Final design of SITELLE, a wide-field imaging Fourier

of 12

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
12 pages
0 downs
Final design of SITELLE, a wide-field imaging Fourier transform spectrometer for the Canada-France-Hawaii telescope F. Grandmont∗a, L. Drissenb, Julie Mandara, S.…
Final design of SITELLE, a wide-field imaging Fourier transform spectrometer for the Canada-France-Hawaii telescope F. Grandmont∗a, L. Drissenb, Julie Mandara, S. Thibaultb, Marc Barilc and the SITELLE Team a b ABB, 585 Boul. Charest est, Suite 300, Québec, Qc, Canada G1K 9H4 Dépt. de physique, de génie physique et d’optique, Université Laval, Québec, Qc, Canada G1V 0A6 and Centre de recherche en astrophysique du Québec (CRAQ) c Canada-France-Hawaii Telescope Corporation, Mamalaoha Highway, Kamuela, HI, 96743 USA ABSTRACT We report here on the current status of SITELLE, an imaging Fourier transform spectrometer to be installed on the Canada-France Hawaii Telescope in 2013. SITELLE is an Integral Field Unit (IFU) spectrograph capable of obtaining the visible (350 nm – 900 nm) spectrum of every pixel of a 2k x 2k CCD imaging a field of view of 11 x 11 arcminutes, with 100% spatial coverage and a spectral resolution ranging from R = 1 (deep panchromatic image) to R > 104 (for gas dynamics). SITELLE will cover a field of view 100 to 1000 times larger than traditional IFUs, such as GMOS-IFU on Gemini or the upcoming MUSE on the VLT. SITELLE follows on the legacy of BEAR, an imaging conversion of the CFHT FTS and the direct successor of SpIOMM, a similar instrument attached to the 1.6-m telescope of the Observatoire du Mont-Mégantic in Québec. SITELLE will be used to study the structure and kinematics of HII regions and ejecta around evolved stars in the Milky Way, emission-line stars in clusters, abundances in nearby gas-rich galaxies, and the star formation rate in distant galaxies. Keywords: Fourier transform spectroscopy, hyperspectral imagery, Integral Field Unit 1. INTRODUCTION SITELLE (Spectromètre Imageur à Transformée de Fourier pour l’Etude en Long et en Large de raies d’Emission, or Imaging FTS for the study of emission lines) has been approved as a guest instrument for the Canada-France-Hawaii telescope (CFHT). Funded by the Canada Foundation for Innovation with additional financial participation by the CFHT, Université Laval, ABB and e2v, SITELLE is an imaging Fourier transform spectrometer (FTS) capable of obtaining spectra in selected wavebands in the visible (from 350 to 850 nm) of every light source in an 11 x 11 arcminute field of view with a spatial sampling of 0.32''. Its spectral resolution is variable, depending on the requirement of the observer, from R = 1 (broad-band image) to R in excess of 104. The spatial resolution is limited by the seeing, resulting in ~ 106 different spectra. The dual output interferometer configuration of SITELLE ensures that virtually every photons collected by the telescope reaches the detector and is analyzed; a by-product of the spectral data cubes is therefore a very deep panchromatic image of the targets. SITELLE's design is based on that of a previous prototype, SpIOMM, attached to the 1.6-m telescope of the Observatoire du Mont Mégantic (Drissen et al., these proceedings). SpIOMM and SITELLE's early development phase was presented by Grandmont, Drissen and collaborators [1-5]. Despite its name and acronym, SITELLE will also be able to perform absorption line studies, like the vast majority of FTS in astronomy, planetary space missions or remote sensing devices. Most of the design and construction work is being performed by ABB Analytical , a Québec-based company specialized in Fourier transform spectrometers and optical sensors. Science lead, optical design and its integration are done at ∗ a class= __cf_email__ href= /cdn-cgi/l/email-protection data-cfemail= aec8dccbcacbdcc7cd80c480c9dccfc0cac3c1c0daeecdcf80cfcccc80cdc1c3 [email protected] /a script data-cfhash='f9e31' type= text/javascript /* ![CDATA[ */!function(t,e,r,n,c,a,p){try{t=document.currentScript||function(){for(t=document.getElementsByTagName('script'),e=t.length;e--;)if(t[e].getAttribute('data-cfhash'))return t[e]}();if(t&&(c=t.previousSibling)){p=t.parentNode;if(a=c.getAttribute('data-cfemail')){for(e='',r='0x'+a.substr(0,2)|0,n=2;a.length-n;n+=2)e+='%'+('0'+('0x'+a.substr(n,2)^r).toString(16)).slice(-2);p.replaceChild(document.createTextNode(decodeURIComponent(e)),c)}p.removeChild(t)}}catch(u){}}()/* ]] */ /script ; phone 1 418 877 2944 x-318; fax 1 418 266 1422 Université Laval, the mechanical design and fabrication of the input and output ports at Université de Montréal, while CFHT takes charge of the detectors' enclosure and cooling system. The advantages and disadvantages of the imaging FTS technique, as well as the relative merit of different approaches to 3-D imagery are discussed by Ridgway & Brault [6], Bennett [7] and, more recently, by Maillard et al.[8]. 2. SCIENCE DRIVERS AND INTRUMENT REQUIREMENTS 2.1 Science drivers While the number of scientific programs for this type of instrument is potentially huge (from the study of individual stars in local star clusters to the search for high-redshift Ly-α emitters), we have chosen a series of typical projects to illustrate its scientific potential by advocating the enormous benefits provided by a systematic, complete 3D mapping of extended emission-line sources. The main science drivers for SITELLE are very similar to those of SpIOMM, a prototype imaging FTS attached to the Observatoire du Mont Mégantic's 1.6-m telescope (see Drissen et al., these proceedings): the study of physical characteristics (temperature, density, kinematics) of nebulae surrounding evolved stars, supernova remnants and the diffuse interstellar gas in the Milky way galaxy; abundance gradients and kinematics of nearby galaxies to understand their evolution; distant galaxies. The much higher throughput and efficiency of SITELLE at CFHT will allow us to map fainter emission lines, to broaden the study of galaxies to the absorption lines due to the presence of an old stellar population, and to extend the study of galaxies to much higher redshift in order to study, for example, the starformation rate across the Universe. Our team has learned a lot from the development of SpIOMM and especially from its use on a regular basis at a telescope. All the potential improvements we have identified will be fully integrated in SITELLE. In particular, we expect: ã A much better sensitivity in the near-UV, mainly because of significant improvements in the interferometer's performances. ã More sensitive CCDs, mostly because of a significant readout noise reduction (from 10e to 3e). ã Much lower dead time between exposures (2 seconds instead of 7), a combination of CCD readout and interferometer performances. 2.2 Instrument requirements The science requirements have defined the following technical requirements of SITELLE which have been used in the design of the instrument: ã Wavelength range - In the local universe, the [OII] 372.7 nm doublet defines the short wavelength requirement. It is used to measure the oxygen abundance in ionized nebulae and the ratio of its two components is an excellent indicator of the electron density in the interstellar medium. Many factors conspire to make this line a real challenge for an Imaging FTS, and in particular the stringent constraints it imposes on the quality of the optical surfaces within the interferometer (mirrors and beamsplitter) as well as the precision of the step-scan and servo mechanisms to which the modulation efficiency is particularly sensitive at short wavelengths. The long wavelength limit is defined by the Ca triplet at 849.8 nm, 854.2 nm and 866.2 nm, which characterizes the old stellar population in galaxies. In the case of the high redshift objects, the wavelength range accessible with the instrument defines the redshift range in which the Lyα line can be detected. The above-mentioned limits (370 – 870 nm) set this range to 2.0 < z < 6.15. ã Spectral resolution – The minimum resolution required for the analysis of the ionized nebula in the Milky way and other galaxies is set by the necessity to separate the [SII] 671.7 / 673.1 nm doublet, the Hα 656.3 nm from its [NII] 654.8 nm and 658.4 nm neighbors, and Hγ 434.1 nm from [OIII] 436.3 nm. This implies a minimum value of R = 1000 over the entire wavelength range. However, kinematics of HII regions and the studies of stars in clusters impose a more stringent requirement of R = 104. ã Field of view – For the study of extended HII regions in the Milky Way, nearby galaxies and the high-redshift Universe, reasonable amounts of observing time require a FOV > 10 arcminutes, which set the constraints for SITELLE. A larger FOV is always welcome but leads to complexity increases that are not linear with the field size beyond this point. The specified value appears as a suit spot considering the available budget. ã Spatial resolution – Most of the projects presented here put the emphasis on the wide field observation of extended, diffuse objects and not on the spatial resolution. However, observations of individual stars in clusters and distant galaxies require that the pixel size be no larger than the typical seeing, 0.6 arcsec. The panchromatic image quality (350 – 900 nm) should be no worse than one arcsec. ã Sensitivity – Two of the most stringent constraints in terms of sensitivity, from the science case, are the ability of SITELLE to detect the faint [OIII] 436.3 line, which is not always detected in HII regions, and its capacity to detect Lyman apha emitters down to a flux of 4.3×10-17 ergs s-1 cm-2 (5 sigma detection) in a reasonable integration time (less than 4 hours). ã High observing efficiency – The readout time, combined with the displacement time of the interferometer between OPD sampling position should be a small fraction of the total on-target exposure time. Considering the read time attainable with low read noise (~3 electrons) on modern CCD, a 2 second interferometer displacement time seems like an acceptable upper limit. For example, 500 images cube acquired over a few hours would limit the dead time to around 15 minutes. ã Autonomy – CFHT now requires that all its instruments be remotely operated for periods of at least one week without direct human intervention. 3. SITELLE DESIGN 3.1 Optical design The design of an IFTS with such emphasis on the imaging/throughput aspect obviously starts by attempting to meet the FOV and image quality requirement over the desired wavelength range simultaneously. This is a challenging task compared to similar astronomical instruments. On one side, wide field camera systems can typically suffer from chromatic aberrations to some degree since band pass filters are typically used and allow for a focus correction. On the other side, Integral field unit spectrometers will typically image a much smaller FOV than SITELLE or, if not, allow for very coarse spatial sampling of it (hence poor image quality). All images acquired with an IFTS are meant to be panchromatic and take full advantage of all the available detector pixels for imagery. Band pass filters can still be used but primarily to reduce the photon noise in the spectra or the number of images required for cube completion and not specifically to improve image quality. The optical design process can be carried out almost independently from the interferometer design as the latter typically adds only reflecting or transmissive flat interfaces which effect on the optical performances are often negligible. The IFTS configuration practically bring one additional requirements to the optical design which is the need to dispose of a collimated beam section of sufficient length to place the interferometer. This latter is ideally centered around a pupil image in the collimated section where the beam envelope is the smallest. The input/output optical design’s intent was to converge to a solution that provided the largest FOV without compromising the wavelength range and image quality threshold as dictated by the science requirements. The pupil size of the collimated section (90mm) was set by the desire to limit as much as possible the size of the interferometer optics which require the high flatness requirement and must be moved by the OPD scanning mechanism. For a given FOV, reducing the interferometer pupil size increases the divergence before the camera lens and further complicates its design. A parametric model of the overall optical design including a metrology channel was established and demonstrated that the dimension of the interferometer had a minima near a pupil of 90 mm. The final design, which is composed of a triplet for input optics and six lenses in the output optics (see Figure 1), allows an unvignetted circular field of view of 5.5 arcminute radius with a full fov of 11' x 11' (0.32'' per pixel) with a 15% vignetting in the corners of the field. The image quality specification is fully met with this final design. The large number of lenses in the optical path (18 surfaces) brings the challenge of finding suitable anti-reflection coatings which will not significantly attenuate the light. Fig. 1 - SITELLE optical design (top) and resulting spot diagrams (bottom) of the 650 - 680 nm wavelength range in the center of the field (left), 5.5' from the center (edge center) and corners of the field (right). 3.2 Interferometer design The interferometer design is largely driven by the desire to obtain high efficiency at near UV wavelengths. A very small number of interferometers are found to have operated in the UV regime in the general literature and there is a good reason for it. The modulation efficiency of any interferometer shows an exponential-like decline in performance as we move toward shorter wavelengths. Many factors affect the modulation efficiency in such a way: 1- Wavefront errors between the two recombining beams at the beamsplitter 2nd pass; 2- Tilt between the two recombining beams at the beamsplitter 2nd pass; 3- Shear between the two recombining beams factored by the beam divergence determined by the detector size; 4- OPD jitters encountered during exposure of the detector. The first parameter essentially depends on the quality of the reflecting surfaces or the transmissive medium (air or glass) encountered between the separation and combination of the science beam. Any configuration limiting the number of surfaces or the length of the cavity (if operated in the presence of air) is favoured. The other parameters impact on modulation efficiency varies strongly depending on the configuration. Fourier transform spectrometers, even imaging versions, are fairly common outside the field of astronomy and are found commercially for over 30 years now. The designs have generally evolved towards the most efficient and reliable architectures over time due to the strong competition between industrial players. Cube corner retro-reflector-based interferometers are by far the most common architecture found in both single pixel and imaging systems nowadays. The reason for this is that they are inherently tilt errors free (item #2) as long as their three mirror surfaces remain orthogonal to one another to the same precision as is acceptable for the tilt error. Cube corners of acceptable quality for infrared interferometry down to 1 um are found commercially in sizes up to 5 cm. The problem is that SITELLE would require cubes with four times better alignment and reflected wavefront errors in size exceeding 20 cm of clear aperture. Moreover, cube corner retroreflectors behave notably badly with temperature such that most systems are temperature-stabilised to maintain the best performance. We wish to operate SITELLE at ambient temperature given that it will be mounted directly on the telescope and that stringent heat dissipation requirement must be met in the dome environment. Our search of potential supplier rapidly concluded to a risky endeavour if at all possible for SITELLE's available budget. Inability to achieve the required performance would result in irrecoverable performance losses at short wavelengths. Considering this, we looked more carefully into the more classical flat mirror Michelson original design. In addition to using simple optical components (plane mirrors) readily available commercially, this approach removes two reflecting surfaces which helps to further reduce the errors between recombining wavefronts. The plane mirror architecture also does not create shears between recombining wavefronts. Its drawback is that a dynamic alignment system must be implemented to correct for the followings contributors: - Residual tilt at interferometer integration; - Tilt induced by thermal distortion over the operating range (interferometer follows ambient); - Tilt induced by the varying instrument orientation due to the Cassegrain mounting; - Inherent tilt profile of the scan mechanism. Contrary to the large high quality cube corner, tip-tilt mechanisms with precision down to the arc second are now common in astronomy thanks to adaptive optics. ABB also has a vast heritage in dynamically aligned plane mirror interferometers, since it carried a high resolution commercial spectrometer for more than 20 years based on that technology. Also, the first dynamically aligned plane mirror interferometer to fly in space in 2011 was produced by ABB based on that same technology platform[9]. The only remaining issue is how to separate the second output port from its input. In a plane mirror interferometer, half of the light goes to one output and the other half is retro-reflected on itself toward the target. This is not acceptable in ground-based astronomy since the sky transparency may affect the source intensity during the scanning of the OPD. The Fourier transform operation finds the associated spectral content for all intensity variations present in the input vector and makes no distinction between true interference and undesired source fluctuation. The access to the second input port is crucial in being able to compensate for the source variation before performing the Fourier transform. To overcome this problem, the solution already implemented in SpIOMM was reused (see Figure 2). It consists of entering the interferometer at a given angle such that the coincident output is angularly separated from the input. The angle is made just large enough to locate the collimator lens and the camera lens barrels side-by-side as depicted in the simplified diagram below. The sketch below showing the whole layout in a 2D plane results in a rather large beamsplitter footprint. A more compact approach is obtained by entering the input beam in a plane perpendicular to the one shown (off-axis input/output rotated at 90 degrees to the interferometer plane). Another advantage of the dual output flat mirror configuration with respect to the cube-corner one is that it results in a much smaller moving mass. A dual output port cube corner interferometer is obtained by entering the science beam in the lower half part of the cube corner which translates it to its upper half. The clear aperture size required for the cube more than doubles the pupil size given that sufficient clearance must be generated in the lateral transfer to account for beam divergence and the longitudinal distance to reach the beamsplitter. A higher moving mass means lower resonance modes and a lower frequency response of the OPD servo system. The SITELLE interferometer final design based on this architecture is shown in Figure 3. Fig. 2 - Simplified representation of the optical arrangement of SITELLE Metrology Injection & Collection OUTPUT 1 INPUT Fix Mirror OUTPUT 2 Beamsplitter Moving & Tip/Tit Mirror Fig. 3 - Exploded view of the interferometer structure The beamsplitter is a key driver in the instrument performance. For SITELLE, we chose a sandwich-type configuration which has the splitting coating sitting in the middle of two identical glass substrates. This ensures identical
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks