High effective time-dependent THz spectroscopy method for the detection and identification of substances with inhomogeneous surface.

We discuss an effective time-dependent THz spectroscopy method for the detection and identification of a substance with an inhomogeneous surface using a broadband THz signal reflected from the substance. We show that a successful and reliable identification can be made using the single long-duration THz signal, which contains not only the main reflected pulse, but also several sub-pulses. The method does not use averaging of the measured THz signals over the viewing angles and scanning over the surface area, which significantly increases the signal processing speed. The identification is based on the method of spectral dynamics analysis together with the integral correlation criteria (ICC). We compare the absorption spectral dynamics of a substance under analysis with the corresponding dynamics for a standard substance from database. For reliable and effective substance detection, we propose to use several ICC simultaneously in different time intervals, which contain not only the main pulse of the reflected THz signal, but also the sub-pulses. This way, one can detect and identify the substance in the sample with high probability. As examples of identification, we used the THz signals reflected from the plastic explosive PWM C4 with both rough and concave surface. We show that the main pulse, reflected from the inhomogeneous surface of the sample, contains information about its absorption frequencies.