Hodes, G.; Kronik, L.; Loo, Y.-L.; Lubomirsky, I.; Marder, S. R.; Mastai, Y.; Miller, J. S.; et al. Hybrid Organic-Inorganic Perovskites (HOIPs): Opportunities and Challenges. Adv. Mater. 2015, 27, 5102−5112. (3) Xiao, Z.; Kerner, R. A.; Zhao, L.; Tran, N. L.; Lee, K. M.; Koh, T.W.; Scholes, G. D.; Rand, B. P. Efficient Perovskite Light-emitting Diodes Featuring Nanometre-sized Crystallites. Nat. Photonics 2017, 11, 108−115. 96

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97

Letter

ACS Energy Letters (4) Cho, H.; Jeong, S.-H.; Park, M.-H.; Kim, Y.-H.; Wolf, C.; Lee, C.L.; Heo, J. H.; Sadhanala, A.; Myoung, N.; Yoo, S.; et al. Overcoming the Electroluminescence Efficiency Limitations of Perovskite Lightemitting Diodes. Science 2015, 350, 1222−1225. (5) You, J.; Hong, Z.; Yang, Y. M.; Chen, Q.; Cai, M.; Song, T.; Chen, C.; et al. Low-Temperature Solution-Processed Perovskite Solar Cells with High Efficiency and Flexibility . ACS Nano 2014, 8, 1674− 1680. (6) Susrutha, B.; Giribabu, L.; Singh, S. P. Recent Advances in Flexible Perovskite Solar Cells. Chem. Commun. 2015, 51, 14696− 14707. (7) Gu, C.; Lee, J.-S. Flexible Hybrid Organic−Inorganic Perovskite Memory. ACS Nano 2016, 10, 5413−5418. (8) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells. J. Am. Chem. Soc. 2009, 131, 6050−6051. (9) Yang, W. S.; Park, B.-W.; Jung, E. H.; Jeon, N. J.; Kim, Y. C.; Lee, D. U.; Shin, S. S.; Seo, J.; Kim, E. K.; Noh, J. H.; et al. Iodide Management in Formamidinium-Lead-Halide−Based Perovskite Layers for Efficient Solar Cells. Science 2017, 356, 1376−1379. (10) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. The Emergence of Perovskite Solar Cells. Nat. Photonics 2014, 8, 506−514. (11) Snaith, H. J. Perovskites: The Emergence of a New Era for LowCost, High-Efficiency Solar Cells. J. Phys. Chem. Lett. 2013, 4, 3623− 3630. (12) Williams, S. T.; Rajagopal, A.; Chueh, C.-C.; Jen, A. K.-Y. Current Challenges and Prospective Research for Upscaling Hybrid Perovskite Photovoltaics. J. Phys. Chem. Lett. 2016, 7, 811−819. (13) Schmidt, T. M.; Larsen-Olsen, T. T.; Carlé, J. E.; Angmo, D.; Krebs, F. C. Upscaling of Perovskite Solar Cells: Fully Ambient Roll Processing of Flexible Perovskite Solar Cells with Printed Back Electrodes. Adv. Energy Mater. 2015, 5, 1500569−1500578. (14) Stamplecoskie, K. G.; Manser, J. S.; Kamat, P. V. Dual Nature of the Excited State in Organic-Inorganic Lead Halide Perovskites. Energy Environ. Sci. 2015, 8, 208−215. (15) Stewart, R. J.; Grieco, C.; Larsen, A. V.; Doucette, G. S.; Asbury, J. B. Molecular Origins of Defects in Organohalide Perovskites and Their Influence on Charge Carrier Dynamics. J. Phys. Chem. C 2016, 120, 12392−12402. (16) Yoon, S. J.; Stamplecoskie, K. G.; Kamat, P. V. How Lead Halide Complex Chemistry Dictates the Composition of Mixed Halide Perovskites. J. Phys. Chem. Lett. 2016, 7, 1368−1373. (17) Rahimnejad, S.; Kovalenko, A.; Fores, S. M.; Aranda, C.; Guerrero, A. Coordination Chemistry Dictates the Structural Defects in Lead Halide Perovskites. ChemPhysChem 2016, 17, 2795−2798. (18) Sharenko, A.; Mackeen, C.; Jewell, L.; Bridges, F.; Toney, M. F. Evolution of Iodoplumbate Complexes in Methylammonium Lead Iodide Perovskite Precursor Solutions. Chem. Mater. 2017, 29, 1315− 1320. (19) Manser, J. S.; Saidaminov, M. I.; Bakr, O. M.; Kamat, P. V.; et al. Making and Breaking of Lead Halide Perovskites. Acc. Chem. Res. 2016, 49, 330−338. (20) Saidaminov, M. I.; Abdelhady, A. L.; Maculan, G.; Bakr, O. M. Retrograde Solubility of Formamidinium and Methylammonium Lead Halide Perovskites Enabling Rapid Single Crystal Growth. Chem. Commun. 2015, 51, 17658−17661. (21) Khlyabich, P. P.; Loo, Y.-L. Crystalline Intermediates and Their Transformation Kinetics during the Formation of Methylammonium Lead Halide Perovskite Thin Films. Chem. Mater. 2016, 28, 9041− 9048. (22) Gardner, K. L.; Tait, J. G.; Merckx, T.; Qiu, W.; Paetzold, U. W.; Kootstra, L.; Jaysankar, M.; Gehlhaar, R.; Cheyns, D.; Heremans, P.; et al. Nonhazardous Solvent Systems for Processing Perovskite Photovoltaics. Adv. Energy Mater. 2016, 6, 1600386−1600394. (23) Stevenson, J.; Sorenson, B.; Subramaniam, V. H.; Raiford, J.; Khlyabich, P. P.; Loo, Y.-L.; Clancy, P. Mayer Bond Order as a Metric of Complexation Effectiveness in Lead Halide Perovskite Solutions. Chem. Mater. 2017, 29, 2435−2444.

(24) Saidaminov, M. I.; Abdelhady, A. L.; Murali, B.; Alarousu, E.; Burlakov, V. M.; Peng, W.; Dursun, I.; Wang, L.; He, Y.; Maculan, G.; et al. High-Quality Bulk Hybrid Perovskite Single Crystals within Minutes by Inverse Temperature Crystallization. Nat. Commun. 2015, 6, 7586. (25) Dirin, D. N.; Dreyfuss, S.; Bodnarchuk, M. I.; Nedelcu, G.; Papagiorgis, P.; Itskos, G.; Kovalenko, M. V. Lead Halide Perovskites and Other Metal Halide Complexes as Inorganic Capping Ligands for Colloidal Nanocrystals. J. Am. Chem. Soc. 2014, 136, 6550−6553. (26) Ahn, N.; Son, D.-Y.; Jang, I.-H.; Kang, S. M.; Choi, M.; Park, N.G. Highly Reproducible Perovskite Solar Cells with Average Efficiency of 18.3% and Best Efficiency of 19.7% Fabricated via Lewis Base Adduct of Lead(II) Iodide. J. Am. Chem. Soc. 2015, 137, 8696−8699. (27) Lee, J. W.; Kim, H. S.; Park, N. G. Lewis Acid-Base Adduct Approach for High Efficiency Perovskite Solar Cells. Acc. Chem. Res. 2016, 49, 311−319. (28) Nayak, P. K.; Moore, D. T.; Wenger, B.; Nayak, S.; Haghighirad, A. A.; Fineberg, A.; Noel, N. K.; Reid, O. G.; Rumbles, G.; Kukura, P.; et al. Mechanism for rapid growth of organic−inorganic halide perovskite crystals. Nat. Commun. 2016, 7, 13303. (29) Huang, Y.; Li, L.; Liu, Z.; Jiao, H.; He, Y.; Wang, X.; Zhu, R.; Wang, D.; Sun, J.; Chen, Q.; et al. The intrinsic properties of FA(1−x)MAxPbI3 perovskite single crystals. J. Mater. Chem. A 2017, 5, 8537−8544. (30) Benesi, H. A.; Hildebrand, J. H. A Spectrophotometric Investigation of the Interaction of Iodine With Aromatic Hydrocarbons. J. Am. Chem. Soc. 1949, 71, 2703−2707. (31) Yan, K.; Long, M.; Zhang, T.; Wei, Z.; Chen, H.; Yang, S.; Xu, J. Hybrid Halide Perovskite Solar Cell Precursors: Colloidal Chemistry and Coordination Engineering behind Device Processing for High Efficiency. J. Am. Chem. Soc. 2015, 137, 4460−4468. (32) Horváth, O.; Mikó, I. Spectra, Equilibrium and Photoredox Chemistry of Tri- and Tetraiodoplumbate(II) Complexes in Acetonitrile. J. Photochem. Photobiol., A 1998, 114, 95−101. (33) Gutmann, V. Coordination Chemistry in Non-Aqueous Solutions; Springer-Verlag: New York, 1968. (34) Cao, X. B.; Li, C. L.; Zhi, L. L.; Li, Y. H.; Cui, X.; Yao, Y. W.; Ci, L. J.; Wei, J. Q. Fabrication of High Quality Perovskite Films by Modulating the Pb−O Bonds in Lewis Acid−Base Adducts. J. Mater. Chem. A 2017, 5, 8416−8422. (35) Hofmann, A.; Simon, A.; Grkovic, T.; Jones, M. Methods of Molecular Analysis in the Life Sciences; Cambridge University Press: Cambridge, U.K., 2014. (36) Etgar, L.; Gao, P.; Xue, Z.; Peng, Q.; Chandiran, A. K.; Liu, B.; et al. Mesoscopic CH3NH3PbI3/TiO2 Heterojunction Solar Cells. J. Am. Chem. Soc. 2012, 134, 17396−17399. (37) Sharenko, A.; Toney, M. F. Relationships between Lead Halide Perovskite Thin-Film Fabrication, Morphology, and Performance in Solar Cells. J. Am. Chem. Soc. 2016, 138, 463−470. (38) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S. Il. Solvent Engineering for High-Performance Inorganic-Organic Hybrid Perovskite Solar Cells. Nat. Mater. 2014, 13, 897−903. (39) Xiao, M.; Huang, F.; Huang, W.; Dkhissi, Y.; Zhu, Y.; Etheridge, J.; Gray-Weale, A.; Bach, U.; Cheng, Y.-B.; Spiccia, L. A Fast Deposition-Crystallization Procedure for Highly Efficient Lead Iodide Perovskite Thin-Film Solar Cells. Angew. Chem. 2014, 126, 10056− 10061. (40) Saliba, M.; Matsui, T.; Seo, J.-Y.; Domanski, K.; Correa-Baena, J.-P.; Nazeeruddin, M. K.; Zakeeruddin, S. M.; Tress, W.; Abate, A.; Hagfeldt, A.; et al. Cesium-containing Triple Cation Perovskite Solar Cells: Improved Stability, Reproducibility and High Efficiency. Energy Environ. Sci. 2016, 9, 1989−1997. (41) Yang, M.; Li, Z.; Reese, M. O.; Reid, O. G.; Kim, D. H.; Siol, S.; Klein, T. R.; Yan, Y.; Berry, J. J.; van Hest, M. F. A. M.; et al. Perovskite Ink with Wide Processing Window for Scalable HighEfficiency Solar Cells. Nat. Energy 2017, 2, 17038.

97

DOI: 10.1021/acsenergylett.7b01057 ACS Energy Lett. 2018, 3, 92−97