Fiber Laser Based Mid-Infrared, Wavelength-Agile Sensor for Monitoring Trace Gas Concentrations in Harsh Environments

该项目开发了一种全光纤传感器，用于测量活塞和航空推进发动机等恶劣环境中的痕量气体浓度(范围为1-10,000 ppm)。该传感器使用中红外吸收光谱来探测目标气体的基本振动带中的跃迁。该传感器可以在4.85 mm附近检测到强烈的一氧化碳转变，并且有可能在1.6-7.5 mm的光谱范围内进行调谐；它还可以检测到大量其他气体，包括一氧化氮(5.2 mm)、二氧化氮(6.1 mm)、一氧化二氮(4.5 mm)、二氧化硫(7.2 mm)、碳氢化合物(3.4 mm)和各种自由基物种。其中许多重要的物种很难在其他光谱范围内获得；例如，电子CO跃迁只存在于真空紫外光中，因此不可能通过单光子技术获得。提出的传感器的核心是一个全光纤光源，用于在中红外产生快速波长扫描。开发这样一个信息源是拟议工作的重要组成部分。这种“波长敏捷型”光源每隔20 ns扫描大约100纳米。该光源由三个元件串联组成：工作在1.56 mm附近的超快光纤激光器，用于将超快脉冲转移到中红外的非线性波长转换元件，以及用于将中红外脉冲安排成快速波长扫描的啁啾光纤布拉格光栅。为了测量痕量气体浓度，波长灵敏的输出被光纤耦合到感兴趣的环境(如发动机)，传输的光由一个小面积的红外探测器监测。在这种方式下，每隔20 ns就可以获得100 nm宽的吸收光谱。可以对连续的吸收光谱进行平均以提高灵敏度；根据应用和痕量气体浓度的不同，每0.2-200毫秒可使用10-10,000个平均值来获得吸收光谱。每个吸收光谱都减少到气体浓度，使传感器能够以足够高的速率持续监测痕量气体，以跟踪大多数瞬时化学现象。传感器的能力可以通过使用相同的超快光纤激光器来扩展，如本提案中所述，利用研究人员最近开发的技术产生中红外和近红外。这种传感器可以监测恶劣环境中的次要物质和主要物质的浓度，以及气体的温度和压力。由于这些环境中存在高压、高温和多种丰富的物种，因此通常无法进行这样的测量。然而，新兴的光纤技术不仅为这一传感挑战提供了解决方案，还为非常紧凑的手持式痕量气体传感器提供了潜力。这里强调一氧化碳(CO)的传感，因为它对健康的影响、监管标准，以及与燃烧、燃料重整和其他过程中的基本化学问题的相关性。

This project develops an all-fiber-optic sensor for measuring trace-gas concentrations (in the range of 1 - 10,000 ppm) in harsh environments such as piston and aeropropulsion engines. The sensor uses mid-infrared absorption spectroscopy to probe transitions in the fundamental vibrational bands of the target gases. The sensor accesses strong carbon-monoxide transitions near 4.85 mm, and has the potential to tune throughout the 1.6 - 7.5 mm spectral range; it can also access a host of other gases including nitric oxide (5.2 mm), nitrogen dioxide (6.1 mm), nitrous oxide (4.5 mm), sulfur dioxide (7.2 mm), hydrocarbons (3.4 mm) and a variety of radical species. Many of these important species are difficult to access in other spectral ranges; for example, electronic CO transitions exist only in the vacuum ultraviolet and are therefore impossible to access by single-photon techniques. The heart of the proposed sensor is an all-fiber-optic source for generating rapid wavelength scans in the mid-infrared. The development of such a source is a significant portion of the proposed work. The "wavelength-agile" source scans through approximately 100 nm every 20 ns. This source consists of three elements connected in series: an ultrafast fiber laser operating near 1.56 mm, a nonlinear wavelength-conversion element for shifting the ultrafast pulses to the mid-infrared, and a chirped fiber Bragg grating for arranging the mid-infrared pulses into rapid wavelength scans. To measure trace-gas concentrations, the wavelength-agile output is fiber-coupled to the environment of interest (such as an engine), and the transmitted light is monitored by a small-area infrared detector. In this fashion, a 100-nm-wide absorption spectra can be obtained every 20 ns. Consecutive absorption spectra can be averaged to improve sensitivity; depending on the application and the trace-gas concentration, 10 - 10,000 averages may be used to obtain absorption spectra every 0.2 - 200 ms. Each absorption spectrum is reduced to gas concentration, enabling the sensor to continuously monitor trace gases at sufficiently high rates to track most transient chemical phenomena. The sensor capabilities can be extended by using the same ultrafast fiber laser to generate both mid-infrared, as described in this proposal, and near-infrared, using techniques recently developed by the investigator. Such a sensor can monitor the concentrations of minor and major species as well as gas temperature and pressure in harsh environments.Measurements such as these have been generally unavailable owing to the high pressures, high temperatures, and multiple abundant species present in these environments. However, emerging fiber-optic technology not only offers a solution to this sensing challenge, but also provides the potential for very compact, handheld trace-gas sensors. Sensing of carbon monoxide (CO) is emphasized here because of the associated health effects, regulatory standards, and relevance to fundamental chemical problems in combustion, fuel reforming, and other processes.