Grism
This page summarizes the design and laboratory spectral performance of the Grism slitless spectrometer for the Wide Field Instrument (WFI), based on Bray et al. (2024).
Grism Specifications (On-axis)
| Parameter | Value |
|---|---|
| Bandpass | 1000 - 1910 nm |
| Resolving Power (R) | ~475 - 900 (2-pixel resolution element) |
| Throughput (peak) | ~65% |
| PSF Quality | Diffraction-limited across full FOV |
| Beam Deviation | Zero at 1550 nm |
Optical Design
Three-element transmissive assembly made entirely of Suprasil 3001.
The first optic contains custom diffraction gratings on both sides.
Designed to maximize (1,1) diffraction efficiency at 1310 nm.
Anti-reflection coatings applied to external surfaces.
Optimized for slitless spectroscopy of emission-line galaxies in the redshift range 1 < z < 3.
Figure 1. Zemax diagram of the Grism optical layout. Gratings are embedded in both sides of the first element. (Bray et al. 2024)
Dispersion Scale
Dispersion scale was measured across multiple field angles (SCA 0–9, 14, 16, 18).
High agreement with optical models: residuals within 0.1%.
Designed variation across the FOV ensures diffraction-limited imaging.
Figure 2. Measured Grism dispersion scale versus wavelength at all field positions. (Bray et al. 2024)
PSF and Encircled Energy
Measurements performed at 4 wavelengths and 13 field positions.
PSFs captured using a microscope objective for high spatial sampling.
50% EE radius matches model within ~15%, attributed partly to detector nonlinearity at low signal.
Figure 3. Best-focused Grism PSFs at various wavelengths. Diffraction-limited performance is maintained across the bandpass. (Bray et al. 2024)
Diffraction Orders and Background
Additional diffraction orders (e.g., (0,0), (2,2)) are well-focused but significantly fainter:
(2,2) typically 5–200× fainter than (1,1)
Background contributions from these orders are wavelength-dependent and increase slightly off-axis.
Ghost images from internal reflections are negligible (integrated flux ~0.014% of parent trace).
Dispersion Clocking
On-axis dispersion is aligned with the +Y direction.
Clocking angle varies predictably with field position; results match model expectations.
Clocking is critical for aligning spectral traces and guiding.
Bandpass Edge Calibration
Blue and red edges were measured across field and polarization states.
Edge wavelength shifts:
Up to several nm off-axis
~0.2 nm variation near field center
Measurement uses a logistic + linear model fit to finely sampled throughput data.
Steep blue edge is by design to improve wavelength assignment and on-chip guiding.
Throughput
Measured throughput of the (1,1) order across 27 wavelengths and 3 SCAs.
Results consistent across field and detectors.
Slight discrepancy with model attributed to scattered light effects not captured in simulation.
Peak throughput: ~83% at 1300–1350 nm.
Test Setup
Performed in the Ellipse Test Bed:
F/7.1 off-axis elliptical mirror
Tunable lasers, comb filters, longpass filters
Motorized hexapods for field angle simulation
Calibration procedures automated with LabVIEW scripting.
References
- Bray et al. 2024, JATIS, 10(1), 014003