The spectral imaging and detection of mid-infrared wavelengths is emerging as an enabling technology of great technical and scientific interest, primarily because important chemical compounds display unique and strong mid-infrared spectral fingerprints that reveal valuable chemical information. Modern quantum cascade lasers have evolved as ideal coherent mid-infrared excitation sources, but simple, low-noise, room-temperature detectors and imaging systems lag behind. We address this need by presenting a novel, field-deployable, upconversion system for sensitive, two-dimensional, midinfrared spectral imaging. A room-temperature dark noise of 0.2 photons/spatial element/second is measured, which is a billion times below the dark noise level of cryogenically cooled InSb cameras. Single-photon imaging and a resolution of up to 200 3 100 spatial elements are obtained with a record-high continuous-wave quantum efficiency of ∼20% for polarized incoherent light at 3 mm. The proposed method is relevant for existing and new mid-infrared applications such as gas analysis and medical diagnostics.O ptical spectroscopy within the ultraviolet, visible and nearinfrared regimes has for decades been an indispensable method for the identification and quantization of chemical analytes. However, emerging mid-infrared applications in environmental gas monitoring or the life sciences call for improved detection systems that challenge today's capabilities in terms of sensitivity and/or imaging functionality. For example, in the face of global warming, mid-infrared detectors capable of measuring minute gas concentrations are required, because important greenhouse gases such as CO 2 , CO, CH 4 and N 2 O have fundamental absorption bands located in the mid-infrared 1 . For example, CO requires a detection sensitivity on the order of 100 ppb (parts per billion) 2 . Monitoring atmospheric trace molecules at these levels provides important inputs for the climate models used for studying global warming and its consequences for life on Earth 3 . In life science, the spectral regime from 0.3 to 2 mm has already been utilized for fundamental studies of breath analysis. However, significant improvements can be expected from using mid-infrared spectroscopy 4,5 . The on-line detection of the numerous different molecules (.1,000) in exhaled human breath may lead to new non-invasive diagnostics tool for doctors. However, such biomarkers are frequently below ppb levels. Indeed, the exhaled concentration of ethane (at 3.3 mm), which is used as a marker for asthma and chronic obstructive pulmonary disease, is found at 100 ppt (parts per trillion) levels, clearly demonstrating the requirement for highly sensitive methods 5 . Similarly, 1-butanol and 3-hydroxy-2-butanone in breath could be useful biomarkers for lung cancer 6 .In the 3-15 mm wavelength regions, two-dimensional mid-infrared spectral imaging demonstrates potential for identifying cancerous tissue, providing a new tool for cancer diagnostics. In this wavelength region, each organic compound an...

The potential for improving the penetration depth of optical coherence tomography systems by using light sources with longer wavelengths has been known since the inception of the technique in the early 1990s. Nevertheless, the development of mid-infrared optical coherence tomography has long been challenged by the maturity and fidelity of optical components in this spectral region, resulting in slow acquisition, low sensitivity, and poor axial resolution. In this work, a mid-infrared spectral-domain optical coherence tomography system operating at a central wavelength of 4 µm and an axial resolution of 8.6 µm is demonstrated. The system produces two-dimensional cross-sectional images in real time enabled by a high-brightness 0.9- to 4.7-µm mid-infrared supercontinuum source with a pulse repetition rate of 1 MHz for illumination and broadband upconversion of more than 1-µm bandwidth from 3.58–4.63 µm to 820–865 nm, where a standard 800-nm spectrometer can be used for fast detection. The images produced by the mid-infrared system are compared with those delivered by a state-of-the-art ultra-high-resolution near-infrared optical coherence tomography system operating at 1.3 μm, and the potential applications and samples suited for this technology are discussed. In doing so, the first practical mid-infrared optical coherence tomography system is demonstrated, with immediate applications in real-time non-destructive testing for the inspection of defects and thickness measurements in samples that exhibit strong scattering at shorter wavelengths.

This article provides an extensive survey of nonlinear parametric upconversion infrared (IR) imaging, starting from its origin to date. Upconversion imaging is a successful innovative technique for IR imaging in terms of sensitivity, speed, and noise performance. In this approach, the IR image is frequency upconverted to form a visible/near-IR image through parametric three-wave mixing followed by detection using a silicon-based detector or camera. In 1968 (50 years back), J. E. Midwinter first demonstrated upconversion imaging from shortwave-IR (1.6 μm) to visible (484 nm) wavelength using a bulk lithium niobate crystal. This technique quickly gained interest, and several other groups demonstrated upconversion imaging further into the mid-and far-IR with significantly improved quantum efficiency. Although a few excellent reviews on upconversion imaging were published in the early 1970's, the rapid progress in recent years merits an updated comprehensive review. The topic includes linear imaging, nonlinear optics, and laser science, and has shown diverse applications. The scope of this article is to provide an in-depth knowledge of upconversion imaging theory. An overview of different phase matching conditions for the parametric process and the sensitivity of the upconversion detection system are discussed. Furthermore, different design considerations and optimization schemes are outlined for application-specific upconversion imaging. The article comprises a historical perspective of the technique, its most recent technological advances, specific outstanding issues, and some cutting-edge applications of upconversion in IR imaging.

Based on enhanced upconversion, we demonstrate a highly efficient method for converting a full image from one part of the electromagnetic spectrum into a new desired wavelength region. By illuminating a metal transmission mask with a 765 nm Gaussian beam to create an image and subsequently focusing the image inside a nonlinear PPKTP crystal located in the high intra-cavity field of a 1342 nm solid-state Nd:YVO 4 laser, an upconverted image at 488 nm is generated. We have experimentally achieved an upconversion efficiency of 40% under CW conditions. The proposed technique can be further adapted for high efficiency mid-infrared image upconversion where direct and fast detection is difficult or impossible to perform with existing detector technologies.

In this work we demonstrate, to the best of our knowledge, a novel wide field-of-view upconversion system, supporting upconversion of monochromatic mid-infrared (mid-IR) images, e.g., for hyperspectral imaging (HSI). An optical parametric oscillator delivering 20 ps pulses tunable in the 2.3-4 μm range acts as a monochromatic mid-IR illumination source. A standard CCD camera, in synchronism with the crystal rotation of the upconversion system, acquires in only 2.5 ms the upconverted mid-IR images containing 64 kpixels, thereby eliminating the need for postprocessing. This approach is generic in nature and constitutes a major simplification in realizing video-frame-rate mid-IR monochromatic imaging. A second part of this paper includes a proof-of-principle study on esophageal tissues samples, from a tissue microarray, in the 3-4 μm wavelength range. The use of mid-IR HSI for investigation of esophageal cancers is particularly promising as it allows for a much faster and possibly more observer-independent workflow than state-ofthe-art histology. Comparing histologically stained sections evaluated by a pathologist to images obtained by either Fourier transform IR or upconversion HSI based on machine learning shows great promise for further work pointing towards clinical translation using the presented mid-IR HSI upconversion system.