Nečas David, Klapetek Petr, Neu Volker, Havlíček Marek, Puttock Robert, Kazakova Olga, Hu Xiukun, Zajíčková Lenka
Plasma Technologies, CEITEC, Masaryk University, Brno, 62500, Czech Republic.
CEITEC, Brno University of Technology, Brno, 63800, Czech Republic.
Sci Rep. 2019 Mar 7;9(1):3880. doi: 10.1038/s41598-019-40477-x.
Magnetic force microscopy has unsurpassed capabilities in analysis of nanoscale and microscale magnetic samples and devices. Similar to other Scanning Probe Microscopy techniques, quantitative analysis remains a challenge. Despite large theoretical and practical progress in this area, present methods are seldom used due to their complexity and lack of systematic understanding of related uncertainties and recommended best practice. Use of the Tip Transfer Function (TTF) is a key concept in making Magnetic Force Microscopy measurements quantitative. We present a numerical study of several aspects of TTF reconstruction using multilayer samples with perpendicular magnetisation. We address the choice of numerical approach, impact of non-periodicity and windowing, suitable conventions for data normalisation and units, criteria for choice of regularisation parameter and experimental effects observed in real measurements. We present a simple regularisation parameter selection method based on TTF width and verify this approach via numerical experiments. Examples of TTF estimation are shown on both 2D and 3D experimental datasets. We give recommendations on best practices for robust TTF estimation, including the choice of windowing function, measurement strategy and dealing with experimental error sources. A method for synthetic MFM data generation, suitable for large scale numerical experiments is also presented.
磁力显微镜在纳米级和微米级磁性样本及器件分析方面具有无与伦比的能力。与其他扫描探针显微镜技术类似,定量分析仍然是一项挑战。尽管该领域在理论和实践上取得了巨大进展,但由于其复杂性以及对相关不确定性和推荐最佳实践缺乏系统理解,目前的方法很少被使用。使用针尖传递函数(TTF)是使磁力显微镜测量实现定量的关键概念。我们对使用具有垂直磁化的多层样本进行TTF重建的几个方面进行了数值研究。我们讨论了数值方法的选择、非周期性和加窗的影响、数据归一化和单位的合适惯例、正则化参数的选择标准以及在实际测量中观察到的实验效应。我们提出了一种基于TTF宽度的简单正则化参数选择方法,并通过数值实验验证了该方法。在二维和三维实验数据集上展示了TTF估计的示例。我们给出了稳健TTF估计的最佳实践建议,包括加窗函数的选择、测量策略以及处理实验误差源。还提出了一种适用于大规模数值实验的合成MFM数据生成方法。