LIRIS: a Long-slit Intermediate Resolution Infrared Spectrograph for the WHT
A. Manchado, F.J. Fuentes, F. Prada, E. Ballesteros, M. Barreto, J.Mª. Carranza, I. Escudero, A. B. Fragoso-López, E. Joven, A. Manescau, M. Pi, L.F. Rodríguez-Ramos, N. Sosa
Instituto de Astrofísica de Canarias, La Laguna, E-38200, Spain

The Instituto de Astrofísica de Canarias (IAC) is undertaking the design and construction of a common-user near infrared spectrograph (LIRIS) for the Cassegrain (f/11) focus of the 4.2 m William Herschel Telescope (WHT) sited at the Observatorio del Roque de Los Muchachos (ORM, La Palma). LIRIS will be a near infrared (0.9-2.5 mm) intermediate-resolution spectrograph designed to operate over a spectral resolution range between 1000 and 5000, with added capabilities for coronographic, multiobject and polarimetric observations. The instrument allows the combination of an adequate spatial resolution (0.25 arcsec/pixel) with a large useful field of view (4.2 arcmin) across the slit, thanks to the use of the new 1024x1024 pixel HgCdTe Hawaii detector manufactured by Rockwell. All the optics and mechanisms situated inside the cryostat will be cooled to below 100 K. The detector will operate at 77 K. Calibration and tracking will be made with the existing Cassegrain A&G Box, into which a near infrared calibration system will be incorporated.


The arrival at the beginning of the 90s of new large-format and low noise level infrared detectors (e.g. NICMOS3) marked a great advance in the development of astrophysics at wavelengths greater than 1 m m and enabled observing programmes that had hitherto been restricted to the visible to be extended into the infrared. The majority of observatories since then have obtained such detectors for direct imaging, although very few have the necessary instrumentation available for carrying out wide-field spectroscopy in the near infrared.

Furthermore, measurements taken in the wavelength range 1- 2.5 m m on the William Herschel Telescope (4.2 m) at the Roque de los Muchachos Observatory (ORM, La Palma) over the last two years have demonstrated the quality of the WHT and of the ORM sky for observing in the near infrared (see Table 1)

LIRIS would enable to carry out infrared spectroscopy in the near-infrared bands with a long slit and spectral resolutions as high as 5000.

LIRIS aims to attain the following objectives:

    1. Maximization of the scientific versatility/technical complexity ratio
    2. To keep overall costs within reasonable limits
    3. To have the instrument on the telescope within about 2 years
    4. To fill a gap in the existing infrared instrumentation at the ORM (La Palma)


Table 1. Measurements of the sky infrared brightness at the ORM (La Palma)
made with WHIRCAM at the WHT during winter and summer periods



LIRIS will be used, mainly in bright time, at the Cassegrain focus of the WHT. Given its common-user status, it should be possible to use it for a wide range of astrophysical disciplines, including the areas of stellar, planetary, extragalactic and cosmological physics. We list below some of the most relevant possible scientific applications:

We describe below some of the projects that would be achievable with LIRIS on the WHT.

2.1 Infrared spectroscopy of proto-planetary nebulae

Proto-planetary nebulae (PPNe) are nebulae in which the central star is highly obscured by dust, there being no visible counterpart in the majority of cases. For this reason, they can be observed only in the infrared. By means of K-band spectroscopy it is possible to investigate the distribution of ionized and molecular gas. This permits an understanding of this rapid transition phase that later gives rise to the formation of a planetary nebula.

2.2 Detection of very low mass secondaries, planets and brown dwarfs

The IR is the most favorable spectral domain for detecting the contribution to the integrated light of a multiple system (binary or planetary) from the least massive components. IR intermediate-dispersion spectroscopy can permit the detection of molecular bands (H2O and CO) from very cool objects superimposed on the continuum of the primary. A particularly interesting case is that of the recently discovered planets at less than 0.5 AU from their stars. The temperature of these planets must be high enough for them to have their emission peak in the near IR, i.e. in the spectral range of LIRIS.

Furthermore, red colours do not constitute a unique identification of brown dwarfs, whereas infrared spectroscopy in the wavelength range of LIRIS would allow the methane bands characteristic of an atmosphere sufficiently cool for a brown dwarf to be stand out.

2.3 IR spectroscopy of nearby and distant galaxies with massive star formation: nearby HII regions and distant starbursts

The existence of a population of galaxies at around z=1 with active star formation, probably of low metallicity and with blue colours has recently been demonstrated by the spectroscopic surveys carried out on the Keck telescope. The possibility that these objects constitute a population of dwarf galaxies in the process of developing the first bursts of star formation is extremely interesting, since they would enable us to find objects of metallicities < 1/100 solar, which might reflect the optimum representation of the pregalactic abundances necessary for deriving a precise value of the primordial abundances. In this way we could intercompare the spectroscopy of our local surroundings with that at other redshifts and thus unravel some of the so far unresolved problems of their chemical evolution.

2.4 Stellar kinematics and populations in spiral galaxies

The presence of dust in the center of spiral galaxies means that the ideal spectral range for their studies is the infrared. In particular, we shall be able to determine kinematic peculiarities and their correlation with the stellar populations at the centers of galaxies without the uncertainty that arises in the visible range. This will enable us to understand the nature and mechanisms of the formation of bulges, discs, bars and dynamical subsystems such as counterrotating discs and bulges.

2.5 Ultraluminous infrared galaxies

Galaxies with high luminosity in the far infrared, denominated "ultraluminous infrared galaxies" (ULIRGs) exhibit violent star formation owing to interaction processes. Such galaxies have huge quantities of dust thereby rendering their study in the visible quite impracticable. LIRIS will enable us to study such galaxies in order to investigate whether they really constitute the progenitors of elliptical galaxies.

2.6 Physical conditions of the gas, and the stellar kinematics and populations of starburst galaxies and AGNs

With LIRIS it is intended to study the physical conditions of the gas, and the kinematics and formation of stars in galaxies with massive star formation and high extinction in their centers. With this project two objectives may be addressed: first, the determination of the origin and age of the bursts of star formation in isolated dwarf irregular galaxies, and secondly the study of nuclear starbursts in galaxies that show bars and/or AGNs. In this context, the study of the underlying stellar populations may be undertaken using the CO absorption band at 2.3 m m, as well as the study of the physics of the gas with lines such as FeII at 1.65 m m, Brg and HeI at 2.06 m m and H2 at 2.12 m m. With the same spectral lines we can determine the gas stellar kinematics, which will give us the dynamics of the nuclear zones since such measurements in the visible wavelength range are severely affected by dust. Such observations will help us to understand the starburst phenomenon.

2.7 The fundamental plane of high-redshift elliptical and S0 galaxies

The Hubble Space Telescope has shown that the morphological classification of galaxy clusters at redshift z=0.5 can be undertaken. The colour-magnitude diagram of one of these clusters reveals elliptical, S0 and a large population of late spiral and irregular galaxies. LIRIS will allow the study of the stellar dynamics and populations of elliptical and S0 galaxies in such clusters. Hence, spectroscopic measurements in the IR could provide the velocities of dispersion and the spectral indices of the stellar component of such galaxies using such species as the CaII triplet. With such measurements, we could obtain the fundamental plane of the high-redshift elliptical and S0 galaxies and compare it with the nearby galaxies in order to understand their formation and evolution.

2.8 Redshift measurements of infrared sources

The large-scale infrared surveys such as 2MASS and ISO will provide an enormous quantity of infrared sources, from stellar objects to galaxies, which will have to be classified. In this context, LIRIS provides us with a very useful tool, since with the expected limiting magnitude we could obtain the spectroscopic counterparts of the said sources. For example, given the limiting magnitude of the 2MASS survey (K=15) we could obtain an S/N=20 spectrum on the WHT with some 10 minutes? integration time.


The aim of this project is to build a common-user near infrared instrument for the Cassegrain (f/11) focus of the WHT. The proposed instrument will have the following characteristics:

The new HgCdTe Hawaii detector will be used to cover the wavelength range from 0.9 to 2.5 m m (the lower limit is desirable from an astronomical point of view because it would be very interesting to be able to work in the range 0.9-1 m m with acceptable efficiency, since this would allow the coverage of a region of the spectrum which up to now has been studied very little, owing to the low efficiencies of CCDs in this wavelength range). The use of a large detector format (1024x1024 array with a pixel size of 18.5 m m) will allow a field of view of 4.2x4.2 arcmin with a plate scale of 0.25 arcsec/pixel at the f/11 telescope. Data about the characteristics and astronomical applications of the Hawaii focal plane array have been published elsewhere1.

Cold coronographic and apodization masks will be used in the focal and pupil planes to reduce the effects of scattered and diffracted light around bright point sources. Slits, coronographic masks and multislit masks will be situated at the telescope focal plane. All the slits and masks will be mounted on the same cold wheel to avoid the problems related with the small depth of focus at the Cassegrain position. Wheel diameter will be designed as large as possible to allocate a minimum of 5 slits, 1 coronographic mask, 9 multislit masks and 1 blank with the required size for imaging.

Grisms will be used to attain spectral resolutions of 1000 and 5000 over the full spectral range. At resolution 1000 it will be possible to observe simultaneously both the Z and J bands and the H and K bands. Grisms will be mounted near the pupil plane. Positions to allocate simultaneously a maximum of 10 grisms will be provided. A total of 19 filters, including narrowband filters and the standard Z, J, H, K and K? broadband photometric filters, will be provided. Filters will be mounted in a maximum of two filter wheels.

A warm half-wave and quarter-wave modulator will be situated in a rotating assembly outside the instrument window and will be used in combination with two Wollanston prisms and four wire-grid analyzers mounted in a cold wheel situated behind the grism wheel for spectropolarimetric measurements. An apodization mask with a central obscuration and spider vanes to block the thermal emission from the telescope structure will be situated at the Lyot position for coronographic measurements. A cold mechanism will be developed to switch between the Lyot stop and the apodization mask. The mask should rotate to compensate for the rotation of the image of the secondary mirror in an altazimuthal telescope.

Since all the mechanisms will be motorized, observers could switch rapidly between spectroscopic, polarimetric, coronographic and direct imaging modes.

One of the great challenges presented in the design of LIRIS is to be able to reach the maximum possible efficiency, which will necessitate a careful optimization of the instrument. A magnitude 17 should be reached in K for a point source with S/N=3, 1800 s of integration, seeing of 0.5 arcsec and a spectral resolution of 1000.

We expect to have the first comissioning at the telescope before the end of 2000.


LIRIS will be a classical collimator/camera design. The optical components of LIRIS can be divided in five subsystems: warm polarimetric plates, window, slits and focal plane masks, collimator, pupil optics (filters, grisms, apodization masks, Wollanston prisms and wire-grid analyzers), camera and detector. The image quality over the entire field-of-view must ensure that the size of the mean seeing disk observed on the ORM (0.5 arcsec in K) will increase less than 10% in the imaging mode and that the FWHM of the 0.5 arcsec slit image will be lesser than 2 pixels in spectroscopic mode. A single camera system will be used to obtain the desired plate scale of 0.25 arcsec/pixel, selected to provide a good sampling of the typical seeing disk at 2.2 m m. Optical system will be based on lenses rather than mirrors to facilitate the design, manufacturing and alignment.

The detector will be mounted in a cold translation mechanism to compensate for non-achromaticity in the observing spectral range. Total translation range will be ± 2.5 mm along the optical axis.

A primary goal for the design of the lens system is to attain the maximum optical efficiency. Lens design (including the selection of materials and anti-reflection coatings) and lens manufacturing will fulfil this main objective. Optical efficiency values for LIRIS optics (including grisms) larger than 30% at a spectral resolution of 1000 in the K band are expected to be obtained.

A pupil lens will be included in the design for direct imaging of the entrance pupil for alignment purposes in the lab and during the first test run at telescope. A Hartmann mask will be included in the grism wheel to focus the instrument.

A stand-alone calibration unit will be designed for LIRIS.


All cold mechanisms will be based on wheels and will be as identical as possible to minimize costs and time of design, manufacturing, integration and verification. The cold plate and the structural support connecting it with the warm walls of the dewar must have the highest possible mechanical stiffness to guarantee a maximum flexure of ±4 m m/hour measured at the detector plane with respect to the telescope focal plane. A detailed finite element analysis will be made to minimize the flexure and the heat conduction of the structural support and to model the gravity induced deflections of the cold plate under operating conditions.

A single wheel should be used at the telescope focal position to mount the slits, coronographic masks and multislit masks. The diameter of this wheel will be optimized during the design to allocate the maximum number of holes. Minimum number of positions is 16, according to §3

LIRIS mechanisms will be completely modular. All mechanisms within LIRIS will be interfaced to the control electronics using the same components: motors, microswitches and connectors. All mechanisms will be driven by stepping motors mounted inside the cryostat to avoid the presence of any mechanical feedthrough.


The cold plate assembly containing the optics, vacuum getter, slit wheel, analyzer wheel, filter wheels, grism wheels, detector focusing mechanism and associate electrical and electronics components will be situated inside the cryostat and cooled to below 100 K. The array will be cooled down to 77 K. The dewar will be designed to make optimum use of the space available at the Cassegrain position of the WHT.

A Gifford-Mc Mahon mechanical cooler will be used for long-term cooling stability of the cold plate, detector and radiation shield. An anti-vibration mounting will be designed to isolate the vibration of the cold head from the instrument cold plate.

The use of a LN2 can or a LN2 open circuit for cooling down LIRIS will be decided during the mechanical conceptual design phase. A heating system would be provided at the cold plate to speed the service time.


LIRIS control system can be divided into two areas, control of mechanisms position and control of the array readout and timing. Although there has been some previous debate over which controller should be used to operate the detector, the decision has been taken to use the San Diego State University (SDSU) CCD controller2 supplied by IR Labs. (Tucson, AZ). This decision has been based in an agreement established between the Instituto de Astrofísica de Canarias (IAC) and the Isaac Newton Group (ING) to develop a common system for the two infrared instruments (LIRIS/IAC and INGRID/ING) that will be sharing the Cassegrain focus of the WHT. Both instruments will interface to the common user instrumentation control system of the telescope.

The data acquisition system and mechanism control will be adapted as far as possible to the basic architecture used in the ING telescopes. This architecture is based on a standard model for control systems: host « network « controllers « I/O « instruments. In particular:



Several software subsystems will be required for LIRIS. The following is brief functional description of these subsystems:

Data Acquisition Software: The implementation of the Data Acquisition System (DAS) will be based on the SDSU solution. Their software will run on the Image Acquisition VME and on the CCD controller. This software will perform the following general functions: control the array operation, read image data from the array in the different modes and send it to the VME memory, image pre-processing on the VME, store images in VME-local disks (accessible through the network) and send images to the host computer for viewing and processing. The system will have the following features:

  1. It will be possible to read one image per second from the detector. For a 1024x1024 array and 16-bits per pixel, the data flow will be 2MB/s.
  2. Several observing modes will be available to the user: movie (for target acquisition), single observation and macros programmed by the user.
  3. Images will be stored in FITS format, and will include a header containing the object information (name, coordinates), telescope information (date, time, airmass) and instrument information (exposure time, readout mode, grating, filter, slit). Connection to the other subsystems will be required.
Mechanism Control Software: This software will run on the Instrument Control VME and will perform the following general functions: control and obtain status of all the mechanisms, receive commands and send mechanism status from/to the user.

Image Processing Software: This software will run on the host computer and will perform the following general functions: data reduction, offset application and obtain statistics.

Access Catalogues Software: The Access Catalogues Software will run on the host computer. This software will be used to select objects for acquisition and guiding in the most important catalogs such as: Digital Sky Survey, SAO, SKYVIEW, Flux standards, 2MASS (when available), DENIS (when available).

Telescope Control Software: LIRIS will interface to the WHT Telescope Control System for acquisition (pointing and offset in all observing modes) and guiding (connection with the WHT autoguider).

Graphic User Interface: The user interface will run on the host computer and will be responsible for reading user commands and displaying acquired infrared data and instrument current status.

User commands will be required for mechanism control, selection of the readout mode, selection of the observing and calibration modes, control of the image, processing and logging systems, accessing catalogues and telescope control

The display will show to the user the image, the spectrum, the statistics, the mechanism status, the current readout mode, the current observing mode, the status of the image processing and logging systems, the information from catalogues, the telescope information and the errors and alarms

An engineering user interface with low-level control capabilities will also be developed.


The authors wish to thank E. Atad-Ettedgui, P. Hasting and T. Purkins (Royal Observatory of Edinburgh), R. Rutten and P. Moore (Isaac Newton Group of Telescopes), C. Aspin (Nordic Optical Telescope) and R. Lenzen and P. Bizenberger (Max-Planck-Institut für Astronomie) for the stimulating discussions and their valuable advice during the early stages of this project.

  1.  K.W. Hodapp, J.L. Hora, D.N.B. Hall, L.L. Cowie, M.Metzger, E. Irwin, K. Vural, L.J. Kozlowski, S.A. Cabelli, C.Y. Chen, D.E. Cooper, G.L. Bostrup, R.B. Bailey and W.E. Kleinhans, "The HAWAII Infrared Detector Arrays: testing and astronomical characterization of prototype and science-grade devices" New Astronomy, vol. 1, p. 177 1996
  2. "SDSU/IR Labs CCD and IR arrays controller information""