CONICA-DP

Optical Design of the Double prism

Used materials
To realize a prism that works simultaneously for the whole spectral region from 1 to 5µm, one high dispersion and a second low dispersion material are required with the following specifications:
        1. High transmission for the whole wavelength region
        2. A pair of materials of equal refraction index near 2.5µm
        3. A pair of materials of maximum and minimum dispersion
        4. An increasing dispersion for larger wavelengths is helpful, to compensate for the             increasing BG within the thermal IR

For the 1.5µm to 5.0µm region, optimum combinations in these aspects are Sapphire/CsBr and MgO/CsBr. For the shorter wavelength region (1.0µm to 1.5µm) however, the dispersion of Sapphire becomes very low. Here SrTiO3/CsBr is the optimum combination. SrTiO3 shows some absorption at 2.9µm that does not influence ground-based astronomical application.

Cameras L27 and L54 are optimized for the 2µm to 5.0µm wavelength region, however they can be used down to 1.5µm without significant loss in resolution.
Cameras S27 and S54 are optimized for 1.0µm to 2.5µm region. As both cameras contain a fused silica lens, they can not used beyond the upper limit.

Fig. 4: transmission and refraction index of SrTiO3

SrTiO3 is a ferroelectric material of high refraction index and high dispersion.

Fig. 5: 3D- representation of Prism spectroscopy in CONICA

Both external surfaces of the double prisms are aligned perpendicular to the optical axis to minimize the displacement of the pupil position (not possible for SrTiO3). Spacing between both components should be minimized, at present we assume 1mm, but realistic spacing will be 0.2mm.

Fig. 7: The wavelength distribution over individual pixels is shown above for a given prism angle of 20deg. Correcting glass prism angle is 20deg as well except for SrTiO3, where the CsBr-prism angle is set to 38deg.

Fig. 8: Spectral resolving power of several double prisms in combination with camera S27 or L27. A 3pixel slit width is assumed.

Responsible: Rainer Lenzen
Last updated: March 11, 2004