美国PAIR Technologies
美国PAIR技术公司开发一种新型传感器“平面阵列红外线光谱仪”，它可以在较低浓度下在液体和气体中识别生物和化学因子，检测时间低于1秒。新的光谱谱仪没有移动部件，依靠焦平面阵列(FPA)探测器。

　　“这是现有的技术的一个良好的替代技术，”该技术的创始人之一大通布鲁斯博士说，“该仪器没有移动部件，轻巧耐用，体积小，便于携带，可以随身携带它到牙医办公室。“

　　目前的检测技术是基于傅立叶变换红外(FT - IR光谱)光谱法，需要数十分钟的化学分子指纹识别。一傅立叶变换红外光谱法(FTIR)是一种重要的分析测试手段。近年来,仪器联用等新技术的不断发展,使FTIR的应用范围日益广泛,成为鉴别未知污染物和环境监测的重要工具。
研究发现，高精度声谱仪能够早期检测疾病、化学武器和环境污染物。

Planar array IR (PA-IR) is a patented technology licensed from the University of Delaware that uses a variety of dispersive elements to separate and display the individual wavelength components from a standard IR continuum light source onto the pixels of a focal plane array (FPA) (e.g., a 256 x 256 MCT) detector. The detected image can be immediately converted to a spectrum by either plotting intensity versus pixel location for a single row or by first co-averaging ("binning") a number of rows before carrying out this operation. Since pixel location within a given row can be previously calibrated to a corresponding frequency range, no Fourier transformation of the data is necessary and spectral information is obtained directly.

Since the PA-IR instrument contains a slit, image curvature occurs but proprietary software has been incorporated so as to remove any frequency variations from row-to-row due to misalignment of the spectra caused by image curvature. Because the PA-IR instrument incorporates an ultrafast FPA, it is capable of obtaining an IR spectrum in less than 100 µsec integration times (using a LN2 cooled 256 x 256 MCT FPA). This enables a whole range of kinetics and dynamics experiments to be carried out that were previously inaccessible with standard FT-IR instruments.

How do PA-IR and FT-IR compare?

Fourier transform IR (FT-IR) instruments use a Michelson interferometer to divide the amplitude of the IR source into 2 beams, which reflect off both a fixed and a moving mirror respectively. When the beams are recombined they generate an interference pattern, which is detected by a single element detector as a variation in intensity as a function of the optical path difference between the two beams. This interferogram must then be Fourier transformed to produce a power spectrum of energy vs. frequency. Two separate experiments must be run: one with no sample in the beam (reference) and a second one with the sample in the beam (sample). During and between the two runs the sample chamber must be purged with N2 gas so as to displace the H2O vapor from the instrument.

The PA-IR instrument passes the IR source through a sample and then uses dispersive optics to break up the IR source into its individual components, which then all impinge on a focal plane array (FPA) detector simultaneously. This multifrequency approach eliminates the necessity of carrying out a Fourier transformation of the data. In addition, PA-IR has been designed in a true double beam configuration so as to take advantage of the large number of pixels available in the FPA. Hence both beams of the double beam instrument can be incident on the FPA simultaneously providing both a "sample" and "reference" spectrum simultaneously. Since the path lengths of each of the two beams is identical, the background H2O vapor can be compensated for directly thereby removing the need for purging. Since the FPA is capable of less than 100 µsec integration times, the current speed of the instrument is governed by the electronic readout time (acquisition time), which is already decreasing from its current value of 17ms. In the near future it is anticipated that with improvements in the readout electronics, the acquisition time will approach the integration time providing an IR spectrum in < 100 µsec. In addition because there are no moving parts in a PA-IR instrument compared to standard FT-IR instruments, the PA-IR instruments are rugged, portable and scaleable providing new opportunities for IR spectroscopy in process monitoring, chemical warfare detection and health care evaluation.

How do PA-IR and FT-IR imaging compare?

Both PA-IR and FT-IR imaging use a focal plane array (FPA) and provide a spatial image. In the latter, an interferogram is produced at each pixel, which in a 128x128 array translates into approximately 18,000 Fourier transforms that must be carried out. However the information obtained represents a 2-dimensional map of chemical variation over the spatial region investigated. The PA-IR image contains less information because it represents essentially only spatial data in 1-dimension since the other dimension represents the wavelength dimension. However because of its speed (integration times < 100 µsec), a PA-IR image can be used to record real time changes in chemical information in that 1-D direction.