为了快速、 准确得到纳米薄膜厚度, 采用Kiessig厚度干涉条纹计算薄膜厚度的线性拟合公式, 计算了不同系列厚度(10~120 nm)的二氧化硅薄膜。 薄膜样品采用热原子层沉积法(T-ALD)制备, 薄膜厚度使用掠入射X射线反射(GIXRR)技术表征, 基于GIXRR得到的反射率曲线系统讨论了线性拟合公式计算薄膜厚度的步骤及影响因素, 同时使用XRR专业处理软件GlobalFit2.0比较了两种方法得到的膜厚, 最后提出一种计算薄膜厚度的新方法-经验曲线法。 结果表明: 峰位级数对线性拟合厚度产生主要影响, 峰位级数增加, 厚度增大; 峰位对应反射角同样对线性拟合厚度有较大影响, 表现为干涉条纹周期增大, 厚度减小。 但峰位级数及其对应反射角在拟合薄膜厚度过程中引入的误差可进一步通过试差法, 临界角与干涉条纹周期的校准来减小。 对任意厚度的同一样品, 线性拟合和软件拟合两种方法得到的薄膜厚度具有一致性, 厚度偏差均小于0.1 nm, 表明线性拟合方法的准确性。 在厚度准确定值的基础上提出薄膜厚度与干涉条纹周期的经验关系曲线, 通过该曲线, 可直接使用干涉条纹周期计算薄膜厚度, 此方法不仅省略了线性拟合过程中确定峰位级数及其对应反射角的繁琐步骤, 而且避免了软件拟合过程中复杂模型的建立, 对快速、 准确获得薄膜厚度信息具有重要的意义。

To obtain nanometer thin film thickness fastly and accurately, a formula of linear fitting method based on the periodic Kiessig fringes for thickness determination is applied, and a series of SiO2 nanometer films on Si substrate with the film thickness ranging from 10 to 120 nm have been calculated with the formula. These samples are prepared with thermal atomic layer deposition (T-ALD) process and film thickness is measured with grazing incidence X-ray reflection (GIXRR) technique, in addition, the linear fitting procedure and several influencing factors among it are studied, all of the work is based on the reflectivity curve from GIXRR experiment. While at the same time, another fitting method based on a soft named Global Fit2.0 is brought into this study to compare the two obtained thicknesses from two kinds of analysis methods. In the end a novel method for film thickness determination-empirical curve is presented. The results show that: during the linear fitting process, the peak position series have a main effect on thickness determination, thickness will increase when the peak position adds up; Besides, any peak’s corresponding reflection angle also has a significant effect on the thickness determination, it is expressed in the form of interference fringe period, thickness will decrease while the interference fringe period increases, however, the errors from either peak series or fringe period can be further weakened with trial and error method, calibration procedure of critical angle and interference fringe period individually. Choosing the same sample with random thickness, no matter using the linear fitting and soft fitting method, the two gained film thicknesses are consistent and the thickness deviation is less than 0.1 nm, which illustrates the accuracy of linear fitting method for thickness determination. An empirical relationship between film thickness and interference fringe period is then put forward on the foundation of the accurate thickness determination, according this curve, the target film thickness is directly got by putting an interference fringe period in the empirical curve. This novel method not only avoids the messy procedure of choosing peak position series or their corresponding angles during linear fitting process, but also avoids the complex task of building a correct structure for soft fitting process; it is of great significance in confirming thin film thickness with quick speed and high accuracy.