Common path interferometer for spectral image generation

Details Common path interferometer for spectral image generation Common path interferometer for spectral image generation

2000-12-14

2004-02-03

Turner, Samuel A. (Department: 2877)

Optics: measuring and testing

By light interference

Spectroscopy

Reexamination Certificate

active

06687007

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to optical instruments which process wavelengths of electromagnetic radiation to produce an interferogram. More particularly, the present invention relates to instruments (e.g., Fourier transform spectrometers) which produce interferograms of a scene, which instruments include an optical system which both splits the incoming wavelengths and spectrally disperses them to produce two sets of spectrally dispersed beams. The dispersion is achieved by a matched pair of gratings or similar system. The instrument is useful in analyzing individual chemical species in absorption or emission spectroscopy where there is a need to image a time and spatially varying scene. This could be, for example, imaging an emission plume for a jet or rocket engine or a smoke-stack.
BACKGROUND OF THE INVENTION
Imaging spectrometers are, broadly speaking, optical instruments which process the electromagnetic radiation from a source into its fundamental components. For instance, an interferometer divides light from a source and interfers it to produce a fringe pattern of interfering light (i.e., an interferogram). The interference pattern can be captured on film or by, for instance, a semi-conductor array detector (e.g., a charged coupled device (CCD)).
There are numerous optical designs. The basic form of the Sagnac (or common path) interferometer is illustrated in FIG.
1
. It is also illustrated in U.S. Pat. No. 4,976,542 to Smith. Other designs include the Mach-Zender interferometer, the Michelson interferometer and Twyman-Green interferometer (See W. L. Wolfe, Introduction to Imaging Spectrometers, SPIE Optical Engineering Press, pp. 60-64, 1997), the Fabry-Perot interferometer (see Wolfe, p. 70-73), the Lloyd's mirror interferometer (see the Smith patent) and, a variation of the common path interferometer (Sagnac) sometimes referred to as the Barnes interferometer (see T. S. Turner Jr., et al., A Ruggedized Portable Fourier Transform Spectrometer for Hyperspectral Imaging Applications, SPIE Vol. 2585 pp 222-232.) There are also dispersive spectrometers such as prism spectrometers and grating spectrometers. (See Wolfe pp. 50-52 and 55-57).
In a non-imaging Fourier transform spectrometer the point source of radiation is split into two virtual points a fixed distance apart to yield a fringe pattern at the detector. If one wants to attain a fine spectral resolution, the distance between the two virtual points should be large; for a course spectral resolution, it should be short. This distance may be controlled by shifting one of the mirrors (typically referred to as lateral shear) of, for instance, the common path interferometer. With this arrangement, a wide spectral range measurement loses resolution, while a high resolution measurement reduces the effective spectral range. In an imaging spectrometer, the point source is replaced with a slit giving the instrument the capability of one-dimensional imaging in the direction perpendicular to the shear.
In a conventional Fourier transform spectrometer, the interferogram records all spectral frequencies from the zero wavenumber to the upper spectral limit of the detector, even though the detector may not be able to sense this entire range. For a system utilizing a silicon detector that responds to wavenumbers from 10,000 to 25,000 cm
−1
(400 to 1000 nm), almost half of the information content in the interferogram is from frequencies that the detector cannot sense. The upper spectral limit of the interferogram (referred to as Nyquist limit of the detector) is determined by the ability of the detector to sample the interferogram properly.
Shear, both lateral and angular, is discussed in Turner, Jr. et al. (supra). For the Sagnac, translation of either mirror in the plane of
FIG. 1
produces lateral shear. Mirror tilt about on axis perpendicular to the drawing plane also produces lateral shear. Conversely, in the Barnes interferometer only angular shear is possible and is produced only by mirror tilt. See
FIGS. 2 and 3
of Turner, Jr., et al.
U.S. Pat. No. 4,976,542 to W. H. Smith discloses a Fourier transform spectrometer which incorporates the common path (or Sagnac) interferometer and in which a charge-coupled device (CCD) is placed in the image plane instead of film. The CCD has pixels aligned along two dimensions to provide both spectral resolution and spatial resolution. The CCD is characterized by greater dynamic range, lower pixel response variation, and is photon nose limited, all of which enhances its use as a detector for a spectrometer. See also Digital Array Scanned Interferometers for Astronomy, W. H. Smith, et al., Experimental Astronomy 1: 389-405, 1991. In these devices, the interferometer introduces a later shear in one direction and a two dimensional camera is aligned so a row of pixels is parallel to this geometric plane. In the perpendicular direction, a set of cylindrical lenses is used to provide an imaging capability along the columns of pixels. A row plot from the detector is an interferogram similar to the interferogram collected in a temporally modulated Michelson interferometer.
In a paper published in 1985, T. Okamoto et al. describe a method for optically improving the resolving power of the photodiode array of a Fourier transform spectrometer by modulating the spatial frequency of the interferogram with a dispersing element. With the use of a dispersing element, particularly an optical parallel, the distance between the two virtual sources varies with the wavenumber (the inverse of wavelength) of the source. Thus, as illustrated in
FIG. 2
of this reference, by placing their optical parallel into the optical path of a common path interferometer, the distance between the virtual source becomes a function of the wavenumber (i.e., the optical parallel refracts the blue beam more than the red beam, yielding a wide distance between S
1
blue
and S
2
blue
and a narrower distance between S
1
red
and S
2
red
). The authors claim that use of the optical parallel greatly enhances the resolution. In principle, the spectrometer can be designed to examine any wavelength band of interest by careful choice of the type of dispersive glass utilized and the thickness of the glass. See “Optical Method for Resolution Enhancement in Photodiode Array Fourier Transform Spectrocopy,” T. Okamoto et al, Applied Optics Vol. 24, No. 23, pp 4221-4225, Dec. 1, 1985.
The approach of Okamoto et al. has a number of drawbacks. First, because of the use of the dispersive block, the system no longer operates with constant wavenumber increments. This is in contrast with conventional Fourier transform spectrometers, which are constant wavenumber devices and are inherently spectrally calibrated. Thus, with Okamoto et al., blue wavelengths have a much smaller spectral resolution than red wavelengths, and the spectral calibration of the instrument becomes a major issue. Another drawback is that the spectral dispersion, while it enhances spectral resolution, adversely affects spatial resolution. Thus, the dispersive element would greatly increase the complexity of an imaging Okamoto et al. spectrometer. Another disadvantage of this technique is that its dependence on a dispersive material restricts its use to wavelengths that can be effectively transmitted through a dispersive element. Finally, the limited glass types that are available restrict the range of spectral enhancements available. While it is theoretically possible to use any dispersive glass and increase the size of the block to achieve the desired spectral enhancement, in practice the size of the block may become so large that the instrument is no longer practical. Also, since the enhancement depends on the glass type and size, the instrument designer has a limited number of parameters to use to optimize the spectrometer design and may not be able to arbitrarily set the lower and upper limits of the spectral region of interest.
In “Spatial Heterdoyne Spectrocopy: A Novel Interferometric Technique for the FUV,” J. Harlander et al., SPIE Vol

Affiliated with

Also associated with

Rating

Common path interferometer for spectral image generation does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Common path interferometer for spectral image generation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Common path interferometer for spectral image generation will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3317484