Hyperspectral imaging with sub-cycle mid-infrared pulses

Abstract

Hyperspectral imaging is a technique that combines imaging and spectroscopy to map the distribution of chemical constituents. Mid-infrared (MIR) hyperspectral technique identifies and maps the chemical composition of an object through molecular vibration. However, the limited number of pixels and the low signal-to-noise ratio of MIR detectors prevent high performance of MIR hyperspectral imaging. To address this issue, up-conversion of the MIR ultrashort pulses to visible or near-infrared light is used. This approach improves the performance of MIR imaging since visible or near-infrared light can be detected with Si-based detectors, which have much higher performance than MIR detectors. In this talk, a new MIR hyperspectral imaging method based on sub-cycle MIR pulses is suggested. This method covers both the functional group and the fingerprint regions with high intensity. Unlike previous methods, the size of the image does not depend on the wavelength, which simplifies the calibration process. The sub-cycle MIR pulses are generated by using four-wave mixing through two-color filamentation. The MIR pulse passes through a sample and is sent to a GaSe crystal, while a chirped 800 nm pulse is also sent to the crystal. The sum frequency signal is then sent to a silicon-based hyperspectral camera and records hyperspectral images. The analytical ability of the new method is tested by imaging and mapping onion cells. The results show that the method has great potential for cell analysis since it can map the cell wall, cytoplasm, and nuclei, with the distribution of nuclei only visible in the MIR hyperspectral image.