1.

INTRODUCTION

The spread of novel coronavirus infections in recent years has led to an increased demand for inexpensively available, uncooled mid-infrared microbolometer array sensors. In addition, mid-infrared spectroscopy is being adapted to everyday life spaces, such as in the example of non-invasive blood glucose level sensors for health care ^[1]. Against this background, we developed a palm-sized mid-infrared imaging-type two-dimensional Fourier spectrometer ^[2]. Our proposed apparatus is shown in Figure 1. The apparatus has a near-common path and a spatial phase shift interferometer. Additionally, it has a multi-slit ^[3] that eliminates the cancellation of the interference intensity between bright spots and has higher sensitivity than conventional devices. Sufficient sensitivity can thus be achieved even with an inexpensive microbolometer. Furthermore, the developed apparatus is constructed with a quasi-common optical path, making it highly robustness against mechanical vibration. Therefore, the proposed apparatus does not require a mechanical vibration suppressor and is portable.

Fig1.

Mid-infrared passive spectroscopic imaging with the apparatus configured Transmission optics

2.

MID-INFRARED PASSIVE SPECTROSCOPIC IMAGING

2.1

Measuring principle

We propose mid-infrared passive spectroscopic imaging ^[4] using the proposed apparatus. Radiation in the mid-infrared region (at a wavelength of approximately 10 μm) is emitted from the object surface with intensity corresponding to the object temperature. The proposed apparatus adopts an uncooled microbolometer array sensor used in infrared cameras and acquires component information from the intensity of the spectral emission of the radiation from an object without a light source. Figure 2 shows the difference between conventional active spectroscopy and passive spectroscopy. In active spectroscopy, a light source irradiates the object to be measured, and spectral characteristics are acquired from the reflected light. Molecular vibrations excited by the light source are measured, and the spectral characteristics are thus absorption spectra of the energy absorbed at the eigenfrequency of the molecule. Passive spectroscopy, in contrast, does not use a light source but acquires spectral characteristics from the radiation of the measurement object itself. In this case, because the eigenfrequency of the molecules in the object itself is detected, the spectral characteristics are the emission spectrum of the eigenfrequency peak. As a result, the wavelengths that are confirmed as absorption wavelengths by active spectroscopy are confirmed as emission wavelengths by passive spectroscopy, which means that active spectroscopy and passive spectroscopy have an opposing negative–positive relationship.

Fig2.

Differences between active and passive spectroscopy

2.2

Demonstration of mid-infrared passive spectroscopic imaging

We measured silicone fluids used in defoamers, polishing agents, and cosmetic additives to verify the opposing relationship of active and passive spectroscopy. Figure 3 shows the silicone measurement results acquired by passive spectroscopy. The white emitting area in the mid-infrared image taken by the spectrometer is the area of the silicone oil coating. The graph shows the radiation intensity calculated at five locations on the silicone oil through passive spectroscopy, with the vertical axis representing the radiation intensity. The absorption spectrum of the silicone fluids measured by active Fourier transform infrared (FT-IR) spectroscopy is also shown, with the vertical axis representing absorbance. A comparison of the emission spectrum of the passive spectrometer with the absorption spectrum of the active spectrometer confirms that peaks exist at the same wavelengths. In addition, the difference processing of the relative intensity between the silicone fluid absorption peak at 1028 cm⁻¹ and the less-absorbent peak at 900 cm⁻¹ confirms the silicone fluid emission on the acquisition mid-infrared image. The above results show that active spectroscopy and passive spectroscopy are inversely related.

Fig3.

Silicone fluids measurement by mid-infrared passive spectroscopic imaging

3.

BROADBAND MID-INFRARED PASSIVE SPECTROSCOPIC IMAGING USING REFLECTIVE OPTICS

The conventional imaging-type two-dimensional Fourier spectrometer based on transmission optics has three Ge lenses (i.e., the front lens, objective lens, and imaging lens), which limits the bandwidth of the spectrometer. Reflective optics were thus constructed using reflective mirrors for the objective and imaging lenses. Figure 4 shows the configuration of the reflective optical system using reflective mirrors and an external view of the system. An uncooled microbolometer array sensor is used as the light-receiving array device. As in the transmissive optical system shown in Figure 1, a multislit is installed on the conjugate plane, which is the image-forming plane of the exchange lens, to improve the interference sharpness. The reflective mirror guarantees high spectral reflectance in a wide bandwidth (3 to 20 μm), thus realizing a measurement bandwidth wider than that of the transmission optics.

Fig4.

Mid-infrared passive spectroscopic imaging with the apparatus configured reflective optics

4.

CONCLUSION

We verified a negative–positive relationship for mid-infrared passive spectroscopic imaging by comparing the spectral emission intensity measured for radiation from silicone fluids and the absorbance acquired by FT-IR spectroscopy, a conventional active spectroscopic method. In addition, we proposed mid-infrared passive spectroscopic imaging using reflective optics that enables broadband measurement. In future work, we will quantitatively measure the glucose concentration in the body through mid-infrared passive spectroscopic imaging using a palm-sized imaging-type two-dimensional Fourier spectrometer.