MIRI offers observing modes for both low-resolution spectroscopy and medium-resolution spectroscopy. Each mode utilizes different spectroscopic elements to produce spectra.
The low-resolution spectrograph (LRS) uses a double prism. The medium resolution spectrometer (MRS) uses dichroics for sub-band selection and gratings for spectral dispersion. The complex and compact design of the spectrometer allows the selection of 3 sets of sub-bands via dichroics and gratings mounted on the axes of 2 wheels. Image slicers are used to spatially reformat the field of view into a series of slits as necessary for the input to the relevant grating.
The LRS shares most of its optics with the imagerexcept the dispersing element and a blocking filter. Its mode (i.e., slit or slitless) is determined solely by the placement location of the target source in the focal plane. Light passing through the focal plane is collimated, forming a pupil image at the filter wheel location that holds a Ge/ZnS double prism (Fischer et al. 2008).
Expected efficiency for the set of prisms is near 80% from 5 to 10 μm (at cold temperatures) but drops below 25% for wavelengths longer than 12 μm if slit losses are included. To mitigate the effect of the fold-over in the dispersion profile, a filter is mounted on the slit to block light at wavelengths short of 4.5 μm.
The MIRI MRS has 4 separate integral field units (IFUs) called channels 1, 2, 3 and 4. Each IFU covers a separate wavelength range between 4.9 to 27.9 μm. An IFU serves primarily to divide a field of view into multiple slitlets suitable for grating dispersion. Its design has several advantages: (1) easier acquisition of point sources, (2) no slit losses due to vignetting, and (3) allowance for spatial variations as a function of wavelength for which a slit cannot account.
The MIRI IFU design consists of several components, including an entrance pupil, an input fold mirror, an image slicer mirror, a mask carrying exit pupils for the individual sliced images, a mask carrying slitlets for the individual images, and an array of reimaging mirrors behind the slitlets.
The IFU itself can be implemented with several different types of technology. The MRS IFU utilizes an image slicer, as shown in Figure 4. The slicer has rows that reflect light from different parts of the field of view into different directions. The image slices are then directed through a regular spectrograph slit and diffracted by a grating, thereby resulting in a spectrum for each row. Spatially, the image is sampled in the dispersion direction by the IFU slicing mirrors and in the slice direction by the detector pixels. Spectrally, the width of the slices defines the spectrometer entrance slit and the width of the image (in pixels) of the slice at the detector defines the width of the spectral sample. In the across slice (dispersion) direction, one slice width is matched to the full-width half-maximum (FWHM) of the JWST point spread function (PSF) at the shortest IFU wavelength.
Words in bold are GUI menus/ panels or data software packages; bold italics are buttons in GUI tools or package parameters.
All 4 MRS channels are observed simultaneously, but each exposure can only cover one-third of the available wavelength range in a single configuration. For complete spectral coverage, 3 different spectral settings must be observed, termed SHORT (A), MEDIUM (B), and LONG (C). Two dichroic filter/grating wheels have 3 working positions to move gratings and dichroics simultaneously. In short, light is separated by the dichroics in the spectrometer pre-optics (SPO) and sent to the 4 IFU channels, where the light is spatially sliced and arranged as a slit. This slit is spectrally dispersed by the gratings and eventually recorded on the detectors.