Effects of surface roughness on terahertz transmission spectra

Terahertz (THz) transmission spectroscopy is widely used to obtain spectra of materials by measuring the electric field of a THz pulse in the time domain. However, time-domain waveforms are easily influenced by the surface roughness, which generates random noise and reduces the detection reliability. In this paper, we target the problem of roughness in the real-life application of THz time-domain spectroscopy. First, we investigate the influence of the sample surface roughness on the THz transmission spectra by establishing an analytical model that relates the sample transmission spectra to the sample surface roughness, refractive index and extinction coefficient. Then, we perform experiments with a THz transmission spectroscopy system using specific samples with different roughness values prepared in a well-controlled manner. By comparing the analytical model to the experimental results, this work accurately evaluates the effects of the surface roughness on THz transmission spectra. We do not address any new physics or any new device, but we provide a deeper understanding of the effects of the surface roughness on THz transmission spectra and a good theoretical reference to calibrate the spectra of rough samples.