Pradeep K G, Herzer G, Raabe D
Max-Planck-Institut für Eisenforschung GmbH, Max-Planck-Str. 1, 40237 Düsseldorf, Germany; Materials Chemistry, RWTH Aachen University, Kopernikusstr. 10, 52074 Aachen, Germany.
Vacuumschmelze GmbH & Co. KG, Grüner Weg 37, 63450 Hanau, Germany.
Ultramicroscopy. 2015 Dec;159 Pt 2:285-91. doi: 10.1016/j.ultramic.2015.04.006. Epub 2015 Apr 10.
A local electrode atom probe has been employed to trace the onset of Cu clustering followed by their coarsening and subsequent growth upon rapid (10s) annealing of an amorphous Fe73.5Si15.5Cu1Nb3B7 alloy. It has been found that the clustering of Cu atoms introduces heterogeneities in the amorphous matrix, leading to the formation of Fe rich regions which crystallizes pseudo-homogeneously into Fe-Si nanocrystals upon annealing. In this paper, we present the data treatment method that allows for the visualization of these different phases and to understand their morphology while still quantifying them in terms of their size, number density and volume fraction. The crystallite size of Fe-Si nanocrystals as estimated from the atom probe data are found to be in good agreement with other complementary techniques like XRD and TEM, emphasizing the importance of this approach towards accurate structural analysis. In addition, a composition driven data segmentation approach has been attempted to determine and distinguish nanocrystalline regions from the remaining amorphous matrix. Such an analysis introduces the possibility of retrieving crystallographic information from extremely fine (2-4 nm sized) nanocrystalline regions of very low volume fraction (< 5 Vol%) thereby providing crucial in-sights into the chemical heterogeneity induced crystallization process of amorphous materials.
采用局部电极原子探针追踪非晶态Fe73.5Si15.5Cu1Nb3B7合金快速(10秒)退火时Cu团簇的起始形成,随后其粗化以及后续的生长过程。已发现Cu原子的团簇在非晶基体中引入了不均匀性,导致形成富Fe区域,该区域在退火时伪均匀地结晶为Fe-Si纳米晶体。在本文中,我们展示了一种数据处理方法,该方法能够可视化这些不同相,并了解它们的形态,同时仍能根据其尺寸、数密度和体积分数对它们进行量化。根据原子探针数据估计的Fe-Si纳米晶体的微晶尺寸与XRD和TEM等其他互补技术的结果吻合良好,强调了这种方法对准确结构分析的重要性。此外,还尝试了一种基于成分驱动的数据分割方法,以从其余非晶基体中确定并区分纳米晶区域。这种分析引入了从极低体积分数(<5体积%)的极细(2 - 4纳米尺寸)纳米晶区域获取晶体学信息的可能性,从而为非晶材料的化学不均匀性诱导结晶过程提供关键见解。