Direction-selective electron beam damage to CH₃NH₃PbI₃ based on crystallographic anisotropy

Since the first application of methylammonium lead iodide (MAPbI₃, MA = CH₃NH₃) as the light absorber in solar cell,¹⁾ organometal halide perovskite solar cells (PSCs) have shown remarkable accomplishments in power conversion efficiency (PCE)^2–9) during a short period of time. Consequently, they have achieved a certified record of 25.2%¹⁰⁾ surpassing records of the former solar cells such as copper indium gallium selenide (CIGS) solar cells. The outstanding performance of PSCs is attributed to the peculiar characteristics of the organometal halide perovskite (OHP) such as long charge carrier diffusion length,^11–13) large absorption coefficient,^14,15) low recombination rate.¹⁶⁾ On the basis of the unprecedented achievements of the PSCs, various structural analyses of the OHP have been conducted and specific structural characteristics of the OHP has been disclosed through various analysis techniques such as X-ray diffraction (XRD).^17–19) Transmission electron microscopy (TEM) also has been used to study the OHP and reported various important features.^20–23) However, TEM analysis accompanies an issue of electron beam damage which influences the structure and the chemistry of the target materials. Unfortunately, the electron beam damage is a considerably negative factor for the microscale investigation of the OHP using TEM observation, because the OHP is vulnerable to external stimuli such as moisture, oxygen, UV light.^24–28) For this reason, it is controversial whether TEM analysis using a strong electron beam is a suitable analysis approach for the OHP.²⁹⁾

When an electron beam of the TEM system is irradiated on target materials, every material is damaged by the irradiated electron beam. Because, however, the degree of electron beam damage to the target materials is strongly variable depending on the condition of the interaction between the target material and the irradiated electron beam, it is critical to understand the major types of electron beam damages on the target material and choose the proper TEM condition which is able to minimize the electron beam damage for reliable TEM observation. To identify the types of the electron beam damage and minimize their negative effects during TEM observation, we have to examine the three principal types of electron beam damage; (1) radiolysis, the damage breaking chemical bonds, (2) knock-on damage, the phenomena inducing point defects in the crystal lattice by displacing atoms from their original site to another site, and (3) heat damage.³⁰⁾ In addition, it should be considered that the electron beam travels in a TEM specimen unidirectionally, when the electron beam penetrates the TEM specimen. Because of its unidirectional behavior, the electron beam interacts with the TEM specimen in an anisotropic way different from isotropic stimuli such as moisture and oxygen. On the basis of the anisotropic characteristics of the electron beam, we can infer that the electron beam damage is direction-selective and that the direction-selectivity should be considered together with the aforementioned three principal electron beam damages when anisotropic compound materials are investigated. Therefore, in the case of the OHPs composed of an anisotropic mixture of organic and inorganic materials, it is obvious that investigation of the direction-selectivity of the electron beam damage in the OHP is necessary to enhance the reliability of their TEM analysis controversially due to its vulnerable characteristics from the external stimuli.

In this letter, we will treat the electron beam damage of the OHP as strongly concerned with the relation between the crystallographic direction of the OHP and the irradiation direction of the electron beam. According to the results, it was identified that the organic part of MAPbI₃ is easily decomposed when the electron beam was irradiated to specific directions and confirmed that the selective decomposition phenomenon is a result of the direction-selective behavior of the electron beam damage.

To investigate the direction-selectivity of the electron beam damage in the OHP, we conducted a TEM observation with a MAPbI₃ single crystal. To obtain the MAPbI₃ single crystal, the inverse temperature crystallization method³¹⁾ was used. 10 mM solution of the equimolar mixture of CH₃NH₃I (95%, Showa chemical industry Co) and PbI₂ (98%, Kanto chemical Co) was dissolved in N,N-dimethylformamide], and heated in an oil bath for 3 h. During the heat process, the solution was sealed. The TEM observation of the MAPbI₃ single crystal was conducted using 300 keV TEM (Hitachi HF-3300). A TEM specimen of the MAPbI₃ single crystal was prepared via the focused ion beam (FIB) technique (Hitachi NB5000) and thinned to less than 100 nm thick. The diameter of the MAPbI₃ single crystal used for the TEM specimen was ∼2 mm and, to irradiate the electron beam along specific crystal direction, the orientation of the MAPbI₃ TEM specimen was investigated via electronic beam diffraction analysis [inset of Fig. 1(a)]. Details of the TEM observation procedure are described in the previous report.²²⁾ The whole TEM analyses were executed based on the tetragonal structure and all the crystallographic directions of this report are in the tetragonal crystal system.

Zoom In Zoom Out Reset image size

As the first step of the TEM analysis for the direction-selectivity of the electron beam damage in the OHP, we conducted TEM observations at the [001] zone axis (corresponding to the [$00\bar{1}$] electron beam irradiation) with the MAPbI₃ single crystal to investigate the electron beam damage along [$00\bar{1}$] and Fig. 1 is the result. The inset of Fig. 1(a) is the [001] zone axis electron diffraction pattern³²⁾ obtained from the MAPbI₃ area of Fig. 1 and shows that the electron beam irradiation was well conducted on MAPbI₃ along the [$00\bar{1}$] as is intended. As is seen in Fig. 1(a), bubble-like flaws in the red solid circle are well identified and they distinguish the damaged region from the non-damaged region (outside of the red solid circle) in the MAPbI₃. The bubble-like flaws on the MAPbI₃ were occurred by a focused electron beam irradiation along [$00\bar{1}$] for high-resolution (HR) TEM observation. Figures 1(b) and 1(c) are magnified TEM images of the black dot rectangle in Fig. 1(a) and clearly demonstrate the emergence of the bubble-like flaws before [Fig. 1(b)] and after [Fig. 1(c)] the focused electron beam irradiation, respectively.

To understand the progress of the bubble-like flaws, we investigated a real-time TEM observation obtained at the black dot rectangle region of Fig. 1(a). Figure 2 is the result. Figure 2 consists of snapshots obtained every 10 s from the real-time TEM observation and they demonstrate the progress of the electron beam damage time-sequentially from 0 s to 50 s. As is seen in the sequence of snapshots, it is identified that the bubble-like flaws in the damaged region are randomly generated from around 10 s electron beam irradiation and that they are growing up as time goes by (colored arrows in Fig. 2). It is worthy to note that the bubble-like flaw is observed when the surfactant materials including organic components are damaged by electron beam during TEM observation and that such kinds phenomenon does not emerge in TEM observation of inorganic materials.³⁰⁾ On the basis of the above results, it is confirmed that the electron beam damage caused by the [$00\bar{1}$] electron beam irradiation starts relatively early timing, around 10 s, and inferred that the electron beam damage along the [$00\bar{1}$] electron beam irradiation is a result of the decomposition of the organic components of the MAPbI₃ based on the emergence of the bubble-like flaws.

Zoom In Zoom Out Reset image size

In the case of the electron beam irradiation along [$\bar{1}1\bar{1}$] in MAPbI₃, we reported details of the electron beam damage at the previous microstructural study for the OHP.²²⁾ In the previous study, a focused electron beam was also irradiated on the MAPbI₃ for the HR TEM observation and zone axis was [$1\bar{1}1$] (corresponding to [$\bar{1}1\bar{1}$] electron beam irradiation). According to the previous study, different from the aforementioned [$00\bar{1}$] electron beam irradiation case (Figs. 1 and 2), no three principal types of the electron beam damage were identified by around 1 min in the microstructural HR TEM observation along [$\bar{1}1\bar{1}$]. The direction-selective phenomenon clearly shows that the electron beam damage in the MAPbI₃ easily occurs along the [$00\bar{1}$] than the [$\bar{1}1\bar{1}$] and directly indicates that crystallographic analysis is necessary to understand the details of the electron beam damage in the OHP. For more information on the electron beam damage in MAPbI₃ caused by the [$\bar{1}1\bar{1}$] electron beam irradiation, the reader refers to Ref. 22.

In order to understand the direction-selectivity of the electron beam damage in the OHP in the view of crystallographic analysis, we simulated and visualized the atomic structure of MAPbI₃. Figures 3(a) and 3(b) are the visualized atomic configurations of MAPbI₃ along [$00\bar{1}$] and [$\bar{1}1\bar{1}$] ([001] and [$1\bar{1}1$] zone axes). Interestingly, the two atomic configurations demonstrated in Figs. 3(a) and 3(b) are a totally different situation for the organic component CH₃NH₃ when the anisotropic or the unidirectional stimulus such as the electron beam irradiation is applied. In Fig. 3(a), we can confirm that the path of the electron beam along [$00\bar{1}$] is freely open to CH₃NH₃ and, thus, we can infer that the organic component of MAPbI₃ is easily attacked and decomposed by the electron beam. In contrast, in the case of electron beam irradiation along [$\bar{1}1\bar{1}$] [Fig. 3(b)], since the path of the [$\bar{1}1\bar{1}$] electron beam is directly blocked by the I atom, and it is difficult for decomposition induced from the chemical bonding breakage of CH₃NH₃ caused by the radiolysis to happen. In addition, it is also hard for atomic displacement inducing the knock-on damage to occur, because the I atom is a considerably heavy atom not displaced by even a 400 keV electron beam.³⁰⁾ Figures 3(c) and 3(d) are the atomic configurations obtained by 90° rotating the yellow cuboids of Figs. 3(a) and 3(b) around [100] and [$\bar{1}12$] axes, respectively, and show the cases of the electron beam path colliding with CH₃NH₃ directly [Fig. 3(c)] and blocked by the I atom [Fig. 3(d)]. On the basis of the crystallographic discussion about the direction-selectivity of the electron beam damage, we can conclude that it is easier for radiolysis and knock-on damage to occur along the [$00\bar{1}$] than along the [$\bar{1}1\bar{1}$] and, surprisingly, the theoretical conclusion is well consistent with the aforementioned our experimental observations.

Zoom In Zoom Out Reset image size

In summary, we investigated the direction-selectivity of the electron beam damage in the OHP by comparing the [$00\bar{1}$] and the [$\bar{1}1\bar{1}$] electron beam irradiation on MAPbI₃ via the TEM observation and the crystallographic discussion. We confirmed that the CH₃NH₃ of the MAPbI₃ is quickly decomposed by the [$00\bar{1}$] electron beam irradiation different from the case of the [$\bar{1}1\bar{1}$] electron beam irradiation and that the direction-selective decomposition phenomenon is well-matched with the crystallographic study of MAPbI₃. This report about the direction-selective damage in the OHP is expected to enhance the understanding of the degradation of the OHP caused by various anisotropic stimuli and to play an important role in the commercialization of PSCs by suggesting a solution.

Acknowledgments