Uniaxial Expansion of the 2D Ruddlesden−Popper Perovskite Familyfor Improved Environmental StabilityIoannis Spanopoulos,† Ido Hadar,† Weijun Ke,† Qing Tu,‡ Michelle Chen,† Hsinhan Tsai,∥

Yihui He,† Gajendra Shekhawat,‡,§ Vinayak P. Dravid,‡,§ Michael R. Wasielewski,†

Aditya D. Mohite,⊥ Constantinos C. Stoumpos,*,†,# and Mercouri G. Kanatzidis*,†

†Department of Chemistry, Northwestern University, Evanston, Illinois 60208, United States‡Department of Materials Science & Engineering, Northwestern University, Evanston, Illinois 60208, United States§Northwestern University Atomic and Nanoscale Characterization Experimental (NUANCE) Center, Northwestern University,Evanston, Illinois 60208, United States∥Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States⊥Department of Chemical and Biomolecular Engineering, Rice University, Houston, Texas 77005, United States

*S Supporting Information

ABSTRACT: The unique hybrid nature of 2D Ruddlesden−Popper (R−P) perovskites has bestowed upon them not onlytunability of their electronic properties but also high-performance electronic devices with improved environmentalstability as compared to their 3D analogs. However, there islimited information about their inherent heat, light, and airstability and how different parameters such as the inorganiclayer number and length of organic spacer molecule affectstability. To gain deeper understanding on the matter we haveexpanded the family of 2D R−P perovskites, by utilizingpentylamine (PA)2(MA)n−1PbnI3n+1 (n = 1−5, PA =C H 3 ( C H 2 ) 4 N H 3

+ , C 5 ) a n d h e x y l a m i n e(HA)2(MA)n−1PbnI3n+1 (n = 1−4, HA = CH3(CH2)5NH3

+, C6) as the organic spacer molecules between the inorganicslabs, creating two new series of layered materials, for up to n = 5 and 4 layers, respectively. The resulting compounds wereextensively characterized through a combination of physical and spectroscopic methods, including single crystal X-ray analysis.High resolution powder X-ray diffraction studies using synchrotron radiation shed light for the first time to the phase transitionsof the higher layer 2D R−P perovskites. The increase in the length of the organic spacer molecules did not affect their opticalproperties; however, it has a pronounced effect on the air, heat, and light stability of the fabricated thin films. An extensive studyof heat, light, and air stability with and without encapsulation revealed that specific compounds can be air stable (relativehumidity (RH) = 20−80% ± 5%) for more than 450 days, while heat and light stability in air can be exponentially increased byencapsulating the corresponding films. Evaluation of the out-of-plane mechanical properties of the corresponding materialsshowed that their soft and flexible nature can be compared to current commercially available polymer substrates (e.g., PMMA),rendering them suitable for fabricating flexible and wearable electronic devices.

■ INTRODUCTIONThe broad class of hybrid halide perovskites AMX3 (A = Cs+,CH3NH3

+ (MA), HC(NH2)2+ (FA); M = Ge2+, Sn2+, Pb2+ ; X

= Cl−, Br−, I−) in solid state solar cells has received hugescientific attention because of the highly promising perform-ance of the corresponding assembled devices as well as thebeneficial electronic properties of these direct band gapsemiconductors.1−4 The power conversion efficiencies(PCEs) of perovskite-based devices that have been achievedso far (>23%)5 meet or exceed those of current marketavailable materials, paving the path for their potentialcommercialization.6,7 Although the performance of hybridhalide perovskites is comparable (but mechanistically differ-ent) to that of their major solar absorber competitors (Si,

GaAs),8 there are a number of obstacles that hinder theentrance of these materials to the market,9 with the mostimportant being inadequate environmental stability.10,11 Asolar cell module has to be air, light, and heat stable,withstanding fluctuations in these parameters over a very widetime period.12,13

Recent studies evaluating the life-cycle environmentalimpact of perovskite/silicon tandem modules, containinglead-based hybrid halide perovskites, show that the emittedlead contributes only 0.27% or less to the total freshwaterecotoxicity and human toxicity, as compared to the other solar

Received: February 3, 2019Published: March 4, 2019

Article

pubs.acs.org/JACSCite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

© XXXX American Chemical Society A DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Dow

nloa

ded

via

NO

RT

HW

EST

ER

N U

NIV

on

Mar

ch 2

2, 2

019

at 1

6:28

:43

(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.

module components.14 These encouraging results clearlyindicate that the remaining hindrance to be overcome is theenvironmental stability of the perovskite-absorbing layer.Previously, we and others have shown that 2D perovskites

(e.g., A′2An−1MnX3n+1,15 where A′ is a monovalent or divalent

organic cation16 acting as a spacer between the perovskitelayers, A = monovalent cation, e.g., Cs+, CH3NH3

+, HC-(NH2)2

+; M = Ge2+, Sn2+, Pb2+; X = Cl−, Br−, I−) tend to havesuperior stability than 3D perovskites.17−19 This can beachieved by introduction, during the synthesis of the 3Dperovskites (e.g., MAPbI3, MA = CH3NH3

+), of organicmolecules containing amine groups that can act both as spacersbetween the inorganic layers and as charge-balancing cations ofthe structure.20 The ability to utilize organic spacer moleculesof diverse chemical nature leads to materials with attractiveproperties, one of them being improved moisture andpotentially improved heat and light stability.19

Although there are numerous reports mentioning theenvironmental stability of mixed phase 2D/3D hybrid halideperovskite films and solar cell devices such as PEA2PbI4/Cs0.1FA0.74MA0.13PbI2.48Br0.3 (PEA = C6H5(CH2)2NH3

+,21

(HOOC(CH2)4NH3)2PbI4/MAPbI3,22 (CA2PbI4/MAP-

bIxCl3−x), (CA: cyclopropylamine)23 doped 3D materialswith bulky organic cations such as BA-FAPbI3 (BA =CH3(CH2)3NH3

+),24 FAxPEA1−xPbI3,25 OA/MAPbI3 (OA =

CH3(CH2)7NH3+),26 and doped 2D materials with inorganic

cations such as (BA)2Cs3x(MA)3−3xPb4I13,27 yet these studies

are not systematic. For example, studies evaluating the stabilityof analogous films and devices based only on pure 2Dperovskites in a comparative sense, where the tunableparameter is the organic cation and the layer thickness, arescarce. These would be beneficial, as the extracted informationwould serve to better understand the origin of environmentalbehavior in a variety of efficient solar cells utilizing 2Dperovskites. Chen et al. used isobutylamine (iso-BA) for thesynthesis of 2D perovskite thin films formulated as (iso-BA)2(MA)3Pb4I13, exhibiting not only a promising PCE of10.63% but also improved moisture (RH = 60%) stability for35 days,28 while Smith et al. synthesized crystals of a 2Dperovskite with formula (PEA)2(MA)2Pb3I10, based on thearomatic spacer phenylethylamine (PEA), where the fabricatedfilms were moisture (RH = 52%) stable for 46 days.18

Utilization of the same partially fluorinated aromatic spacer bySlavney et al. for the synthesis of the material(FPEA)2(MA)2Pb3I10 (FPEA = 4-FC6H4(CH2)2NH3

+)showed improved moisture (RH = 66%) stability for morethan 5 days.29 Passarelli et al. used bulky polyaromatic linkerssuch as pyrene-o-propylamine for the synthesis of materialswith chemical formula (R-NH3)2PbI4, where the assembledfilms exhibited superior water stability, as they maintained theircrystallinity after immersion into liquid water for up to 5 min.30

The majority of the reported studies so far provide limitedstability information (only air or heat stability) and examineonly one film composition, e.g., (iso-BA)2(MA)3Pb4I13, n = 4layers; therefore, there is no information about the effect of thenumber of the inorganic layers in the environmental stability ofa 2D material. Additionally there is no comparative studyexamining the effect on stability of the systematic modificationof the organic spacer molecule in the same family ofhomologous materials and among the same inorganic layerthickness under identical experimental conditions.31

The most studied 2D hybrid halide perovskite belongs to theRuddlesden−Popper perovskite family, developed by our

group, with chemical formula (BA)2(MA)3Pb4I13.32−34 In this

case the organic spacer molecule is butylamine (BA), analiphatic amine molecule containing a linear four-carbon chain,while the inorganic part consists of four layers of corner-sharing [PbI6]

4− octahedra that are charge balanced bymethylammonium cations. A solar cell device based on thisspecific material exhibits not only one of the highest PCEamong the 2D-based perovskite compounds of 12.5%35 butalso improved light and moisture stability as compared to themost studied 3D perovskite MAPbI3.

36−38

Taking all of these into account we used targeted synthesisto expand this promising 2D perovskite family of materials andsystematically evaluate their inherent environmental stability asa function of spacer cation as well as perovskite layer thickness(value of n). For this purpose, we used pentylamine (PA) andhexylamine (HA) as the organic spacer molecules, increasingin this way by one and two carbon atoms, respectively, the sideorganic chain as compared to the previously published materialcontaining butylamine (BA, four carbon atoms). Thismodification of the organic part is expected to increase thehydrophobicity of the resulting materials and thus theirmoisture stability. By performing contact angle measurements,we verified our initial hypothesis, providing experimentalevidence of the increase of the contact angle, by increasing thenumber of carbon atoms in the organic spacer molecules.Specifically, we synthesized nine compounds, six of thempresented here for the first time, with chemical formulas(PA)2(MA)n−1PbnI3n+1 (n = 1−5, PA = CH3(CH2)4NH3

+, C5)and (HA)2 (MA) n− 1Pb n I 3 n + 1 (n = 1−4 , HA =CH3(CH2)5NH3

+, C6). For simplicity the notation C5N1−5and C6N1−4 is used throughout this work, where the numberafter the (C) corresponds to the number of carbon atoms ofthe side organic chain and the number after (N) correspondsto the number of the inorganic layers (i.e., layer thickness).The resulting materials were extensively characterized withsingle-crystal XRD and powder XRD (PXRD) studies. High-resolution variable-temperature PXRD measurements usingsynchrotron radiation verified the phase transition temperatureof these perovskites, which were consistent with DSCmeasurements. TGA analysis shed light on the thermalstability of the crystals, while UV−vis spectroscopy, photo-luminescence (PL), and TRPL measurements evaluated theoptical properties of these compounds. Ambient-pressurephotoemission spectroscopy (APS) was used for experimentaldetermination of the valence band maximum of the crystals,allowing accurate determination of the band structures.Following the characterization of the new 2D perovskite

materials we evaluated their environmental (heat, light, andair) stability by performing an extensive series of tests with andwithout encapsulation and compared it to the previouslyreported compound based on butylamine as well as their 3Dcounterpart, MAPbI3. In order for these studies to be as closeas possible to commercial device applications we fabricatedthin films of these perovskites on different substrates(amorphous glass, crystalline SiO2, and polycrystalline Si)and used them for the stability tests. Two different film-castingmethods were used (one step (OS) and hot casting (HC)) aswell as different film concentrations to cover a wider range ofexperimental parameters. The new compounds exhibit indeedimproved thermal (4 h at 100 °C, for C5N4) and light stability(10 h continuously under 1 sun illumination, for C5N5) in air,as compared to the previously reported 2D analogue (C4N4),and much higher than the 3D analogue (MAPbI3) which

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

B

contained a significant amount of PbI2 treated under identicalconditions. At the same time, most of the materials maintaintheir structural integrity in air and desired orientation for morethan a year, depending on the used substrate and castingmethod (C5N3 on polycrystalline (100) Si). Interestingly, ifthe corresponding thin films are encapsulated (protected frommoisture and O2) then their heat (12 days at 100 °C in air, forC5N4, 90 days continuously in the glovebox for C5N3) andlight (60 h under 1 sun illumination in air, for C5N4) stabilityis dramatically increased as compared to the unencapsulatedones. This finding is especially important as it may offer apotential solution to their inherent heat and light instabilityunder ambient conditions, which is obsolete if the films aresufficiently protected from air. Lastly, we evaluated the bulkmechanical properties of the resulting crystalline materials bymeasuring the Young’s modulus and hardness. Apparently, byincreasing the side carbon chain length of the spacermolecules, one can adjust the softness of the material. Thesemeasurements provide insight for the potential utilization ofthose materials not only in static large solar cell assemblies butalso in flexible and wearable electronic devices, an unexploredbut potentially fruitful field for the perovskite materials.

■ RESULTS AND DISCUSSIONWe first discuss the synthetic aspects of this family of materialsand describe their refined crystal structures, thermal analysis,and temperature-induced phase transitions. We then presenttheir optical absorption and photoluminescence properties andstudy of their mechanical properties. Finally, we present thefilm assembly process, which gives uniform highly orientedfilms, which were used for systematic investigation of theirtemporal stability under a broad set of environmentalconditions.Synthetic Aspects. Both new families of 2D perovskites

(PA)2(MA)n−1PbnI3n+1 and (HA)2(MA)n−1PbnI3n+1 were syn-thesized using a HI/H3PO2 solvent mixture, which canproduce uniform and high-quality single crystals. One wouldexpect that since the difference of those materials with thepreviously published ones, (BA)2(MA)n−1PbnI3n+1, is only oneand two carbon atoms using PA and HA, respectively, theoptimum reaction conditions would be if not identical at leastalmost the same.32 However, the resulting synthetic parame-ters, such as the organic spacer and reactants concentration,varied significantly with increasing the carbon chain length ofthe organic spacer molecule. In the case of PA and HA analogs,the synthetic procedure was simplified, both by the use ofmuch cheaper starting materials, such as PbO and MACl,instead of PbI2 and MAI, and by the modification of thereaction steps. Initially, the same amount of PbO (10 mmol)was dissolved under heating in a suitable amount of solvent(16 and 17 mL of HI for PA and HA, respectively), then astoichiometric amount of MACl was added, leading toformation of the 3D analogue (MAPbI3) that was fullydissolved, and then the optimum spacer concentration wasadded to the H3PO2 solution at room temperature (RT).Because of the milder reaction of the amine with the H3PO2compared to the highly exothermic reaction with the HI acid inthe case of synthesis of the BA analogs, use of an ice bath is notneeded, simplifying the reaction conditions. Lastly, the H3PO2solution with the spacer is added to the hot reaction slowly,and the clear yellow solution is left to cool slowly at RT. If thesolution is cooled immediately at RT then a mixture ofproducts, containing different number of inorganic layers, is

formed. This clearly indicates the delicate nature of thecrystallization process of these hybrid materials, where thesystem needs time to arrange in the solution all of the differentreaction components for the acquisition of only onethermodynamically stable phase.In targe t ing a spec ific layer th ickness (e .g . ,

(PA)2(MA)3Pb4I13 = C5N4) the following general syntheticstrategy was developed and implemented in all cases. Thestoichiometric ratio Pb/MA, based on the chemical formula,was utilized, and the concentration of PA was varied until aproduct of a uniform layer number was obtained. Interestingly,although the targeted material was in this example the C5N4,based on the stoichiometric ratio of Pb/MA (4:3), theresulting products were determined by the amount of PAspacer, starting from C5N2 (80% main product) for thehighest amount of PA, to the C5N5 for the lowest amount, seeFigure S2. After the optimum amount of spacer for eachnumber of layers was determined experimentally, a second setof experiments was carried out. In this case the amount ofspacer was fixed, targeting, e.g., the C5N4, and the ratio Pb/MA was modified targeting different number of layers, e.g., 5,6, and 7. As it is shown in Figure S3, the resulting material inall cases was the C5N4, governed by the specific amount ofspacer only. In the final round, after determining the optimumamount of spacer for each specific number of layers, thestoichiometric Pb/MA ratio was used, targeting each numberof inorganic layers, which led to the synthesis of pure, uniformlayer compounds. These results clearly indicate that the criticalparameter toward acquiring a pure product is the concen-tration of spacer in solution and to a much lesser degree thePb/MA ratio. The same trend was also observed in the case ofHA. Notably, for the same amount of Pb2+ cations (10 mmol)the optimum amount of spacer (BA, PA, HA) decreases withincreasing the length of the organic chain (see Figure S4), andalso the overall concentration of the reactants decreases (byincreasing the amount of the HI solvent) because of thedecrease of the solubility of the resulting materials.In Figure S4 we present the dependence of the spacer

concentration toward the number of layers for all 3 families ofmaterials, based on BA, PA, and HA. In all cases the sametrend is observed regarding the concentration of spacer withincreasing number of layers, where it reaches a plateau for n >5. This clearly indicates the synthetic challenge in theacquisition of pure, n > 5 materials, as the difference in theoptimum concentration for high layers is small. Furthermore,considering these trends, it is possible to estimate the optimumconcentration of spacer for specific materials based on spacermolecules with longer or shorter organic chains. This tailor-made synthetic methodology, presented here for the first timein these compounds, can be a unique and general tool forfurther expansion of the 2D perovskite family.

Structural Properties. Single-crystal XRD analysis re-vealed that the resulting materials (7 solved single-crystalstructures, 6 of them new) (PA)2(MA)n−1PbnI3n+1 and(HA)2(MA)n−1PbnI3n+1 belong to the 2D Ruddlesden−Popperperovskite family, and they constitute a uniaxial expansion ofthe previously reported (BA)2(MA)n−1PbnI3n+1 compoundsalong the layer stacking axis. Room-temperature PXRDmeasurements verified the phase purity of the synthesizedmaterials, as there are no residual low-angle peaks (<13° 2θ)which would be an indication of coprecipitation of more thanone inorganic layer in each case (see Figure 1). Furthermore,comparison of the calculated PXRD patterns based on the

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

C

solved single-crystal structures and the experimentally recordedones validate the uniform composition and phase purity of allcompounds (Figures S5−13). The structures of the C5N1,C6N1, and C5N2 materials have been presented before;39,40

however, we synthesized them here again for evaluation oftheir optical properties and for comparison purposes with theother members. SEM studies revealed the uniform rectangularplate shape of the single crystals for both families (PA and HA)of materials, see Figures 2 and S14, a distinctive characteristicfor the 2D Ruddlesden−Popper perovskites.20,41

The single-crystal structures for the (PA)2(MA)n−1PbnI3n+1(n = 1−5) perovskites are presented in Figure 3. The structureof the first member of the series (n = 1) consists of 2Dinorganic layers of corner-shared [PbI6]

4− octahedra which areseparated by two layers of organic pentylammonium cations,where through dispersion interactions (between the carbonchains) and electrostatic interactions (between the ammoniumheads and the anionic perovskite layers) the structure isstabilized, and propagates in 2D. For n > 1 methylammoniumcations reside into the cavities formed by the inorganicoctahedra, charge balancing the framework, in a similar manneras in the case of the 3D perovskites.42

Taking a closer look at the structures of the compounds withthe same number of layers, e.g., n = 4, but with differentorganic spacers, C4N4, C5N4, and C6N4, it is possible toidentify the effect of the change of the organic spacer moleculeon the structural characteristics of the materials. As it isexpected, with increasing the length of the organic spacermolecules the distance between the inorganic layers increasesfrom 7.0 Å for the BA compounds to 8.5 and 9.4 Å for the PAand HA, respectively (without taking into account the van der

Waals radii).43 The BA-based 2D perovskites crystallize inorthorhombic space groups, whereas the PA and HA analogscrystallize in monoclinic space groups. Taking a closer look atthe upper axial Pb−I bond lengths of the octahedra of theouter inorganic layers (those facing the organic layer), for n = 4the distance varies from 3.044(2) to 3.069(4) and 3.050(4) Åfor the BA, PA, and HA analogs, while the lower axial Pb−Idistances vary from 3.3091(18) to 3.302(4) and 3.217(5) Å,respectively, thus progressively decreasing the polarity of thelayers along the confinement direction.Regarding the (Pb−I−Pb)011 in-plane equatorial tilting

angles, in the case of (PA)2(MA)n−1PbnI3n+1 series (n = 2−5) there is a slight decrease with increasing n, with values of162.32(9)°, 164.12(6)°, 162.98(14)°, and 161.4(2)° forC5N2, C5N3, C5N4, and C5N5, respectively. This trend isalmost ironed out by calculating the average Pb−I−Pb angles(axial and equatorial) which exhibit values of 163.1°, 165.1°,1 6 4 . 3 ° , a n d 1 6 3 . 2 ° , r e s p e c t i v e l y . F o r t h e(HA)2(MA)n−1PbnI3n+1 series (n = 2−4), the correspondingaverage values range from 167.6° to 165.2° and 163.4° forC6N2, C6N3, and C6N4, respectively. Additionally, there isnot a clear trend in the (BA)2(MA)n−1PbnI3n+1 perovskite seriesalso (n = 2−5) with values of 164.9°, 170.7°, 169.2°, and165.4° for the C4N2, C4N3, C4N4, and C4N5 materials,respectively. These findings indicate that with increasingthickness of the inorganic layers (i.e., value of n) there isonly a small impact from the organic spacer molecules towardthe distortion of the perovskite structure, as it was alsoobserved in the case of the isostructural 2D hybrid Dion−Jacobson perovskites (3AMP)(MA)n−1PbnI3n+1 and (4AMP)-(MA)n−1PbnI3n+1 (3AMP = 3-(aminomethyl)piperidinium,4AMP = 4-(aminomethyl)piperidinium).44 These observationssupport the trend in the optical properties of the three differenthomologous perovskite families described here (see below).All (PA)2(MA)n−1PbnI3n+1 and (HA)2(MA)n−1PbnI3n+1 com-

pounds crystallize in monoclinic space groups at roomtemperature (Table 1), except for C6N1, which crystallizesin an orthorhombic space group.39 This is verified by thecomparison of the calculated and experimental PXRD patterns,

Figure 1. (A) Comparison of PXRD patterns of the as-made crystalsof materials C5N1−5 and (B) C6N1−4 for different thickness ofinorganic layers.

Figure 2. Representative SEM images of compounds C6N1−4.

Figure 3. Two-dimensional structures of the (PA)2(MA)n−1PbnI3n+1family of materials with increasing thickness of inorganic layers from n= 1 (C5N1) to n = 5 (C5N5). Colored octahedra represent the[PbI6]

4− moieties. For inorganic layers n > 1, methylammoniumcations (CH3NH3

+) are residing among the inorganic layers to chargebalance the framework.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

D

which match exactly (Figure S10). Interestingly, the material(FPEA)2(MA)2Pb3I10, containing the bulky aromatic spacerFPEA, also crystallizes in a monoclinic space group at RT,29

(PEA)2(MA)2Pb3I10 is triclinic at RT,18 while all

(BA)2(MA)n−1PbnI3n+1 compounds are orthorhombic at RT.Temperature-dependent phase transitions are a common

phenomenon in hybrid halide perovskites, not only in thethinner layer 2D C5N1, C6N1, C7N1, and C12N1 perov-skites39,45,46 but also in the case of the 3D analogs MAPbI3

47

FAPbI3,48 MASnCl3,

49 and MAPbBr3.50 Therefore, it is likely

that in the case of higher layer (n > 1) Ruddlesden−Popperhybrid perovskites a similar behavior is observed. To verify thiswe performed high-resolution variable-temperature PXRD aswell as differential scanning calorimetry (DSC) measurements.From the DSC measurements it was possible to identify theexact temperature of the phase transitions occurring above RT.For the (PA)2(MA)n−1PbnI3n+1 (n = 1−5) family of materialsthe transition temperature starts at 47 °C, heating cycle, for thelowest n member compound C5N1, close to the previouslyreported one by Billing et al.39 As the value of n is increased,the transition temperature increases to 78 °C for the C5N2analogue, and then for the C5N3, C5N4, and C5N5 materialsthe temperature remains almost the same with values of 80 and81 °C, respectively (Figure 4a). On the other hand, for the(HA)2(MA)n−1PbnI3n+1 (n = 1−4) analogs, the lowest nmember C6N1 has one main and one smaller endothermic

DSC peak at 82 and 97 °C, in perfect agreement with thepreviously reported ones at 81.7 and 98.1 °C, respectively.39

However, with increasing layer thickness there is only onetransition which shifts to higher temperatures, starting at 96 °Cfor the C6N2 material, 103 °C for the C6N3 material, and 123°C for the C6N4 material (Figure 4b).There are two trends stemming from the above analyses,

with the first one being that the phase transition temperaturedepends on the organic spacer but does not follow a specific

Table 1. Crystal and Structure Refinement Data for C5N2−5 and C6N2−4 Compoundsa

C5N2 C5N3 C5N4 C5N5

cryst syst monoclinic monoclinic monoclinic monoclinicspace group Cc Pc Cc Pcunit cell dimens a = 41.3091(13) Å, α = 90° a = 26.9783(2) Å, α = 90° a = 66.269(5) Å, α = 90° a = 39.4540(7) Å, α = 90°

b = 8.9764(3) Å,β = 95.0826(18)°

b = 8.9540(2) Å,β = 95.5426(9)°

b = 8.9480(5) Å,β = 93.179(4)°

b = 8.9418(6) Å,β = 93.698(4)°

c = 8.8719(3) Å, γ = 90° c = 8.8757(6) Å, γ = 90° c = 8.8839(6) Å, γ = 90° c = 8.893(3) Å, γ = 90°vol 3276.83(19) Å3 2134.02(15) Å3 5259.8(6) Å3 3130.9(10) Å3

Z 4 2 4 2density (calcd) 3.0631 g/cm3 3.3165 g/cm3 3.4741 g/cm3 3.5758 g/cm3

no. of independent reflns 7876 [Rint = 0.0401] 11 948 [Rint = 0.0364] 11 362 [Rint = 0.061] 9569 [Rint = 0.0543]completeness to θ = 29.33° 98% 98% 98% 98%data/restraints/parameters 7876/19/128 11948/20/171 11362/21/214 9569/22/260goodness-of-fit 1.42 1.92 1.50 2.51final R indices [I > 2σ(I)] Robs = 0.0421, wRobs = 0.1111 Robs = 0.0423, wRobs = 0.1214 Robs = 0.0707, wRobs = 0.1840 Robs = 0.1041, wRobs = 0.2677R indices [all data] Rall = 0.0588, wRall = 0.1194 Rall = 0.0560, wRall = 0.1261 Rall = 0.1347, wRall = 0.2083 Rall = 0.1335, wRall = 0.2776largest diff. peak and hole 1.25 and −0.82 e·Å−3 1.71 and −1.88 e·Å−3 2.99 and −3.70 e·Å−3 8.27 and −3.32 e·Å−3

C6N2 C6N3 C6N4

cryst syst monoclinic monoclinic monoclinicspace group Cc Pc Ccunit cell dimens a = 45.3552(8) Å, α = 90° a = 28.1426(2) Å, α = 90° a = 68.752(8) Å, α = 90°

b = 8.9209(2) Å, β = 98.2088(8)° b = 8.94130(10) Å, β = 94.5632(10)° b = 8.9445(9) Å, β = 93.770(5)°c = 8.8062(2) Å, γ = 90° c = 8.9025(5) Å, γ = 90° c = 8.8984(9) Å, γ = 90°

vol 3526.56(13) Å3 2233.05(13) Å3 5460.3(10) Å3

Z 4 2 4density (calcd) 2.899 g/cm3 3.2112 g/cm3 3.3807 g/cm3

no. of independent reflns 8556 [Rint = 0.0384] 12 162 [Rint = 0.0411] 6685 [Rint = 0.0538]completeness to θ = 29.33° 98% 98% 98%data/restraints/parameters 8556/23/134 12162/24/177 6685/25/220goodness-of-fit 1.63 1.47 2.50final R indices [I > 2σ(I)] Robs = 0.0443, wRobs = 0.1173 Robs = 0.0524, wRobs = 0.1375 Robs = 0.0805, wRobs = 0.2178R indices [all data] Rall = 0.0579, wRall = 0.1242 Rall = 0.0818, wRall = 0.1500 Rall = 0.0986, wRall = 0.2242largest diff. peak and hole 1.62 and −1.01 e·Å−3 2.11 and −1.52 e·Å−3 3.61 and −5.59 e·Å−3

aR = Σ||Fo| − |Fc||/Σ|Fo|, wR = (Σ[w(|Fo|2 − |Fc|2)2]/Σ[w(|Fo|4)])1/2, and iw = 1/(σ2(I) + 0.0004I2).

Figure 4. (A) DSC curves for all C5N1−5 materials and (B) C6N1−4 recorded under helium flow. The small peak of sample C6N1 (at 97°C) is not an C6N2 impurity, as there is no methylammonium cationsin the synthesis that could lead to the formation of the two layeredstructure and is consistent to what was observed before.Corresponding sharp endothermic peaks are indicative of the phasetransition temperature of these materials.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

E

order. For the (PA)2(MA)n−1PbnI3n+1 series the temperatureramps up to ∼80 °C and remains nearly constant irrespectiveof n, while in (HA)2(MA)n−1PbnI3n+1 the transition temper-ature increases in a stepwise manner showing a strongdependence on the n number. For (BA)2(MA)n−1PbnI3n+1,the phase transitions appear to be closer to the trend observedfor (PA)2(MA)n−1PbnI3n+1, with all end members converging toa phase transition at ∼10 °C.51 Unlike the trend observed forlonger organic chains analogs, the premelting phase transitionwas not observed in the studied homologous series.52 Thesecond trend is that the phase transition temperature increaseswith increasing layer thickness in (PA)2(MA)n−1PbnI3n+1 and(HA)2(MA)n−1PbnI3n+1, unlike (BA)2(MA)n−1PbnI3n+1wherethe transition temperature is nearly constant. These observa-tions can be explained by the fact that during a phase transitionadjacent inorganic layers shift relative to each other, as theorganic molecules adopt a different orientation, followed by atilting of the inorganic octahedra, as recorded for the C5N1and C6N1 compounds.39

High-resolution variable-temperature PXRD measurementsusing synchrotron radiation (11BM) confirmed the trends andtransition temperatures of the DSC results. In the case ofC5N1 and C6N1 the materials crystallize in a monoclinic(P21/a) and in an orthorhombic (Pbca) space group,respectively, as the experimental RT PXRD patterns (recordedin-house) match exactly the calculated ones from the single-crystal XRD measurements (see Figures S5 and S10).Regarding the (PA)2(MA)n−1PbnI3n+1 (n = 2−4) series, allmaterials overgo a phase transition between 70 and 90 °C(C5N2−4), in agreement with the DSC phase transitiontemperatures (78−81 °C) (Figures 4, S15, and S16). Therecorded PXRD patterns at 70 °C are the same as the RTmeasured ones, while above 80 °C, the corresponding patternsshift and shrink to lower 2θ values, followed by disappearanceof the (400) peak of the monoclinic phase and appearance ofthe (080) peak of the orthorhombic phase (see Figure 5). At69 °C in the cooling cycle PXRD pattern, the phasetransformation exhibits a hysteresis, as both phases existsimultaneously, which is evident by the presence of bothcharacteristic (400) and (080) peaks (see Figures 5 and S15and S16). All phase transitions were fully reversible in theexamined temperature range (29−180 °C), as the initial andthe final target temperature diffraction patterns match exactly.S im i l a r r e s u l t s we r e a l s o ob s e r v ed i n t h e

(HA)2(MA)n−1PbnI3n+1 (n = 2−4) series of materials. TheC6N2 and C6N4 perovskites exhibit a phase transitionbetween 90 and 110 and 110 and 130 °C, in agreement withthe DSC phase transition temperatures of 96 and 123 °C,respectively (see Figures S17−19). These results are consistentwith the crystallographic data for the (BA)2(MA)n−1PbnI3n+1series of materials, where because of the shorter organic linkerthe phase transitions (apparently from monoclinic toorthorhombic space groups) take place at temperatures lowerthan 25 °C for all structures (n = 1−5) which at RT crystallizein orthorhombic space groups.53

Indexing of the synchrotron PXRD data suggests that thematerials undergo a space group change from monoclinic (Ccand Pc for even and odd number of n, respectively) at RT toorthorhombic (Cc2m and C2cb) at high temperatures (seeFigures S20−26), which is accompanied by a drastic increasein the unit cell volume (because of an increase in the size of thelayer stacking axis). These high-resolution PXRD studies alsoshed light on the high-purity phase of the synthesized

materials. It is evident that the current synthetic conditionsafforded compounds C5N2−4 of pure phase (Figures S20−22), while C5N5 contains 1% C5N6 and 1% C5N4 phases,based on indexed PXRD patterns (Figure S23). Similarly,C6N2 was acquired in pure phase (see Figure S24), whereasC6N3 contains 1% C6N2 impurity (see Figure S25), andC6N4 contains 1% C6N3 impurity (see Figure S26).Markedly, because of the high-quality data in the case of

C5N4 it was possible to determine the crystal structurethrough the Rietveld refinement method at high temperature(453 K) (see Figure 6). The structure crystallizes in theorthorhombic Cc2m space group (see Table S1) with a 3.2%slightly expanded unit cell volume as compared to the RTmonoclinic structure because of thermal expansion. The (Pb−I−Pb)011 in-plane equatorial tilting angles of the inorganicoctahedra are 170.9(5)°, higher than the RT structure withvalues of 162.98(14)°, revealing a much less distortedstructure, which is also verified by the perfect ordering of theplanar pentylammonium cations among the layers (see Figure6), as observed in the case of C5N1 orthorhombic phasestructure.39 Furthermore, looking across the b axis in theorthorhombic C5N4 structure the adjacent octahedra in thesame slab eclipse each other, while they are perfectly staggeredto the octahedra of the adjacent slab (see Figure S27).However, in the case of the monoclinic C5N4 structure,adjacent octahedra in the same slab and among adjacent slabsare slightly displaced from the ideal staggered arrangement.This is in contrast to what was observed in the case of themonoclinic C5N1 structure, where the adjacent octahedra

Figure 5. High-resolution variable-temperature PXRD patterns ofcompound C5N3, consisting of one heating and one cooling cycle.Sample undergoes a reversible phase transition, which takes placebetween 70 and 90 °C, from monoclinic Pc phase to orthorhombicC2cb. Highlighted area is enlarged to better show the shift of thediffraction peaks with increasing temperature.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

F

were in an eclipsed arrangement. The same type ofdisplacement was observed for all monoclinic C5N2−5compounds.Thermogravimetric analysis (TGA) measurements shed

light on the thermal stability of the new materials. Bothfami l ies (PA)2(MA) n− 1Pb n I 3 n+1 (n = 1−5) and(HA)2(MA)n−1PbnI3n+1 (n = 1−4) are thermally stable up to270 °C, just 25 °C below their 3D analogue MAPbI3, where itsfirst weight loss step appears at 295 °C.54 There are twodistinct decomposition steps at ∼270 and ∼450 °C (FiguresS28−36). The first weight loss corresponds to decompositionof the organic part of the structure and HI; the second stepcorresponds to evaporation of the inorganic part, PbI2, aboveits melting temperature. The organic content of the 2Dperovskites undergoes mass loss at the first decompositionstep, which decreases gradually with increasing layer thickness,from 39% for the C5N1 to 32% for C5N5. Similarly, theinorganic step weight loss gradually increases from 57% to60%, respectively.

Optical Absorption and Photoluminescence. Becausethe 2D hybrid perovskite structures are natural quantum wellswith semiconducting slabs alternating with dielectric slabs,55,56

the band gap and the PL peak positions shift to lower energieswith increasing layer thickness because of the quantum anddielectric confinement effects.57−59 The signature character-istic, deriving from these properties, is the presence of sharpexcitonic peaks that gradually diminish with increasing numberof layers. In the case of (PA)2(MA)n−1PbnI3n+1 (n = 1−5)materials, the band gap ranges from 2.39 eV for the C5N1 to1.76 eV for the C5N5 perovskites (based on the absorptionedge), while for the HA analogs these values range from 2.34to 1.85 eV for C6N1 to C6N4, respectively (Table 2 andFigure 7a and 7b). These results are similar for the n = 2−4

layers in both series, exhibiting only a small difference of 0.05eV for the n = 1 materials, indicating that it is primarily theinorganic perovskite part of the structure that governs theoptical properties of the materials. The main difference amongthem can be attributed to the distortion of the structure,represented by the Pb−I−Pb equatorial tilting angles. Forexample, C6N1 and C4N1 have almost the same energy bandgaps of 2.34 and 2.35 eV, respectively, and tilting angles with

Figure 6. Comparison of the two C5N4 structures before and afterthe phase transition. In the high-temperature orthorhombic phase thetilting angle of the adjacent octahedra increases whereas the tiltingangle of the organic chains decreases, signaling a much less distortedinorganic part and a highly oriented organic part.

Table 2. Measured Band Gap Energy (eV) from the Absorption Spectra, Measured Valence Band Maxima (eV), andCorresponding PL Peak Positions (eV) for Both Families (PA)2(MA)n−1PbnI3n+1 (C5N1−5) and (HA)2(MA)n−1PbnI3n+1(C6N1−4)a

(PA)2(MA)n−1PbnI3n+1 (HA)2(MA)n−1PbnI3n+1

no. of layers(n)

band gap (abs.edge)(eV)

band gap (excit.peak)(eV)

PL peak(eV)

VBM(eV)

band gap (abs.edge)(eV)

band gap (excit.peak)(eV)

PL peak(eV)

VBM(eV)

1 2.39 2.47 2.42 5.61 2.34 2.41 2.36 5.512 2.09 2.16 2.14 5.45 2.10 2.16 2.14 5.403 1.97 2.01 2.00 5.39 1.96 2.01 2.00 5.374 1.85 1.91 1.90 5.37 1.85 1.91 1.90 5.315 1.76 1.83 1.84 5.34MAPbI3 1.52 1.62 5.44

aThe band gap energy is calculated based on both the absorption edge and the exciton peak of the optical absorption spectra, while the PL energy iscalculated by the PL peak position of the optical emission spectra.

Figure 7. Systematic evolution of optical absorption spectra ofcompounds (A) C5N1−5 and MAPbI3 and (B) C6N1−4. Emissionspectra of compounds (C) C5N1−5 and MAPbI3 and (D) C6N1−4.With increasing thickness of the inorganic layers the band gap and thePL shifts to lower energies, followed by the gradual diminishing of theexciton peak.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

G

values of 155.65(5)° and 155.08(2)°, whereas the C5N1material with a more distorted structure (tilting angle153.68(3)°) has a higher band gap of 2.39 eV, as it was alsoobserved in the GAMAPbI4 perovskite.

60 The band gap valuesbased on the excitonic peak positions are the same in all threeRP families for BA-, PA-, and HA-based analogs (n = 2−4),verifying that indeed the nature of the insulating organic layerhas a negligible effect on the optical properties of n > 1 layeredperovskite structures but plays a more important role for the n= 1 structures.61,62

The presence of stable excitons at RT leads to strong PLlight emission for all of the title materials, ranging from 2.42 eV(512 nm) for C5N1 to 1.84 eV (674 nm) for C5N5 and 2.36eV (525 nm) for C6N1 to 1.90 eV (652 nm) for C6N4 (Table2 and Figure 7c and 7d). These values match closely theexcitonic peak positions of the corresponding absorptionspectra and match with the relative PL values of the BA-basedmaterials (for n > 1). Time-resolved PL measurements shedlight on the lifetime of the excitons, which lay in the rangefrom 113 to 310 ps for the PA materials and from 113 to 303ps for the HA analogs (Figures S37 and 38). These values areconsistent with other multilayered 2D perovskites, such as theBA-based Ruddlesden−Popper materials, with values ofaround 200 ps,63 the AMP-based Dion−Jacobson with valuesin the range of 100−280 ps,44 and the PEA-basedRuddlesden−Popper with values of 158−189 ps.64

Considering the importance of knowing the exact bandenergy structure of a semiconducting material that poses as acandidate in photovoltaic applications,65,66 we used ambient-pressure photoemission spectroscopy (APS) to gain insightinto the electronic band alignment of the synthesizedperovskite materials.67 The position of the valence bandmaxima (VBM) was determined through APS measurements,while the conduction band minimum (CBM) was determinedby subtracting the corresponding band gap values (based onthe absorption edge) from the measured VBM energies. TheVBM values of the PA crystalline analogs range from 5.61 eVfor C5N1 to 5.34 eV for C5N5, while for the HA materialsthese values vary from 5.51 eV for C6N1 to 5.31 eV for C6N4(Figure 8). The error in these values is ±0.05 eV. Thecorresponding value for MAPbI3 single crystals was also

determined, as an internal method validation proof, and foundto be 5.44 eV, matching to the widely accepted VBM value of5.43−5.46 eV.68,69

Mechanical Properties. When it comes to the imple-mentation of a semiconductor material into real applications(photovoltaic, electronic, etc.), except for their environmentalstability issues, another trait that must be taken into account istheir mechanical properties.70 The inherent ability of a materialto withstand mechanical stress and/or strain in plane and out-of-plane can determine to which extend and in whichcommercial applications these materials can be utilized.71

Toward this end we evaluated the out-of-plane mechanicalproperties of C4N4, C5N4, and C6N4 2D perovskite singlecrystals through determination of Young’s modulus (E) andHardness (H) by performing nanoindentation measure-ments.72 The Young’s modulus E and hardness H of C4N4crystals (E = 6.02 ± 0.57 GPa and H = 0.76 ± 0.09 GPa) arehigher than those of C4N3 (E = 4.40 ± 0.57 GPa and H = 0.43± 0.07 GPa) but lower than those of C4N5 (E = 9.03 ± 1.88GPa and H = 1.06 ± 0.29 GPa) in accordance with ourprevious measurements (Figure 9).73 As an extra proof of

validation of our method we synthesized large MAPbI3 singlecrystals and determined the same out-of-plane mechanicalproperties. Apparently, the E and H values of the bulk MAPbI3crystals are 23.92 ± 3.63 GPa and 1.10 ± 0.19 GPa,respectively, which are much higher than the correspondingvalues of the 2D perovskites. Furthermore, these values areconsistent with the reported literature values of E = 23.9 ± 1.3GPa,74 E = 20 ± 1.5 GPa and H = 1 ± 0.1 GPa,75 and E =23.68 GPa.76

The E and H values of C5N4 (E = 4.88 ± 0.59 GPa and H =0.54 ± 0.08 GPa) and C6N4 (E = 3.64 ± 0.60 GPa and H =0.35 ± 0.05 GPa) are higher than the corresponding values ofC5N1 (E = 2.65 ± 0.43 GPa and H = 0.20 ± 0.07 GPa) andC6N1 (E = 2.10 ± 0.14 GPa and H = 0.21 ± 0.02 GPa),respectively, but lower than those of C4N4. It is clear that asthe alkyl chain length of the spacer molecules increases fromfour carbon atoms (BA) to six atoms (HA), both E and Hvalues decrease. This is reasonable as the amount of the softerorganic part in the structure increases and the stiffer inorganicpart decreases.In the particular case of C6N4 perovskite, the E value of 3.64

± 0.60 GPa can be well compared to the Young’s modulusvalues of every day used polymers, such as Nylon-6 E = 3.15GPa, PS (polystyrene) E = 3.72 GPa, and PMMA (poly-(methyl methacrylate) E = 3.83 GPa.77 This increasedstructural flexibility and soft nature of the examined 2D

Figure 8. Comparative band energy diagram of all synthesizedperovskite crystals C5N1−5, C6N1−4, as well as MAPbI3 forcomparison purposes.

Figure 9. (A) Determined Young’s moduli and (B) hardness for thecrystals of C4N4, C5N4, and C6N4 as well as MAPbI3 forcomparison purposes. With increasing amount of the organic part(for the same number of inorganic layers) both values diminishaccordingly.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

H

perovskite materials is beneficial as it could allow theirutilization in flexible and wearable electronic devices,78 wheretheir major competitors Si and GaAs with very high E values of174.879 and 87 GPa,80 respectively, are much more stiff andunfavorable.Film Assembly. The majority of the reported film

fabrication protocols extensively utilize mixtures of precursorcomponents (e.g., PbI2, MAI, FAI, BAI, etc.) for film assemblyin a glovebox.81,82 In the present study we used solutions madefrom single crystals to fabricate the corresponding films in air.When casting 2D perovskite films a critical parameter that

must be taken into account is the crystalline orientation of theperovskite on the substrate. As the structure of a 2D perovskiteis anisotropic the inorganic layer will lay either parallel orperpendicular to the substrate. If parallel then charge transferand collection is blocked leading to poor photovoltaicperformance.83 Fortunately, the natural tendency of 2Dperovskites with n > 2 is to deposit with the inorganic layersperpendicular to the substrate, leading to good performance, aswe first reported.19,35 In the case where the inorganic layers areoriented vertical to the substrate the resulting PXRD pattern isdominated by 2 diffraction peaks corresponding to the (011)and (022) set of planes (monoclinic space groups).Considering, e.g., the C6N3 material, these peaks will occurat 14.1° 2θ and 28.4° 2θ (Cu Rad), respectively. Indeed, thecalculated PXRD pattern assuming preferred orientation(March−Dollase)84 along these specific set of planes matchesexactly the experimentally observed ones (Figure S39).However, not all layered materials can be oriented along thisspecific direction, as the PXRD patterns of the low layeredC5N2, C6N2, and C6N3 compounds using the one-stepmethod (OS) (see SI) are consistent with randomly orientedpolycrystalline samples (Figure S40a,e,f), similar to the C4N2material.19

The thin film crystallinity and vertical orientation of theinorganic layers can be enhanced by using the hot-casting(HC) method where the dissolved crystals’ solution is spin caston a hot substrate.35 In the case of C6N3 and C5N5compounds, this leads to a significantly improved preferredorientation of the layers and film crystallinity (Figure S41) ascompared to the conventional one-step method (OS), wherethe solution is cast on the substrate at RT and then thesubstrate is postannealed at a specific temperature.85 The highcrystallinity and vertical orientation of the 2D C5N3 andC6N3 films obtained from the HC method are furtherconfirmed by synchrotron grazing-incidence wide-angle X-rayscattering (GIWAXS) measurements, where distinct Braggspot features are observed predominantly in the out-of-plane(qz) direction (see Figure 10e and 10f).Contact angle measurements on the fabricated perovskite

films MAPbI3, C4N3, C5N3, and C6N3 from DMF solutionson glass substrates revealed the hydrophobicity trend of the 2Dmaterials. The 3D film upon contact with the water dropletexhibits a very low average contact angle of 37.5°, which isconsistent with previously reported measurements.86 On theother hand, for the 2D perovskites the contact angle increaseswith increasing the length of the carbon chain of the organicspacer molecules, ranging from 68.9° for the C4N3 to 78.3° forthe C6N3 perovskite film, validating the increased hydro-phobicity of the materials (Figure 10a−d).Film Stability Studies. Notably, most reported film and

device environmental stability tests evaluate only the moisturestability for short periods of time and to a lesser extent include

heat and light stability of the corresponding films anddevices.87 Here we provide a more comprehensive set ofresults under a variety of testing conditions for an extensiveseries of compounds. Examination of all three parameters (air,heat, and light stability) is very important as many of the testprotocols for evaluation of solar cells aiming at commercializa-tion, such as ISOS-D-3 (damp heat) and ISOS-L-3, require notonly high levels of humidity (RH 50−85%) but also hightemperatures (65−85 °C) and 1 sun illumination (AM1.5G).88,89 Therefore, in order for our studies to be close tothese stability parameters we fabricated thin films based on thenew PA (C5N2−5) and HA (C6N2−4) perovskite materialsand performed the following tests: (a) heat treatment at 100°C in air, no encapsulation, RH 50%, (b) light stability underambient light, no encapsulation, RH 50%, (c) light treatmentunder 1 sun in air, no encapsulation, RH 50%, (d) air stabilityin the dark, no encapsulation, RH 50−80%, (e) heat treatmentat 100 °C in air, with encapsulation, and (f) light treatmentunder 1 sun in air, with encapsulation. Furthermore, twodifferent film fabrication methods were used and evaluated,one step (OS) and hot casting (HC), as well as different filmconcentrations (0.35 M for HC only and 1 M for HC and OS),and for one material the air stability in the dark was tested fordifferent substrates. For more details see the SI.

Figure 10. (Top) Contact angles between the perovskite filmsMAPbI3 (A), C4N3 (B), C5N3 (C), and C6N3 (D) and a drop ofwater. Contact angle increases with increasing length of the carbonchain of the organic spacer molecules in the 2D perovskites, resultingin increased hydrophobicity. Examined films were decomposed;however, this measurement offers a good indication of the effect ofthe organic spacer molecule on the hydrophobic properties of thematerials. (Bottom) Synchrotron grazing-incidence wide-angle X-rayscattering (GIWAXS) data of the HC-prepared C5N3 (E) and C6N3(F) films on amorphous glass substrate.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

I

In order for the stability tests to be comparable among thedifferent materials tested, the same synthetic conditions andsolution concentrations (1M) were used for both filmdeposition methods. In particular, using the HC castingmethod another lower concentration was utilized (0.35 M) forcomparison purposes. Profilometry measurements wereperformed in all cases to shed light on the film thickness, asthis could be an important parameter for stability evaluation.The thickness of the films fabricated with the HC-O.35Mmethod ranged between 185 and 250 nm, for the HC-1Mmethod these values were between 780 and 900 nm, while forthe OS-1M they ranged from 430 to 600 nm (Table S2).Effect of Deposition Method. It is clear from the solution

deposition behavior of the films that the HC leads to bettergrowth orientation for most of the materials. For example,comparing C5N3−5 and C6N3 films from HC and OSmethods (Figures S40−41) the diffraction peaks are narrowerfor HC and there are no low-angle diffraction peaks.Furthermore, judging from the full width at half-maximum(fwhm) of the (011) peak of the C5N4 OS and HC films(0.21° versus 0.16° respectively) there is a slight improvementon film crystallinity for the HC method.Comparing the effect of the two methods on the heat

stability of the examined films at 100 °C in air, under ambientlight, at RH of 50%, the HC-1M-based compounds exhibitbetter stability, as the amount of formed PbI2, which is themajor decomposition product of Pb-based perovskites in air(see PXRD patterns in Figures S42−60),90,91 is much less thanin the corresponding OS-1M fabricated materials.Effect of Film Thickness. Three different film thickness

ranges (see above) have been tested for the heat stability testsof the examined materials at 100 °C in air, under ambient light,at RH of 50%. The film stability increases with film thickness

for almost all examined materials, as judged by the decreasingamount of PbI2 (Figures S42−67).

Effect of Film Substrate. Perovskite C5N3 was cast usingthe HC method on 5 different substrates (amorphous glass,single-crystalline SiO2, PEDOT:PSS, FTO, and polycrystallineSi). The desired film orientation and very high crystallinity wasobserved in all cases, indicating that this fabrication methodgives high-quality thin films on a variety of substrates, withdifferent chemical substrate compositions (Figure S68).Furthermore, for the C5N3 we tested the stability in air inthe dark (RH 50−80%) cast on glass, single-crystalline SiO2,and polycrystalline Si (polycr.Si) (Figures 11a, 11c, and S69).The film fabricated on polycrystalline Si exhibited the beststability in air, as there was no formation of low-anglediffraction peaks with increasing time exposure. In contrast, inthe case of glass-deposited film there was formation of thehydrate MAPbI3·H2O phase92,93 after 120 days. This impliesthat crystalline substrates may lead to films with highercrystallinity since the fwhm(011) values for the polycrystalline Siat 0.17° are lower than the 0.24° for the amorphous glass.They are also more oriented than those cast on amorphoussubstrates.

Heat Stability in Air, No Encapsulation. In this set ofexperiments the inherent thermal stability of films of MAPbI3,C4N4, C5N2−5, and C6N2−4 was examined after heating thecorresponding films on a hot plate, at 100 °C, in air, underambient light, with RH 50%, for 4 h. The effect of filmthickness based on the different fabrication methods wasdiscussed in the previous section. Therefore, in this sectioneach deposition method (HC-1M, HC-0.35, and OS-1M) willbe examined separately.In the case of HC-1M-based films, visual examination

suggests that the heat stability increases with increasing

Figure 11. (A) Comparison of the PXRD patterns of the fresh C5N3 film synthesized with the OS method (1M) on polycrystalline siliconsubstrate and after 90, 300, and 450 days exposure to air in the dark (RH = 20−80% over the whole year). (C) Same material cast on amorphousglass substrate freshly prepared and after 120 days exposure in air in the dark (RH = 50−80%). After 120 days, because of the very high humiditylevels, there is trace formation of the hydrated MAPbI3·H2O phase. Blue asterisk indicates the 2θ peak position of the strongest diffraction peak ofthis hydrated phase. (B) Comparison of the PXRD patterns of the fresh C5N4 film synthesized with the HC method (1M) on amorphous glasssubstrate and after 10 h continuous illumination of 1 sun in air (RH = 50%). Green asterisks indicate the 2θ peak positions of the strongestdiffraction peaks of PbI2, which can also be observed by the color change of the film. (D) Same material was encapsulated and exposed for 60 h inair (RH = 50%). Film is structurally intact because of the protection from moisture.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

J

thickness of inorganic layer (n). This is based on the emergingyellow color of the film indicative of formation of PbI2, whereit covers more area in the film of the lower n materials.However, PXRD measurements of the heat-treated films donot exhibit a clear trend, revealing that C5N4 and C5N5exhibited the best thermal stability (Figures S47 and S48),although there is some formation of PbI2 based on PXRDmeasurements yet there is no visual color change in the heat-treated films. Comparing C4N4, C5N4, and C6N4 films, it isclear based on PXRD that C5N4 and C6N4 exhibit muchbetter thermal stability than C4N4, according to the trendC5N4 > C6N4 > C4N4 (Figures S47, S51, S43, and S44). Bycomparison, the 3D MAPbI3 film shows a significant amountof PbI2, revealing the inferior heat stability of the 3D analogue,although there is no color change on the heat-treated 3D film(Figure S42). A rather peculiar finding is the fact that althoughthe C6N2 and C5N2 films had the same initial film crystalliteorientation (inorganic slabs parallel to the substrate), theC6N2 decomposes almost completely to PbI2 (Figure S49)while the C5N2 film seems intact with almost no color changeafter 4 h (Figure S45). The corresponding PXRD pattern ofthe latter (C5N2) shows that the there is a change in the filmcrystal orientation upon annealing at 100 °C for 4 h, orientingthe inorganic layers vertical to the substrate, as in the case ofhigher n number films (n ≥ 3). This change in orientation(from the undesired parallel to the desired vertical) leads to animproved thermal stability as a result of the more efficient heatdistribution through the film. The formation of a small amountof PbI2 renders the stability of this n = 2 film comparable to thehigher layer materials, C6N4, C5N4, and C5N5. The fact thatthe film orientation can be adjusted by further annealing it forlonger time may allow the use of C5N2 material as the upperabsorbing layer in tandem devices (Eg = 2.09 eV) or otherelectronic applications.94

Examining the OS-1M-based films, an increase in thethermal stability with increasing number of inorganic layerswas observed clearly only in the case of C6N2-, C6N3-, andC6N4-based films. Comparing the films of the n = 4 basedmaterials, the stability increases again in the order of C5N4 >C6N4 > C4N4, based on the amount of formed PbI2 (FiguresS52, S55, and S59). This means that C5N4 is the most stable,as it contained the least amount of PbI2. Concerning the C5N2behavior, the same result was also observed as in the case ofthe HC method. The orientation of the perovskite changedsignificantly (from the undesired parallel to the desiredvertical) followed by improved thermal stability (Figure S53).Regarding the HC-0.35M-based films, the effect of film

thickness is clear, as these exhibit the lowest stability amongthe three used methods. Nonetheless, among C4N4, C5N4,and C6N4, C5N4 seems to exhibit the least thermal damage,as it is evident from PXRD measurements where the PbI2formation is limited (Figures S60, S63, and S67). Notably theC5N2 films decomposed in the same manner as the otherfilms, without alteration in the film crystalline orientation asobserved in both of the previous cases above. It is possible thatbecause of the much thinner film for 0.35 M as compared to 1M the system could not easily reorient and decomposedrapidly.Light Stability in Air under Ambient Light, No

Encapsulation. In order to identify whether the 2D perovskitematerials are sensitive even to ambient light, this set ofexperiments was designed to evaluate the air stability ofcompounds MAPbI3, C4N4, C5N2−5, and C6N2−4 under

continuous ambient light, with RH 50%, for up to 27 days.This examination involved the test of two different castingmethods, HC-0.35M and OS-1M.Regarding the HC-0.35M-based films, for the 3D MAPbI3

analogue there is formation of a considerable amount of PbI2based on the recorded PXRD pattern (Figure S82); however,there is no change in film color. The stability increases asC5N4 > C4N4 > C6N4, the same trend as in the case of 1 sunlight stability experiments (see below) (Figures S83, S87, andS90). C6N4 has the highest amount of PbI2, while in C5N4there was no detectable PbI2 based on PXRD. This absence ofPbI2 signals the superior stability of C5N4 among all materialsunder the current examined conditions.Because the OS-1M-based films are thicker than those from

HC-0.35M solutions, they exhibit much better stability, up to27 days in air, under continuous ambient light. For n = 4, thestability increases in the following order C5N4 > C6N4 >C4N4. For that reason, C5N4 showed the best stability underthese examined conditions.

Light Stability in Air under 1 Sun Illumination, NoEncapsulation. The inherent light stability of compoundsMAPbI3, C4N4, C5N2−5, and C6N2−4 was also examined byexposing the films (HC-1M) under light, 1 sun solar simulator(AM 1.5G), with RH 50%, for 10 h. The 3D perovskiteMAPbI3 decomposed almost completely after 6 h based onPXRD measurements, where the diffraction pattern isdominated by formation of PbI2 (Figure S72). The stabilitytrend is C5N4 > C4N4 > C6N4. From the correspondingPXRD patterns, C6N4 tends to lose its optimum orientation,as there is the appearance of additional diffraction peaks fromdifferent diffraction planes, accompanied by formation ofhigher amounts of PbI2 (Figure S81), whereas for C5N4 andC4N4 formation of PbI2 is small (Figures 11b and S73).Additionally, in the case of C4N4 there is formation of thehydrated MAPbI3·H2O phase (Figures S73−74), which couldbe a result of the less hydrophobic nature of the film because ofthe shorter segment of the hydrophobic organic part.Comparing all materials, C5N5 exhibited superior lightstability according to the PXRD patterns with no formationof PbI2 and no loss of orientation (Figure S78).Interestingly, although most of the diffraction patterns of the

examined materials contain either a very small amount of PbI2or none as in the case of C6N3 (Figure S80), a large portion ofthe film has changed color, turning yellow, characteristic ofPbI2 and/or other solvated phases.Notably, under illumination in air and no encapsulation the

C5N2 and C6N2 behaved differently from the thicker layeredmaterials. Not only C5N2 but also C6N2 films changedorientation upon light treatment (Figure S79), as observed forthe C5N2 heat-treated sample, (see above). The orientation ofthe perovskites changed significantly, as the low-anglediffraction peaks (<12° 2θ) disappear; however, C5N2exhibited much better stability than C6N2, which based onPXRD analysis not only contains a higher amount of PbI2 butalso has lower crystallinity, as evidenced by the increase in thesignal-to-noise ratio in the recorded pattern (Figure S79). Tothe best of our knowledge, this behavior of crystal reorientationin the film upon exposure to intense light or heat treatment isreported here for the first time in 2D perovskite-based films,and it may have serious implications on the device perform-ance over extended periods of time.

Air Stability in the Dark, No Encapsulation. Taking intoaccount the fact that perovskite compounds are air and light

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

K

sensitive,95 based on the above results, we also performed airstability tests in the dark for MAPbI3, C4N4, C5N2−5, andC6N2−4 with RH ranging from 20% in the winter to 80% inthe summer for more than 1 year. In this examination the OS-1M method was adopted for most of the materials, while in thecase of C5N3 and C6N4, different substrates (amorphousglass, crystalline Si, and crystalline SiO2) were used.In particular, except for MAPbI3 which decomposes rapidly

in air, all other materials are stable in air in the dark for at least120 days. There is no formation of PbI2 nor change in filmcolor; however, because of the high humidity levels in thesummer time (∼80%) there was formation of a very smallamount of the MAPbI3 hydrate phase, marked by the blueasterisk in the corresponding PXRD patterns, Figures S99−104. Interestingly, the fabricated C5N3 films on crystalline Si(450 days in air), SiO2 (150 days in air) (Figure S69), andC6N4 on SiO2 (150 days in air) (Figure S70) exhibit muchbetter air stability than on amorphous glass-deposited ones(120 days in air) (see Figures 11c and S104, respectively), asthere is absence of the hydrate MAPbI3 phase. The uniformsurface of the crystalline substrates leads to better filmcrystallinity (see also Effect of Film Substrate), which resultsin enhanced air stability.Heat Stability in Air, with Encapsulation. It has been

established that if a perovskite film is protected from moistureand O2 the stability is increased exponentially.96 This is usuallyachieved through encapsulation of the corresponding filmusing various methods, such as thermoplastic sealants andthermally or UV curable epoxys.97,98 For that reason we used athermally curable thermoplastic sealant (Surlyn) to encapsulatethe MAPbI3-, C4N4-, C5N4-, C6N4-, and C5N3-based films(OS-1M on amorphous glass and crystalline SiO2) andevaluate their thermal stability by heating the films on a hotplate, at 100 °C, in air, under ambient light, with RH 50%, for12 days continuously. Details about the encapsulation processare given in the SI.C4N4, C5N4, and C6N4 were chosen as benchmark

materials considering the fact that C4N4 has demonstratedhigh PCE solar cells and the expectation that the longer-chainanalogs may also exhibit high PCE, accompanied by improvedenvironmental stability. Examining the heat-treated encapsu-lated films on amorphous glass, it is clear that yellow PbI2forms but only at the edges (Figures S105−108). This couldbe because of the insufficient protection near the edges leadingto some degradation.Comparing the materials that were cast on amorphous glass

and encapsulated, MAPbI3 was structurally intact after 12 daysof continuous heat treatment, whereas in the 2D analogs afterthat time the stability follows the trend C5N4 > C4N4 >C6N4, based on the amount of detected PbI2 from PXRDmeasurements of the uncovered films (removal of glass cover)after treatment. It is possible that this sealing method did notprovide enough protection from the moisture, leading to someformation of PbI2. To test if this were the case, we fabricatednew films based on C5N3 on crystalline SiO2 and performedthe same test inside a nitrogen-filled glovebox in the absence ofmoisture. These films were encapsulated with the sameprotocol as above. The films were then split into two groups.One group was placed on the hot plate continuously at 100 °Cfor 90 days, while the other group stayed for 40 days only 8 hper day at the same temperature (100 °C). When theencapsulation of the films was removed, PXRD measurementsrevealed that the film which stayed for 40 days 8 h per day was

structurally intact, without traces of PbI2 or loss of crystallinityand orientation, while the film which was continuously exposedto 100 °C for 90 days exhibited a trace amount of PbI2 and novisual color change (Figure 12). This again could be a result of

the loss of the sealant protection ability rather than an intrinsiceffect. The degradation of the sealant or loss of sealantprotection is therefore minimized in the absence of moisture,providing absolute protection to the perovskite film. Theseresults clearly indicate that if the perovskite is perfectlyprotected from moisture and O2 then its stability is increasedsubstantially (tested under extreme conditions, 90 days at 100°C continuously, see Figure S109), rendering these materialssuitable for real photovoltaic applications. The results alsosuggest that inherently these 2D perovskites are stable to lightand heat and any instability derives from interactions withexternal environmental factors (e.g., moisture, oxygen).

Light Stability in Air, Under 1 Sun Illumination withEncapsulation. In a similar manner, as in the previous test, thelight stability of MAPbI3, C4N4, C5N4, and C6N4 wasexamined (OS-1M) by exposing the films to light, 1 sun solarsimulator (AM 1.5G), with RH 50%, for 4 h per day for 15days (60 h in total). Comparing all films, MAPbI3 was the leaststable; we observed severe damage, which lead to theformation of a huge amount of PbI2 based on PXRDmeasurements (Figure S112). On the contrary, all 2D filmsexhibited far better light stability than MAPbI3 in the followingorder C5N4 > C4N4 > C6N4, exactly the same order as in theunencapsulated experiment (light stability in air). C5N4exhibited superior stability with zero formation of PbI2 (Figure11d). In the case of C4N4 and C6N4, however, there weresome traces of yellow PbI2 spots on the protected area of thefilm (inner area defined by the line of the sealant around theedges of the film) and on the unprotected edges of the films(Figures S110 and S111). Considering that the same stabilitytrend is observed for films with and without encapsulation, weconclude that indeed C5N4 exhibits the best light stabilityamong this family of 2D perovskite films and that proper

Figure 12. Comparison of powder X-ray diffraction patterns (PXRD)for the as-made C5N3 synthesized with the OS method (1M) and 2encapsulated films that were left at 100 °C inside a glovebox for 90days continuously (blue pattern) and 8 h per day for 40 days (redpattern). The film that was left for 90 days continuously exhibits atrace amount of PbI2, indicated by the green asterisk, while the oneleft for 8 h per day for 40 days is intact. This clearly verifies that if thefilms are sufficiently protected from moisture then their heat stabilityis very high.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

L

encapsulation of the films leads to much longer light stability,far superior to the 3D perovskites.

■ CONCLUSIONS

The library of Ruddlesden−Popper-type 2D perovskitematerials with general formula (cation)2(MA)n−1PbnI3n+1 hasbeen expanded by the use of pentylammonium (PA) andhexylammonium (HA) cations as organic spacer molecules (n= 1−5). These materials have been synthesized using anoptimized and universal methodology, which gave rise touniform high-purity compounds, as verified by extensivecharacterization techniques. This represents a uniaxialexpansion of the R−P family of materials along the layerstacking direction and goes two steps further from thebutylamine (C4N1−5)-based layered materials to pentylamine(C5N1−5) and hexylamine (C6N1−4) analogs. This gen-eration of new perovskites allowed us to carry out a systematicand comparative study of comprehensive stability tests amongits members. Utilization of high-resolution PXRD measure-ments shed light on the thermal stability, phase purity, andphase transitions of these higher layer 2D perovskites for thefirst time in the literature, revealing an increase in crystalsymmetry with increasing temperature. The new materials havemaintained the unique optical characteristics of the BAanalogs, while at the same time they exhibit high air, heat,and light stability (especially for n = 4). Our extensive studiesshow the following: (a) the HC fabrication method leads toimproved crystallinity and orientation in all of the higher layermaterials and in all different examined substrates, revealing itsgeneral applicability, (b) the thickness of the film is critical, asthe stability increases with increasing layer thickness, (c)crystalline substrates are essential for improved orientation andlong-term air stability, (d) in terms of heat stability (100 °C)in air, for the same number of layers, the stability follows thetrend C5N4 > C6N4 > C4N4 > MAPbI3, regardless of thefabrication method and film thickness, (e) in terms of lightstability (1 sun) in air the trend is C5N4 > C4N4 > C6N4 >MAPbI3, (f) in terms of air stability under ambient light thetrend is C5N4 > C4N4 > C6N4 > MAPbI3 for the HC methodand C5N4 > C6N4 > C4N4 for the OS method, (g) regardingthe air stability in the dark all 2D materials are stable for morethan 120 days when cast on amorphous glass substrates, whilestability increases when cast on crystalline substrates, e.g.,C5N3-based film on crystalline silicon is stable in air for morethan 1 year (450 days so far), C6N4 and C5N3 films oncrystalline SiO2 are stable for more than 150 days in air, and(h) for heat and light stability under encapsulation the stabilitydecreases with the following order C5N4 > C4N4 > C6N4.One of the most important findings is that if the films areefficiently protected from O2 and moisture through a suitablesealant assembly the heat (e.g., 90 days continuously at 100°C) and light stability increases drastically, validating that thesematerials are indeed inherently heat and light stable. Thisstability coupled with their soft and flexible nature, as verifiedby evaluating their mechanical properties, renders themsuitable for wearable and flexible electronic devices, expandingtheir utilization beyond photovoltaic applications. Futurestudies should focus on the fabrication of optoelectronicdevices based on this broad library of 2D perovskite halidematerials as well as determination of the source of the superiorenvironmental stability of pentylamine-based compounds, ascompared to butylamine- and hexylamine-based analogs.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/jacs.9b01327.

Materials and methods, synthetic details, additionalsupplementary figures and tables about materialcharacterization, SEM images, X-ray diffraction measure-ments, photoluminescence measurements, thermogravi-metric analysis, profilometry measurements, and filmstability studies (PDF)C r y s t a l d a t a f o r c o m p o u n d(CH3(CH2)4NH3)2(MA)3Pb4I13 (C5N4) (CIF)C r y s t a l d a t a f o r c o m p o u n d(CH3(CH2)4NH3)2(MA)4Pb5I16 (C5N5) (CIF)Crystal data for compound (CH3(CH2)5NH3)2(MA)-Pb2I7 (C6N2) (CIF)C r y s t a l d a t a f o r c o m p o u n d(CH3(CH2)5NH3)2(MA)2Pb3I10 (C6N3) (CIF)C r y s t a l d a t a f o r c o m p o u n d(CH3(CH2)5NH3)2(MA)3Pb4I13 (C6N4) (CIF)Crystal data for compound (CH3(CH2)4NH3)2(MA)-Pb2I7 (C5N2) (CIF)C r y s t a l d a t a f o r c o m p o u n d(CH3(CH2)4NH3)2(MA)2Pb3I10 (C5N3) (CIF)

■ AUTHOR INFORMATIONCorresponding Authors*m-kanatzidis@northwestern.edu*cstoumpos@materials.uoc.grORCIDIoannis Spanopoulos: 0000-0003-0861-1407Ido Hadar: 0000-0003-0576-9321Weijun Ke: 0000-0003-2600-5419Qing Tu: 0000-0002-0445-6289Yihui He: 0000-0002-1057-6826Gajendra Shekhawat: 0000-0003-3497-288XVinayak P. Dravid: 0000-0002-6007-3063Michael R. Wasielewski: 0000-0003-2920-5440Constantinos C. Stoumpos: 0000-0001-8396-9578Mercouri G. Kanatzidis: 0000-0003-2037-4168Present Address#Department of Materials Science and Technology, Universityof Crete, Heraklion GR-70013, Greece.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by ONR Grant N00014-17-1-2231.APS measurements were carried out with equipment acquiredby ONR grant N00014-18-1-2102. It was supported in part bythe LEAP Center, an Energy Frontier Research Center fundedby the U.S. Department of Energy, Office of Science, andOffice of Basic Energy Sciences under Award DE-SC0001059(evaluation of optical properties). This work made use of theSPID, EPIC, and NUFAB facilities of Northwestern Uni-versity’s NUANCE Center, as well as the IMSERC facilities,which have received support from the Soft and HybridNanotechnology Experimental (SHyNE) Resource (NSFECCS-1542205), the MRSEC program (NSF DMR-1720139) at the Materials Research Center, the International

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

M

Institute for Nanotechnology (IIN), the Keck Foundation, andthe State of Illinois through the IIN. This research used theresources of the Advanced Photon Source, a U.S. Departmentof Energy (DOE) Office of Science User Facility operated forthe DOE Office of Science by Argonne National Laboratoryunder Contract No. DE-AC02-06CH1135. Work at LosAlamos National Laboratory was supported by the LDRDprogram (XWPG). The authors acknowledge the help of Dr.Kevin Yager and Dr. Ruipeng Li with GIWAXS setup. Thisresearch used the 11-BM of the National Synchrotron LightSource II, a U.S. Department of Energy (DOE) Office ofScience User Facility operated for the DOE Office of Scienceby Brookhaven National Laboratory under Contract No. DE-SC0012704.

■ REFERENCES(1) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.;Seok, S. I. High-performance photovoltaic perovskite layers fabricatedthrough intramolecular exchange. Science 2015, 348, 1234.(2) Jeon, N. J.; Noh, J. H.; Yang, W. S.; Kim, Y. C.; Ryu, S.; Seo, J.;Seok, S. I. Compositional engineering of perovskite materials for high-performance solar cells. Nature 2015, 517, 476.(3) Manser, J. S.; Christians, J. A.; Kamat, P. V. IntriguingOptoelectronic Properties of Metal Halide Perovskites. Chem. Rev.2016, 116, 12956.(4) Li, X.; Bi, D.; Yi, C.; Decoppet, J.-D.; Luo, J.; Zakeeruddin, S.M.; Hagfeldt, A.; Gratzel, M. A vacuum flash−assisted solutionprocess for high-efficiency large-area perovskite solar cells. Science2016, 353, 58.(5) Research Cell Record Efficiency Chart, 2019; National RenewableEnergy Laboratory, 2019; https://www.nrel.gov/pv/assets/pdfs/pv-efficiency-chart.20190103.pdf (accessed Jan 12, 2019).(6) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E.D. Solar cell efficiency tables (version 48). Prog. Photovoltaics 2016,24, 905.(7) Ramanujam, J.; Singh, U. P. Copper indium gallium selenidebased solar cells - a review. Energy Environ. Sci. 2017, 10, 1306.(8) Egger, D. A.; Bera, A.; Cahen, D.; Hodes, G.; Kirchartz, T.;Kronik, L.; Lovrincic, R.; Rappe, A. M.; Reichman, D. R.; Yaffe, O.What Remains Unexplained about the Properties of Halide Perov-skites? Adv. Mater. 2018, 30, 1800691.(9) Stranks, S. D.; Snaith, H. J. Metal-halide perovskites forphotovoltaic and light-emitting devices. Nat. Nanotechnol. 2015, 10,391.(10) Pellet, N.; Gao, P.; Gregori, G.; Yang, T.-Y.; Nazeeruddin, M.K.; Maier, J.; Gratzel, M. Mixed-Organic-Cation Perovskite Photo-voltaics for Enhanced Solar-Light Harvesting. Angew. Chem., Int. Ed.2014, 53, 3151.(11) Park, N.-G.; Gratzel, M.; Miyasaka, T.; Zhu, K.; Emery, K.Towards stable and commercially available perovskite solar cells. Nat.Energy 2016, 1, 16152.(12) Berhe, T. A.; Su, W.-N.; Chen, C.-H.; Pan, C.-J.; Cheng, J.-H.;Chen, H.-M.; Tsai, M.-C.; Chen, L.-Y.; Dubale, A. A.; Hwang, B.-J.Organometal halide perovskite solar cells: degradation and stability.Energy Environ. Sci. 2016, 9, 323.(13) Sherwani, A. F.; Usmani, J. A.; Varun. Life cycle assessment ofsolar PV based electricity generation systems: A review. RenewableSustainable Energy Rev. 2010, 14, 540.(14) Snaith, H. J. Present status and future prospects of perovskitephotovoltaics. Nat. Mater. 2018, 17, 372.(15) Mitzi, D. B. Templating and structural engineering in organic−inorganic perovskites. J. Chem. Soc., Dalton Trans. 2001, 0, 1.(16) Yao, Z.; Zhou, Y.; Yin, X.; Li, X.; Han, J.; Tai, M.; Zhou, Y.; Li,J.; Hao, F.; Lin, H. Role of alkyl chain length in diaminoalkane linked2D Ruddlesden−Popper halide perovskites. CrystEngComm 2018, 20,6704.

(17) Quan, L. N.; Yuan, M.; Comin, R.; Voznyy, O.; Beauregard, E.M.; Hoogland, S.; Buin, A.; Kirmani, A. R.; Zhao, K.; Amassian, A.;Kim, D. H.; Sargent, E. H. Ligand-Stabilized Reduced-DimensionalityPerovskites. J. Am. Chem. Soc. 2016, 138, 2649.(18) Smith, I. C.; Hoke, E. T.; Solis-Ibarra, D.; McGehee, M. D.;Karunadasa, H. I. A Layered Hybrid Perovskite Solar-Cell Absorberwith Enhanced Moisture Stability. Angew. Chem. 2014, 126, 11414.(19) Cao, D. H.; Stoumpos, C. C.; Farha, O. K.; Hupp, J. T.;Kanatzidis, M. G. 2D Homologous Perovskites as Light-AbsorbingMaterials for Solar Cell Applications. J. Am. Chem. Soc. 2015, 137,7843.(20) Stoumpos, C. C.; Soe, C. M. M.; Tsai, H.; Nie, W.; Blancon, J.-C.; Cao, D. H.; Liu, F.; Traore, B.; Katan, C.; Even, J.; Mohite, A. D.;Kanatzidis, M. G. High Members of the 2D Ruddlesden-PopperHalide Perovskites: Synthesis, Optical Properties, and Solar Cells of(CH3(CH2)3NH3)2(CH3NH3)4Pb5I16. Chem. 2017, 2, 427.(21) Cho, K. T.; Grancini, G.; Lee, Y.; Oveisi, E.; Ryu, J.; Almora,O.; Tschumi, M.; Schouwink, P. A.; Seo, G.; Heo, S.; Park, J.; Jang, J.;Paek, S.; Garcia-Belmonte, G.; Nazeeruddin, M. K. Selective growthof layered perovskites for stable and efficient photovoltaics. EnergyEnviron. Sci. 2018, 11, 952.(22) Grancini, G.; Roldan-Carmona, C.; Zimmermann, I.; Mosconi,E.; Lee, X.; Martineau, D.; Narbey, S.; Oswald, F.; De Angelis, F.;Graetzel, M.; Nazeeruddin, M. K. One-Year stable perovskite solarcells by 2D/3D interface engineering. Nat. Commun. 2017, 8, 15684.(23) Ma, C.; Leng, C.; Ji, Y.; Wei, X.; Sun, K.; Tang, L.; Yang, J.;Luo, W.; Li, C.; Deng, Y.; Feng, S.; Shen, J.; Lu, S.; Du, C.; Shi, H.2D/3D perovskite hybrids as moisture-tolerant and efficient lightabsorbers for solar cells. Nanoscale 2016, 8, 18309.(24) Wang, F.; Geng, W.; Zhou, Y.; Fang, H. H.; Tong, C. J.; Loi, M.A.; Liu, L. M.; Zhao, N. Phenylalkylamine Passivation of OrganoleadHalide Perovskites Enabling High-Efficiency and Air-Stable Photo-voltaic Cells. Adv. Mater. 2016, 28, 9986.(25) Li, N.; Zhu, Z.; Chueh, C.-C.; Liu, H.; Peng, B.; Petrone, A.; Li,X.; Wang, L.; Jen, A. K.-Y. Mixed Cation FAxPEA1−xPbI3 withEnhanced Phase and Ambient Stability toward High-PerformancePerovskite Solar Cells. Adv. Energy Mater. 2017, 7, 1601307.(26) Jung, M.; Shin, T. J.; Seo, J.; Kim, G.; Seok, S. I. Structuralfeatures and their functions in surfactant-armoured methylammoniumlead iodide perovskites for highly efficient and stable solar cells. EnergyEnviron. Sci. 2018, 11, 2188.(27) Zhang, X.; Ren, X.; Liu, B.; Munir, R.; Zhu, X.; Yang, D.; Li, J.;Liu, Y.; Smilgies, D.-M.; Li, R.; Yang, Z.; Niu, T.; Wang, X.; Amassian,A.; Zhao, K.; Liu, S. Stable high efficiency two-dimensional perovskitesolar cells via cesium doping. Energy Environ. Sci. 2017, 10, 2095.(28) Chen, Y.; Sun, Y.; Peng, J.; Zhang, W.; Su, X.; Zheng, K.;Pullerits, T.; Liang, Z. Tailoring Organic Cation of 2D Air-StableOrganometal Halide Perovskites for Highly Efficient Planar SolarCells. Adv. Energy Mater. 2017, 7, 1700162.(29) Slavney, A. H.; Smaha, R. W.; Smith, I. C.; Jaffe, A.; Umeyama,D.; Karunadasa, H. I. Chemical Approaches to Addressing theInstability and Toxicity of Lead−Halide Perovskite Absorbers. Inorg.Chem. 2017, 56, 46.(30) Passarelli, J. V.; Fairfield, D. J.; Sather, N. A.; Hendricks, M. P.;Sai, H.; Stern, C. L.; Stupp, S. I. Enhanced Out-of-Plane Conductivityand Photovoltaic Performance in n = 1 Layered Perovskites throughOrganic Cation Design. J. Am. Chem. Soc. 2018, 140, 7313.(31) Etgar, L. The merit of perovskite’s dimensionality; can thisreplace the 3D halide perovskite? Energy Environ. Sci. 2018, 11, 234.(32) Stoumpos, C. C.; Cao, D. H.; Clark, D. J.; Young, J.; Rondinelli,J. M.; Jang, J. I.; Hupp, J. T.; Kanatzidis, M. G. Ruddlesden−PopperHybrid Lead Iodide Perovskite 2D Homologous Semiconductors.Chem. Mater. 2016, 28, 2852.(33) Leng, K.; Abdelwahab, I.; Verzhbitskiy, I.; Telychko, M.; Chu,L.; Fu, W.; Chi, X.; Guo, N.; Chen, Z.; Chen, Z.; Zhang, C.; Xu, Q.-H.; Lu, J.; Chhowalla, M.; Eda, G.; Loh, K. P. Molecularly thin two-dimensional hybrid perovskites with tunable optoelectronic propertiesdue to reversible surface relaxation. Nat. Mater. 2018, 17, 908.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

N

(34) Quintero-Bermudez, R.; Gold-Parker, A.; Proppe, A. H.; Munir,R.; Yang, Z.; Kelley, S. O.; Amassian, A.; Toney, M. F.; Sargent, E. H.Compositional and orientational control in metal halide perovskites ofreduced dimensionality. Nat. Mater. 2018, 17, 900.(35) Tsai, H.; Nie, W.; Blancon, J.-C.; Stoumpos, C. C.; Asadpour,R.; Harutyunyan, B.; Neukirch, A. J.; Verduzco, R.; Crochet, J. J.;Tretiak, S.; Pedesseau, L.; Even, J.; Alam, M. A.; Gupta, G.; Lou, J.;Ajayan, P. M.; Bedzyk, M. J.; Kanatzidis, M. G.; Mohite, A. D. High-efficiency two-dimensional Ruddlesden−Popper perovskite solar cells.Nature 2016, 536, 312.(36) Noh, J. H.; Im, S. H.; Heo, J. H.; Mandal, T. N.; Seok, S. I.Chemical Management for Colorful, Efficient, and Stable Inorganic−Organic Hybrid Nanostructured Solar Cells. Nano Lett. 2013, 13,1764.(37) Chu, Z.; Yang, M.; Schulz, P.; Wu, D.; Ma, X.; Seifert, E.; Sun,L.; Li, X.; Zhu, K.; Lai, K. Impact of grain boundaries on efficiencyand stability of organic-inorganic trihalide perovskites. Nat. Commun.2017, 8, 2230.(38) Zhang, Y.; Li, F.; Jiang, K.-J.; Huang, J.-H.; Wang, H.; Fan, H.;Wang, P.; Liu, C.-M.; Zhang, L.-P.; Song, Y. From 2D to 3D: a facileand effective procedure for fabrication of planar CH3NH3PbI3perovskite solar cells. J. Mater. Chem. A 2018, 6, 17867.(39) Billing, D. G.; Lemmerer, A. Synthesis, characterization andphase transitions in the inorganic-organic layered perovskite-typehybrids [(CnH2n+1NH3)2PbI4], n = 4, 5 and 6. Acta Crystallogr., Sect.B: Struct. Sci. 2007, 63, 735.(40) Shiguo, H.; Peng, W.; Jing, Z.; Xitao, L.; Zhihua, S.; Xiaoying,H.; Lina, L.; Chengmin, J.; Weichuan, Z.; Bing, T.; Weida, H.;Maochun, H.; Junhua, L. Exploring a Polar Two-dimensional Multi-layered Hybrid Perovskite of (C5H11NH3)2(CH3NH3)Pb2I7 forUltrafast-Responding Photodetection. Laser Photonics Rev. 2018, 12,1800060.(41) Chen, H.; Xu, X. Ruddlesden-Popper compounds in thedouble-perovskite family Sr2FeTaO6(SrO)n (n = 0, 1 and 2) and theirphotocatalytic properties. Appl. Catal., B 2017, 206, 35.(42) Stoumpos, C. C.; Malliakas, C. D.; Kanatzidis, M. G.Semiconducting Tin and Lead Iodide Perovskites with OrganicCations: Phase Transitions, High Mobilities, and Near-InfraredPhotoluminescent Properties. Inorg. Chem. 2013, 52, 9019.(43) Mantina, M.; Chamberlin, A. C.; Valero, R.; Cramer, C. J.;Truhlar, D. G. Consistent van der Waals Radii for the Whole MainGroup. J. Phys. Chem. A 2009, 113, 5806.(44) Mao, L.; Ke, W.; Pedesseau, L.; Wu, Y.; Katan, C.; Even, J.;Wasielewski, M. R.; Stoumpos, C. C.; Kanatzidis, M. G. HybridDion−Jacobson 2D Lead Iodide Perovskites. J. Am. Chem. Soc. 2018,140, 3775.(45) Lemmerer, A.; Billing, D. G. Synthesis, characterization andphase transitions of the inorganic-organic layered perovskite-typehybrids [(CnH2n+1NH3)2PbI4], n = 7, 8, 9 and 10. Dalton Trans 2012,41, 1146.(46) Billing, D. G.; Lemmerer, A. Synthesis, characterization andphase transitions of the inorganic-organic layered perovskite-typehybrids [(CnH2n+1NH3)2PbI4] (n = 12, 14, 16 and 18). New J. Chem.2008, 32, 1736.(47) Whitfield, P. S.; Herron, N.; Guise, W. E.; Page, K.; Cheng, Y.Q.; Milas, I.; Crawford, M. K. Structures, Phase Transitions andTricritical Behavior of the Hybrid Perovskite Methyl AmmoniumLead Iodide. Sci. Rep. 2016, 6, 35685.(48) Weber, O. J.; Ghosh, D.; Gaines, S.; Henry, P. F.; Walker, A. B.;Islam, M. S.; Weller, M. T. Phase Behavior and Polymorphism ofFormamidinium Lead Iodide. Chem. Mater. 2018, 30, 3768.(49) Yamada, K.; Kuranaga, Y.; Ueda, K.; Goto, S.; Okuda, T.;Furukawa, Y. Phase Transition and Electric Conductivity of ASnCl3(A = Cs and CH3NH3). Bull. Chem. Soc. Jpn. 1998, 71, 127.(50) Wang, K.-H.; Li, L.-C.; Shellaiah, M.; Wen Sun, K. Structuraland Photophysical Properties of Methylammonium Lead Tribromide(MAPbBr3) Single Crystals. Sci. Rep. 2017, 7, 13643.(51) Gelvez-Rueda, M. C.; Hutter, E. M.; Cao, D. H.; Renaud, N.;Stoumpos, C. C.; Hupp, J. T.; Savenije, T. J.; Kanatzidis, M. G.;

Grozema, F. C. Interconversion between Free Charges and BoundExcitons in 2D Hybrid Lead Halide Perovskites. J. Phys. Chem. C2017, 121, 26566.(52) Barman, S.; Venkataraman, N. V.; Vasudevan, S.; Seshadri, R.Phase Transitions in the Anchored Organic Bilayers of Long-ChainAlkylammonium Lead Iodides (CnH2n+1NH3)2PbI4; n = 12, 16, 18. J.Phys. Chem. B 2003, 107, 1875.(53) Liu, G.; Gong, J.; Kong, L.; Schaller, R. D.; Hu, Q.; Liu, Z.; Yan,S.; Yang, W.; Stoumpos, C. C.; Kanatzidis, M. G.; Mao, H.-k.; Xu, T.Isothermal pressure-derived metastable states in 2D hybrid perov-skites showing enduring bandgap narrowing. Proc. Natl. Acad. Sci. U.S. A. 2018, 115, 8076.(54) Spanopoulos, I.; Ke, W.; Stoumpos, C. C.; Schueller, E. C.;Kontsevoi, O. Y.; Seshadri, R.; Kanatzidis, M. G. Unraveling theChemical Nature of the 3D “Hollow” Hybrid Halide Perovskites. J.Am. Chem. Soc. 2018, 140, 5728.(55) Ishihara, T.; Hong, X.; Ding, J.; Nurmikko, A. V. Dielectricconfinement effect for exciton and biexciton states in PbI4-based two-dimensional semiconductor structures. Surf. Sci. 1992, 267, 323.(56) Kumagai, M.; Takagahara, T. Excitonic and nonlinear-opticalproperties of dielectric quantum-well structures. Phys. Rev. B: Condens.Matter Mater. Phys. 1989, 40, 12359.(57) Pedesseau, L.; Sapori, D.; Traore, B.; Robles, R.; Fang, H.-H.;Loi, M. A.; Tsai, H.; Nie, W.; Blancon, J.-C.; Neukirch, A.; Tretiak, S.;Mohite, A. D.; Katan, C.; Even, J.; Kepenekian, M. Advances andPromises of Layered Halide Hybrid Perovskite Semiconductors. ACSNano 2016, 10, 9776.(58) Tsai, H.; Nie, W.; Blancon, J.-C.; Stoumpos, C. C.; Soe, C. M.M.; Yoo, J.; Crochet, J.; Tretiak, S.; Even, J.; Sadhanala, A.; Azzellino,G.; Brenes, R.; Ajayan, P. M.; Bulovic, V.; Stranks, S. D.; Friend, R.H.; Kanatzidis, M. G.; Mohite, A. D. Stable Light-Emitting DiodesUsing Phase-Pure Ruddlesden−Popper Layered Perovskites. Adv.Mater. 2018, 30, 1704217.(59) Blancon, J. C.; Stier, A. V.; Tsai, H.; Nie, W.; Stoumpos, C. C.;Traore, B.; Pedesseau, L.; Kepenekian, M.; Katsutani, F.; Noe, G. T.;Kono, J.; Tretiak, S.; Crooker, S. A.; Katan, C.; Kanatzidis, M. G.;Crochet, J. J.; Even, J.; Mohite, A. D. Scaling law for excitons in 2Dperovskite quantum wells. Nat. Commun. 2018, 9, 2254.(60) Soe, C. M. M.; Stoumpos, C. C.; Kepenekian, M.; Traore, B.;Tsai, H.; Nie, W.; Wang, B.; Katan, C.; Seshadri, R.; Mohite, A. D.;Even, J.; Marks, T. J.; Kanatzidis, M. G. New Type of 2D Perovskiteswith Alternating Cations in the Interlayer Space, (C(NH2)3)-(CH3NH3)nPbnI3n+1: Structure, Properties, and Photovoltaic Perform-ance. J. Am. Chem. Soc. 2017, 139, 16297.(61) Booker, E. P.; Thomas, T. H.; Quarti, C.; Stanton, M. R.;Dashwood, C. D.; Gillett, A. J.; Richter, J. M.; Pearson, A. J.; Davis, N.J. L. K.; Sirringhaus, H.; Price, M. B.; Greenham, N. C.; Beljonne, D.;Dutton, S. E.; Deschler, F. Formation of Long-Lived Color Centersfor Broadband Visible Light Emission in Low-Dimensional LayeredPerovskites. J. Am. Chem. Soc. 2017, 139, 18632.(62) Quarti, C.; Marchal, N.; Beljonne, D. Tuning the Optoelec-tronic Properties of Two-Dimensional Hybrid Perovskite Semi-conductors with Alkyl Chain Spacers. J. Phys. Chem. Lett. 2018, 9,3416.(63) Blancon, J.-C.; Tsai, H.; Nie, W.; Stoumpos, C. C.; Pedesseau,L.; Katan, C.; Kepenekian, M.; Soe, C. M. M.; Appavoo, K.; Sfeir, M.Y.; Tretiak, S.; Ajayan, P. M.; Kanatzidis, M. G.; Even, J.; Crochet, J.J.; Mohite, A. D. Extremely efficient internal exciton dissociationthrough edge states in layered 2D perovskites. Science 2017, 355,No. 1288.(64) Peng, W.; Yin, J.; Ho, K.-T.; Ouellette, O.; De Bastiani, M.;Murali, B.; El Tall, O.; Shen, C.; Miao, X.; Pan, J.; Alarousu, E.; He, J.-H.; Ooi, B. S.; Mohammed, O. F.; Sargent, E.; Bakr, O. M. UltralowSelf-Doping in Two-dimensional Hybrid Perovskite Single Crystals.Nano Lett. 2017, 17, 4759.(65) Zhang, J.; Chen, R.; Wu, Y.; Shang, M.; Zeng, Z.; Zhang, Y.;Zhu, Y.; Han, L. Extrinsic Movable Ions in MAPbI3 Modulate EnergyBand Alignment in Perovskite Solar Cells. Adv. Energy Mater. 2018, 8,1701981.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

O

(66) Guerrero, A.; Bou, A.; Matt, G.; Almora, O.; Heumuller, T.;Garcia-Belmonte, G.; Bisquert, J.; Hou, Y.; Brabec, C. Switching OffHysteresis in Perovskite Solar Cells by Fine-Tuning Energy Levels ofExtraction Layers. Adv. Energy Mater. 2018, 8, 1703376.(67) Harwell, J. R.; Baikie, T. K.; Baikie, I. D.; Payne, J. L.; Ni, C.;Irvine, J. T. S.; Turnbull, G. A.; Samuel, I. D. W. Probing the energylevels of perovskite solar cells via Kelvin probe and UV ambientpressure photoemission spectroscopy. Phys. Chem. Chem. Phys. 2016,18, 19738.(68) Jung, H. S.; Park, N. G. Perovskite Solar Cells: From Materialsto Devices. Small 2015, 11, 10.(69) Ryu, S.; Noh, J. H.; Jeon, N. J.; Chan Kim, Y.; Yang, W. S.; Seo,J.; Seok, S. I. Voltage output of efficient perovskite solar cells withhigh open-circuit voltage and fill factor. Energy Environ. Sci. 2014, 7,2614.(70) Rolston, N.; Printz, A. D.; Tracy, J. M.; Weerasinghe, H. C.;Vak, D.; Haur, L. J.; Priyadarshi, A.; Mathews, N.; Slotcavage, D. J.;McGehee, M. D.; Kalan, R. E.; Zielinski, K.; Grimm, R. L.; Tsai, H.;Nie, W.; Mohite, A. D.; Gholipour, S.; Saliba, M.; Gratzel, M.;Dauskardt, R. H. Effect of Cation Composition on the MechanicalStability of Perovskite Solar Cells. Adv. Energy Mater. 2018, 8,1702116.(71) Wang, C.; Wang, C.; Huang, Z.; Xu, S. Materials and Structurestoward Soft Electronics. Adv. Mater. 2018, 30, 1801368.(72) Tu, Q.; Spanopoulos, I.; Yasaei, P.; Stoumpos, C. C.;Kanatzidis, M. G.; Shekhawat, G. S.; Dravid, V. P. Stretching andBreaking of Ultrathin 2D Hybrid Organic−Inorganic Perovskites.ACS Nano 2018, 12, 10347.(73) Tu, Q.; Spanopoulos, I.; Hao, S.; Wolverton, C.; Kanatzidis, M.G.; Shekhawat, G. S.; Dravid, V. P. Out-of-Plane MechanicalProperties of 2D Hybrid Organic−Inorganic Perovskites by Nano-indentation. ACS Appl. Mater. Interfaces 2018, 10, 22167.(74) Ferreira, A. C.; Letoublon, A.; Paofai, S.; Raymond, S.; Ecolivet,C.; Ruffle, B.; Cordier, S.; Katan, C.; Saidaminov, M. I.; Zhumekenov,A. A.; Bakr, O. M.; Even, J.; Bourges, P. Elastic Softness of HybridLead Halide Perovskites. Phys. Rev. Lett. 2018, 121, 085502.(75) Spina, M.; Karimi, A.; Andreoni, W.; Pignedoli, C. A.; Nafradi,B.; Forro, L.; Horvath, E. Mechanical signatures of degradation of thephotovoltaic perovskite CH3NH3PbI3 upon water vapor exposure.Appl. Phys. Lett. 2017, 110, 121903.(76) Yu, J.; Wang, M.; Lin, S. Probing the Soft and NanoductileMechanical Nature of Single and Polycrystalline Organic−InorganicHybrid Perovskites for Flexible Functional Devices. ACS Nano 2016,10, 11044.(77) Jee, A.-Y.; Lee, M. Comparative analysis on the nano-indentation of polymers using atomic force microscopy. Polym. Test.2010, 29, 95.(78) Park, M.; Kim, H. J.; Jeong, I.; Lee, J.; Lee, H.; Son, H. J.; Kim,D.-E.; Ko, M. J. Mechanically Recoverable and Highly EfficientPerovskite Solar Cells: Investigation of Intrinsic Flexibility ofOrganic−Inorganic Perovskite. Adv. Energy Mater. 2015, 5, 1501406.(79) Chen, C. P.; Leipold, M. H. Fracture toughness of silicon. Am.Ceram. Soc. Bull. 1980, 59, 469.(80) Hjort, K.; Soderkvist, J.; Schweitz, J. A. Gallium arsenide as amechanical material. J. Micromech. Microeng. 1994, 4, 1.(81) Liu, G.; Zheng, H.; Xu, X.; Zhu, L.-Z.; Zhang, X.; Pan, X.Design of High-Efficiency and Environmentally Stable Mixed-Dimensional Perovskite Solar Cells Based on Cesium-FormamidiniumLead Halide Component. Chem. Mater. 2018, 30, 7691.(82) Hsu, H. L.; Hsiao, H. T.; Juang, T. Y.; Jiang, B. H.; Chen, S. C.;Jeng, R. J.; Chen, C. P. Carbon Nanodot Additives Realize High-Performance Air-Stable p−i−n Perovskite Solar Cells ProvidingEfficiencies of up to 20.2%. Adv. Energy Mater. 2018, 8, 1802323.(83) Soe, C. M. M.; Nie, W.; Stoumpos, C. C.; Tsai, H.; Blancon, J.-C.; Liu, F.; Even, J.; Marks, T. J.; Mohite, A. D.; Kanatzidis, M. G.Understanding Film Formation Morphology and Orientation in HighMember 2D Ruddlesden−Popper Perovskites for High-EfficiencySolar Cells. Adv. Energy Mater. 2018, 8, 1700979.

(84) Zolotoyabko, E. Determination of the degree of preferredorientation within the March-Dollase approach. J. Appl. Crystallogr.2009, 42, 513.(85) Ke, W.; Stoumpos, C. C.; Spanopoulos, I.; Chen, M.;Wasielewski, M. R.; Kanatzidis, M. G. Diammonium Cations in theFASnI3 Perovskite Structure Lead to Lower Dark Currents and MoreEfficient Solar Cells. ACS Energy Lett. 2018, 3, 1470.(86) Xiong, H.; Rui, Y.; Li, Y.; Zhang, Q.; Wang, H. Hydrophobiccoating over a CH3NH3PbI3 absorbing layer towards air stableperovskite solar cells. J. Mater. Chem. C 2016, 4, 6848.(87) 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.(88) Reese, M. O.; Gevorgyan, S. A.; Jørgensen, M.; Bundgaard, E.;Kurtz, S. R.; Ginley, D. S.; Olson, D. C.; Lloyd, M. T.; Morvillo, P.;Katz, E. A.; Elschner, A.; Haillant, O.; Currier, T. R.; Shrotriya, V.;Hermenau, M.; Riede, M.; R. Kirov, K.; Trimmel, G.; Rath, T.;Inganas, O.; Zhang, F.; Andersson, M.; Tvingstedt, K.; Lira-Cantu,M.; Laird, D.; McGuiness, C.; Gowrisanker, S.; Pannone, M.; Xiao,M.; Hauch, J.; Steim, R.; DeLongchamp, D. M.; Rosch, R.; Hoppe,H.; Espinosa, N.; Urbina, A.; Yaman-Uzunoglu, G.; Bonekamp, J.-B.;van Breemen, A. J. J. M.; Girotto, C.; Voroshazi, E.; Krebs, F. C.Consensus stability testing protocols for organic photovoltaicmaterials and devices. Sol. Energy Mater. Sol. Cells 2011, 95, 1253.(89) Zhu, J.; Gottschalg, R.; Koehl, M.; Hoffmann, S.; Berger, K. A.;Zamini, S.; Bennett, I.; Gerritsen, E.; Malbranche, P.; Sicot, L.;Pugliatti, P.; Stefano, A. D.; Aleo, F.; Paletta, F.; Roca, F.; Graditi, G.;Pellegrino, M.; Zubillaga, O.; Cano, P.; Tselepis, S.; Pozza, A.;Sample, T. Towards Service Lifetime Prediction of Photovoltaic Modules,Towards Service Lifetime Prediction of Photovoltaic Modules; Photo-voltaic European Research Infrastructure (SOPHIA), 2014; https://www.nrel.gov/pv/assets/pdfs/2014_pvmrw_59_zhu.pdf.(90) Brunetti, B.; Cavallo, C.; Ciccioli, A.; Gigli, G.; Latini, A. Onthe Thermal and Thermodynamic (In)Stability of MethylammoniumLead Halide Perovskites. Sci. Rep. 2016, 6, 31896.(91) Conings, B.; Drijkoningen, J.; Gauquelin, N.; Babayigit, A.;D’Haen, J.; D’Olieslaeger, L.; Ethirajan, A.; Verbeeck, J.; Manca, J.;Mosconi, E.; Angelis, F. D.; Boyen, H. G. Intrinsic Thermal Instabilityof Methylammonium Lead Trihalide Perovskite. Adv. Energy Mater.2015, 5, 1500477.(92) Hao, F.; Stoumpos, C. C.; Liu, Z.; Chang, R. P. H.; Kanatzidis,M. G. Controllable Perovskite Crystallization at a Gas−Solid Interfacefor Hole Conductor-Free Solar Cells with Steady Power ConversionEfficiency over 10%. J. Am. Chem. Soc. 2014, 136, 16411.(93) Leguy, A. M. A.; Hu, Y.; Campoy-Quiles, M.; Alonso, M. I.;Weber, O. J.; Azarhoosh, P.; van Schilfgaarde, M.; Weller, M. T.; Bein,T.; Nelson, J.; Docampo, P.; Barnes, P. R. F. Reversible Hydration ofCH3NH3PbI3 in Films, Single Crystals, and Solar Cells. Chem. Mater.2015, 27, 3397.(94) Lee, J.-W.; Hsieh, Y.-T.; De Marco, N.; Bae, S.-H.; Han, Q.;Yang, Y. Halide Perovskites for Tandem Solar Cells. J. Phys. Chem.Lett. 2017, 8, 1999.(95) Liu, F.; Dong, Q.; Wong, M. K.; Djurisic, A. B.; Ng, A.; Ren, Z.;Shen, Q.; Surya, C.; Chan, W. K.; Wang, J.; Ng, A. M. C.; Liao, C.; Li,H.; Shih, K.; Wei, C.; Su, H.; Dai, J. Is Excess PbI2 Beneficial forPerovskite Solar Cell Performance? Adv. Energy Mater. 2016, 6,1502206.(96) Han, Y.; Meyer, S.; Dkhissi, Y.; Weber, K.; Pringle, J. M.; Bach,U.; Spiccia, L.; Cheng, Y.-B. Degradation observations of encapsulatedplanar CH3NH3PbI3 perovskite solar cells at high temperatures andhumidity. J. Mater. Chem. A 2015, 3, 8139.(97) Dong, Q.; Liu, F.; Wong, M. K.; Tam, H. W.; Djurisic, A. B.;Ng, A.; Surya, C.; Chan, W. K.; Ng, A. M. C. Encapsulation ofPerovskite Solar Cells for High Humidity Conditions. ChemSusChem2016, 9, 2597.(98) Cheacharoen, R.; Boyd, C. C.; Burkhard, G. F.; Leijtens, T.;Raiford, J. A.; Bush, K. A.; Bent, S. F.; McGehee, M. D. Encapsulating

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

P

perovskite solar cells to withstand damp heat and thermal cycling.Sustainable Energy Fuels 2018, 2, 2398.

Journal of the American Chemical Society Article

DOI: 10.1021/jacs.9b01327J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Q