[9p-W321-1~14] 23.1 Joint Session N "Informatics"

[9p-W321-10] Quantitative prediction of crystal orientation fluctuation by using CNN

〇Masashi Ishii¹, Tetsuya ozawa² (1.NIMS, 2.Rigaku Co.)

Keywords:AI, XRD, Crystal orientation fluctuation

By using 2D X-ray diffraction patterns of Mg₂O₆Si₂, we constructed a CNN for prediction of crystal orientation fluctuation (COF). After learning diffraction patterns with various COF from perfect single crystalline (COF=0 deg.) to polycrystalline (COF=90 deg.), the CNN well predicted COF within accuracy of < 0.1deg. The CNN after learning may be applied to high-speed and quantitative “in-situ” analyses for time-resolved XRD.