With the rapid development of ultraintense and ultrafast lasers, strong terahertz (THz) sources and their applications are progressing significantly. THz electro-optic sampling (THz-EOS) is an effective means for THz coherent detection. However, for measuring irreversible or low-repetition processes, such as protein denaturation, material damage, structural phase transitions, and ultrafast dynamics, traditional THz-EOS methods face challenges owing to their reliance on multiple scanning. Unfortunately, strong THz sources are often driven by ultrafast laser systems with low repetition rates. Therefore, various THz single-shot detection methods have been developed. Among these detection methods, the spectral-coding method based on time-frequency mapping can realize the single-shot measurement of a THz waveform with a simple structure. However, this method is hindered by a low signal-to-noise ratio (SNR) and limited time resolution. The common-path self-reference spectral interference, which is based on spectral coding, can overcome the “over-rotation” problem that manifests in the traditional intensity-modulated spectral-coding method under strong THz fields. Simultaneously, two interfered pulses in the common-path scheme are always collinear transmissions and thereby effectively avoid stability issues caused by external factors. In the present study, we report a single-shot detection technique for THz time-domain spectroscopy that is based on two-dimensional spectral interference. By using two-dimensional spectral interference, this method significantly increases the information capacity of single-shot detection, which enables an improved SNR for THz pulse measurements and facilitates the detection of multiple THz-related ultrafast events. Therefore, the results of this study provide an accurate method with a high signal-to-noise ratio for single-shot measurements in strong-field THz time-domain spectroscopy.

We employ a Ti∶sapphire amplifier with an average power of up to 2.7 W as the light source. The output laser pulse is divided into two pulses by a beam splitter. One beam with 99% energy is incident into a LiNbO₃ crystal to generate a vertically polarized THz field via the tilted-pulse-front technique. The other beam is stretched up to 12 ps by a pair of prisms as a probe pulse. The broadened probe pulse width is sufficient to cover the generated THz field. The probe pulse is horizontally polarized by a polarizer and then focused onto a 1 mm-thick (110) ZnTe crystal with the THz pulse. The crystal is an electro-optic crystal for electro-optical sampling, and its [0,0,1] axis is arranged horizontally. To introduce the appropriate spectral interference fringe density between the two orthogonal polarization components of the probe pulse, the beam passes through a 1 mm-thick α-crystal with its optical axis set at 45° to the horizontal direction. More importantly, we use an imaging spectrometer to achieve two-dimensional spectral interference. After passing through a second polarizer (P2), the probe pulse is directed into the imaging spectrometer, where the spectral interference fringes are recorded. The optical axis of P2 is aligned vertically to maximize fringe modulation.

Based on a theoretical analysis of the THz electro-optical effect, this study optimizes the probe pulse polarization direction and electro-optic (EO) crystal angle for spectral interference with common-path self-reference (Fig. 2). For the vertically polarized THz pulse, the optimization is achieved when the polarization direction of the probe pulse and EO crystal axis [0,0,1] are both horizontal, as is shown in Fig. 4 (b). Owing to the intrinsic inhomogeneity of the probe pulse intensity, the recorded interference spectrum exhibits a nonuniform intensity distribution (Fig. 6). This uneven intensity affects the signal-to-noise ratio of the measured THz time-domain waveforms, as weaker interference spectra result in lower SNR values under the same system noise conditions. Consequently, the SNRs of THz time-domain waveforms vary across different lines on the imaging spectrometer (Fig. 8). The imaging spectrometer provides a spatially resolved visualization of the THz waveform distribution (Fig. 7), thereby offering a novel method for the single-shot detection of THz time-domain spectra across multiple ultrafast events. By averaging the time-domain waveforms measured across all lines of the imaging spectrometer [Fig. 9 (a)], the SNR of the THz signals reaches 166.6∶1, which is 2.2 times the maximum SNR achieved by a single line on the charge coupled device (CCD) panel of an imaging spectrometer.

In this study, a single-shot detection for THz time-domain spectroscopy based on two-dimensional spectral interference is proposed. This technique further improves THz detection efficiency by optimizing the polarization direction of the probe pulse and rotation angle of the detection crystal. The optimized structure enhances THz detection efficiency by 13.6%, achieves a higher optical modulation depth, and simplifies the alignment process of the detection setup. Additionally, imaging spectrometers replace traditional one-dimensional spectrometers to enable two-dimensional spectral interference and significantly increase the information capacity of each measurement. The experimental results demonstrate that, compared with the spectral interference scheme with common-path self-reference, the introduction of two-dimensional spectral interference improves the detection SNR by nearly 2.2 times, resulting in a maximum SNR of 166.7∶1. This technique can overcome the “over-rotation” problem that manifests in traditional electro-optic sampling as well as offer a high SNR and optimal detection efficiency. This study provides an effective detection means for strong THz radiation sources with single-shot or low-repetition frequency. Moreover, two-dimensional spectral interference shows potential for developing multi-channel THz time-domain spectrometers.