Quantitative Study of Electron Beam Damage in Metal Halide Perovskites using Nanobeam Diffraction

Metal halide perovskites have been extensively studied for applications in photovoltaics, optoelectronics, and semiconductors. Nanoscopic defect characterization plays a critical role in developing these electronic systems and devices. Significant efforts have been made to apply transmission electron microscopy (TEM) to metal halide perovskites despite the challenges posed by their extreme beam sensitivity [1, 2]. Understanding beam damage mechanisms is crucial for optimizing characterization conditions and ensuring data fidelity. To this end, systematic studies have examined damage parameters, including acceleration voltage, electron dose rate, temperature, and surface coating [1, 2]. Many studies have reached a consensus on degradation pathways, key damage parameters, and reliable analysis conditions while also identifying potential pitfalls that could lead to misinterpretation [3, 4]. Nevertheless, existing reports often show inconsistencies and remain primarily qualitative, largely due to challenges in controlling specimen variables [1, 2]. Since most damage studies have been conducted on a grain-by-grain basis, it has been difficult to control key variables such as thickness, crystallinity, and crystal orientation, which can significantly influence beam sensitivity estimations [5]. In this study, electron beam damage was investigated within single grains of formamidinium lead iodide (FAPbI3) along the <001> zone axis using defocused nanobeam diffraction. Time Series of diffraction patterns were acquired at multiple locations under varying damage parameters (Fig 1a). Thickness variations within a grain were addressed by applying electron energy loss spectroscopy (EELS). To quantitatively assess beam sensitivity, virtual masks were applied to diffraction patterns to distinguish different crystalline phases and amorphous components (Fig 1b). The characteristic doses, Dc (0.75) and Dc (0.5), which are defined as the dose at which the intensity of the pure perovskite peak decreases to 75% and 50% of its original value respectively, were measured under various parameters such as dose rate, specimen thickness, acceleration voltage, and temperature. This approach enables a quantitative evaluation of perovskite structural degradation and phase transitions under electron beam exposure. A key finding is the accelerated yet localized beam damage at lower temperatures. Diffraction patterns were monitored at 296 K, 170 K, and 100 K as a function of total dose with a dose rate of 2.2 e/Ų·s (Fig 2). At cryogenic temperatures, amorphization rather than transition into a secondary phase was observed, consistent with previous reports [6]. Notably, our results reveal lower characteristic doses at lower temperatures, indicating that perovskite framework degradation accelerates as temperature decreases (Fig 3). This low-temperature vulnerability may result from suppressed ‘healing’ effects and from limited ion migration which constrains the system to a more fragile degradation pathway, supported by 4D-STEM and EELS analyses, which reveal more localized phase transitions at lower temperatures. By determining characteristic doses across various damage parameters, we identify optimal analysis conditions to minimize beam-induced artifacts. These findings not only provide a quantitative assessment of halide perovskite beam sensitivity but also establish a framework applicable to other beam-sensitive materials, including metal-organic frameworks, molecular systems, polymers, and biological specimens [7]. Beam Damage Experiments Using Nanobeam Diffraction. (a) Schematic of the damage experiment illustrating the irradiation of a defocused nanobeam with the convergence angle of approximately 2 mrad and acquisition of time-series diffraction pattern at multiple spots. A large defocus (tens of μm) was required to reduce the dose rate below 1 e−/sŲ and to minimize drift sensitivity. The specimen used was a FAPbI3 perovskite film directly spin-coated onto a C-film. (b) An experimental diffraction pattern with applied virtual masks is displayed alongside simulated diffraction patterns of FAPbI3 and PbI2 along the <001> and <-441> zone axes, respectively. Four types of virtual masks were used to distinguish phase contributions: (1) ‘PbI2 + Perovskite overlap’, (2) ‘center beam’, (3) ‘pure perovskite’, and (4) ‘background’. The averaged intensities of diffraction spots corresponding to each phase—PbI2, perovskite FAPbI3, and amorphous material—were tracked by applying the respective masks. Temperature Effects on Beam Damage. (a) A high-angle annular dark field (HAADF) scanning transmission electron microscopy (STEM) image of an FAPbI3 grain where nanobeam diffraction (NBD) damage experiments were conducted. Below, thickness map (the value of t/λ, where t is thickness and λ is inelastic mean free path) for two damaged spots at 296 K (left) and 100 K (right) are displayed. These maps were acquired using electron energy loss spectroscopy (EELS) at room temperature. (b) NBD patterns recorded over time at three different temperatures (296 K, 170 K, and 100 K) with a dose rate (DR) of 2.2 e−/Ų·s are shown (top), along with corresponding accumulated electron dose values. Each diffraction dataset was acquired at the location marked by an arrow in the HAADF-STEM image in (a). Two distinct initial crystalline phases were observed: cubic FAPbI3 at 296 K and orthorhombic FAPbI3 at 100 K. Scale bar = 2 nm-1. Quantification of temperature effects on beam damage. (a) The integrated intensity of the diffraction mask is plotted as a function of the total electron dose. Specifically, the intensity difference between the perovskite and background is plotted (top) serving as an estimate of beam-induced damage. Additionally, the background intensity (measured using a mask that excludes crystalline peaks but may include amorphous contributions) is plotted to quantify amorphization (bottom). (b) The characteristic electron dose (Dc) measured from top in (a) is plotted as a function of temperature, showing that the Dc decreases as temperature decreases.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.