The role of PbI₂ in CH₃NH₃PbI₃ perovskite stability, solar cell parameters and device degradation

Abstract

We report a systematic investigation on the role of excess PbI₂ content in CH₃NH₃PbI₃ perovskite film properties, solar cell parameters and device storage stability. We used the CH₃NH₃I vapor assisted method for the preparation of PbI₂-free CH₃NH₃PbI₃ films under a N₂ atmosphere. These pristine CH₃NH₃PbI₃ films were annealed at 165 °C for different time intervals in a N₂ atmosphere to generate additional PbI₂ in these films. From XRD measurements, the excess of PbI₂ was quantified. Detailed characterization using scanning electron microscopy, X-ray diffraction, UV-Visible and photoluminescence for continuous aging of CH₃NH₃PbI₃ films under ambient condition (50% humidity) is carried out for understanding the influence of different PbI₂ contents on degradation of the CH₃NH₃PbI₃ films. We find that the rate of degradation of CH₃NH₃PbI₃ is accelerated due to the amount of PbI₂ present in the film. A comparison of solar cell parameters of devices prepared using CH₃NH₃PbI₃ samples having different PbI₂ contents reveals a strong influence on the current density–voltage hysteresis as well as storage stability. We demonstrate that CH₃NH₃PbI₃ devices do not require any residual PbI₂ for a high performance. Moreover, a small amount of excess PbI₂, which improves the initial performance of the devices slightly, has undesirable effects on the CH₃NH₃PbI₃ film stability as well as on device hysteresis and stability.

---

Introduction

Organometal halide perovskite solar cells have achieved great progress in the last few years. The power conversion efficiency (PCE) of CH₃NH₃PbI₃ perovskite based thin film photovoltaic devices has already reached a high value of 22% in just a few years.¹ This rapid progress was achieved by optimizing both the material and device architecture and understanding the underlying photoinduced charge separation mechanisms of these devices.^2–11 Major efforts have been made to achieve high efficiency using various approaches comprising passivation of defects¹² and tuning and controlling numerous parameters in perovskite layers such as the surface morphology,¹³ grain size,¹⁴ different halogen ratio^15,16 and different cation compositions (mixture of CH₃NH₃/HC(NH₂)₂).¹⁷ Additionally, both the charge extraction layers, the electron transport layer (ETL)⁷ and the hole transport layer (HTL),¹⁸ have also been studied and optimized. Several recent papers reveal that one of the preferred techniques to improve the PCE of CH₃NH₃PbI₃ is using excess PbI₂, which is not converted to CH₃NH₃PbI₃. For example, Chen et al. demonstrated the role of PbI₂ in CH₃NH₃PbI₃ for the improvement of device performance by annealing after preparing pristine CH₃NH₃PbI₃ using the CH₃NH₃I (MAI)-vapor assisted method.¹² Kim et al. demonstrated that 5.7 wt% PbI₂ in the (CH(NH₂)₂PbI₃)_0.85(CH₃NH₃PbBr₃)_0.15 perovskite prepared using the solvent engineering method improved the PCE as well as reduced the hysteresis.¹⁷ Roldan-Carmona et al. observed that 10 wt% additional PbI₂ in CH₃NH₃PbI₃ using the solvent engineering method is beneficial for the improvement of PCE of devices.¹⁹ It was also reported that 2 wt% residual PbI₂ in CH₃NH₃PbI₃ is beneficial to improve the device efficiency.²⁰ Moreover, the benefits of PbI₂ in CH₃NH₃PbI₃ towards an improved performance can be inferred from the fact that most of the high PCE perovskite solar cells prepared using a variety of methods exhibited unintentionally a small amount of excess PbI₂.^21,22 Jacobsson et al.²³ also confirmed that the hysteresis index in (CH(NH₂)₂PbI₃)_0.85(CH₃NH₃PbBr₃)_0.15 devices is higher for samples having a residual PbI₂ content compared to the PbI₂-deficient samples. Based on this, it was suggested that excess PbI₂ in CH₃NH₃PbI₃ can help to passivate defects at surfaces and grain boundaries¹² and a small amount of residual PbI₂ in the perovskite supports reduction of charge recombination and improves open circuit voltage (V_oc) and the fill factor (FF).¹⁷ It has also been suggested that the small amount of excess PbI₂ in the perovskite influences the morphology and increases the size as well as the uniformity of perovskite crystals using the solvent engineering method.^19,24

On the other hand, it is noteworthy that for PbI₂-free CH₃NH₃PbI₃ solar cells using a modified MAI-vapor assisted method for CH₃NH₃PbI₃ crystallization, we clearly demonstrated a reproducible performance with ∼15% PCE.²⁵ Li et al. have a similar observation that a high PCE of 14.2% can be achieved for PbI₂-free CH₃NH₃PbI₃ prepared using a low pressure MAI-vapor assisted method.²⁶ While considering the excess PbI₂ for high PCE perovskite solar cells, one should not overlook the fact that PbI₂ in perovskites exhibits intrinsic instability under illumination.²⁷ Liu et al. demonstrated that residual PbI₂ accelerates the degradation of CH₃NH₃PbI₃ films upon exposure to illumination.²⁷ Encapsulated devices have the same stability performance like excess PbI₂-free CH₃NH₃PbI₃ devices. Thus, the influence of PbI₂ in perovskite films on solar cells is moderately known but their influence on degradation or stability is ambiguous. It is therefore of great interest to study systematically the effect of PbI₂ in CH₃NH₃PbI₃ not only on the PCE of solar cells, but also on their photophysics and stability, since both long term stability and a high PCE are essential for practical applications. For a comprehensive investigation of the effect of PbI₂, it is necessary to consider separately the effects caused by PbI₂ in CH₃NH₃PbI₃ on absorption/emission, morphology, decomposition of the perovskite layer, device parameters and the overall device stability. Thus the open questions that we address are (1) what is the influence of residual PbI₂ on the stability of the perovskite films and morphology? (2) Can we observe any quantifiable trend in the PL properties and PbI₂ content? (3) How far is the beneficial effect of PbI₂ counteracting the performance and stability of the devices? and (4) Is it possible to separate the different positive and negative effects to make conclusion on the net effect of residual PbI₂?

We used the MAI-vapor assisted method for the preparation of CH₃NH₃PbI₃ films under a N₂ atmosphere.²⁵ With this method we guarantee uniform film formation with similar grain sizes and high reproducibility and stability of the CH₃NH₃PbI₃ films.^25,28 There are three ways to introduce PbI₂ in such films: (a) by prolonged thermal annealing of a PbI₂-free film¹² or (b) by purposefully adding non-stoichiometric amounts of precursors to obtain residual non-converted PbI₂ in the final perovskite^17,19 or (c) partial conversion of PbI₂ to CH₃NH₃PbI₃.^21,22 We adopted the first method to obtain samples with different amounts of PbI₂ by annealing the samples at 165 °C for different time intervals in a N₂ atmosphere. Characterization using scanning electron microscopy (SEM), X-ray diffraction (XRD), UV-Visible and photoluminescence (PL) with aging of CH₃NH₃PbI₃ films under ambient conditions (50% humidity) is carried out for understanding the influence of residual PbI₂ on the stability of CH₃NH₃PbI₃ itself. A comparison of solar cell parameters for the different CH₃NH₃PbI₃ samples with varying PbI₂ contents is also performed. We demonstrate that a solar cell device with a very small amount of excess PbI₂ can exhibit a slightly higher PCE, but this small amount of excess PbI₂ has undesirable effects on the CH₃NH₃PbI₃ film stability, current density–voltage (J/V) hysteresis and device stability. The presence of a small amount of excess PbI₂ results in the faster degradation of the CH₃NH₃PbI₃ layer in a humid air atmosphere and the degradation rate increases with the amount of additional PbI₂ in CH₃NH₃PbI₃ films. Thus, this work gives insights into the overall consequences of residual PbI₂ on different aspects of CH₃NH₃PbI₃ solar cells. This work confirms the initial positive influence of excess PbI₂ on CH₃NH₃PbI₃ solar cell devices but simultaneously also proves that it accelerates film degradation and therefore limits device lifetime.

Experimental section

Materials

All starting materials were purchased from Sigma-Aldrich and used as received. Spiro-OMeTAD was purchased from Merck Chemicals.

MAI synthesis

MAI was synthesized by reacting 24 mL of methylamine (33 wt% in absolute ethanol) and 10 mL of hydroiodic acid (57 wt% in water) in a round-bottom flask at room temperature (RT) for 2 h under Ar with stirring. The raw precipitate was recovered by removing the solvent in a rotary evaporator at 40 °C. The raw product was washed with ethyl ether, dried in a vacuum at RT and redissolved in absolute ethanol. The pure MAI recrystallized upon cooling is filtered and dried at RT in a vacuum oven for 24 h. The purity of MAI is very crucial for the synthesis of a pure perovskite.

Device fabrication

Fluorine doped tin oxide (FTO)-coated glass sheets (17 Ω □⁻¹) were etched with zinc powder and HCl (2 M) to obtain the required electrode pattern. The substrates were then cleaned with a detergent followed by sonication in deionized water, acetone and ethanol for 10 min each, and dried with clean dry air. A 50 nm compact-TiO₂ (c-TiO₂) blocking layer was deposited by spray pyrolysis of titanium(IV) bis(acetoacetonato)–di(isopropanoxylate) diluted in ethanol at 450 °C on FTO-coated glass substrates and annealed at 450 °C for 1 h. After cooling, the substrates were transferred in a glovebox under a N₂ atmosphere. For CH₃NH₃PbI₃ formation, we adapted a published procedure^12,25,28 and optimized it as follows. PbI₂ (1 M) was dissolved in N,N-dimethyl formamide overnight under stirring conditions at 100 °C and 80 μL solution was spin coated on the FTO/c-TiO₂ substrates at 2000 rpm for 50 s, and dried at 100 °C for 5 min. 100 mg MAI powder was spread out around the PbI₂ coated substrates with a Petri dish covering on the top and heated at 165 °C for 13 h for full conversion. Subsequently, the as-prepared samples were annealed at 165 °C for different time intervals to generate residual PbI₂. The HTL deposition solution comprised of 2,2′,7,7′-tetrakis-(N,N-di-p-methoxyphenylamine) 9,9′-spirobifluorene (spiro-OMeTAD) in chlorobenzene (102.85 mg mL⁻¹), 40 mL of 4-tert-butylpyridine and 37 mL (520 mg mL⁻¹) solution of bis(trifluoromethane) sulfonimide lithium salt (LiTFSI) in acetonitrile. This solution was spin coated at 2000 rpm for 30 s under N₂ conditions and stored in dry air overnight. No oxidants were used. This was followed by thermal deposition of gold to form the back contact. The active areas of the devices fabricated in this work are 0.09 and 0.16 cm².

Characterization

The structural and phase characterization of the CH₃NH₃PbI₃ layer deposited on the FTO/c-TiO₂ substrate was carried out in the reflection mode XRD using a Panalytical X'pert pro diffractometer with Cu Kα₁ radiation (λ = 0.15405 nm), operated at 40 kV and 40 mA. The diffractograms were recorded in the 2θ range of 10 to 45° with a 2θ step size of 0.008°. UV-Vis absorption spectra were collected using a V-630 UV-Vis Jasco Spectrophotometer coupled with an integrating sphere. The surface morphology and cross-section were characterized by field emission SEM using a Zeiss 1530 instrument with an accelerating voltage of 3.0 kV. PL spectra and PL quantum yield were recorded of CH₃NH₃PbI₃ deposited on quartz substrates with an integrating sphere coupled via optical fiber to a Oriel spectrograph MS125 and a Andor Idus CCD camera. The excitation source is a 485 nm laser diode LDH-D-C-485, PicoQuant GmbH. The used intensity was about 450 mW cm⁻². Each sample is measured after 2 min of laser exposure in order to account for the “soaking effect”.^29,30 The exposure time is kept constant for all measurements. The influence of the soaking effect was investigated for each sample at day 1 (see Fig. S1 in the ESI†). The J/V plots of solar cell devices were obtained using an Oriel solar simulator under AM 1.5, 1000 W m⁻² conditions. The J/V data under these conditions were obtained using a Keithley model 2400 source meter. All J/V measurements were performed in a N₂ atmosphere. Both forward (from −1 V to +2 V) and backward (from +2 V to −1 V) sweeps were recorded to understand the hysteresis. All the devices were measured at a sweep rate of 100 mV s⁻¹. Each device was equilibrated by multiscans in reverse mode under illumination till the short circuit current density (J_sc) was stabilized. Each device type was fabricated in two batches (4 devices each) and analyzed to understand the scattering of device parameters.

Results and discussion

XRD analysis of the residual PbI₂ content and the influence on the morphology

For generation of PbI₂ in CH₃NH₃PbI₃ films, we used thermal annealing of pure CH₃NH₃PbI₃ films at 165 °C for different time intervals (0, 15, 30, 45 and 60 min) and the five samples obtained are mentioned in the following as A, B, C, D and E, respectively. To monitor the generation of PbI₂ by annealing and to verify the comparative amount of PbI₂ in the processed CH₃NH₃PbI₃ material, XRD experiments have been performed for all samples including pure PbI₂ (Fig. 1a). The X-ray patterns confirm the presence of the tetragonal perovskite phase of CH₃NH₃PbI₃ in all of the prepared films. In sample A, XRD pattern shows that the films composed of a single pristine perovskite phase with strong diffraction peaks (110), (220) and (114), characteristic of CH₃NH₃PbI₃ in the tetragonal phase. No peaks of PbI₂ could be detected. The PbI₂ phase in the pure perovskite can be identified by the characteristic PbI₂ peak located at 2θ ∼ 12.7°. Accordingly, for sample B, a small PbI₂ peak becomes apparent at 2θ of 12.7°. With increasing annealing time, the characteristic PbI₂ peak increases in intensity from sample B to E on the concomitant expense of the intensity of the (110) peak of the perovskite at about 15°. Fig. 1b shows that the amount of PbI₂ present in the pervoskite films and PbI₂ wt% rapidly increase with an increase in the annealing time, confirming this interconversion during thermal annealing. An increase of PbI₂ amounts in the films as obtained from XRD is confirmed by PbI₂ wt% calculated from optical measurements and both values agree very well. As per XRD measurements, the samples A to E have 0, 21, 34, 46 and 48 wt% PbI₂ respectively. Detailed calculations of PbI₂ wt% in CH₃NH₃PbI₃ films from XRD and optical spectroscopy (Fig. S2 and S3) are explained in the ESI.†

Fig. 1 (a) XRD patterns of CH₃NH₃PbI₃ samples on compact TiO₂ annealed at 165 °C for (A) 0 min, (B) 15 min, (C) 30 min, (D) 45 min and (E) 60 min; (b) PbI₂ wt% and CH₃NH₃PbI₃ wt% with annealing time. XRD peaks are labeled for CH₃NH₃PbI₃ (black dot), PbI₂ (open triangle) and substrate (black star).

SEM images obtained from the surface analysis of the CH₃NH₃PbI₃ films A, B, C and E are shown in Fig. 2, which illustrate the changes in the grain-size upon annealing. Since there is no detectable difference for samples D and E, the image of sample D is given in Fig. S4 (ESI†). From the surface view of sample A, it is clear that the CH₃NH₃PbI₃ films are composed of micrometer-sized grains that exhibit a densely packed morphology and are free from pinholes/pores. The average grain size is 1000 nm. From sample B to E, compactness of the samples decreases as well as edges of grains vanish from sample B to E. Especially in sample E, one can observe loss of matter at the grain boundaries, which suggest that with a prolonged annealing time, the PbI₂ content increases by evaporation of MAI at grain boundaries and probably at the surface of the film as well. The cross-sectional SEM of samples A to E (Fig. S4, ESI†) does not show any clear separation/gap between the grains in the bulk of the film. According to Chen et al., CH₃NH₃PbI₃ decomposed upon prolonged annealing at 150 °C and conversion from CH₃NH₃PbI₃ to PbI₂ occurred where CH₃NH₃I species escaped from the perovskite film to form the PbI₂ phase mostly at grain boundaries and at the relevant interfaces.¹² Xie et al. also confirmed the existence of PbI₂ on the surface formed by decomposition of CH₃NH₃PbI₃ by annealing at 150 °C.³¹ However, in our case it seems most decomposition happens at the surface, but it is difficult to exactly depict the location regarding the presence/distribution of PbI₂ in these films just from SEM analysis.

Fig. 2 SEM images of CH₃NH₃PbI₃ samples A, B, C and E at magnification 10 KX. The SEM image of sample D is very similar to that of sample E and is shown in Fig. S4 (ESI†).

Influence of PbI₂ on the optical properties

Fig. 3a shows the absorption spectra of samples A to E. The initial thicknesses of all samples were similar and lie in the range of 300 to 325 nm. The position of band-edge absorption of all samples is almost the same but the absolute absorbance at 700 nm for samples A to E indicates that the optical density decreases. It is obvious that, from samples A to E, the CH₃NH₃PbI₃ amount in the film decreases with an increase in the PbI₂ content. In the absorption graph for samples B to E, a kink (shoulder) is observed at 500 nm and its intensity increased from sample B to E. Upon comparison to the PbI₂ absorption (Fig. 3a), this is indicative of an increased amount of the PbI₂ content in the CH₃NH₃PbI₃ films. To obtain the band gap (E_g), these data were further analyzed using the classical Tauc relation, for direct optical gap semiconductors: α = A(hν − E_g)ⁿ/hν, where α is the absorption coefficient, A is the proportionality constant, n = 0.5 for direct band gap materials and hν = photon energy. The plots of (αhν)²versus hν of the CH₃NH₃PbI₃ films are shown in Fig. 3b. This graph gives E_g, when a straight portion of (αhν)² against the hν plot is extrapolated to the point α = 0. For all samples it is around 1.6 eV, thus independent of the excess PbI₂ content and consistent with recent literature reports.^32,33

Fig. 3 (a) UV-Visible absorption spectra of samples A to E and PbI₂ film; (b) plots of (αhν)² against hν of all CH₃NH₃PbI₃ samples; (c) steady-state photoluminescence (PL) for all samples.

Fig. 3c shows the steady-state PL spectra of the different samples on quartz substrates without any protection layers on top and measured in humid air (50% humidity). While sample A does not show any PL, the PL intensity increases from sample B to E and the peak position stays constant around 780 nm (1.59 eV).

Influence of PbI₂ on degradation of perovskite films

A second series of XRD measurements were performed to explore the film stability and the degree of degradation of the pristine CH₃NH₃PbI₃ and PbI₂ containing CH₃NH₃PbI₃ films over time in humid air. Fig. 4 shows the typical XRD patterns of CH₃NH₃PbI₃ samples A (no PbI₂), B (least PbI₂) and E (highest PbI₂) over time in a controlled 50% humid atmosphere. For sample A, strong characteristic peaks of the CH₃NH₃PbI₃ tetragonal structure are observed on the first day, after 15 days and after 30 days of storage. There is no observable change in the XRD pattern between the pristine CH₃NH₃PbI₃ film and the CH₃NH₃PbI₃ film after 15 days of storage at 50% humidity as well as after 30 days. For sample B, the diffraction characteristic peaks of the CH₃NH₃PbI₃ structure (at 2θ = 14.2°) along with PbI₂ characteristic peak at a 2θ value of 12.7° are observed on day 1. After storage of 15 days in 50% humidity, the intensity of the (110) peak at a 2θ value of 14.2° is slightly decreased and the PbI₂ peak intensity is increased. After 30 days of storage, the peak intensity of PbI₂ is significantly increased with a radical decrease of the CH₃NH₃PbI₃ (110) peak intensity. For sample E (the highest PbI₂ content with respect to CH₃NH₃PbI₃), the CH₃NH₃PbI₃ diffraction characteristic (110) peak intensity is lower than the PbI₂ characteristic peak on the first day. The (110) peak intensity is drastically decreased with an increase in the PbI₂ peak after 15 days of storage of sample E. After 30 days, the PbI₂ peak intensity is increased and CH₃NH₃PbI₃ peaks appear with a very low intensity.

Fig. 4 XRD patterns of pure and PbI₂ containing CH₃NH₃PbI₃ samples obtained upon air exposure for different periods (at 50% humidity), (a) samples A, (b) sample B and (c) sample E. The peaks assigned: black dot for CH₃NH₃PbI₃ and open triangle for PbI₂.

The degradation mechanism of CH₃NH₃PbI₃ in the presence of water, high and low relative humidity has been already suggested by different authors in reactions (1)–(3) below.^34–40

According to the literature, pure CH₃NH₃PbI₃ likely reacts with water from humid air and forms non-soluble PbI₂ and aqueous CH₃NH₃I (reaction (1)).^35,36 At higher (>80%) humidity CH₃NH₃PbI₃ reacts with water and forms (CH₃NH₃)₄PbI₆·2H₂O (reaction (2)).^34,37,38 In the presence of lower (<50%) humidity CH₃NH₃PbI₃ decomposes into CH₃NH₂, HI, PbI₂ with an irreversible reaction as given in eqn (3).^39,40

In our experiment, the CH₃NH₃PbI₃ films (sample A) are highly air-stable robust materials that do not show any PbI₂ peaks due to degradation for up to 30 days, similar to the observations of Christians et al.³⁴ However, when PbI₂ is present in CH₃NH₃PbI₃, most probably a non-reversible reaction (4) takes place producing volatile (CH₃NH₂). Thus, as observed in our experiments, this degradation process is faster for the excess PbI₂ content in CH₃NH₃PbI₃ compared to the PbI₂-free CH₃NH₃PbI₃ samples. Based on these observations, it is clear that PbI₂ acts as a catalyst for the degradation of CH₃NH₃PbI₃ as we suggest in the reaction (4).

To understand the stability of the CH₃NH₃PbI₃ films stored under ambient condition with 50% humidity, we tracked the normalized absorbance of the film at 700 nm for samples A, B, C, D and E for a month (Fig. 5a). The complete spectral change for each sample is given in Fig. S5 (ESI†). Absorbance measurements made on CH₃NH₃PbI₃ films with different amounts of PbI₂ content generated by annealing of the CH₃NH₃PbI₃ sample show the typical spectral features of CH₃NH₃PbI₃ with PbI₂ on first day. Thus the rate of decrease of the absorbance intensity at 700 nm increases from sample B to E and it is the highest for sample E. To summarize the changes in absorption with storage time, we have additionally plotted the change in optical absorbance with storage time in 50% humidity for the two extreme cases; for sample A without any residual PbI₂ and E having an appreciable amount of PbI₂ (Fig. 5b). For sample A, overlap of the optical spectra for the first 15 days is observed and only a negligible change is observed after 30 days. The optical absorption curves in Fig. 5a and Fig. S5a (ESI†) do not show characteristic peaks from PbI₂, however the absorption curves show a slight decline after 30 days, even though no degradation could be confirmed in XRD studies. Thus we believe that the observed negligible decline of absorbance after 30 days could be due to the surface interaction of the perovskite with humidity in those samples during optical measurements. Additionally, at the very first degradation stage, where only negligible PbI₂ is present, PbI₂ may not be detected by our used XRD set up, but it may be sensitive enough for optical measurements. In contrast, for sample E the decrease is much stronger and the shoulder at 500 nm (arising out of PbI₂ absorption, see Fig. 3a) gets more pronounced with storage time. Sample A shows the least change in the optical spectra. The increased rate of absorbance loss indicates that PbI₂ in the samples B to E accelerates the rate of degradation.

Fig. 5 (a) Normalized absorbance at 700 nm of CH₃NH₃PbI₃ samples A to E stored under 50% RH as a function of time (for 0–30 days); dotted lines are linear fits to guide the eye. (b) Comparative decay of UV-Visible absorption of samples A and E stored under 50% relative humidity for 0, 15 and 30 days. (c) Integrated PL intensity for the different samples over storage time under 50% relative humidity. Dotted lines are 2nd order polynomial fits to serve as guide to the eye (see ESI,† Table S2 for fit parameters).

The PL emission studies on samples without and with PbI₂ content give further insight into the role of PbI₂ in degradation. All the samples were stored under the same conditions of 50% humidity for about 60 days. We observed that samples A to E show different evolutions in the integrated PL intensity over time (Fig. 5c). The pure CH₃NH₃PbI₃ (sample A) with hardly any initial PL emission (Fig. 3c) continuously increased in the PL intensity with a storage time of up to 58 days.

For sample B, PL emission is observed from the first day onwards and it increased up to 38 days; and afterwards decreased continuously. A similar trend is observed for sample C, yet the PL peak emission is reached already after 20 days. Sample D (not shown here) and E have similar PL emission spectra over time. For both samples, PL emission decreased from the first day onwards. The integrated intensity first increases and reaches its maximum even before 20 days for both D and E samples. The peak position shifts to shorter times with increasing excess PbI₂ and the peak intensity reduces. This indicates that whenever samples have some amount of PbI₂, it helps to increase PL emission, but also deteriorates the PL efficiency for a long term. The actual mechanisms of PbI₂ in PL enhancement are still topics of ongoing discussion. As widely accepted in the literature, the main mechanism is related to trap-filling effects.⁴¹ Degradation of the perovskite structure is accompanied by a blue shift of the PL peak position, which is also seen here (see S6 in ESI†).⁴² Moreover, the evolution of the PL intensity with time correlates with the content of PbI₂. This suggests that a small fraction of PbI₂ reduces non-radiative recombination, a higher fraction leads to a decrease in the PL intensity. The relative change in PL is a sensitive tool that indicates the degradation of the CH₃NH₃PbI₃ films.

Influence of the PbI₂ content on solar cell parameters and hysteresis

To comparatively study the influence of the PbI₂ content on the photovoltaic parameters, simple planar devices (glass/FTO/c-TiO₂/CH₃NH₃PbI₃/spiro-OMeTAD/Au) without any mesoporous TiO₂ layers were prepared using samples A, B, C, D and E under N₂ and investigated. The J/V plots of the best devices are illustrated in Fig. 6. All of the data presented here are obtained in the reverse scan from +2.0 to −1.0 V at 100 mV s⁻¹. The solar cell parameters of the best-performing devices and average PCE values (of 8 devices for each type from 2 batches) are listed in Table 1. The scattering of solar cell parameters for devices with samples A to E is given in Table S3 (ESI†). The device based on sample A exhibits a J_sc of 15.18 mA cm⁻², a V_oc of 0.90 V, and 69% FF with a maximum PCE of 10.80% (average: 9.12% and constant output PCE at V_MPP = 9.4%). For B, where even though a small amount of CH₃NH₃PbI₃ has been consumed for the formation of PbI₂, the J_sc (17.22 mA cm⁻²) and V_oc (0.94 V) values are improved and consequently a higher PCE of 11.00% (average: 10.17%) is obtained. For devices based on samples C to E (with an increase in the PbI₂ content), J_sc decreases from 16.15 to 13.30 mA cm⁻², V_oc from 0.94 to 0.87 V as well as the FF from 66.88 to 56.37%. The decrease of J_sc from sample B to E is obvious because of a considerable reduction of the CH₃NH₃PbI₃ material in the active layer by decomposition and formation of PbI₂ upon annealing. It can be observed from Fig. 6 that CH₃NH₃PbI₃ solar cells fabricated from both samples, A without PbI₂ as well as B with PbI₂, exhibit excellent performance. The difference in average values for all the parameters is not considerable, even though the sample B exhibits the best performance. This is in agreement with numerous reports, which suggested to add a small amount of PbI₂ to improve the solar cell performance. Furthermore, the series and shunt resistances (R_s and R_sh) of the various devices were determined from illuminated J/V curves at V_oc and J_sc, respectively, and are listed in Table 1. From sample A to C, the R_s value decreased from 8.33 to 5.78 Ω cm², and it again increases from sample C to E, with an increase in the PbI₂ content. This has an obvious detrimental effect on charge transport as observed in the low J_sc and FF values. Among samples A to D, the device using sample A has the lowest R_sh value, which suggests that sample A has the highest carrier recombination within the device for the given geometry and contacts.

Fig. 6 (a) J/V characteristics of CH₃NH₃PbI₃ solar cell for samples A to E, (b) influence of scanning directions on J/V characteristics of the CH₃NH₃PbI₃ solar cell for samples A and E. For the forward bias from −1 V to +2 V and for the reverse bias from +2 V to −1 V measured under simulated AM 1.5 100 mW cm⁻² sun light at scan rate 100 mV s⁻¹; the hysteresis indices for CH₃NH₃PbI₃ solar cells employing samples A to E with increasing PbI₂ content are given in Table 1. The J/V assigned: filled symbol for reverse scan and open symbol for forward scan.

Table 1 Photovoltaic parameters of the devices prepared from different CH₃NH₃PbI₃ samples. The devices were measured directly after preparation under a nitrogen atmosphere. The average of 8 devices is given in brackets for all parameters. All the devices were measured under 100 mW cm⁻² AM 1.5 irradiation at a scan rate of 100 mV s⁻¹ in reverse bias. HI: hysteresis index

Samples	J sc (mA cm^−2) ±σ	V oc (V)	FF (%)	PCE (%)	R s (Ω cm^2)	R sh (Ω cm^2)	HI
A (0 min)	16.7 (15.1 ± 0.8)	0.94 (0.92 ± 0.03)	68 (65 ± 4)	10.8 (9.1 ± 1.1)	8.33	831	0.22
B (15 min)	17.2 (16.3 ± 0.8)	0.94 (0.94 ± 0.02)	68 (67 ± 2)	11.0 (10.2 ± 0.5)	7.41	2219	0.27
C (30 min)	16.6 (16.2 ± 0.4)	0.95 (0.94 ± 0.01)	67 (64 ± 4)	10.6 (9.7 ± 0.7)	5.78	1875	0.34
D (45 min)	14.9 (14.8 ± 0.3)	0.90 (0.90 ± 0.02)	68 (61 ± 6)	9.1 (8.2 ± 0.7)	7.87	7013	0.40
E (60 min)	12.81 (13.3 ± 0.8)	0.92 (0.87 ± 0.04)	59 (56 ± 6)	7.0 (6.5 ± 0.5)	8.33	2146	0.43

---

For perovskite solar cells containing perovskite/TiO₂ interfaces, the J/V data obtained depend on the scan rate, forward or backward sweep directions as well as the voltage of equilibration under illumination. This observed hysteresis causes usually drastic changes in the FF depending on the measurement protocol. Consequently the PCE calculated from the reverse scan (from +2.0 to −1.0 V) seems to be higher than that for the forward scan (from −1.0 to +2.0 V). In this regard, various measurement protocols have been recommended.^43,44 For the device application point of view, it is important to know why perovskite solar cells have hysteresis and how it can be reduced. Therefore, we examined, if there is an influence of generated PbI₂ on J/V hysteresis. In the course of J/V studies of devices with increasing PbI₂ content, we observe an enhancement in photovoltaic hysteresis (Fig. 6). From sample A to E, a decrease in J_sc mainly results from decreasing absorption due to gradual increasing conversion of CH₃NH₃PbI₃ to PbI₂. But the FF in the reverse scan increases as a consequence of the sweep direction. Additionally, we observed an increase of hysteresis from sample A to E. Hysteresis is measured in terms of the dimensionless hysteresis index factor as defined by Sanchez et al. to compare the different solar cells by taking the photocurrent values at 50% of V_oc.⁴⁵ Kim et al. modified this index by considering the fact that hysteresis has a pronounced effect of change in photocurrent at 80% of V_oc.⁴⁶ We used following hysteresis index to correlate the PbI₂ content in CH₃NH₃PbI₃ with J/V hysteresis as calculated from eqn (5).

where J_R(0.8V_oc) and J_F(0.8V_oc) are photocurrent values for reverse and forward scans at 80% of the V_oc value. The corresponding hysteresis indices for devices A to E are shown in Table 1. Upon comparing devices with similar J_sc and PCE (devices A and B), the hysteresis index is smaller for device A without PbI₂. It is higher for devices C, D and E. For comparable devices, thus the hysteresis index increased with an increase in the PbI₂ content in CH₃NH₃PbI₃. Here, prolongation of the annealing time may enhance the density of the defects and deteriorate the interface of TiO₂/CH₃NH₃PbI₃, which would also result in the hysteresis. Yu et al.⁴⁷ observed that the hysteresis index in CH₃NH₃PbI₃ devices is higher for samples having the residual PbI₂ content compared to samples without residual PbI₂. Jacobsson et al.²³ also observed that the hysteresis was generally higher in the PbI₂-rich samples whereas it was rather small in the stoichiometric and in most of the PbI₂-deficient samples which was attributed to difference in ion migration at the grain boundaries. Meanwhile, there's hysteresis for pure CH₃NH₃PbI₃ solar cells, indicating that there is still hysteresis in the devices even though the PbI₂ is absent. Although the origin of hysteresis effects is still unclear, one of the suggested reasons for this hysteretic behavior is dominantly associated with the perovskite material itself.⁴⁸ We firmly believe that by reducing the concentration of defects and their migration, the hysteresis can be suppressed considerably.

Device storage stability

The solar cell devices were stored under a N₂ atmosphere at room temperature without any further encapsulation. The J/V characteristics were measured for samples A to E for more than a month and the corresponding solar cell parameters are presented in Fig. 7. In general, the influence on V_oc and the FF as a function of time is not predominant and not considerably different for the different devices. But both J_sc and PCE exhibit considerable differences. For device A, J_sc increased from 15.18 to 16.10 mA cm⁻² and V_oc increased from 0.91 to 0.96 V, resulting in a slight improvement of the PCE from 9.4 to 9.6% for the first seven days. After 14 days, the overall PCE slightly decreased to 9.03%. At the end of 35 days (840 h), the device parameters J_sc, V_oc, FF and PCE are almost constant and show a negligible decrease, which demonstrates the excellent durability of PbI₂-free samples.

Fig. 7 Stacked plot of the normalized average performance parameters with different storage times for solar cells prepared from samples A (black dot), B (blue triangle), C (red square), D (gray square) and E (orange triangle).

For device B, J_sc shows a minor decrease (17 to 16.5 mA cm⁻²) and V_oc remains almost the same, resulting in a slight decrease of the PCE from 11.00 to 9.88% for the first seven days. After 14 days, the overall PCE decreased drastically to 8.60%. At the end of 35 days, all the device parameters J_sc, V_oc, FF and PCE decreased up to 12% of their original values. For devices C to E, J_sc was decreased and V_oc fluctuated in the range of 0.91 to 0.94 V, resulting in a slight decrease of the PCE for all devices for the first seven days. After 14 days, the overall PCE decreased drastically with time. J_sc of sample A remains almost constant and consequently the PCE after 30 days remains almost constant. But for the devices C, D and E, J_sc decreases up to 70% and accordingly PCE also goes down (in device C about 56% of the starting value). We did see a negative influence of the PbI₂ content on device storage stability of CH₃NH₃PbI₃ devices stored in a N₂ atmosphere. In contrast, Liu et al. did not observe any drastic decrease of the performance for encapsulated CH₃NH₃PbI₃ with and without residual PbI₂ devices.²⁷

Conclusions

We have demonstrated that different amounts of excess PbI₂ generated by annealing of pure CH₃NH₃PbI₃ prepared using the MAI-vapor assisted method have detrimental effects on the CH₃NH₃PbI₃ film stability, even though CH₃NH₃PbI₃ devices with small excess PbI₂ can exhibit a higher power conversion efficiency. The XRD, absorbance spectroscopy and PL measurement techniques were employed for a qualitative fundamental study on the stability of CH₃NH₃PbI₃ films with and without PbI₂. We show that the CH₃NH₃PbI₃ samples without PbI₂ degrade very slowly compared to CH₃NH₃PbI₃ with additional PbI₂. The rate of degradation is increased with an increase in the PbI₂ content in CH₃NH₃PbI₃. Pure CH₃NH₃PbI₃ without additional PbI₂ shows a positive impact on the material stability as well as device storage stability, suggesting a route toward perovskite solar cells with prolonged device lifetimes and resistance to humidity. We have shown that storage stability of the devices in N₂ atmospheres and the degree of J/V hysteresis (difference in the J/V curves depending on scan directions) in solar cells are dependent on the PbI₂ content in CH₃NH₃PbI₃. This study provides elucidation of fundamental pathways in the perovskite film decomposition process, which can be avoided to get more stable materials and more commercially sustainable devices.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Financial support by the German Federal Ministry of Education and Research BMBF (Project: 03SF0484C), the German National Science Foundation DFG (SFB 840, TH 807/6-1 and GRK 1640) and the Bavarian State Ministry of Science, Research, and the Arts for the Collaborative Research Network ‘‘Solar Technologies go Hybrid’’, are gratefully acknowledged.

References

1. NREL Solar Cells Efficiency Chart, http://www.nrel.gov/ncpv/images/efficiency_chart.jpg, 04/2017.
2. A. Ng, Z. Ren, Q. Shen, S. H. Cheung, H. C. Gokkaya, G. Bai, J. Wang, L. Yang, S. K. So, A. B. Djurišić, W. W.-f. Leung, J. Hao, W. K. Chan and C. Surya, J. Mater. Chem. A, 2015, 3, 9223–9231.
3. K.-L. Wu, A. Kogo, N. Sakai, M. Ikegami and T. Miyasaka, Chem. Lett., 2015, 44, 321–323.
4. T. Leijtens, G. E. Eperon, S. Pathak, A. Abate, M. M. Lee and H. J. Snaith, Nat. Commun., 2013, 4, 2885.
5. Y. Han, S. Meyer, Y. Dkhissi, K. Weber, J. M. Pringle, U. Bach, L. Spiccia and Y.-B. Cheng, J. Mater. Chem. A, 2015, 3, 8139–8147.
6. Y. H. Lee, J. Luo, R. Humphry-Baker, P. Gao, M. Grätzel and M. K. Nazeeruddin, Adv. Funct. Mater., 2015, 25, 3925–3933.
7. H. Zhou, Q. Chen, G. Li, S. Luo, T. B. Song, H. S. Duan, Z. Hong, J. You, Y. Liu and Y. Yang, Science, 2014, 345, 542–546.
8. S. Lv, S. Pang, Y. Zhou, N. P. Padture, H. Hu, L. Wang, X. Zhou, H. Zhu, L. Zhang, C. Huang and G. Cui, Phys. Chem. Chem. Phys., 2014, 16, 19206–19211.
9. F. X. Xie, D. Zhang, H. Su, X. Ren, K. S. Wong, M. Gratzel and W. C. H. Choy, ACS Nano, 2015, 9, 639–646.
10. Y. Zhou, M. Yang, W. Wu, A. L. Vasiliev, K. Zhu and N. P. Padture, J. Mater. Chem. A, 2015, 3, 8178–8184.
11. W. Nie, H. Tsai, R. Asadpour, J. C. Blancon, A. J. Neukirch, G. Gupta, J. J. Crochet, M. Chhowalla, S. Tretiak, M. A. Alam, H. L. Wang and A. D. Mohite, Science, 2015, 347, 522–525.
12. Q. Chen, H. Zhou, T. B. Song, S. Luo, Z. Hong, H. S. Duan, L. Dou, Y. Liu and Y. Yang, Nano Lett., 2014, 14, 4158–4163.
13. G. E. Eperon, V. M. Burlakov, P. Docampo, A. Goriely and H. J. Snaith, Adv. Funct. Mater., 2014, 24, 151–157.
14. J. Chen, T. Shi, X. Li, B. Zhou, H. Cao and Y. Wang, Appl. Phys. Lett., 2016, 108, 053302.
15. G. E. Eperon, S. D. Stranks, C. Menelaou, M. B. Johnston, L. M. Herz and H. J. Snaith, Energy Environ. Sci., 2014, 7, 982–988.
16. P. Docampo, F. C. Hanusch, S. D. Stranks, M. Döblinger, J. M. Feckl, M. Ehrensperger, N. K. Minar, M. B. Johnston, H. J. Snaith and T. Bein, Adv. Energy Mater., 2014, 4, 1400355.
17. Y. C. Kim, N. J. Jeon, J. H. Noh, W. S. Yang, J. Seo, J. S. Yun, A. Ho-Baillie, S. Huang, M. A. Green, J. Seidel, T. K. Ahn and S. I. Seok, Adv. Energy Mater., 2016, 6, 1502104.
18. J. Liu, S. Pathak, T. Stergiopoulos, T. Leijtens, K. Wojciechowski, S. Schumann, N. Kausch-Busies and H. J. Snaith, J. Phys. Chem. Lett., 2015, 6, 1666–1673.
19. C. Roldán-Carmona, P. Gratia, I. Zimmermann, G. Grancini, P. Gao, M. Graetzel and M. K. Nazeeruddin, Energy Environ. Sci., 2015, 8, 3550–3556.
20. D. H. Cao, C. C. Stoumpos, C. D. Malliakas, M. J. Katz, O. K. Farha, J. T. Hupp and M. G. Kanatzidis, APL Mater., 2014, 2, 091101.
21. J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Gratzel, Nature, 2013, 499, 316–319.
22. W. A. Laban and L. Etgar, Energy Environ. Sci., 2013, 6, 3249–3253.
23. T. J. Jacobsson, J. P. Correa-Baena, E. Halvani Anaraki, B. Philippe, S. D. Stranks, M. E. Bouduban, W. Tress, K. Schenk, J. Teuscher, J. E. Moser, H. Rensmo and A. Hagfeldt, J. Am. Chem. Soc., 2016, 138, 10331–10343.
24. J. Xiong, B. Yang, R. Wu, C. Cao, Y. Huang, C. Liu, Z. Hu, H. Huang, Y. Gao and J. Yang, Org. Electron., 2015, 24, 106–112.
25. T. P. Gujar and M. Thelakkat, Energy Technol., 2016, 4, 449–457.
26. Y. Li, J. K. Cooper, R. Buonsanti, C. Giannini, Y. Liu, F. M. Toma and I. D. Sharp, J. Phys. Chem. Lett., 2015, 6, 493–499.
27. F. Liu, Q. Dong, M. K. Wong, A. B. Djurišić, A. Ng, Z. Ren, Q. Shen, C. Surya, W. K. Chan, J. Wang, A. M. C. Ng, C. Liao, H. Li, K. Shih, C. Wei, H. Su and J. Dai, Adv. Energy Mater., 2016, 6, 1502206.
28. F. Panzer, S. Baderschneider, T. P. Gujar, T. Unger, S. Bagnich, M. Jakoby, H. Bässler, S. Hüttner, J. Köhler, R. Moos, M. Thelakkat, R. Hildner and A. Köhler, Adv. Opt. Mater., 2016, 4, 917–928.
29. P. Docampo, J. M. Ball, M. Darwich, G. E. Eperon and H. J. Snaith, Nat. Commun., 2013, 4, 2761.
30. S. D. Stranks, V. M. Burlakov, T. Leijtens, J. M. Ball, A. Goriely and H. J. Snaith, Phys. Rev. Appl., 2014, 2, 034007.
31. H. Xie, X. Liu, L. Lyu, D. Niu, Q. Wang, J. Huang and Y. Gao, J. Phys. Chem. C, 2016, 120, 215–220.
32. C. Wehrenfennig, M. Liu, H. J. Snaith, M. B. Johnston and L. M. Herz, Energy Environ. Sci., 2014, 7, 2269–2275.
33. D. Bi, A. M. El-Zohry, A. Hagfeldt and G. Boschloo, ACS Photonics, 2015, 2, 589–594.
34. J. A. Christians, P. A. Miranda Herrera and P. V. Kamat, J. Am. Chem. Soc., 2015, 137, 1530–1538.
35. G. Niu, W. Li, F. Meng, L. Wang, H. Dong and Y. Qiu, J. Mater. Chem. A, 2014, 2, 705–710.
36. B. Philippe, B.-W. Park, R. Lindblad, J. Oscarsson, S. Ahmadi, E. M. J. Johansson and H. Rensmo, Chem. Mater., 2015, 27, 1720–1731.
37. J. Yang, B. D. Siempelkamp, D. Liu and T. L. Kelly, ACS Nano, 2015, 9, 1955–1963.
38. A. M. A. Leguy, Y. Hu, M. Campoy-Quiles, M. I. Alonso, O. J. Weber, P. Azarhoosh, M. van Schilfgaarde, M. T. Weller, T. Bein, J. Nelson, P. Docampo and P. R. F. Barnes, Chem. Mater., 2015, 27, 3397–3407.
39. L. Zhang and P. H. L. Sit, J. Phys. Chem. C, 2015, 119, 22370–22378.
40. J. M. Frost, K. T. Butler, F. Brivio, C. H. Hendon, M. van Schilfgaarde and A. Walsh, Nano Lett., 2014, 14, 2584–2590.
41. H. He, Q. Yu, H. Li, J. Li, J. Si, Y. Jin, N. Wang, J. Wang, J. He, X. Wang, Y. Zhang and Z. Ye, Nat. Commun., 2016, 7, 10896.
42. D. Yamashita, T. Handa, T. Ihara, H. Tahara, A. Shimazaki, A. Wakamiya and Y. Kanemitsu, J. Phys. Chem. Lett., 2016, 7, 3186–3191.
43. J. A. Christians, J. S. Manser and P. V. Kamat, J. Phys. Chem. Lett., 2015, 6, 852–857.
44. E. Zimmermann, K. K. Wong, M. Müller, H. Hu, P. Ehrenreich, M. Kohlstädt, U. Würfel, S. Mastroianni, G. Mathiazhagan, A. Hinsch, T. P. Gujar, M. Thelakkat, T. Pfadler and L. Schmidt-Mende, APL Mater., 2016, 4, 091901.
45. R. S. Sanchez, V. Gonzalez-Pedro, J. W. Lee, N. G. Park, Y. S. Kang, I. Mora-Sero and J. Bisquert, J. Phys. Chem. Lett., 2014, 5, 2357–2363.
46. H. S. Kim and N. G. Park, J. Phys. Chem. Lett., 2014, 5, 2927–2934.
47. H. Yu, H. Lu, F. Xie, S. Zhou and N. Zhao, Adv. Funct. Mater., 2016, 26, 1411–1419.
48. M. L. Petrus, J. Schlipf, C. Li, T. P. Gujar, N. Giesbrecht, P. Müller-Buschbaum, M. Thelakkat, T. Bein, S. Hüttner and P. Docampo, Adv. Energy Mater., 2017, 1700264.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7cp04749e

---