Preprocessing for robust estimation of material parameters by continuous wave THz spectroscopy

Benedikt Friederich, University of Duisburg-Essen, Duisburg, Germany
Kevin Kolpatzeck, University of Duisburg-Essen, Duisburg, Germany
Xuan Liu, University of Duisburg-Essen, Duisburg, Germany
Throsten Schultze, University of Duisburg-Essen, Duisburg, Germany
Jan C. Balzer, University of Duisburg-Essen, Duisburg, Germany
Andreas Czylwik, University of Duisburg-Essen, Duisburg, Germany
Ingolf Willms, University of Duisburg-Essen, Duisburg, Germany

Terahertz (THz) spectroscopy has been widely studied for the determination of material parameters like thickness, refractive index, and absorption. THz time-domain spectroscopy (TDS) has traditionally been favored due to its large bandwidth. However, frequency-domain – i.e. continuous wave (cw) – THz spectroscopy has recently gained attention due to its lower cost, smaller footprint, potential for monolithic integration, and high spectral resolution.

The high spectral resolution leads to an undesired effect in most measurements: standing waves between transmitter and receiver can be resolved and lead to a sharp ripple in the measured spectra. A practical method to overcome this problem is the placement of absorbers in the THz path. However, these deteriorate the measurement system’s noise figure and introduce unnecessary dispersion.

As a lossless alternative, we propose a computationally efficient signal-processing solution based on Tukey windowing in the time-domain. This approach is demonstrated by estimating the frequency-dependent refractive index of a Polymethylmethacrylat (PMMA) sample with a thickness of 750 µm.

« back