Probing structural deformations of the Kagome metal CsV₃Sb₅ using X-ray diffraction temperature clustering (X-TEC)

The recently discovered Kagome metal CsV3Sb5 proves to be an exciting playground to study charge density waves and superconductivity. Scanning tunneling microscopy experiments have indicated the emergence of unidirectional stripe order, chirality, and broken time-reversal symmetry, at temperatures much below the onset of CDW. To investigate the CDW and other potential phases in the bulk, we study the temperature-dependent X-ray diffraction data from single crystal CsV3Sb5. To analyze the large volume (~48GB) of data spanning 2000 Brillouin zones, we utilize the recently introduced unsupervised machine learning framework called X-ray Temperature Clustering (XTEC) [1]. Using X-TEC, we observe slightly different onset temperatures for the 2x2x2 (Tc ~ 90K) and 2x2x4 (Tc ~ 85K) CDW order. Further, we observe a subtle change in the height and width of the Bragg and CDW peaks onsetting below a lower temperature (~55K). We also discuss the effect of the cooling profile on these observations.

[1] J. Venderley et al., Harnessing Interpretable and Unsupervised Machine Learning to Address Big Data from Modern X-Ray Diffraction, Proceedings of the National Academy of Sciences 119, e2109665119 (2022).

Acknowledgment: Work supported by the U.S. DOE, Office of Science, Basic Energy Sciences, Materials Science and Engineering Division.