Barrier Design to Prevent Metal-InducedDegradation and Improve Thermal Stability inPerovskite Solar CellsCaleb C. Boyd,† Rongrong Cheacharoen,† Kevin A. Bush,† Rohit Prasanna,† Tomas Leijtens,†

and Michael D. McGehee*,†,‡

†Materials Science and Engineering, Stanford University, Stanford, California 94305, United States‡Department of Chemical and Biological Engineering, University of Colorado, Boulder, Colorado 80309, United States

*S Supporting Information

ABSTRACT: Metal-contact-induced degradation and escape of volatile species fromperovskite solar cells necessitate excellent diffusion barrier layers. We show thatmetal-induced degradation limits thermal stability in several perovskite chemistrieswith Au, Cu, and Ag gridlines even when the metal is separated from the perovskite bya layer of indium tin oxide (ITO). Channels in a sputtered ITO layer that align withperovskite grain boundaries are pathways for metal and halide diffusion into or out ofthe perovskite. Planarizing the perovskite morphology with a spin-cast organic charge-transport layer results in a subsequently deposited ITO layer that is uniform andimpermeable. We show that it is critical to seal the edges of the active layers to preventescape of volatile species. We demonstrate 1000 h thermal stability at 85 °C inCH3NH3PbI3 solar cells with complete-coverage silver contacts. Our barrier layerdesign enables long-term thermal stability of perovskite solar cells, a critical step tocommercialization.

Perovskite solar cells (PSCs) have the potential tocombine high efficiencies with cost-efficient manufac-turing techniques for terawatt energy production,

especially when used in low-cost tandem solar cells.1−9

However, to realize the benefits of easy-to-manufacture,solution-processable PSCs, their stability must be improvedto enable at least a 20−30 year lifetime.10 One of the mostimportant obstacles is metal-contact-induced degradation dueto the ubiquitous reactivity of metals and halides and the desireto make the entire electrode out of metal or metal gridlines toreduce series resistance. Although some metals may be stablewith respect to the perovskite structure,11 almost all metalspecies, even platinum, are energetically disposed to formhalides.12 Thus, metal contacts act as sinks for halide orhalogen species that escape from the perovskite under thestressors of moisture, oxygen, light, and heat, causingirreversible degradation of both the metal contacts and theperovskite active layer.13−23 Gold has also been shown todiffuse through an organic transport layer into the perovskitelayer, setting up a two-fold degradation mechanism in whichhalogen/halide species can diffuse up to the metal and themetal can diffuse into the perovskite (Figure 1A).24

Nonmetal contact layers such as carbon electrodes haveenabled excellent PSC stability, but they limit solar cellperformance by increasing series resistance losses.25 Trans-parent conducting oxide (TCO) electrodes such as ITO are

often used in single-junction thin-film modules and perovskitetandem solar cells not only because it is necessary for theelectrode that is deposited on top of the perovskite to betransparent but also to improve stability.4,26 However, ITO has2 orders of magnitude higher resistivity than silver27 and canonly be used to carry current for approximately 1 cm withoutcausing the series resistance to be unacceptably high.28 Thereis considerable interest in coating 6 in. wide silicon solar cellswith a perovskite cell to make a tandem. Such a cell must havemetal gridlines on top of the TCO.To improve PSC stability while maintaining performance,

conductive barrier layers must be developed that prevent bothmetal and halide species from diffusing in or out of theperovskite active layer. Potential barrier layers includenanostructured carbon layers,29 cross-linked charge-transportlayers,30 copper phthalocyanine (CuPC),31 chromium oxide/metal bilayers,32 molybdenum oxide/aluminum bilayers,33−35

aluminum zinc oxide/tin oxide bilayers,36−39 tantalum-dopedtungsten oxide/conjugated polymer bilayers,40 and TCOs suchas ITO.26 Designing a barrier that can be broadly applied tomultiple PSC architectures, prevents decomposition of theperovskite layer, and allows for metal contacts while

Received: June 3, 2018Accepted: July 2, 2018Published: July 2, 2018

LetterCite This: ACS Energy Lett. 2018, 3, 1772−1778

© 2018 American Chemical Society 1772 DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

Dow

nloa

ded

via

UN

IV O

F C

OL

OR

AD

O B

OU

LD

ER

on

Aug

ust 1

3, 2

018

at 1

7:32

:42

(UT

C).

Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

demonstrating impressive thermal stability has proven to be achallenge for the community.Au, Cu, and Ag Contacts Degrade PSCs. In the first stage of

this study, we fabricated PSCs on glass that are identical tothose used in world record perovskite−silicon tandem cellspreviously reported.4 They had an “inverted” p−i−narchitecture ITO/NiOx/FA0.83Cs0.17Pb(I0.83Br0.17)3/evaporatedC60/SnO2/ITO/Ag (details in the Experimental Methods)section with semitransparent ITO contacts. We chose thisperovskite composition because the use of formamidinium(FA, CH2(NH2)2) and Cs at the A site in ABX3 is known toimprove thermal stability.41−43 For this study, we added metalfingers on top of these cells to reduce the series resistance(Figure 1). We found that the introduction of the metal fingerssubstantially accelerated thermal degradation at 85 °C, acommon temperature for accelerated testing of solar cells. Wevisually saw degradation of the perovskite active layeremanating from Au, Cu, or Ag fingers deposited on the ITOcontacts, corresponding to changes in the JV curves of the agedcells and overall PCE loss (Figures 1B,C, S1, and S2 and TableS1). Identical cells without the metal gridlines showed novisible signs of degradation and little to no change in theircurrent−voltage characteristics (Figure S1). We found similarresults when we aged MAPbI3 and FA0.75Cs0.25Sn0.5Pb0.5I3devices with metal gridlines (Figure S1 and Tables S2 andS3). It is very clear that the metal accelerates the degradationbecause we observed many times that devices made with threedifferent perovskite compositions were much more stable

without metal gridlines on top of the ITO, and we could easilysee the perovskite film turn yellow near the gridlines.Because it is not surprising that a reaction between a metal

and perovskites would be undesirable, we will not attempt toexplain in this Letter precisely what reaction took placebetween the metals and the perovskites or why the JV curveschanged in the way they did. We speculate that metal diffusinginto the perovskite layer may react with the perovskite at grainboundaries, creating an insulating layer that leads to decreasesin JSC and slight increases in VOC via passivation. We leave thisto future work and focus here on demonstrating a highlyeffective solution for improving the barrier properties of theTCO to prevent the reaction between perovskites and themetal that is needed to reduce series resistance.Ag Dif fuses through the ITO Layer into the Perovskite Layer

and Volatile Halide Species Penetrate the ITO and React with Ag.We were surprised by the persistence of metal-induceddegradation through both an 8 nm thick atomic-layer-deposited (ALD) SnO2 layer and a 150 nm thick sputteredITO layer. To confirm that degradation shown in the previoussection was due to metal−perovskite reactions, we performed aseries of experiments using X-ray photoelectron spectroscopy(XPS) to characterize diffusion of Ag and halide species in andout of the perovskite active layer.Cells aged at 85 °C in a dark, nitrogen environment for 1000

h with Ag contacts were depth profiled to determine whetherAg diffused into the perovskite layer or not. We used tape topeel off the SnO2, ITO, and metal contacts before depthprofiling down into the perovskite layer to avoid diffusion of

Figure 1. Au, Cu, and Ag contacts degrade PSCs. (A) Diagram of devices in this study showing possible degradation mechanisms of metaland iodide diffusion through C60/PCBM, SnO2, and ITO layers. (B) FA0.83Cs0.17Pb(I0.83Br0.17)3 cell with Cu contacts before and afterheating for 1000 h at 85 °C in N2 in the dark. The light areas indicate degradation of the perovskite active layer, which originates around themetal fingers or bus bars. (C) JV curves of FA0.83Cs0.17Pb(I0.83Br0.17)3 cells with no metal and Ag, Cu, and Au contacts before (blue) and after(red) thermal testing at 85 °C for 1000 h in N2. All cells with metal degrade similarly, while cells without metal do not. See also Tables S1−S3 and Figures S1 and S2.

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1773

Figure 2. Ag diffuses through the ITO layer into the perovskite layer and volatile halide species penetrate the ITO and react with Ag. (A,C)XPS depth profile to detect Ag present in the perovskite. Tape was used to peel away the top layers of the stack to avoid diffusion of metalinto the active layer during sputtering and thus false positives. (C) Ag 3d5 counts measured in the XPS depth profile shown in (A) throughtwo MAPbI3 PSCs after peeling off the top SnO2/ITO/Ag layers, one kept at 85 °C and one kept at room temperature for 1000 h in a darkN2 environment. (B,D) XPS with in situ heating to look for halide diffusion to the metal electrode. (D) XPS spectra on the surface of Agcontacts on MAPbI3 solar cells heated in situ at 150 °C with evaporated PCBM, an ALD SnO2 layer, and sputtered ITO covering the surfaceof the perovskite, showing an increase in iodine species at the surface of the solar cell over the course of 4 h. See also Figures S3−S5.

Figure 3. Planarizing the PSC prevents Ag diffusion. (A,B) Cross-sectional SEM images of ITO barrier layers on MAPbI3 cells withevaporated (A) and spun (B) PCBM, with red arrows showing diffusion channels in evaporated PCBM that propagate along the roughperovskite grain boundaries. These channels are not present in the layer on spun PCBM. (C,D) Top-down SEM of ITO on MAPbI3 withevaporated (C) and spun PCBM (D). The channels in the ITO layer match grain boundaries of the perovskite layer. The spun layereffectively smooths the rough perovskite morphology. See also Figures S6−S10.

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1774

metal into the active layer during sputtering (Figure 2A). Weconfirmed that the peel method delaminates the solar cell inthe C60 or PCBM layer through XPS, as previously reported.44

We positively identified Ag peaks in the perovskite active layerso f FA 0 . 8 3C s 0 . 1 7 P b ( I 0 . 8 3 B r 0 . 1 7 ) 3 , MAPb I 3 , a n dFA0.75Cs0.25Sn0.5Pb0.5I3 cells that exhibited degradation afteraging at 85 °C (Figures 2B and S3−S5). We used secondaryion mass spectroscopy (SIMS) with depth profiling to furtherconfirm the presence of Ag in the perovskite layer after heatingand validate the use of XPS depth profiling to assess thepresence of silver (Figure S3). We note that it is difficult to saywith certainty where in the perovskite layer the Ag lies due toconvoluting effects of roughness during ion milling, ion knockon, and the fact that the grains in the MAPbI3 do not alltraverse through the entire film. However, we did not detectAg with XPS in the perovskite in fresh cells of any architecturestudied in this Letter, in cells that were kept at roomtemperature for 1000 h, or in cells that were heated for 1000 hwithout Ag contacts (Figures 2B, S4, and S5). We cantherefore strongly conclude that Ag did penetrate theperovskite film aged for 1000 h at 85 °C and that the Agsignal observed by XPS is not an experimental artifact.While metal diffusion into the active layer is one source of

potential degradation, iodide/iodine species escaping theactive layer is potentially a more ubiquitous and challengingdegradation mechanism under both heat and light.15,22,23 Weperformed in situ heating at 150 °C on fresh MAPbI3 solarcells with evaporated PCBM, ALD SnO2, and sputtered ITOlayers with opaque silver contacts (complete active areacoverage) in an XPS tool, taking high-sensitivity scans on themetal contact surfaces over the course of 4 h to look fordiffusion of elements out of the perovskite active layer (Figure2B,D). We chose MA because it is the most volatile A-sitecation used in ABX3 metal halide perovskites, and we wanted

to characterize the worst-case scenario. Iodine (or iodide)presence on the surface of the Ag contacts increased drasticallyduring heating, indicating that iodine or iodide species werediffusing out of the active layer to the silver surface. Thus, notonly does Ag diffuse into the active layer, but the Ag contactalso acts as a sink for iodine from the perovskite, aiding indecomposition of the active layer. Both of these diffusionprocesses are clearly harmful for device operation and must beprevented for stable PSCs.Planarizing the PSC for an Impermeable ITO Layer. ITO has

been shown to effectively prevent metal, oxygen, and watervapor diffusion,45−48 contrary to our evidence of degradationdue to metal and iodine/iodide diffusion. Because it is notlikely that diffusion is occurring through the ITO bulk,improving the morphology of the ITO to prevent diffusion atgrain boundaries or through pin holes is the key to improvingthermal stability. We tried replacing ITO with the amorphousTCO indium zinc oxide (IZO), which has no grain boundariesand has been shown to be more mechanically stable andthermally stable than ITO, but found that it did not improvethe barrier morphology or resistance to metal-contact-induceddegradation (Figures S6 and S7).27,49,50 Using scanningelectron microscopy (SEM) and atomic force microscopy(AFM), we saw that the morphology of both TCOs drasticallychanged with planarization of the perovskite layer throughspin-coating of the electron-transport layer (ETL) versusevaporating the ETL (Figures 3A−D, S6, and S8). Evaporationconformally coats the perovskite surface, and the sputteredTCO process cannot fill in the deep perovskite grainboundaries. Instead, the perovskite grain boundaries propagatethrough the TCO layer as channels, as can clearly be seen inthe cross-sectional and top-down SEM images in Figure 3A−D. We observed this grain boundary propagation even with a500 nm thick TCO (Figure S9). A spun PCBM layer, in

Figure 4. 1000 h thermal stability of MAPbI3 solar cells with an impermeable TCO. (A−D) Schematics of barrier layers evaluated in thisstudy showing degradation mechanisms that occur with each design. (C,D) Barrier layers evaluated in long-term thermal stability testing,with results shown in (E−G). (E) Normalized plot of the power conversion efficiency during thermal stability testing of MAPbI3 solar cellswith an ITO barrier layer and Ag electrodes at 85 °C in a dark N2 environment. Points are averages for each data set, and error bars boundthe range of all cells tested. Cells with spun PCBM have an impermeable ITO layer that effectively blocks Ag diffusion and iodide diffusionand enables 1000 h thermal stability, while cells with evaporated PCBM have a permeable ITO layer that allows for Ag and iodide diffusion,causing a drop in PCE. (F,G) Current−voltage curves of MAPbI3 cells with evaporated (F) and spun (G) PCBM before and after aging. Seealso Figures S11−S15.

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1775

contrast, fills the valleys of the perovskite grains, allowing for adense, uniform TCO layer. Not coincidentally, many of thedevices in other studies that report good stability under lightsoaking or heating with metal contacts contain barrier layersdeposited on solution-processed transport layers such asnanoparticle layers or hole-transport layers.26,29,31,32,35,36,40

Using the same XPS depth profiling experiment as that in theprevious section with silver contacts that completely coverMAPbI3 devices to examine a device in which metal−perovskite reactions would be most likely to occur, we foundthat the improved TCO layer morphology successfully blockedsilver diffusion into the perovskite layer (Figure S10).While a spun PCBM layer successfully blocked Ag diffusion

into the perovskite layer, in situ XPS heating showed that theiodide/iodine signal increased at the surface for cells with bothspun and evaporated PCBM that had ITO only on top of thesolar cell (Figures 4A,B and S11). To test the hypothesis thatthe volatile halide species were able to escape from theuncovered edges of the active layer, we sealed the edges of theactive layer with ITO (Figure 4C,D). We observed that theamount of iodide/iodine signal that accumulated during in situheating decreased drastically for the cell with evaporatedPCBM and almost completely disappeared for the cell withspun PCBM (Figures S11 and S12). Thus, in addition tohaving an impermeable TCO layer on top of the perovskiteactive layer, it is also necessary to seal the edges of theperovskite layer, especially for perovskite compounds thatcontain volatile organic species such as MA.1000 h Thermal Stability of MAPbI3 Solar Cells with an

Impermeable TCO. To test the stability of our improved barrierlayer in actual devices with a notoriously volatile perovskite, wefabricated MAPbI3 solar cells with either spun or evaporatedPCBM ETLs. We completely covered the active area of the cellwith metal to put the barrier properties of the TCO throughthe toughest test. We deposited an 8 nm thick ALD SnO2 layerand a 150 nm thick sputtered ITO layer, making sure to extendthe ITO layer over the edges of the perovskite active layer toprevent release of the volatile MA species. We used anevaporated MgF2 insulating strip between the top and bottomITO layers on one edge of the solar cell in order to preventshunting and allow for complete ITO coverage of the activelayer (Figure S13). We kept three unencapsulated cells of eachtype with initial efficiencies of 10.1−13.1% on a hot plate in aN2 glovebox in the dark for 1000 h at 85 °C.51 We note thatthe ALD process had been optimized for a different perovskitecompound that produces record efficiency perovskite/silicontandem solar cells,4 and we used MAPbI3 as a model case totruly challenge the effectiveness of the designed barrier. It is forthis reason that the efficiency of our cells is lower than isfrequently observed for MAPbI3. We saw that all of the cellswith evaporated PCBM experienced degradation due to metaldiffusion, but remarkably all of the cells with spun PCBMsuffered no loss in efficiency (Figures 4 and S14 and Table S2).We also heated the four MAPbI3 architectures shown in

Figure 4 at 150 °C in ambient air (25 °C, ∼40% RH) for 4 h.These test parameters are relevant for packaging commercialsolar modules, which often use a 10−20 min, 150 °Cencapsulation step. All architectures showed considerablevisual degradation except the one with a spun PCBM layerand ITO on the edges as well as the top of the perovskiteactive layer, which effectively sealed in volatile componentsand kept moisture and oxygen out of the cell (Figure S15).

In this Letter, we have shown that if metal is used to reducethe series resistance of the electrodes in PSCs then it isnecessary to have a dense, channel-free barrier layer to protectthe metal from the perovskite and contain volatile componentsof the perovskite. If there are channels between grains of theperovskite, then it is necessary to solution-cast a contactmaterial to fill in the channels and disrupt the perovskitemorphology before sputtering a TCO electrode that serves thedual purposes of carrying current and acting as a diffusionbarrier. It is also critically important to seal the edges of thedevice. If modules are made with laser scribes between cells, itwill be necessary to make sure that metal in the vias isprotected from the perovskite. We have shown that solar cellsmade with even one of the most thermally unstable perovskitescan be stable for 1000 h at 85 °C even when Ag, which reactseasily with perovskites, is on top of the TCO. We stillrecommend, however, using the more thermally stableperovskites, a dense TCO barrier layer, and high-qualitypackaging to maximize the chances of having nearly 100% ofthe cells operated outside survive for >25 years.In this study, the only stressor that we subjected the solar

cells to was heat. We suspect, however, that the barrier layersdescribed in this Letter will also improve stability under lightbecause a recent paper shows that light accelerates reactionsbetween perovskites and metal contacts by creating mobilehalogen vacancies through a process in which photogeneratedholes oxidize halogens, making them smaller and allowingthem to become interstitials.23 In addition to increasinghalogen mobility, the photogenerated vacancies allow anyoxygen that might be present to penetrate the film andultimately react with it.14 We therefore predict that properlyprotecting perovskites with barrier layers will allow them toreturn to their original state unharmed after photoexcitationand are currently testing this hypothesis.

■ EXPERIMENTAL METHODSSee the Supporting Information.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsenergy-lett.8b00926.

Detailed methods, tables showing JV characteristics ofthe species studied, and figures showing pictures of theperovskite cells with different metals and active layers,external quantum efficiencies, depth profiles, XPSspectra, AFM and SEM images, optical microscopeimages, and the acrhitectures of the cells (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: michael.mcgehee@colorado.edu.

ORCIDCaleb C. Boyd: 0000-0001-7408-7901Kevin A. Bush: 0000-0003-1813-1300Rohit Prasanna: 0000-0002-9741-2348Tomas Leijtens: 0000-0001-9313-7281Michael D. McGehee: 0000-0001-9609-9030

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1776

Author ContributionsC.C.B., R.C., and M.D.M. designed the experiments andanalyzed the data. C.C.B. and R.C. performed the experiments.K.A.B., R.P., and T.L. provided expertise and feedback. C.C.B.and M.D.M. wrote the paper. All authors discussed the resultsand commented on the manuscript.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

The information, data, and work presented herein were fundedin part by the U.S. Department of Energy (DOE) PVRD2program under Award Number DE-EE0008154 and by theOffice of Naval Research under Award Number N00014-17-1-2525. C.C.B. acknowledges support by the National ScienceFoundation Graduate Research Fellowship under Grant No.DGE-1656518. R.C. acknowledges support from the Ministryof Science and Technology of Thailand Fellowship. Part of thiswork was performed at the Stanford Nano Shared Facilities(SNSF), supported by the National Science Foundation underAward ECCS-1542152. The authors would like to thankChuck Hitzman and Juliet Jamtgaard for insightful discussionsand expertise on spectroscopy techniques and analysis. Wethank Farhad Moghadam and Wen Ma of Sunpreme fordepositing the ITO electrodes. We thank Zhengshan Yu andProf. Zachary Holman of ASU for depositing IZO electrodes,James Raiford and Axel Palmstrom for depositing SnO2 layers,and Wanliang Tan for fruitful discussion.

■ REFERENCES(1) 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. IodideManagement in Formamidinium-Lead-Halide-Based PerovskiteLayers for Efficient Solar Cells. Science 2017, 356 (6345), 1376−1379.(2) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. The Emergence ofPerovskite Solar Cells. Nat. Photonics 2014, 8 (7), 506−514.(3) Noh, J. H.; Im, S. H.; Heo, J. H.; Mandal, T. N.; Seok, S. Il.Chemical Management for Colorful, Efficient, and Stable Inorganic-Organic Hybrid Nanostructured Solar Cells. Nano Lett. 2013, 13 (4),1764−1769.(4) Bush, K. A.; Palmstrom, A. F.; Yu, Z. J.; Boccard, M.;Cheacharoen, R.; Mailoa, J. P.; McMeekin, D. P.; Hoye, R. L. Z.;Bailie, C. D.; Leijtens, T.; et al. 23.6%-Efficient MonolithicPerovskite/Silicon Tandem Solar Cells with Improved Stability.Nat. Energy 2017, 2 (4), 17009.(5) Fu, F.; Feurer, T.; Weiss, T. P.; Pisoni, S.; Avancini, E.; Andres,C.; Buecheler, S.; Tiwari, A. N. High-Efficiency Inverted Semi-Transparent Planar Perovskite Solar Cells in Substrate Configuration.Nat. Energy 2016, 2 (1), 16190.(6) Shen, H.; Duong, T.; Peng, J.; Jacobs, D.; Wu, N.; Gong, J.; Wu,Y.; Karuturi, S. K.; Fu, X.; Weber, K.; et al. Mechanically-StackedPerovskite/CIGS Tandem Solar Cells with Efficiency of 23.9% andReduced Oxygen Sensitivity. Energy Environ. Sci. 2018, 11, 394−406.(7) Eperon, G. E.; Leijtens, T.; Bush, K. A.; Prasanna, R.; Green, T.;Wang, J. T.-W.; McMeekin, D. P.; Volonakis, G.; Milot, R. L.; May,R.; et al. Perovskite-Perovskite Tandem Photovoltaics with OptimizedBand Gaps. Science 2016, 354 (6314), 861−865.(8) Steirer, K. X.; Schulz, P.; Teeter, G.; Stevanovic, V.; Yang, M.;Zhu, K.; Berry, J. J. Defect Tolerance in Methylammonium LeadTriiodide Perovskite. ACS Energy Lett. 2016, 1 (2), 360−366.(9) Li, Z.; Klein, T. R.; Kim, D. H.; Yang, M.; Berry, J. J.; van Hest,M. F. A. M.; Zhu, K. Scalable Fabrication of Perovskite Solar Cells.Nat. Rev. Mater. 2018, 3 (4), 18017.(10) Chang, N. L.; Yi Ho-Baillie, A. W.; Basore, P. A.; Young, T. L.;Evans, R.; Egan, R. J. A Manufacturing Cost Estimation Method with

Uncertainty Analysis and Its Application to Perovskite on GlassPhotovoltaic Modules. Prog. Photovoltaics 2017, 25 (5), 390−405.(11) Zhao, J.; Zheng, X.; Deng, Y.; Li, T.; Shao, Y.; Gruverman, A.;Shield, J.; Huang, J. Is Cu a Stable Electrode Material in HybridPerovskite Solar Cells for a 30-Year Lifetime? Energy Environ. Sci.2016, 9 (12), 3650−3656.(12) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements, 2nded.; Butterworth-Heinemann: Oxford, U.K., 1997.(13) Leijtens, T.; Bush, K.; Cheacharoen, R.; Beal, R.; Bowring, A.;McGehee, M. D. Towards Enabling Stable Lead Halide PerovskiteSolar Cells; Interplay between Structural, Environmental, andThermal Stability. J. Mater. Chem. A 2017, 5 (23), 11483−11500.(14) Aristidou, N.; Eames, C.; Sanchez-Molina, I.; Bu, X.; Kosco, J.;Islam, M. S.; Haque, S. A. Fast Oxygen Diffusion and Iodide DefectsMediate Oxygen-Induced Degradation of Perovskite Solar Cells. Nat.Commun. 2017, 8, 15218.(15) Conings, B.; Drijkoningen, J.; Gauquelin, N.; Babayigit, A.;D’Haen, J.; D’Olieslaeger, L.; Ethirajan, A.; Verbeeck, J.; Manca, J.;Mosconi, E.; et al. Intrinsic Thermal Instability of MethylammoniumLead Trihalide Perovskite. Adv. Energy Mater. 2015, 5 (15), 1500477.(16) Leijtens, T.; Eperon, G. E.; Pathak, S.; Abate, A.; Lee, M. M.;Snaith, H. J. Overcoming Ultraviolet Light Instability of SensitizedTiO2 with Meso-Superstructured Organometal Tri-Halide PerovskiteSolar Cells. Nat. Commun. 2013, 4, 2885.(17) Han, Y.; Meyer, S.; Dkhissi, Y.; Weber, K.; Pringle, J. M.; Bach,U.; Spiccia, L.; Cheng, Y.-B. Degradation Observations ofEncapsulated Planar CH3NH3PbI3 Perovskite Solar Cells at HighTemperatures and Humidity. J. Mater. Chem. A 2015, 3 (15), 8139−8147.(18) Kato, Y.; Ono, L. K.; Lee, M. V.; Wang, S.; Raga, S. R.; Qi, Y.Silver Iodide Formation in Methyl Ammonium Lead IodidePerovskite Solar Cells with Silver Top Electrodes. Adv. Mater.Interfaces 2015, 2 (13), 1500195.(19) Li, J.; Dong, Q.; Li, N.; Wang, L. Direct Evidence of IonDiffusion for the Silver-Electrode-Induced Thermal Degradation ofInverted Perovskite Solar Cells. Adv. Energy Mater. 2017, 7 (14),1602922.(20) Besleaga, C.; Abramiuc, L. E.; Stancu, V.; Tomulescu, A. G.;Sima, M.; Trinca, L.; Plugaru, N.; Pintilie, L.; Nemnes, G. A.; Iliescu,M.; et al. Iodine Migration and Degradation of Perovskite Solar CellsEnhanced by Metallic Electrodes. J. Phys. Chem. Lett. 2016, 7 (24),5168−5175.(21) Lee, H.; Lee, C. Analysis of Ion-Diffusion-Induced InterfaceDegradation in Inverted Perovskite Solar Cells via Restoration of theAg Electrode. Adv. Energy Mater. 2018, 8 (11), 1702197.(22) Shlenskaya, N. N.; Belich, N. A.; Gratzel, M.; Goodilin, E. A.;Tarasov, A. B. Light-Induced Reactivity of Gold and HybridPerovskite as a New Possible Degradation Mechanism in PerovskiteSolar Cells. J. Mater. Chem. A 2018, 6, 1780−1786.(23) Kim, G. Y.; Senocrate, A.; Yang, T.; Gregori, G.; Gratzel, M.;Maier, J. Large Tunable Photoeffect on Ion Conduction in HalidePerovskites and Implications for Photodecomposition. Nat. Mater.2018, 17, 445−449.(24) Domanski, K.; Correa-Baena, J. P.; Mine, N.; Nazeeruddin, M.K.; Abate, A.; Saliba, M.; Tress, W.; Hagfeldt, A.; Gratzel, M. Not AllThat Glitters Is Gold: Metal-Migration-Induced Degradation inPerovskite Solar Cells. ACS Nano 2016, 10 (6), 6306−6314.(25) Grancini, G.; Roldan-Carmona, C.; Zimmermann, I.; Mosconi,E.; Lee, X.; Martineau, D.; Narbey, S.; Oswald, F.; De Angelis, F.;Graetzel, M.; et al. One-Year Stable Perovskite Solar Cells by 2D/3DInterface Engineering. Nat. Commun. 2017, 8, 15684.(26) Bush, K. A.; Bailie, C. D.; Chen, Y.; Bowring, A. R.; Wang, W.;Ma, W.; Leijtens, T.; Moghadam, F.; McGehee, M. D. Thermal andEnvironmental Stability of Semi-Transparent Perovskite Solar Cellsfor Tandems by a Solution-Processed Nanoparticle Buffer Layer andSputtered ITO Electrode. Adv. Mater. 2016, 28 (20), 3937−3943.(27) Gordon, R. G. Criteria for Choosing Transparent Conductors.MRS Bull. 2000, 25 (8), 52−57.

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1777

(28) Jacobs, D. A.; Catchpole, K. R.; Beck, F. J.; White, T. P. A Re-Evaluation of Transparent Conductor Requirements for Thin-FilmSolar Cells. J. Mater. Chem. A 2016, 4 (12), 4490−4496.(29) Bi, E.; Chen, H.; Xie, F.; Wu, Y.; Chen, W.; Su, Y.; Islam, A.;Gratzel, M.; Yang, X.; Han, L. Diffusion Engineering of Ions andCharge Carriers for Stable Efficient Perovskite Solar Cells. Nat.Commun. 2017, 8, 15330.(30) Zhu, Z.; Zhao, D.; Chueh, C. C.; Shi, X.; Li, Z.; Jen, A. K. Y.Highly Efficient and Stable Perovskite Solar Cells Enabled by All-Crosslinked Charge-Transporting Layers. Joule 2018, 2 (1), 168−183.(31) Kim, Y. C.; Yang, T.-Y.; Jeon, N. J.; Im, J.; Jang, S.; Shin, T. J.;Shin, H.-W.; Kim, S.; Lee, E.; Kim, S.; et al. Engineering InterfaceStructures between Lead Halide Perovskite and Copper Phthalocya-nine for Efficient and Stable Perovskite Solar Cells. Energy Environ.Sci. 2017, 10 (10), 2109−2116.(32) Kaltenbrunner, M.; Adam, G.; Głowacki, E. D.; Drack, M.;Schwodiauer, R.; Leonat, L.; Apaydin, D. H.; Groiss, H.; Scharber, M.C.; White, M. S.; et al. Flexible High Power-per-Weight PerovskiteSolar Cells with Chromium Oxide-Metal Contacts for ImprovedStability in Air. Nat. Mater. 2015, 14 (10), 1032−1039.(33) Zhao, Y.; Nardes, A. M.; Zhu, K. Effective Hole ExtractionUsing MoOx-Al Contact in Perovskite CH3NH3PbI3 Solar Cells. Appl.Phys. Lett. 2014, 104 (21), 213906.(34) Sanehira, E. M.; Tremolet De Villers, B. J.; Schulz, P.; Reese, M.O.; Ferrere, S.; Zhu, K.; Lin, L. Y.; Berry, J. J.; Luther, J. M. Influenceof Electrode Interfaces on the Stability of Perovskite Solar Cells:Reduced Degradation Using MoOx/Al for Hole Collection. ACSEnergy Lett. 2016, 1 (1), 38−45.(35) Christians, J. A.; Schulz, P.; Tinkham, J. S.; Schloemer, T. H.;Harvey, S. P.; Tremolet de Villers, B. J.; Sellinger, A.; Berry, J. J.;Luther, J. M. Tailored Interfaces of Unencapsulated Perovskite SolarCells for > 1,000 h Operational Stability. Nat. Energy 2018, 3, 68−74.(36) Brinkmann, K. O.; Zhao, J.; Pourdavoud, N.; Becker, T.; Hu,T.; Olthof, S.; Meerholz, K.; Hoffmann, L.; Gahlmann, T.; Heiderhoff,R.; et al. Suppressed Decomposition of Organometal HalidePerovskites by Impermeable Electron-Extraction Layers in InvertedSolar Cells. Nat. Commun. 2017, 8, 13938.(37) Hoffmann, L.; Brinkmann, K. O.; Malerczyk, J.; Rogalla, D.;Becker, T.; Theirich, D.; Shutsko, I.; Gorrn, P.; Riedl, T. SpatialAtmospheric Pressure Atomic Layer Deposition of Tin Oxide asImpermeable Electron Extraction Layer for Perovskite Solar Cellswith Enhanced Thermal Stability. ACS Appl. Mater. Interfaces 2018,10 (6), 6006−6013.(38) Zhao, J.; Brinkmann, K. O.; Hu, T.; Pourdavoud, N.; Becker,T.; Gahlmann, T.; Heiderhoff, R.; Polywka, A.; Gorrn, P.; Chen, Y.;et al. Self-Encapsulating Thermostable and Air-Resilient Semi-transparent Perovskite Solar Cells. Adv. Energy Mater. 2017, 7 (14),1602599.(39) Hu, T.; Becker, T.; Pourdavoud, N.; Zhao, J.; Brinkmann, K.O.; Heiderhoff, R.; Gahlmann, T.; Huang, Z.; Olthof, S.; Meerholz,K.; et al. Indium-Free Perovskite Solar Cells Enabled by ImpermeableTin-Oxide Electron Extraction Layers. Adv. Mater. 2017, 29 (27),1606656.(40) Hou, Y.; Du, X.; Scheiner, S.; McMeekin, D. P.; Wang, Z.; Li,N.; Killian, M. S.; Chen, H.; Richter, M.; Levchuk, I.; et al. A GenericInterface to Reduce the Efficiency-Stability-Cost Gap of PerovskiteSolar Cells. Science 2017, 358, 1192−1197.(41) McMeekin, D. P.; Sadoughi, G.; Rehman, W.; Eperon, G. E.;Saliba, M.; Horantner, M. T.; Haghighirad, A.; Sakai, N.; Korte, L.;Rech, B.; et al. A Mixed-Cation Lead Mixed-Halide PerovskiteAbsorber for Tandem Solar Cells. Science 2016, 351 (6269), 151−155.(42) Li, Z.; Yang, M.; Park, J.-S.; Wei, S.-H.; Berry, J. J.; Zhu, K.Stabilizing Perovskite Structures by Tuning Tolerance Factor:Formation of Formamidinium and Cesium Lead Iodide Solid-StateAlloys. Chem. Mater. 2016, 28 (1), 284−292.(43) Tan, W.; Bowring, A. R.; Meng, A. C.; Mcgehee, M. D.;Mcintyre, P. C. Thermal Stability of Mixed Cation Metal HalidePerovskites in Air. ACS Appl. Mater. Interfaces 2018, 10, 5485−5491.

(44) Cheacharoen, R.; Rolston, N. J.; Harwood, D.; Bush, K. A.;Dauskardt, R. H.; McGehee, M. D. Design and Understanding ofEncapsulated Perovskite Solar Cells to Withstand TemperatureCycling. Energy Environ. Sci. 2018, 11, 144−150.(45) Nicolet, M. A. Diffusion Barriers in Thin Films. Thin Solid Films1978, 52, 415−443.(46) Kolawa, E.; Garland, C.; Tran, L.; Nieh, C. W.; Molarius, J. M.;Flick, W.; Nicolet, M. A.; Wei, J. Indium Oxide Diffusion Barriers forAl/Si Metallizations. Appl. Phys. Lett. 1988, 53 (26), 2644−2646.(47) Liu, C. M.; Liu, W. L.; Chen, W. J.; Hsieh, S. H.; Tsai, T. K.;Yang, L. C. ITO as a Diffusion Barrier Between Si and Cu. J.Electrochem. Soc. 2005, 152 (3), G234−G239.(48) Henry, B. M.; Erlat, A. G.; Mcguigan, A.; Miyamoto, T.;Noguchi, N.; Niijima, T.; et al. Characterization of TransparentAluminum Oxide and Indium Tin Oxide Layers on PolymerSubstrates. Thin Solid Films 2001, 382, 194−201.(49) Kang, J.-W.; Jeong, W.-I.; Kim, J.-J.; Kim, H.-K.; Kim, D.-G.;Lee, G.-H. High-Performance Flexible Organic Light-Emitting DiodesUsing Amorphous Indium Zinc Oxide Anode. Electrochem. Solid-StateLett. 2007, 10 (6), J75−J78.(50) Jung, Y. S.; Seo, J. Y.; Lee, D. W.; Jeon, D. Y. Influence of DCMagnetron Sputtering Parameters on the Properties of AmorphousIndium Zinc Oxide Thin Film. Thin Solid Films 2003, 445 (1), 63−71.(51) IEC 61646. Thin Film Terrestrial Photovoltaic (PV) Modules−Design Qualification and Type Approval; 2008.

ACS Energy Letters Letter

DOI: 10.1021/acsenergylett.8b00926ACS Energy Lett. 2018, 3, 1772−1778

1778