Probing fractal abnormal grain growth at the atomistic level using APT

虽然目前的模型在解释常规多晶体中的正常晶粒生长方面相当成功，但我们对纳米晶（NC）材料中的晶粒生长的理解仍然相当初级。在纳米尺度下，粗化在本质上通常是异常的，其中少数晶粒比周围NC基质中的相邻晶粒大几个数量级。在由惰性气体冷凝产生的NC Pd 90 Au 10样品中，显微结构粗化是双重异常的，因为快速生长的晶粒向基体中发出“触角”，首先包围附近的晶粒，然后消耗它们。所得到的晶粒的周长可以用大约1.2的分形维数来很好地描述（而不是期望的接近1的值）。广泛接受的基于曲率的晶界（GB）迁移机制与这种“分形”生长不相容，但即使是“普通的”异常生长--如在Fe-Si钢等工业合金中观察到的--仍然是一个未解之谜。Hwang等人最近提出的一项建议，然而，是第一个在“固态润湿”的单一机制下将这些不同形式的异常生长subsidiary。在这幅图中，GB的分形迁移是润湿的极端表现，使得NC Pd Au成为验证概念的理想模型系统。没有Au，NC Pd中的晶粒生长正常进行;因此，我们推测，分形GB迁移的结果来自与NC Pd Au中Au向GB偏析相关的热力学和/或动力学效应。这些效应可以抑制基体晶粒的曲率驱动的生长，同时促进（或至少不阻碍）固态润湿。检验这一假设的关键步骤是测量分形生长晶粒附近的Au浓度，比较分形/基体和基体/基体晶界处的Au偏析量。原子探针层析成像（APT）是注定要完成这一任务，能够量化的局部成分在一个3D的空间分辨率相媲美的原子interatomic spacing.In这个项目中，Au原子的浓度在NC Pd Au将被改变，以最大限度地提高分形维数。为了从GB迁移的动力学方面去卷积热力学，将向样品中添加具有已知的分离成GB的趋势的三元元素（氢）。FIB研磨将用于提取感兴趣的GB，其将通过APT表征并与各种形式的电子显微镜相关联。Au偏析的测量值将输入到相场模型中，用于模拟同时的固态润湿和曲率驱动的生长。模拟晶粒的形态将与实验进行验证，并结合Au和H偏析的局部研究，以确定固态润湿是否是NC Pd Au中分形异常晶粒生长的起源。

Although current models are reasonably successful at explaining normal grain growth in conventional polycrystals, our understanding of grain growth in nanocrystalline (NC) materials is still quite rudimentary. At the nanoscale, coarsening is often abnormal in nature, with a few grains growing orders of magnitude larger than neighboring grains in the surrounding NC matrix. In samples of NC Pd90Au10 produced by inert gas condensation, microstructural coarsening is doubly abnormal, as the rapidly growing grains send forth “tentacles” into the matrix, first encircling nearby grains and then consuming them. The perimeters of the resulting grains are well described by a fractal dimension of about 1.2 (rather than the expected value close to unity).The widely accepted curvature-based mechanism for grain boundary (GB) migration is incompatible with such “fractal” growth, but even “ordinary” abnormal growth—as observed in technical alloys like Fe-Si steels—remains an unsolved mystery. A recent proposal by Hwang et al., however, is the first to subsume these disparate forms of abnormal growth under the single mechanism of “solid-state wetting.” In this picture, the fractal migration of GBs is an extreme manifestation of wetting, rendering NC PdAu an ideal model system for validating the concept.Without Au, grain growth in NC Pd proceeds normally; therefore, we speculate that fractal GB migration results from thermodynamic and/or kinetic effects associated with Au segregation to GBs in NC PdAu. These effects may suppress curvature-driven growth of the matrix grains while, at the same time, promoting (or at least not hindering) solid-state wetting. The key step in testing this hypothesis is to measure the Au concentration in the vicinity of fractally growing grains, comparing the amounts of Au segregation at fractal/matrix and matrix/matrix GBs. Atom-probe tomography (APT) is predestined for this task, capable of quantifying local compositions in a 3D specimen at a spatial resolution comparable to that of the interatomic spacing.In this project, the concentration of Au atoms in NC PdAu will be varied to maximize the fractal dimension. To deconvolute thermodynamic from kinetic aspects of GB migration, a ternary element (hydrogen) having a known tendency to segregate to GBs will be added to the samples. FIB milling will be employed to extract GBs of interest, which will be characterized by APT and correlated with various forms of electron microscopy. Measured values for Au segregation will be input into a phase field model for simulating simultaneous solid-state wetting and curvature-driven growth. The morphology of the simulated grains will be validated against experiment and combined with the local studies of Au and H segregation to determine whether solid-state wetting is the origin of fractal abnormal grain growth in NC PdAu.