New imaging monochromator for the far-ultraviolet
In the ultraviolet, especially the far-ultraviolet (FUV), the construction of narrow-band filters is seriously limited by the lack of available optical materials. However, relatively narrow-band FUV reflective filters have been produced.1 The out-of-band wavelength suppression for FUV imaging in the daytime requires a large number of reflective filter elements, resulting in large, cumbersome instruments. To provide comparable performance, these UV-filter instruments must be similar in mass and size to grating instruments, which allow great flexibility in selecting and tuning the wavelength band(s).
Imaging monochromators that produce simultaneous images in several wavelength bands have numerous applications in upper atmospheric research. For example, multiwavelength imaging is often used to identify emitting species, measure constituent densities that enhance or inhibit the received radiation, or establish the spectroscopic properties of the emitting species, such as the rotational temperature of molecules.
For aurora, we can use the ratios of the FUV N2 (Lyman– Birge–Hopfield, or LBH, band) emissions in two different FUV wavelength regions to estimate O2 absorption and thus the O2 density. With this information, we can make inferences about the collision altitude and the energy of the particles causing the aurora. In the sunlit upper atmosphere, measurement of the same LBH bands yields the solar photoelectron production rate of the emission feature from atmospheric N2. The simultaneous observation of the OI (un-ionized molecular oxygen) 135.6nm emission provides the same information for atomic oxygen, and the ratio of the two emission intensity measurements constitutes a remote sensing tool to estimate the O to N2 mixing ratio, which is known to be highly variable due to magnetic activity. Furthermore, absolute measurements of 135.6nm emission can be used to estimate the dominant O+ ion density in the upper atmosphere. This is possible because at night, electron and O+ ion recombination produces 135.6nm OI emission that is approximately proportional to the square of the ionospheric electron density.
The first FUV imaging monochromator was the Spectrographic Imager (SI) flown on the National Aeronautics and Space Administration (NASA) Imager for Magnetopause-to-Aurora Global Exploration (IMAGE) spacecraft.2 This Wadsworth-type instrument used an on-axis collimator, and thus required a hole in the middle of the grating for the input slit. The Czerny–Turner configuration presented here has the major advantage of being intrinsically an off-axis design with two concave mirrors in symmetrical opposition, compensating for each other's off-axis aberrations. In the FUV wavelength regions at >130nm, reflectors can be made with 80% or higher efficiency, so adding an extra surface needed for the Czerny–Turner configuration was considered a minor disadvantage.
The new concept is illustrated schematically in Figure 1. The light coming through the entrance slit (dimensions 6×32mm) at the focal plane of the collimator mirror is essentially parallel when it arrives at the grating. The grating disperses the light according to wavelength, illustrated as different colors, and the second mirror focuses the dispersed parallel light where the exit slits are placed. The collimator also acts as an imaging camera, forming an intermediate image at the grating from parallel light entering the instrument. The back imager, shown schematically in Figure 2, refocuses the image on the detector using light that passed through the monochromator. In this design, the grating is slightly curved (1.7m radius of curvature), with 4300 holographically ruled grooves per millimeter. Both mirrors are also spherical. There are two exit slits, one for the 135.6nm OI nearest to the grating, and one for the 150nm region of the LBH. The 135.6nm channel transmission versus wavelength profile is shown in Figure 3. Each back imager camera is a two-mirror decentered inverse Cassegrain.
These Czerny–Turner imaging monochromator configurations offer several advantages. For example, the intrinsically off-axis system permits high performance without obscuration. Also, the self-compensating property permits improvements in resolution and throughput. Furthermore, by adding some power to the grating, a third surface can be used for optimizing optical performance, which allows for increased system speed and a larger field of view (FOV). The FOV of the IMAGE SI was 16.2 degrees, corresponding to a collimator speed of F/3.5. The new design opens the field of view to 24 degrees, with a corresponding speed increase to F/2.35.
The configurations have several disadvantages. Because another surface is required, there is a corresponding additional reflection loss. Also, the second mirror comes with an added mass penalty. Although this mirror requires further resources (cost mass), the resulting three-mirror system uses simple, low-cost spherical surfaces and may actually provide a net cost saving.
The SI on the IMAGE satellite operated for more than 5 years and produced a global image database of the Earth's proton and electron auroras of unprecedented quality. The newer design presented in this paper will provide higher spatial resolution and better image quality suitable for observing the upper atmosphere on future spacecraft missions.
The authors are grateful for many helpful discussions with Michael Lampton and Jerry Edelstein of the University of California (UC), Berkeley.
Stephen Mende is a senior space fellow and a group leader directing research in aurora-magnetospheric imaging studies and lightning-induced upper atmospheric phenomena. He was lead scientist for the FUV instrument of the NASA IMAGE satellite launched in March 2000, and his group made the first global imaging observations of proton auroras from a satellite, using an FUV imaging monochromator.
