Terahertz time domain spectroscopy - present and future modalities

Abstract

We overview and compare free-space and onchip terahertz time-domain spectroscopy systems. Example spectroscopic applications are given for both types of system: vibrational spectroscopy of crystalline compounds for freespace systems, and assay of dielectric films for on-chip systems.Index Terms -Submillimeter wave imaging, Submillimeter wave propagation.I. THz TIME DOMAIN SPECTROSCOPY Terahertz time domain spectroscopy (THz-TDS) allows the optical properties of materials to be extracted in the frequency range -50 GHz to 30 THz. A typical system is shown in Fig. 1. It comprises a visible continuous-wave laser pumping a Ti:sapphire oscillator, which emits sub-100 fs duration pulses of near-infrared (NIR) radiation at a centre frequency of -800 nm, and at a repetition rate of -80 MHz. The laser pulse-train is split into two: one part is focused onto a semiconductor surface; low-temperaturegrown gallium arsenide (LT-GaAs) is widely chosen as this semiconductor since it has high carrier mobility and short (sub-picosecond) photoconductive carrier lifetime [1]. The photoexcited carriers (electrons and holes), generated at the semiconductor surface, are accelerated in the bias field created by a lateral antenna. This results in a transient dipole which radiates pulses with a broad spectral content extending well over 1 THz [2]. The THz pulses are then collected and collimated using off-axis parabolic mirrors, before being focused to a (diffraction-limited) spot size of a few hundred microns at the sample position. The frequency range of the pulses can be maximized by collecting them from the front surface of the LT-GaAs wafer, which reduces the effect of absorption within the GaAs substrate. When combined with a suitably fast laser system for excitation (-10 fs), this extends the usable bandwidth of the system to above 25 THz [3].After passing THz pulses through, or reflecting them from, the sample surface, the modified THz pulses are collected and collimated, before being focused onto a detection crystal. The second NIR beam is given a variable time-delay using a retroreflector mounted on a linear translation stage, and is also focused onto the detection crystal. Two different methods of detection can be employed. The first, photoconductive detection, uses a similar scheme to the emitter, based on an antenna deposited on LT-GaAs; the incident THz radiation acts as a bias field for the carriers generated by the second NIR beam beam strip-line off-axis splitter antenna parabolics Fig. 1. Schematic of typical THz-TDS system, showing the paths of the NIR (dark grey lines) and THz (light grey)beams. Electrooptic detection is achieved using a ZnTe crystal, Wollaston prism, and photodiodes as described in the text.at the detector, and the resulting current is measured as a function of time delay applied to the NIR beam. In this way, it is possible to map the time-evolution of the electric field which constitutes the THz pulse. A second detection mechanism exploits the ac Pockels effect in an electro-optic crystal such as zinc ...