Tailoring the Interfacial Chemical Interaction for High-EfficiencyPerovskite Solar CellsLijian Zuo,†,‡ Qi Chen,†,‡ Nicholas De Marco,†,‡ Yao-Tsung Hsieh,†,‡ Huajun Chen,† Pengyu Sun,†

Sheng-Yung Chang,† Hongxiang Zhao,†,‡ Shiqi Dong,† and Yang Yang*,†,‡

†Department of Materials Science and Engineering and ‡California NanoSystems Institute, University of California, Los Angeles,California 90095, United States

*S Supporting Information

ABSTRACT: The ionic nature of perovskite photovoltaicmaterials makes it easy to form various chemical interactionswith different functional groups. Here, we demonstrate thatinterfacial chemical interactions are a critical factor indetermining the optoelectronic properties of perovskite solarcells. By depositing different self-assembled monolayers(SAMs), we introduce different functional groups onto theSnO2 surface to form various chemical interactions with theperovskite layer. It is observed that the perovskite solar celldevice performance shows an opposite trend to that of the energy level alignment theory, which shows that chemical interactionsare the predominant factor governing the interfacial optoelectronic properties. Further analysis verifies that proper interfacialinteractions can significantly reduce trap state density and facilitate the interfacial charge transfer. Through use of the 4-pyridinecarboxylic acid SAM, the resulting perovskite solar cell exhibits striking improvements to the reach the highest efficiencyof 18.8%, which constitutes an ∼10% enhancement compared to those without SAMs. Our work highlights the importance ofchemical interactions at perovskite/electrode interfaces and paves the way for further optimizing performances of perovskite solarcells.

KEYWORDS: Perovskite solar cells, interfacial chemical interactions, trap states passivation, coordination bond,self-assembly monolayer, energy level alignment

The advent of perovskite solar cells has rapidly jumped intothe hot seat of academic research and industrial interest

due to the recent surges in device performances and thepotential low-cost solution-based manufacturing.1−6 The per-ovskite structure of the active layers can be described as ABX3,where the A is a cation, e.g. methylammonium (MA+),formamidinium (FA+), Cs+, etc., B is a metal cation, e.g.,Pb2+, Sn2+, etc., and the X is an anion, e.g., I−, Br−, Cl−, SCN−,etc.2,4,7 One of the most striking features distinguishing theperovskite solar cells from other PV materials is its mixedionic−covalent bonding peroskite structures, which enables thecombined advantages of both solution processability andextraordinary optoelectronic properties, such as high dielectricconstant,8,9 large absorption coefficient,10 long carrier life-time,11 high tolerance for trap states,12 and more.The device architecture of perovskite solar cells is typically a

layer-by-layer configuration with a perovskite layer sandwichedbetween electrodes.13−16 Interfacial layers are applied betweenthe perovskite and electrode layers in order to facilitate chargecarrier transport and collection and suppress interfacial chargerecombination.13,14,17 The theory of energy level alignment18,19

was established as the guiding rule for the design of interfaciallayers in solar cells.20,21 Electron transporting layers (ETLs)with low work functions are favored to align with theconduction band of perovskite (∼3.75 eV for MAPbIxCl3−x)

for efficient electron charge carrier extraction22 and likewise forhole transporting layers (HTLs) at the anode (valence band∼5.31 eV for MAPbIxCl3−x).

13 A variety of efficient ETLs havebeen developed in recent years, including organic materials(e.g., [6,6]-phenyl-C61-butyric acid methyl ester (PC61BM),23

poly [(9,9-bis(3′-(N,N-dimethylamino)propyl)-2,7-fluorene)-alt-2,7-(9,9-dioctylfluorene)] (PFN),24 polyethylenimineethoxylated (PEIE),13 etc.25), metal oxides (TiO2,

13 SnO2,26

ZnO,27 etc.), and more. Strong electron coupling is observedbetween these ETLs and the perovskite layer, which enableefficient electron carrier extraction and high device perform-ance.13 Aside from the energy level alignment, it should benoted that different chemical interactions between theperovskite film and the interfacial layers can form due to theionic nature of the perovskite.28,29 The formation of differentsurface chemical interactions or the different chemicalterminations can dramatically change the electron density andenergy level structures at the surface of the perovskite crystals,as studied in previous works.30−32 Therefore, chemicallytailoring the interactions at the perovskite/electrode interfaces

Received: September 26, 2016Revised: December 2, 2016Published: December 12, 2016

Letter

pubs.acs.org/NanoLett

© XXXX American Chemical Society A DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

should also play a critical role in the interfacial optoelectronicproperties but unfortunately has been rarely explored.The self-assembled monolayer (SAM) has been demon-

strated as an effective strategy for fine-tuning the interfacialoptoelectronic properties by bonding the perovskite layer to theelectrode substrates.27,29,33−35 In using SAMs, the substratework function can be adjusted due to the permanent dipolemoment resulting from the ordered molecular alignment, whichhas been widely studied in organic electronics.36 Mostimportantly, the deposition of SAMs on substrates offersdiverse linkages between the substrates and the ionic perovskitefilms. Although a variety of SAMs have been developed toenhance device performances of perovskite solar cells, theeffects of chemical interactions between perovskite films andthe substrates were neglected, as limited by the difficulty indistinguishing the role of chemical interactions from othercontributing factors, e.g., energy level alignment and surfaceinduced morphological variations.27,37

In this work, we systematically studied the independent roleof interfacial chemical interactions for highly efficient perovskitesolar cells. The SAMs with different terminal groups wereanchored onto a SnO2 ETL to form different chemicalinteractions with the perovskite film. The chemical structuresand schematic diagram for the various interactions are shown inthe Figure 1a. van der Waals interactions with benzoic acid(BA) and dipolar interactions with the 4-pyridinecarboxylicacid (PA), 3-aminopropanoic acid (C3), 4-aminobenzoic acid(ABA), and 4-cyanobenzoic acid (CBA) were explored. In

contrast to the energy level alignment theory, an opposite trendwas observed in device performance of perovskite solar cellsusing different SAMs, which showed that interfacial chemicalinteractions play a dominant role in determining the interfacialoptoelectronic properties and device performance (devicestructure shown in Figure 1a and b). The best deviceperformances were obtained with the deposition of PA, andthe highest device performance was able to reach 18.8% for asingle scan and a steady state power output of 17.7%,composing an approximate 10% improvement compared todevices without SAMs. Further analysis showed that theenhanced device performances was a result of proper chemicalinteractions that reduced trap state density and facilitatedinterfacial charge transfer. Our results highlight the importanceof chemical interactions for achieving highly efficient perovskitesolar cells, which should be taken into consideration for futuredesigns.

Results and Discussion. The deposition of different SAMson the SnO2 surfaces was confirmed by the existence of Nelements as shown in the X-ray photoelectron spectroscopy(XPS) in Figure S1. Compared to the bare SnO2 and BA-SAMmodified SnO2 surfaces, obvious N 1s signals appeared at abinding energy of ∼399.5 eV with PA, CBA, ABA, and C3modification. The deposition of different SAMs formeddifferent functional groups on the SnO2 surfaces. It has beenreported that substrates properties can significantly influencethe morphologies of perovskite films processed from the one-step solution method due to the surface induction effect.27,37 Inthe one-step solution process, the precursors, e.g., MAI andPbI2, are evenly mixed and casted onto the substrates.Crystallization is controlled by the nucleation of the perovskitefilms on the substrate, which can be affected by the substrateproperties. Since the film morphology has been demonstratedto significantly influence the device performances of perovskitesolar cells, it is necessary to exclude the influences ofmorphological surface variations in order to judiciously evaluatethe role of chemical interactions on device performance. Here,the two-step solution process38,39 was adopted for theperovskite film formation to avoid this effect. In a two-stepsolution process, the PbI2 and MAI precursors are castedsequentially, and the crystallization is governed by theinterdiffusion of PbI2 and MAI. An SEM image of theperovskite film on the pristine SnO2 substrate is shown inFigure 1c. The perovskite film exhibits grain sizes of severalhundreds of nanometers, which are closely packed to form a flatfilms without any pin-holes.26 Figure S2 shows the surfacemorphology of the polycrystalline perovskite films on differentSAM-modified SnO2 substrates. The morphology of theseperovskite films on different SAM-modified SnO2 substrates areshown to be similar in terms of surface morphology, e.g., crystalgrain sizes and flat pinhole free textures. The X-ray diffraction(XRD) pattern of perovskite on SnO2 is shown in Figure 1d,where the diffraction peaks at 14.1° and 28.6° are attributed tothe <110> and <220> faces of the perovskite structure. Thetiny peak at 12.7° results from a slight excess of PbI2, theexistence of which has been demonstrated to have a positiveeffect for grain boundary passivation.40 Figure S3 shows theXRD patterns of perovskite films on different SAM-modifiedSnO2 substrates. As can be seen, the perovskite films exhibitsimilar XRD peak positions and intensities. These resultsindicate the similar crystalline properties of perovskite filmregardless of different substrate properties. It is thus safe toconclude that the perovskite film morphology shows little

Figure 1. (a) Schematic diagram of the PVSKSC device structure(left) and the SAM between the SnO2 and perovskite film. (BA isbenzoic acid, PA is 4-pyridine carboxylic acid, CBA is 4-cyanobenzoicacid, ABA is 4-aminobenzoic acid, and C3 is 3-propanoic acid). (b)SEM cross-sectional image of the perovskite solar cell. (c) SEM imageof perovskite films on SnO2 substrate. (d) XRD pattern of perovskitefilm on SnO2 substrate. The perovskite films on different SAMsmodified SnO2 substrates have a similar surface morphology (fromSEM images) and XRD patterns (Figures S1 and S2).

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

B

dependence on substrate surface properties with different SAMmodification. As a result, the changes in device performancescan be independently attributed to the variations in interfacialoptoelectronic properties of the perovskite solar cells.The work function of the interfacial layer is critical for

achieving optimal electrical contact at interfaces of perovskitesolar cells.21 To distinguish the influence of work functionvariation from that of interfacial chemical interactions, weexamined the work function changes of the SnO2 ETL withmodification of different SAMs via the ultraviolet photoelectronspectroscopy (UPS) measurement.41 As shown in Figure 2a,the work function of bare SnO2 is ∼4.21 eV. After surfacemodification with PA, BA, CBA, ABA, and C3, the surface workfunction can be tuned to 4.17, 4.16, 4.29, 4.05, and 4.07 eV,respectively. The trend of work function variation is inaccordance with the direction of surface dipole moment.According to the energy level alignment theory, a favorablephotocarrier collection at the cathode is expected with a lowerwork function of the ETL. Therefore, the SAM-modified SnO2substrate is expected to be more suitable as an ELT in thesequence of CBA, none, PA, BA, C3, and ABA.On the SAM-modified SnO2/ITO glass substrates, we

fabricated perovskite solar cells with the device architectureshown in Figure 1a and b. In this structure, the SnO2

26 and2,2′,7,7′-tetrakis(N,N-di-4-methoxyphenylamino)-9,9′-spirobi-fluorene (Spiro-OMeTAD)42 are applied as an ETL and HTL,respectively. Figure 2b depicts the I−V characteristic curves ofperovskite solar cells with or without different SAM

modifications. The corresponding device parameters aresummarized in Table 1. As shown, the control PVSKSC witha bare SnO2 ETL shows a high device performance of 17.19%,with a VOC of 1.06 V, JSC of 21.65 mA/cm2, and FF of 0.749,which is similar to previous reports.26 However, the PVSKSCswith the different SAM modifications show an opposite trendcompared to that observed in the work function variation(Figure 2a). For example, the PVSKSCs with a CBAmodification show a higher PCE of 18.27% compared to thatof the control device, regardless of the increased work function.On the other hand, PVSKSCs with ABA show an obvious dropin device performance (16.50%) compared to that without anySAM, despite our expectation that the low work function at theinterface should improve the device performance compared tothat of bare SnO2. Also, the PVSKSCs with the C3 show noenhancement in device performance compared to that on bareSnO2. Here, we attribute the abnormal performance variationsto the strong chemical interactions between the terminalfunctional groups of the SAMs and the ionic perovskite layers,where an intimate contact formed between the SAM andperovskite film induces a change of interfacial properties. Forexample, an amino or the cyano group can form bipolar bondsto the Pb2+ ion by sharing the electron pair with the emptyorbitals of Pb2+. Actually, strong interactions between thesefunctional groups, e.g., pyridine, have been shown toimmobilize metal oxide nanoparticles with modified surfaces.43

Haruyama et al.30 and Geng et al.31,32 independently simulatedthe effects of different terminations on the surfaces of

Figure 2. (a) Ultraviolet photoelectron spectroscopy of SnO2, (b) I−V characteristic curves of perovskite solar cells without or with different SAMsmodification, (c) EQE spectra and integrated JSC of champion perovskite solar cells, (d) simulated EQE and IQE spectra of perovskite solar cellswith PA modification. (e) Hysteresis I−V characteristic curves of perovskite solar cells without any SAM and with PA modification. (f) Steady-statephotocurrent and output power of perovskite solar cells with and without PA modification.

Table 1. Device Performance of Organic and Perovskite Solar Cells with Different SAMs Modification

organic solar cells perovskite solar cells

SAMs Ef (eV) JSC (mA/cm2) VOC (V) PCE* (%) FF JSC (mA/cm2) VOC (V) PCEa (%) FF

none 4.21 16.64 0.72 4.96 (4.52 ± 0.21) 0.414 21.65 1.06 17.19 (16.21 ± 0.89) 0.749CBA 4.29 16.82 0.60 4.51 (4.14 ± 0.26) 0.447 21.66 1.08 18.27 (17.24 ± 0.68) 0.781PA 4.17 16.52 0.77 7.16 (6.87 ± 0.19) 0.563 22.03 1.10 18.77 (18.01 ± 0.57) 0.774BA 4.16 16.24 0.77 7.33 (6.95 ± 0.22) 0.586 21.88 1.11 18.11 (17.14 ± 0.64) 0.746C3 4.07 16.85 0.78 7.80 (7.67 ± 0.05) 0.593 21.48 1.08 17.11 (16.30 ± 0.49) 0.738ABA 4.05 16.80 0.80 8.00 (7.84 ± 0.11) 0.595 22.00 1.04 16.50 (15.21 ± 0.89) 0.721

aThe PCE values outside of the parentheses are the best, and those in parentheses are the averaged values with deviations.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

C

perovskite crystals. They concluded that the optoelectronicproperties of the perovskite grain boundaries varied substan-tially with different terminations, e.g., the MA+ rich or the I−

rich grain surface. The interfacial structures between differentterminations of perovskite and TiO2 surfaces were alsosimulated to dramatically influence device performances.32 Inour case, modified SnO2 surfaces have different chemical groupsto interact with the surfaces of the perovskite films. Theoptoelectronic property changes are induced at the surface ofperovskite grains adjacent to the SnO2 substrates. Also, theorigin of reduced work function with SAMs relied on theirpermanent dipole moment, which have different electrondensity distributions. The coordination of SAMs with perov-skite might affect the electron density distribution of SAMs and,thus, the work function. Thereafter, a device variation contraryto work function change of SnO2/SAM is observed.Remarkably, PVSKSCs on PA modified SnO2 substrates

show the most striking improvements compared to the deviceswith bare SnO2 ETL layer, with a best PCE of 18.77%, VOC of1.10 V, JSC of 22.03 mA/cm2, and FF of 0.774. This deviceperformance is among the highest achieved for planarheterojunction perovskite solar cells.13 Notably, a significantpassivation effect was observed through use of pyridineaccording to a previous report.44 As such, a similar passivationeffect can be expected at the interface of SnO2 and perovskite.Figure 2c shows the external quantum efficiency (EQE) of thePVSKSCs. As can be seen, the maximum EQE is ∼94% andrepresents one of the highest in literature,5,45 indicating theefficacy of optimal interfacial chemical interactions for efficientphoton-to-electron conversion. The integrated JSC (21.76 mA/cm2) from the EQE spectra matches well to that of themeasured JSC (22.03 mA/cm2) from I−V measurements.Optical simulations were carried out to calculate the internalquantum efficiency of the perovskite solar cells.46,47 Detailedinformation about the optical simulations can be found in theSupporting Information. As shown in Figure 2d, the simulatedEQE of the perovskite solar cells (or the light absorbingfraction of perovskite within the whole device) shows a similarprofile to that of the measured EQE. The appearance of peaksand valleys in the EQE spectra ranging from 550 to 800 nm canbe attributed to the optical microcavity or the opticalinterference effects as proven by the optical intensitydistribution of the whole device (Figure S4). The IQE isobtained by dividing the experimental EQE with the simulatedEQE, which depicts the absorption fraction of the perovskitelayer in the device. As can be seen, the calculated IQE spectra isover 90% in almost the whole absorption region of theperovskite solar cells. This result verifies the efficacy of thedesigned device architecture with PA modification for highlyefficient internal photon-to-electron conversion.One of the issues with perovskite solar cells is the hysteresis

problem,48,49 where I−V characteristic curves can vary withdifferent scanning directions, e.g., the forward scanning (FW,from negative bias to positive bias) or the reverse scanning (RS,from positive bias to the negative bias). The existence ofhysteresis makes it difficult to accurately evaluate the deviceperformance of perovskite solar cells. Alternatively, the steadyoutput power has been regarded as a more reliable method toprecisely characterize the true device efficiency of perovskitesolar cells.45 The hysteresis phenomenon is also observed inour case, which is shown in Figure 2e. The forward scan ofperovskite solar cells without SAMs shows an efficiency of15.6% and improved to 17.2% with deposition of PA. It is hard

to determine whether the hysteresis issue is alleviated with PA,because both the forward and reverse scan device efficiencieswere improved. By applied a bias voltage near the maximumpower output point, we recorded the variation of thephotocurrent density to measure the steady state output ofthe perovskite solar cells under AM 1.5 G 1 sun intensity. Asshown in Figure 2f, the output curves of perovskite solar cellswithout any surface modification show a steady output powerof 15.8 mW/cm2, with bias of 0.88 V and a photocurrent of∼17.9 mA/cm2, while for the PVSKSCs with PA, the steadyoutput show the steady state output power of 17.7 mA/cm2,with the bias of 0.88 V and the photocurrent of 20.1 mA/cm2.These results clearly verify the improvement of deviceperformance with PA modification and confirm the influenceof interfacial chemical interactions on device performance.We also deposited the BA onto the SnO2 surface in order to

form an aromatic rich surface. No chemical interaction waspresent except for a weak van der Waals force formed betweenthe substrates and perovskite, which is even weaker than thatwith the pristine SnO2. The device performance shows anappreciable improvement with a PCE of 18.11%. The improveddevice performance can be attributed to the lowered workfunction according to the energy level alignment theory.Furthermore, we applied the SAMs to organic solar cells,

where the interaction between the active layers and thesubstrates is only van der Waals type. As such, the influences ofchemical interactions can be excluded in organic solar cells. Thestate-of-the-art poly[4,8-bis(5-(2-ethylhexyl)thiophen-2-yl)-benzo[1,2-b;4,5-b′]dithiophene-2,6-diyl-alt-(4-(2-ethylhexyl)-3-fluorothieno[3,4-b]thiophene)-2-carboxylate-2−6-diyl):[6,6]-phenyl-C71-butyric acid methyl ester (PTB7-Th:PC71BM)composite film was selected as the active layer, and MoO3/Ag was adopted as the anode (Figure 3a). The I−V

Figure 3. Organic solar cells with different SAMs modification. (a)Device architecture, (b) I−V characteristics.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

D

characteristic curves of organic solar cells with or withoutdifferent SAM modification are depicted in Figure 3b, and thedevice parameters are summarized in Table 1. As can be seen,the organic solar cells with bare SnO2 as an ETL show poorefficiency of 4.95%, with a low VOC of 0.72 V and a low FF of0.41. With the BA, PA, C3, and ABA modification, the deviceperformance show significant improvement with the highestefficiency up to 8.00%. The VOC is significantly increased to0.77−0.80 V, and the FF also shows appreciable improvementwith BA, PA, C3, and ABAs, owing to the lowered workfunction of SnO2 as revealed by UPS measurements. With theCBA modification, the device PCE and VOC droppedremarkably, which is consistent with the higher work functionof the CBA modified surfaces. The device performancevariations conform to the work function changes with differentSAMs, which abides to the energy level alignment theory. Thesharp contrast in the device performance variations ofperovskite solar cells and organic solar cells with differentSAMs further stresses the importance of chemical interactionson interfacial properties for perovskite solar cells.Device performances are closely correlated to charge

dynamics in perovskite solar cells. Photoluminescence (PL)properties of the perovskite film provides a direct observationof photocarrier dynamics.11,22,40,50 As shown in Figure 4a, thesteady state PL intensity of perovskite on ABA is mostlyannihilated compared to the perovskite films without any SAM.The PL intensity of perovskite films on the C3 or BA showshigher than that on ABA, but still lower compared to thatwithout SAM deposition. Little PL quenching is observed onperovskite films deposited on PA or the CBA compared to thaton pristine SnO2 substrate. Figure 4b shows the transient PL ofperovskite on SnO2 with different SAM modification. Asshown, transient PL shows an obvious biexponential decaybehavior. The fast decay PL species (τ1) can be attributed toquenching via trap states or interfacial charge transfer, and the

longer PL species (τ2) is attributed to the bimolecularrecombination via the light emission.22,40,50 It can be seenthat the perovskite film on the bare SnO2 shows the longestlifetime of 150 ns (τ1), and the PL lifetimes are sequentiallyshortened to 62, 75, 83, 92, and 99 ns with deposition of C3,ABA, BA, PA, and CBA. The PL quenching and shortened PLlifetime with deposition of any SAM can be attributed to theenhanced electrical coupling (or faster charge transfer) betweenthe perovskite and the SnO2 ETL or the increased interfacialtrap states. However, the two PL quenching mechanisms canlead to significantly different carrier fates, where chargerecombination occurs with traps but photocurrent is generatedwith interfacial charge transfer. To distinguish the twomechanisms, photocarrier dynamics in a real device weremonitored via transient photovoltage decay measurements.During the operation of a real device, photogenerated

carriers, either free carriers or excitons, are mostly separated atthe interfaces of Spiro-OMeTAD/perovksite or perovskite/SnO2, resulting in strong PL annihilation by charge transfer atthe interfaces. The separated charges can either be recombinedvia surface trap states or be swept out of device for electricalpower. In this scenario, the technique of transient photovoltage(TPV) can provide a clear picture of photocarriers dynamicsduring device operation.50,51 For the TPV measurement, thedevices were operated at open-circuit voltage condition (a 6MΩ resistance was applied) at a light intensity of 1 sun, and alaser pulse was generated to perturb the device and cause arapid decay of photovoltage. The carrier lifetime is estimated tobe the decay time at which the intensity become 1/e of theinitial intensity. As shown in Figure 4c, the decay of transientphotovoltage exhibits monoexponential decay. The carrierlifetime without SAMs is 4.1 μs. With deposition of BA, PA,and CBAs, the carrier lifetime is significantly improved to 5.4,6.6, and 7.6 μs, respectively. With deposition of C3, the carrierlifetime decreases to 3.1 μs and becomes even shorter (1.5 μs)

Figure 4. Photocarrier dynamics in perovskite films and perovskite devices. (a) Steady state and (b) transient PL spectra of perovskite film on SnO2with different SAMs. (c) Transient photovoltage and of perovskite solar cells with different SAMs. (d) Schematic diagram for the charge dynamics atperovskite/SnO2 interface in the perovskite solar cells: (1) Photoluminescence process, (2) PL quenching via trap states, (3) charge transfer process,(4) charge recombination via trap states, (5) power generation.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

E

with ABA. Since the deposition of SAM at the interface haslittle influence on the bulk properties of perovskite film, weattributed the change in the carrier lifetime to the modificationof the surface or interface properties with SAM layer. Due tothe open-circuit condition, all of the photocarriers either in theperovskite bulk or transferred at the interfaces can only berecombined via trap states. Thus, long carrier lifetimes indicatereduced trap states in real perovskite devices, and it isconcluded that the deposition of CBA, PA, and BA-SAM cansignificantly reduce the surface trap states. Combining thereduced carrier lifetimes from transient PL measurements andthe little changes in the PL intensity of the perovskite films withCBA and PA modification, we can see that the electroniccoupling between the perovskite and SnO2 layers wereenhanced via deposition of these SAMs. In the case of ABAand C3, it seems that the trap state density at the interface wasincreased according to the reduced carrier lifetimes from TPVmeasurements, which also contribute to the reduced PLintensity. These results confirm that the chemical interactionsat the electrode interfaces have a dramatic influence on thephotocarrier dynamics and the interface trap states.A schematic diagram is drawn in Figure 4d to show the

carrier dynamics and the mechanism of enhanced deviceperformance with deposition of PA. As illustrated, thephotogenerated carriers can be relaxed to ground state viathe light emission and trap state recombination. To generatethe photocurrent, the photocarriers need to be collected by theelectrode via charge transfer at the perovskite/electrodeinterfaces. The separated photocarriers need to bypass theinterfacial recombination via the surface trap states in order tocontribute to the output power. With the PA deposition, thesurface trap states were suppressed as evidenced by theenhanced TPV decay time. The charge transfer at theperovskite/SnO2 is enhanced due to the lowered work functionand the improved interaction of perovskite and SnO2 substratewith PA deposition, as evidenced by the shortened PL lifetime.The improved electronic coupling and suppressed interfacialtraps contributed to the improved device performance.Conclusion. In summary, we demonstrated that the

interfacial chemical interactions between perovskite and chargetransporting layers are important for device performance ofperovskite solar cells. The ionic−covalent nature of perovskitefilms makes it easy to form various chemical interactions withdifferent chemical groups, which can significantly influenceinterfacial optoelectronic properties, e.g., the surface trap statedensity. Through delicate control of the interfacial chemicalinteractions via self-assembled monolayers, we achieved highlyefficient perovskite solar cells with device efficiencies reaching18.8%, showing a 10% improvement compared to the thatwithout SAM modification. We stress that the interfacialchemical interactions are an important consideration indesigning highly efficient charge transporting layers for thefuture development of highly efficient perovskite solar cells.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.nano-lett.6b04015.

Details regarding the experiments along with thesupporting figures (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: yangy@ucla.edu.ORCIDHuajun Chen: 0000-0003-1150-2623Yang Yang: 0000-0001-8833-7641NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was financially supported by a grant from The AirForce Office of Scientific Research (AFOSR, grant number:FA2386-15-1-4108, Program Manager: Dr. Charles Lee),UCLA Internal Funds, UC-Solar Program (Fund NumberMRPI 328368), and the Enli Tech (in Taiwan) for donating theEQE measurement system to UCLA. We also like toacknowledge Dr. Yajie Jiang and Prof. Anita Ho-Baillie forassistance in optical modelling.

■ REFERENCES(1) McGehee, M. D. Nature 2013, 501 (7467), 323−325.(2) Jung, H. S.; Park, N.-G. Small 2015, 11 (1), 10−25.(3) Stranks, S. D.; Nayak, P. K.; Zhang, W.; Stergiopoulos, T.; Snaith,H. J. Angew. Chem., Int. Ed. 2015, 54 (11), 3240−3248.(4) Chen, Q.; De Marco, N.; Yang, Y.; Song, T.-B.; Chen, C.-C.;Zhao, H.; Hong, Z.; Zhou, H.; Yang, Y. Nano Today 2015, 10 (3),355−396.(5) Yang, W. S.; Noh, J. H.; Jeon, N. J.; Kim, Y. C.; Ryu, S.; Seo, J.;Seok, S. I. Science 2015, 348 (6240), 1234.(6) Snaith, H. J. J. Phys. Chem. Lett. 2013, 4 (21), 3623−3630.(7) Cheng, Z.; Lin, J. CrystEngComm 2010, 12 (10), 2646−2662.(8) Lin, Q.; Armin, A.; Nagiri, R. C. R.; Burn, P. L.; Meredith, P. Nat.Photonics 2014, 9 (2), 106−112.(9) Juarez-Perez, E. J.; Sanchez, R. S.; Badia, L.; Garcia-Belmonte, G.;Kang, Y. S.; Mora-Sero, I.; Bisquert, J. J. Phys. Chem. Lett. 2014, 5 (13),2390−2394.(10) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. Nat. Photonics 2014, 8(7), 506−514.(11) Stranks, S. D.; Eperon, G. E.; Grancini, G.; Menelaou, C.;Alcocer, M. J. P.; Leijtens, T.; Herz, L. M.; Petrozza, A.; Snaith, H. J.Science 2013, 342 (6156), 341.(12) Oga, H.; Saeki, A.; Ogomi, Y.; Hayase, S.; Seki, S. J. Am. Chem.Soc. 2014, 136 (39), 13818−13825.(13) Zhou, H.; Chen, Q.; Li, G.; Luo, S.; Song, T.-b.; Duan, H.-S.;Hong, Z.; You, J.; Liu, Y.; Yang, Y. Science 2014, 345 (6196), 542.(14) Juarez-Perez, E. J.; Wuβler, M.; Fabregat-Santiago, F.; Lakus-Wollny, K.; Mankel, E.; Mayer, T.; Jaegermann, W.; Mora-Sero, I. J.Phys. Chem. Lett. 2014, 5 (4), 680−685.(15) Liu, D.; Kelly, T. L. Nat. Photonics 2013, 8 (2), 133−138.(16) Liu, M.; Johnston, M. B.; Snaith, H. J. Nature 2013, 501 (7467),395−398.(17) Li, Y.; Zhao, Y.; Chen, Q.; Yang, Y.; Liu, Y.; Hong, Z.; Liu, Z.;Hsieh, Y.-T.; Meng, L.; Li, Y.; Yang, Y. J. Am. Chem. Soc. 2015, 137(49), 15540−15547.(18) Greiner, M. T.; Helander, M. G.; Tang, W.-M.; Wang, Z.-B.;Qiu, J.; Lu, Z.-H. Nat. Mater. 2011, 11 (1), 76−81.(19) Zuo, L.; Yao, J.; Li, H.; Chen, H. Sol. Energy Mater. Sol. Cells2014, 122, 88−93.(20) Schulz, P.; Whittaker-Brooks, L. L.; MacLeod, B. A.; Olson, D.C.; Loo, Y.-L.; Kahn, A. Adv. Mater. Interfaces 2015, 2 (7), 1400532.(21) Hu, Q.; Wu, J.; Jiang, C.; Liu, T.; Que, X.; Zhu, R.; Gong, Q.ACS Nano 2014, 8 (10), 10161−10167.(22) Bi, D.; Tress, W.; Dar, M. I.; Gao, P.; Luo, J.; Renevier, C.;Schenk, K.; Abate, A.; Giordano, F.; Correa Baena, J.-P.; Decoppet, J.-D.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Gratzel, M.; Hagfeldt, A.Sci. Adv. 2016, 2 (1), e1501170.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

F

(23) Jeng, J.-Y.; Chiang, Y.-F.; Lee, M.-H.; Peng, S.-R.; Guo, T.-F.;Chen, P.; Wen, T.-C. Adv. Mater. 2013, 25 (27), 3727−3732.(24) Tripathi, N.; Shirai, Y.; Yanagida, M.; Karen, A.; Miyano, K. ACSAppl. Mater. Interfaces 2016, 8 (7), 4644−4650.(25) Min, J.; Zhang, Z.-G.; Hou, Y.; Ramirez Quiroz, C. O.; Przybilla,T.; Bronnbauer, C.; Guo, F.; Forberich, K.; Azimi, H.; Ameri, T.;Spiecker, E.; Li, Y.; Brabec, C. J. Chem. Mater. 2015, 27 (1), 227−234.(26) Ke, W.; Fang, G.; Liu, Q.; Xiong, L.; Qin, P.; Tao, H.; Wang, J.;Lei, H.; Li, B.; Wan, J.; Yang, G.; Yan, Y. J. Am. Chem. Soc. 2015, 137(21), 6730−6733.(27) Zuo, L.; Gu, Z.; Ye, T.; Fu, W.; Wu, G.; Li, H.; Chen, H. J. Am.Chem. Soc. 2015, 137 (7), 2674−2679.(28) Li, X.; Ibrahim Dar, M.; Yi, C.; Luo, J.; Tschumi, M.;Zakeeruddin, S. M.; Nazeeruddin, M. K.; Han, H.; Gratzel, M. Nat.Chem. 2015, 7 (9), 703−711.(29) Liu, L.; Mei, A.; Liu, T.; Jiang, P.; Sheng, Y.; Zhang, L.; Han, H.J. Am. Chem. Soc. 2015, 137 (5), 1790−1793.(30) Haruyama, J.; Sodeyama, K.; Han, L.; Tateyama, Y. J. Phys.Chem. Lett. 2014, 5 (16), 2903−2909.(31) Geng, W.; Tong, C.-J.; Tang, Z.-K.; Yam, C.; Zhang, Y.-N.; Lau,W.-M.; Liu, L.-M. J. Materiom. 2015, 1 (3), 213−220.(32) Geng, W.; Tong, C.-J.; Liu, J.; Zhu, W.; Lau, W.-M.; Liu, L.-M.Sci. Rep. 2016, 6, 20131.(33) Mei, A.; Li, X.; Liu, L.; Ku, Z.; Liu, T.; Rong, Y.; Xu, M.; Hu, M.;Chen, J.; Yang, Y.; Gratzel, M.; Han, H. Science 2014, 345 (6194), 295.(34) Chang, C.-Y.; Chang, Y.-C.; Huang, W.-K.; Liao, W.-C.; Wang,H.; Yeh, C.; Tsai, B.-C.; Huang, Y.-C.; Tsao, C.-S. J. Mater. Chem. A2016, 4 (20), 7903−7913.(35) Abrusci, A.; Stranks, S. D.; Docampo, P.; Yip, H.-L.; Jen, A. K.Y.; Snaith, H. J. Nano Lett. 2013, 13 (7), 3124−3128.(36) Yip, H.-L.; Hau, S. K.; Baek, N. S.; Ma, H.; Jen, A. K. Y. Adv.Mater. 2008, 20 (12), 2376−2382.(37) Gu, Z.; Zuo, L.; Larsen-Olsen, T. T.; Ye, T.; Wu, G.; Krebs, F.C.; Chen, H. J. Mater. Chem. A 2015, 3 (48), 24254−24260.(38) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao,P.; Nazeeruddin, M. K.; Gratzel, M. Nature 2013, 499 (7458), 316−319.(39) Xiao, Z.; Bi, C.; Shao, Y.; Dong, Q.; Wang, Q.; Yuan, Y.; Wang,C.; Gao, Y.; Huang, J. Energy Environ. Sci. 2014, 7 (8), 2619−2623.(40) Chen, Q.; Zhou, H.; Song, T.-B.; Luo, S.; Hong, Z.; Duan, H.-S.;Dou, L.; Liu, Y.; Yang, Y. Nano Lett. 2014, 14 (7), 4158−4163.(41) Wang, C.; Liu, X.; Wang, C.; Xiao, Z.; Bi, C.; Shao, Y.; Huang,J.; Gao, Y. J. Vac. Sci. Technol. B 2015, 33 (3), 032401.(42) Kim, H.-S.; Lee, C.-R.; Im, J.-H.; Lee, K.-B.; Moehl, T.;Marchioro, A.; Moon, S.-J.; Humphry-Baker, R.; Yum, J.-H.; Moser, J.E.; Gratzel, M.; Park, N.-G. Sci. Rep. 2012, 2, 591.(43) Malynych, S.; Luzinov, I.; Chumanov, G. J. Phys. Chem. B 2002,106 (6), 1280−1285.(44) Noel, N. K.; Abate, A.; Stranks, S. D.; Parrott, E. S.; Burlakov, V.M.; Goriely, A.; Snaith, H. J. ACS Nano 2014, 8 (10), 9815−9821.(45) Jeon, N. J.; Noh, J. H.; Yang, W. S.; Kim, Y. C.; Ryu, S.; Seo, J.;Seok, S. I. Nature 2015, 517 (7535), 476−480.(46) Jiang, Y.; Almansouri, I.; Huang, S.; Young, T.; Li, Y.; Peng, Y.;Hou, Q.; Spiccia, L.; Bach, U.; Cheng, Y.-B.; Green, M. A.; Ho-Baillie,A. J. Mater. Chem. C 2016, 4 (24), 5679−5689.(47) Chen, C.-W.; Hsiao, S.-Y.; Chen, C.-Y.; Kang, H.-W.; Huang, Z.-Y.; Lin, H.-W. J. Mater. Chem. A 2015, 3 (17), 9152−9159.(48) Meloni, S.; Moehl, T.; Tress, W.; Franckevicius, M.; Saliba, M.;Lee, Y. H.; Gao, P.; Nazeeruddin, M. K.; Zakeeruddin, S. M.;Rothlisberger, U.; Graetzel, M. Nat. Commun. 2016, 7, 10334.(49) Snaith, H. J.; Abate, A.; Ball, J. M.; Eperon, G. E.; Leijtens, T.;Noel, N. K.; Stranks, S. D.; Wang, J. T.-W.; Wojciechowski, K.; Zhang,W. J. Phys. Chem. Lett. 2014, 5 (9), 1511−1515.(50) Marco, N. D.; Zhou, H.; Chen, Q.; Sun, P.; Liu, Z.; Meng, L.;Yao, E.-P.; Liu, Y.; Schiffer, A.; Yang, Y. Nano Lett. 2016, 16 (2),1009−1016.(51) Shao, Y.; Xiao, Z.; Bi, C.; Yuan, Y.; Huang, J. Nat. Commun.2014, 5, 5784.

Nano Letters Letter

DOI: 10.1021/acs.nanolett.6b04015Nano Lett. XXXX, XXX, XXX−XXX

G