In this paper, an improved quantification technique for STEM/EDX measurements of 1D dopant profiles based on the Cliff-Lorimer equation is presented. The technique uses an iterative absorption correction procedure based on density models correlating the local mass density and composition of the specimen. Moreover, a calibration and error estimation procedure based on linear regression and error propagation is proposed in order to estimate the total measurement error in the dopant density. The proposed approach is applied to the measurement of the As profile in a nanodevice test structure. For the calibration, two crystalline Si specimens implanted with different As doses have been used, and the calibration of the Cliff-Lorimer coefficients has been carried out using Rutherford Back Scattering measurements. The As profile measurement has been carried out on an FinFET test structure, showing that quantitative results can be obtained in the nanometer scale and for dopant atomic densities lower than 1%. Using the proposed approach, the measurement error and detection limit for our experimental setup are calculated and the possibility to improve this limit by increasing the observation time is discussed.

中文翻译：

---

一种改进的 STEM/EDX 纳米级掺杂剂分析定量方法

在本文中，提出了一种改进的量化技术，用于基于 Cliff-Lorimer 方程的一维掺杂剂分布的 STEM/EDX 测量。该技术使用基于与样本的局部质量密度和成分相关的密度模型的迭代吸收校正程序。此外，提出了一种基于线性回归和误差传播的校准和误差估计程序，以估计掺杂剂密度的总测量误差。所提出的方法适用于测量纳米器件测试结构中的 As 曲线。对于校准，使用了注入不同 As 剂量的两个晶体硅样品，并使用卢瑟福背散射测量进行了 Cliff-Lorimer 系数的校准。在 FinFET 测试结构上进行了 As 轮廓测量，表明可以在纳米级和低于 1% 的掺杂剂原子密度下获得定量结果。使用所提出的方法，计算了我们实验装置的测量误差和检测限，并讨论了通过增加观察时间来改善这一限制的可能性。