Answers are grounded in this paper and cite their sources.
Large-area, high-quality graphene obtained through SiC pyrolysis was imaged using terahertz (THz) time-domain spectroscopy. The time waveform at a position offset from the large focus was measured, simultaneously capturing information from both the graphene sample and the surrounding region without the sample. The time waveform of the THz wave and the corresponding power spectrum were analyzed using principal component analysis to explore the possibility of sample identification based on these analyses. Machine learning was employed to classify the samples, and the accuracy of identification was evaluated for each method and dataset used in the analysis. The findings suggest that samples can be identified without precisely examining the time waveforms or power spectra. In addition, the results suggest the possibility of estimating the relative contribution of regions within the THz spot. These findings demonstrate the potential of THz imaging and machine learning for nondestructive material characterization and quality assessment in industrial and scientific applications.
No takes yet. Share an insight, caveat, or question.
Zhao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: