Technical aspects of the new AAO/UKST Htex2html_wrap_inline182 interference filter

Quentin A Parker , Joss Bland-Hawthorn, PASA, 15 (1), 33
The html and gzipped postscript versions of this paper are in preprint form.
To access the final published version, download the pdf file
.

Next Section: Basic features of the
Title/Abstract Page: Technical aspects of the
Previous Section: Basic filter options considered
Contents Page: Volume 15, Number 1

Practical considerations with interference filters

Specification of a narrow-band interference filter mounted at the telescope focal surface is complicated by environmental and optical considerations which have implications for the chosen filter design. If used in converging beams the inherent properties of interference filters lead to significant blueward shifts in the measured wavelength of the transmitted beam on and off axis while also affecting the filter bandpass shape. Comprehensive details of the main effects are given by Elliot & Meaburn (1976), Miller (1978) and from filter manufacturers (e.g. the `Photonics design and applications handbook', 1994). A brief description is included here.

Temperature effects
Thermal variations cause changes in the refractive indices of the spacer layers in any interference filter which leads to small central wavelength shifts in converging beams. The design temperature of interference filters is typically tex2html_wrap_inline250C while night time temperatures at the UKST are tex2html_wrap_inline252C less. Small blueward shifts in the filter's central wavelength occur whose magnitude depends on the number of filter cavities and the refractive indices of the di-electric layers of the multi-layer stack defining the interference filter. When combined with other effects the resultant blueward shifts could have implications for the intended scientific use. For example if too narrow a bandpass is selected it may not be possible to fully cover the adjacent [NII] 6548.1 Å, Htex2html_wrap_inline182 6562.8 Å, [NII] 6583.6 Åemission lines for low redshift extragalactic projects over the expected operating temperature range. These effects have been accounted for in the final specifications adopted for the filter.

Humidity effects
Narrow-band interference filters have a finite lifetime due to the effects of constant thermal variations and changing humidity. Moisture eventually penetrates the hygroscopic di-electric layers causing localised delamination and gross image defocussing. A process known as scribing at the filter edges offers a degree of protection but interference filters should still be protected from prolonged exposure to large temperature variations and high humidity. The new UKST filter is stored in a specially constructed container purged with dry nitrogen when not in use to minimise these problems.

Effects with collimated or uncollimated incident flux

Collimated flux
Ideally an interference filter should be illuminated with collimated flux. If the flux is not normal to the filters surface then the central wavelength of the filter passband is shifted to the blue. For angles tex2html_wrap_inline256 degrees the shift is given by the formula:

equation24
where tex2html_wrap_inline258 is the shift with incident angle tex2html_wrap_inline260, tex2html_wrap_inline262 is the chosen central wavelength of the filter bandpass and tex2html_wrap_inline264 is the refractive index of the spacer layers (generally tex2html_wrap_inline266).

Uncollimated Flux - converging beams
With an uncollimated beam the situation is more complicated as rays can enter the filter through a range of angles leading to angle-dependent wavelength shifts and, to a lesser extent, a broadening of the bandwidth and depression of peak transmittance. Near the UKST's focal surface the situation with the uncollimated f/2.48 cone encountering the flat filter at different angles from the field centre is given by Elliot & Meaburn (1976). The values for wavelength shifts depend on the precise filter specification. Choosing a high value for tex2html_wrap_inline264 leads to much lower sensitivity to such shifts.

In Figure.1 we simulate how the bandpass varies for the UKST f/2.48 beam with a 1% filter (i.e. a filter whose bandpass is about 1% the value of the central wavelength) both for the on-axis case (solid curve) and at an off-axis position (thin curve) at the edge of a 356 mm square filter (a full size UKST filter). Note the trend in broadening (particularly the full width zero intensity), skewing and the loss in mean transmission. The ideal filter is shown as a dotted line and is assumed to have a mean refractive index tex2html_wrap_inline270. The maximum centroid shift from centre to edge is 38 Å. This necessitates a redward shift of the central wavelength from Htex2html_wrap_inline182 to minimize the beam effect. The resulting bandpass is slightly asymmetric. On-axis broadening is, however, tex2html_wrap_inline274.

  figure30
Figure 1: UKST bandpass simulation for a 1% filter both on-axis (solid curve) and off-axis (thin curve) c.f. an ideal filter (dotted line). Note the transmission only extends to 1.2 for the purposes of clarity of the underlying curve.

Below we present the basic specifications and features of the filter. Many of these details were independently confirmed by the CSIRO National Measurement Laboratory.


Next Section: Basic features of the
Title/Abstract Page: Technical aspects of the
Previous Section: Basic filter options considered
Contents Page: Volume 15, Number 1

Welcome... About Electronic PASA... Instructions to Authors
ASA Home Page... CSIRO Publishing PASA
Browse Articles HOME Search Articles
© Copyright Astronomical Society of Australia 1997
ASKAP
Public