Long-wavelength Infrared Upconversion Spectroscopy and Imaging

Abstract

Optical frequency conversion opens new horizons in the context of using siliconbased detectors for long-wavelength infrared (LWIR) detection, enabling tremendous scientific, technological advancement within diverse fields, such as cancer diagnostics, exhaled breath analysis, and environmental gas sensing. An infrared spectroscopic system is the fundamental tool for probing characteristic spectral signatures of molecules, e.g. in biological or gas samples, which requires sensitive and lownoise detection of LWIR signals. Mercury cadmium telluride (MCT) detectors and bolometers are the standard choices for direct LWIR detection, yet even with cooling their signal-to-noise ratio (SNR) is several orders of magnitude below that of their silicon counterparts. This thesis explores a LWIR upconversion detector, converting LWIR radiation at 9.4 µm to 12 µm into the near-infrared (NIR) range of 958nm to 980nm which is suitable for a silicon-based photodetector. The LWIR upconversion detector is combined with a quantum cascade laser (QCL) and a home-built microscopy unit (an X-Y translational stage) to implement a raster-scanning upconversion microscopy imaging system for analyzing microcalcifications present in a ductal carcinoma in situ (DCIS) breast cancer biopsy. This work involves the theoretical modeling and experimental characterization of the LWIR upconversion detector and the analysis of the DCIS biopsy in a comparison with a commercial Fourier-transform infrared (FTIR) imaging spectroscopy system. Firstly, the LWIR upconversion detector is presented, converting the LWIR signal to the NIR, which can be acquired using a silicon detector. The LWIR upconversion detector is characterized in terms of conversion efficiency and acceptance parameters. It is subject to a theoretical modeling, which takes into account the finite beam size as well as diffraction and absorption of the infrared signal in the nonlinear crystal. A comparison between the integral approach and the plane-wave approximation is made. We show the capabilities of upconversion detection for LWIR sensing in a wide wavelength range, promising fast acquisition speeds and a good SNR even for single-pulse detection. The use of silicon detectors at room temperature can benefit a wide variety of applications employing LWIR spectroscopy. Secondly, the upconversion detector is combined with a QCL source and a microscopy unit facilitating X-Y micro-movement for LWIR imaging detection of DCIS breast cancer biopsy samples containing microcalcifications. In contrast to traditional diagnosis for cancers through examination of samples that are required to be endogenously stained, one interesting alternative way to identify a cancerous invasion in breast lesions is spectral analysis of microcalcifications without the need of endogenous staining of samples. We demonstrate an upconversion microscopy imaging system capable of providing chemical images at wavenumbers below 900cm=1. Using a Micro-FTIR imaging system (Agilent 670) to detect LWIR images recorded below 900cm=1 is unreliable due to unavoidable dark noise originating from the finite temperature of the LWIR detector. In contrast, the silicon detector-based LWIR upconversion imaging system enables measuring LWIR signals down to 830cm=1, promising a better SNR in this spectral range than direct LWIR detection. The results show excellent agreement between upconversion raster scanning microscopy and Micro-FTIR imaging in terms of image structure and spectral features of breast microcalcifications. Discrete wavelength tuning of the QCL source to only relevant wavelengths with the biggest discrimination factor can substantially reduce the acquisition time. Thirdly, a self-referencing system is being investigated with the purpose of noise reduction of the LWIR detection. The approach is to split the LWIR beam into two parts and use one as an unperturbed common-mode reference to remove the noise from the sample beam – the signal beam, which is transmitted through the sample. The experimental results show a promising degree of correlation between the signal and the reference beams, potentially allowing to improve the sensitivity and stability of the LWIR upconversion detection schemes.