Getting to the Heart of the Matter: Control over the Photolysis of PbI₂ Through Partial Lead Substitution

Abstract

A crucial problem of the photoinduced degradation of perovskite semiconductors based on complex lead halides has been addressed here by suppressing PbI₂ photolysis to metallic lead. The systematic screening of >30 modifying cations introduced as substituents for 5% of Pb²⁺ in the PbI₂ composition has revealed their tremendous effects on the rate of material degradation under light exposure. Thus, the most successful stabilizing cations could maintain a high absorbance of the Pb_0.95M_0.1/nI₂ films and block Pb⁰ formation after 400 h of continuous illumination, when the non-modified PbI₂ films completely decomposed to Pb⁰ and I₂. The obtained results present a promising solution for the problem of metallic lead formation in the active layer of perovskite solar cells during their operation, which can pave the way for the development of a new generation of highly efficient and stable perovskite photovoltaics.

1. Introduction

Complex lead halides with the perovskite crystal structure APbX₃, where A is a univalent cation of methylammonium (MA), formamidinium (FA), or cesium, and X is the halide anion (I⁻, Br⁻, or Cl⁻), have attracted tremendous attention from researchers and industry due to their excellent semiconductor properties combined with unprecedented defect tolerance. The perovskite solar cells (PSCs) based on these materials emerged rapidly within the last decade and demonstrated certified power conversion efficiencies exceeding 26% [1]. Furthermore, PSCs have the benefits of the relatively simple and cost-effective technology of their production and abundance of all raw materials, which are the major prerequisites of third-generation photovoltaic (PV) technology [2,3,4,5,6]. However, the low stability of complex lead halides with respect to light, elevated temperatures, and humidity hinders the commercialization of PSCs [7].

The photoinduced degradation of the absorber materials produced at the first stage of binary lead iodide PbI₂ does not impair the photovoltaic performance of PSCs very much [8]. On the contrary, PbI₂ is intentionally introduced into the perovskite absorber layer to improve the performance of PSCs, particularly those assembled using the n-i-p configuration [9]. This is believed to passivate defect states at the grain boundaries of the perovskite absorber material and thus suppresses the non-radiative recombination of charge carriers [10,11,12]. Furthermore, PbI₂ could react back with the volatile organic products formed by photoinduced or thermal decomposition of complex lead halides and thus regenerate the perovskite absorber material [8]. This effect is largely responsible for the self-healing behavior of perovskite solar cells [13]. However, PbI₂ itself undergoes facile photolysis and produces metallic lead Pb⁰ [8,14,15], which induces the massive recombination of charge carriers, leading to rapid decay in PSC performance [16]. Furthermore, another PbI₂ photolysis product, the molecular iodine I₂, is exceptionally corrosive, even in the form of the triiodide anion I₃⁻, and it aggressively reacts with the top electrode material, just forming an insulating metal iodide interlayer and leading finally to PSC device failure [17,18]. Thus, blocking the photolysis of PbI₂ in the absorber layer of perovskite solar cells represents an urgent and crucially important task for achieving the required long-term stability of PSCs.

This challenge has been addressed recently by combining PbI₂ with RbCl in n-i-p PSCs, which resulted in the formation of some complex that was more stable than pristine PbI₂, so such a modification enhanced the operational stability of PSCs [19]. Alternatively, some organic additives were employed to improve the resistance of PbI₂ to light, e.g., semicarbazide hydrochloride [20], 2-thiophenecarboxamide [21], or tryptophan [22]. However, similar reports are very scarce, and the performance of the proposed PbI₂ stabilization approaches has not been verified directly in intrinsic photostability tests.

Herein, we present an alternative approach to control the photolysis of PbI₂ through the partial substitution of Pb²⁺ cations in their composition by other rationally chosen metal ions. We have demonstrated previously that the photostability of all inorganic CsPbI₃ and hybrid MAPbI₃ could substantially be enhanced due to the introduction of certain substituent metal ions, which were found to be different for different material compositions [23,24]. We observed that in some cases, the formation of metallic lead was considerably or even completely suppressed, which inspired us to search for material compositions that would solve the problem of PbI₂ photolysis.

2. Results and Discussion

Altogether, we screened >30 different substituent metal ions in thin films of Pb_0.95M_0.1/nI₂ (where Mⁿ⁺ is the substituent cation) compositions. It should be emphasized that the vast majority of the introduced modifying Mⁿ⁺ cations could not occupy Pb²⁺ positions in the crystal lattice or can only be included in the perovskite lattice to a low extent due to the ion size mismatch and/or for other reasons; therefore, the corresponding MI_n phase is localized mostly on the surface of PbI₂ grains or at the grain boundaries. The localization of the introduced Mⁿ⁺ cations has not been investigated thoroughly in this work since we were focused more on achieving the practically important result, i.e., enhancing the photostability of PbI₂. However, interested readers can refer to our recent study of lead substitution in MAPbI₃, where the substituent ion localization has been analyzed in detail [24]. In the case of the structural replacement of Pb²⁺ in the PbI₂ lattice at low substitution levels (about 1%), when the cations have an oxidation state of 1+ (M⁺), the electroneutrality can be achieved by the local compensation of the negative charge by positively charged vacancies of I⁻ or by h⁺ capture, and in the case of M^3+/4+, it is possible by interstitial I⁻ or e⁻ capture. At high substitution levels (>1%), when the oxidation state of the substituting cation is +3, the electroneutrality can be preserved by replacing Pb²⁺ with M³⁺ due to the formation of Pb²⁺ vacancies (V_Pb²⁻). In addition, there are previous works showing that layered crystalline systems (PbI₂)_1−x(MI_n)_x, where n = 1–3, often represent non-mixing alternating individual nanoclusters or layers enriched with lead iodide and metal iodide, which can impart their tunable semiconductor properties [25,26,27].

The intrinsic photostability of the Pb_0.95M_0.1/nI₂ films was investigated in an inert atmosphere of pure nitrogen (O₂ and H₂O levels below 0.1 ppm) under exposure to the white light provided by an LED array with an incident intensity of 85 mW/cm². However, considering the spectral irradiance of the light source and the absorption spectrum of PbI₂, the number of the photons absorbed by the samples under the used test conditions corresponds to the value obtained under the ~150 mW/cm² irradiation with the true AM1.5G solar spectrum. The equilibrium sample temperature was quite low (32 °C), so the heat did not play a major role in accelerating the PbI₂ photolysis reaction [28].

The photoinduced aging of the Pb_0.95M_0.1/nI₂ films was monitored by periodically measuring the UV-Vis optical spectra of the films (Figures S1–S8, Electronic Supplementary Materials (ESMs)). The films of pristine PbI₂ undergo rapid photobleaching, as shown in Figure 1a. However, some of the material formulations demonstrated much superior resistance to light, as illustrated by the behavior of Pb_0.95Fe_0.05I₂ in Figure 1b. A wavelength of 360 nm was used to analyze the time dynamics of the photolysis reaction by plotting the normalized absorbance as a function of the aging time t (see below). The values of the normalized absorbance of Pb_0.95M_0.1/nI₂ films obtained after 400 h of aging (A₄₀₀/A₀) are presented in Figure 1c for each of the studied metal cations. Interestingly, all the studied substituents, except for Hg²⁺, improve the photostability of the PbI₂ films. The highest photostability was revealed for the films modified with Bi³⁺, Eu^2+/3+, Pt²⁺, Zn²⁺, Fe²⁺, Co²⁺, and Ag⁺ cations. The analysis of the phase composition of the aged Pb_0.95M_0.1/nI₂ films using X-ray diffraction (XRD) revealed the absence of the crystalline phase of metallic lead (Pb⁰) in all the systems that were found to be the most stable according to the UV-Vis spectroscopy data (Figure 1d, Figures S9–S11, ESMs). It is worth mentioning that the stabilization effect of Ag⁺ introduced in the form of AgI is very surprising since silver iodide is known for its exceptionally high light sensitivity and is widely exploited in the film photography industry.

We have analyzed the dynamics of the photoinduced decomposition of Pb_0.95M_0.1/nI₂ films and plotted the normalized absorbance of the films A_t/A₀ as a function of the aging time, as shown in Figure S12 (ESMs), for all studied systems and in Figure 2a for a selected group of cations. One could notice that the thin films of the reference PbI₂ and Pb_0.95Hg_0.05I₂ degrade considerably more rapidly than other material formulations presented in Figure 2a. It has been found that their aging behavior is well described by the third-order reaction kinetics, as can be concluded from the linear dependence of (A_t/A₀)⁻² vs. aging time (Figure 2d).

Similar behavior among all the studied systems was found only for Pb_0.95Y_0.1/3I₂, but this material shows much slower degradation dynamics manifested in a smaller slope of the fitted curve in Figure 2d and, hence, a lower rate constant as compared to that of PbI₂ and Pb_0.95Hg_0.05I₂. The formal third order of the photolysis reaction points to its complicated mechanism and suggests that the aging products catalyze the degradation of the starting material. Such an observation has been made previously for the photoinduced degradation of MAPbI₃ films and was supported by several microscopy techniques [14]. The second-order reaction kinetics have been observed for the samples modified by Yb²⁺, In³⁺, and Lu³⁺ (Figure 2c). Multiple systems, including the most stable materials incorporating Eu^2+/3+ and Bi³⁺ cations, have demonstrated first-order reaction kinetics (Figure 2b). For other studied systems, the aging kinetics cannot be fitted with any reasonable accuracy with such simple equations, so the reaction order represents some fractional number.

The obtained results demonstrated the tremendous effects of the modifying metal cations on the photolysis rate of the Pb_0.95M_0.1/nI₂ films. This is clearly seen in the absorption dynamics shown in Figure 2a (and Figure S12) and also in the very different slopes of the kinetic curves presented, for example, in Figure 2b. Thus, the introduced substituent cations can block some of the elementary acts of PbI₂ photolysis and thus decelerate this undesired process. Even though we cannot study the elementary acts of the PbI₂ photolysis reaction with the available experimental techniques, we could speculate about some of the possible action mechanisms for the most efficient PbI₂ stabilizers: Eu^2+/3+ and Bi³⁺ cations. In the case of europium, it has the unique Eu²⁺/Eu³⁺ redox potential, matching that of Pb²⁺/Pb⁰ in the perovskite precursor solution, so mixing pure Eu²⁺ results in its partial oxidation to Eu³⁺ accompanied by the Pb²⁺ reduction to Pb⁰. However, all these processes are reversible, so Eu²⁺/Eu³⁺ acts as a redox shuttle, promoting the recombination of the photogenerated Pb⁰ and I₂ (I₃⁻) and extending the lifetime of the perovskite films [29,30]. We believe that similar redox chemistry involving the Eu²⁺/Eu³⁺ pair is responsible for the observed stabilization of PbI₂ under light exposure.

Another stabilizing cation, Bi³⁺, is known for its ability to trap molecular iodine and form polyiodides [31], which leads, as we believe, to the realization of a very similar stabilization pathway as Eu²⁺/Eu³⁺. Thus, the photolysis of non-modified PbI₂ produces molecular iodine that evaporates from the aged film to the environment, thus making the reaction irreversible. On the contrary, in the modified films, the photogenerated I₂ could efficiently be trapped either by Eu²⁺ (the redox reaction produces Eu³⁺ and I⁻) or by BiI₃ (the reaction produces polyiodide complexes), so the formed species stay in the film and have enough time to react back with Pb⁰, reforming Pb²⁺ and thus regenerating PbI₂.

3. Materials and Methods

3.1. Materials

Borosilicate glass slides (25 × 25 mm) used as substrates were purchased from Isolab GmbH, (Eschau, Germany). Anhydrous dimethyl formamide (DMF) used as solvent was purchased from Sigma-Aldrich and used as received inside nitrogen glove boxes. The following anhydrous reagents and solvents were purchased from Sigma-Aldrich (St. Louis, MO, USA): PbI₂, CaI₂, SrI₂, BaI₂, PtI₂, EuI₂, SnI₄, SnI₂, MnI₂, FeI₂, CoI₂, CuI, AgI, ZnI₂, HgI₂, CdI₂, and SbI₃ (purity 99.999%); YI₃, NdI₃, DyI₂, YbI₂, and LuI₃ (purity 99.9%); GeI₂ (purity 99.8%); MgI₂, BiI₃, and InI₃ (purity 99.998%); and NiI₂, LaI₃, CeI₃, ErI₃, TbI₃, and GdI₃ (purity 99.99%).

3.2. Preparation of the PbI₂ and Pb_0.95M_0.1/nI₂ Films

3.2.1. Substrate Preparation

The glass substrates were cleaned sequentially with deionized water, acetone, and toluene followed by the RF air plasma treatment for 5 min. All further manipulations were performed inside nitrogen glove boxes under a well-controlled atmosphere.

3.2.2. Preparation of Precursor Solutions Based on Lead Iodide

The Pb_0.95M_0.1/nI₂ precursor solutions with a concentration of 0.3 M were obtained by dissolving together 0.285 mmol of PbI₂ and 0.3/n mmol of MI_n in 1 mL of anhydrous DMF under continuous stirring at room temperature.

3.2.3. Deposition of the Films Based on Lead Iodide

The 45 µL aliquots of the corresponding precursor solutions were dropped on glass substrates rotating at 3000 rpm. The obtained films were annealed at 90 °C for 30 min in an inert atmosphere of the glove box to ensure the elimination of the solvent molecules and the proper crystallization of the modified PbI₂ film.

3.3. Photostability Testing

The aging tests were carried out under continuous illumination using a special setup placed inside MBraun gloveboxes with an inert nitrogen atmosphere (O₂ and H₂O < 1 ppm). White light with a power of 85 ± 10 mW/cm² was provided by LED arrays. The temperature of the samples was within the range of 32 ± 2 °C.

3.4. Pb_0.95M_0.1/nI₂ Film Characterization

The UV-Vis absorption spectra were obtained using an AvaSpec-2048-2 UV-VIS fiber spectrometer integrated inside a glove box (Advantest corporation, Tokyo, Japan). The X-ray diffraction (XRD) patterns were collected using an Aeris instrument (Malvern PANalytical B.V., Malvern, UK) with the CuKα source.

4. Conclusions

To summarize, we have explored, for the first time, the potential of ion substitution chemistry as a simple but efficient approach to control the photoinduced decomposition of PbI₂ films. The obtained results demonstrated a tremendous enhancement of PbI₂ stability due to the introduction of optimum substituent ions such as Bi³⁺, Eu^2+/3+, Pt²⁺, Zn²⁺, Fe²⁺, Co²⁺, and Ag+ cations. The modified films maintained high absorbance and showed no metallic lead formation detectable by XRD after 400 h of continuous illumination, while pristine non-modified PbI₂ completely decomposed under these conditions. The obtained results are crucially important in the context of improving the photostability of the lead halide perovskite absorber materials and pave the way to the development of perovskite solar cells with long operational lifetimes.