“…As illustrated in Figure a,d, the red shift gradually appeared with increasing excitation power density, which agrees well with previous works. − ,− We also demonstrated that there was a threshold of excitation power density for phase segregation, and phase segregation could occur only if the excitation power density was higher than the threshold. ,,, An increase in iodine content was likely to cause lattice mismatches and trigger more defect vacancies, resulting in a smaller threshold . Furthermore, it is worth mentioning that the PL peak position changes two trends: only a red shift (Figure d) or a red shift first and then recovered to the original peak (Figure a), which is consistent with reported research studies. ,,,− We repeated the measurements to confirm the reproducibility of the illumination of variable power densities. Both repeated measurement results of two microspheres with different power densities almost coincided (Figure b,e).…”