高能电子在穿过物质时对局部结构极为敏感，能够收集大量关于晶格缺陷与应变、电磁性质、化学成分及电子结构的信息。这种灵敏度被应用在透射电子显微镜（TEM）中来研究结构与性质的关系，人们在提高分辨率、加快成像速度和开发新的成像方式等方面也不懈努力。高动态范围直接电子探测器的出现引发了一个真正的范式转变，它不再依赖于将电子转化为光子，使空间分辨衍射成像成为可能，为四维扫描透射电子显微镜（4D-STEM）的高分辨率测量提供了新的机会4D-STEM的一个显著优势在于出色的信息密度，在每个探针位置上都会获得动态散射电子的图像。
然而，由于缺乏能够充分解释多种动态散射过程的经验模型以及变化的信噪比，4D-STEM数据的分析常常受到挑战。近年来，机器学习方法在显微学中的应用呈指数级增长，用于加速各种科学任务，包括实时数据缩减、分割和自动化实验。此外，机器学习还可以用于解开多模态纳米光谱成像中的特征，且具有更好的统计意义。通过精心设计的机器学习架构和定制的正则化策略，可以从多模态成像中以纳米尺度空间分辨率统计地分离和解释功能材料的结构特性。
来自挪威科技大学材料科学与工程系的Ursula Ludacka等，展示了一种在纳米尺度上成像和表征铁电材料的方法。作者应用4D-STEM，研究了单轴铁电氧化物中的铁电畴、畴壁和涡旋结构，利用电子散射同时进行了高分辨成像和局部结构分析。通过将自定义正则化的卷积自编码器应用于在从模型系统Er(Mn,Ti)O3上获得的扫描电子衍射SED数据，统计性地解开了与铁电畴、畴壁的特定结构畸变以及畴壁电荷状态相关的特征，获得了铁电畴、畴壁及其涡旋状交汇点的实空间图像。
Fig. 3 | Structure of the custom CA.
类似于对Er(Mn,Ti)O3数据集进行专门训练，这一模型也可以针对其他系统进行训练和定制。该模型的核心元素，包括其架构、正则化技术和超参数调优方法，能广泛适用于高维成像模式，而不仅限于铁电材料。
Fig. 4 | Comparison of measured and simulated SED diffraction patterns.
该方法提供了一种强大的手段，将纳米级成像与结构解卷积相结合，为改进结构-性质关系的研究、提高测量精度以及自动化科学实验开辟了新的路径。相关论文发表于npj Computational Materials 10: 106 (2024)。
Editorial Summary
High-energy electrons traveling through matter are highly sensitive to the local structure, collecting a multitude of information about lattice defects and strain, electric and magnetic properties, as well as chemical composition and electronic structure. This sensitivity is utilized in transmission electron microscopy (TEM) to study structure-property relations, and there are continuous efforts to increase resolution, enhance imaging speeds, and enable new imaging modalities. A real paradigm shift was triggered by the advent of high dynamic-range direct electron detectors (DED), which no longer rely on converting electrons into photons. DEDs enable spatially resolved diffraction imaging, providing additional opportunities for high-resolution measurements known as four-dimensional scanning transmission electron microscopy (4D-STEM). A significant advantage of 4D-STEM is the outstanding information density; an image of the dynamically scattered electrons is acquired at every probe position. The analysis of 4D-STEM data, however, is often challenged by a lack of empirical models that can fully explain the multitude of dynamic scattering processes, as well as varying signal-to-noise ratios. Recently, exponential increases in the deployment of machine learning methods in microscopy have been applied to accelerate a variety of scientific tasks, including real-time data reduction, segmentation, and automated experiments. Furthermore, they can be used to disentangle features in multimodal nanoscale spectroscopic imaging with improved statistical significance. Through careful design of machine learning architectures and custom regularization strategies, it is now possible to statistically disentangle and interpret structural properties of functional materials with nanoscale spatial resolution from multimodal imaging.
Ursula Ludacka et al from the Department of Materials Science and Engineering, Norwegian University of Science and Technology, demonstrated a powerful pathway for imaging and characterizing ferroelectric materials at the nanoscale. The authors applied 4D-STEM to investigate domains, domain walls, and vortex structures in a uniaxial ferroelectric oxide, utilizing the scattering of electrons for simultaneous high-resolution imaging and local structure analysis. By applying a custom-designed convolutional autoencoder to scanning electron diffraction data gained on the model system Er(Mn,Ti)O3, they statistically disentangled features in the diffraction patterns that correlated with the distinct structural distortions in the ferroelectric domains and domain walls, as well as the domain wall charge state, and readily gained real-space images of ferroelectric domains, domain walls. Analogous to the training specifically performed for the Er(Mn,Ti)O3 datasets, the model could be trained and specifically tailored to other systems. The core elements of the model—including its architecture, regularization techniques, and hyperparameters tuning methods—are broadly applicable to high-dimensional imaging modalities, not only in ferroelectrics. This method provides a powerful method that combines nanoscale imaging and structural deconvolution—opening a pathway towards improved structure-property correlations, increased fidelity, and automated scientific experiments. This article was recently published in npj Computational Materials 10: 106 (2024).
原文Abstract及其翻译
Imaging and structure analysis of ferroelectric domains, domain walls, and vortices by scanning electron diffraction（使用扫描电子衍射对铁电畴、畴壁和涡进行成像和结构分析）
Ursula Ludacka, Jiali He, Shuyu Qin, Manuel Zahn, Emil Frang Christiansen, Kasper A. Hunnestad, Xinqiao Zhang, Zewu Yan, Edith Bourret, István Kézsmárki, Antonius T. J. van Helvoort, Joshua Agar & Dennis Meier
Abstract Direct electron detectors in scanning transmission electron microscopy give unprecedented possibilities for structure analysis at the nanoscale. In electronic and quantum materials, this new capability gives access to, for example, emergent chiral structures and symmetry-breaking distortions that underpin functional properties. Quantifying nanoscale structural features with statistical significance, however, is complicated by the subtleties of dynamic diffraction and coexisting contrast mechanisms, which often results in a low signal-to-noise ratio and the superposition of multiple signals that are challenging to deconvolute. Here we apply scanning electron diffraction to explore local polar distortions in the uniaxial ferroelectric Er(Mn,Ti)O3. Using a custom-designed convolutional autoencoder with bespoke regularization, we demonstrate that subtle variations in the scattering signatures of ferroelectric domains, domain walls, and vortex textures can readily be disentangled with statistical significance and separated from extrinsic contributions due to, e.g., variations in specimen thickness or bending. The work demonstrates a pathway to quantitatively measure symmetry-breaking distortions across large areas, mapping structural changes at interfaces and topological structures with nanoscale spatial resolution.
摘要扫描透射电子显微镜中的直接电子探测器为纳米尺度上的结构分析提供了前所未有的可能性。在电子和量子材料领域，这一新兴技术使得我们获取例如新出现的手性结构和支撑功能特性的对称破缺畸变成为可能。然而，由于动态衍射的微妙性和共存的对比机制，量化纳米级结构特征的统计显著性变得复杂，常常导致信噪比低以及多重信号的叠加，这些信号难以解卷积。在这里，我们应用扫描电子衍射来探索单轴铁电Er(Mn, Ti)O3中的局部极性畸变。通过应用定制的卷积自编码器结合特定设计的正则化方法，我们证明了铁电畴、畴壁及涡旋结构的散射特征的微小变化可以被统计分离，并且可以将其从样品厚度变化或弯曲等外在因素引起的信号中区分出来。本项工作展示了一种定量测量大范围内对称性破缺畸变的方法，并能够以纳米级空间分辨率映射界面和拓扑结构的结构变化，为材料的结构-性能关系研究提供了新的视角。
