X-ray Diffraction Spectral Denoising Based on Empirical Mode Decomposition and t-Test

X-ray diffraction (XRD) has been widely used in the field of analytical chemistry because it can quickly analyze the composition of materials and the structure and morphology of atoms or molecules inside materials. However, due to the influence of instrument vibration, electromagnetic interference and other factors, the XRD spectrum measured by X-ray diffractometer is extremely noisy. Therefore, in this study, empirical mode decomposition (EMD) combined with t-test was introduced for XRD spectral denoising. Firstly, the XRD spectrum was decomposed by EMD to obtain a series of intrinsic mode function (IMF) components. The high frequency components represented noise and the low frequency components represented useful information. However, sometimes the noise was indistinguishable from useful information. Therefore, the statistical t-test method was introduced in this study to determine the significant difference between the mean of IMFs and zero. Finally, the components with no significant difference were deleted, and the components with significant difference were reconstructed to obtain denosied XRD spectrum. The feasibility of this method was verified by a simulated XRD spectrum and two measured XRD spectrum. The results showed that EMD combined with t-test could effectively remove the noise in XRD spectrum in comparison with Savitzky-Golay (SG) smoothing.