When heat flows in bulk samples and across an interface, thermal conductivity k and interfacial thermal conductance G are two crucial measurables. However, scalar quantities offer limited microscopic knowledge. Interestingly, k and G are microscopically determined by energy-dependent relaxation time τ(E) and transmission T(E). τ(E) and T(E) contain much rich information but have been hard (or, considered impossible) to measure. In this work, Zhantao et al developed a framework that targets the determination of τ(E) and T(E) by using scientific machine learning to analyze ultrafast diffraction data. The real-time in ps-scale, real-space in nm-scale, and frequency-resolved phonon propagation can then be portrayed in great detail, hence we termed such capability as “panoramic mapping.”
