Interferometry with the Large Binocular Telescope

J. Roger P. Angel , J. M. Hill , P.A. Strittmatter , P. Salinari and Gerd Weigelt

Steward Observatory , The University of Arizona , Tucson, AZ 85721

TABLE OF CONTENTS

1. Introduction
2. Scientific Programs
3. Adaptive Optics at tht LBT
4. Wide Field Aperture Synthesis
5. Bracewell Nulling
6. Imaging at 350 µm
7. Conclusion
8. Acknowledgements
9. References

Abstract:

The Large Binocular Telescope (LBT) has been designed for optical/infrared interferometry that combines high sensitivity and resolution. Key scientific projects will be deep, wide field infrared images of the Hubble Deep Field, with nearly ten times the resolution of the Hubble telescope, and the study of planets and dust in extra-solar systems, from their formation onward. A basic requirement for interferometry of faint objects is that the aberrations across the two 8.4 m telescopes be corrected for atmospheric phase errors. This will be done at the telescopes' secondary mirrors, so as to preserve the very low emissivity of the direct beam combination optics. Sodium lasers projected co-axially from above each secondary will allow wavefront sensing for correction of even the faintest objects. The two telescopes are rigidly mounted close together on a single alt-azimuth mount, to cover a large fraction of the u-v plane in a single exposure, with baselines continuous from 0 to 23 m. Field rotation during the night completes the cover, to allow recovery of images with the full resolution of a diffraction limited 23 m telescope. The beam combining optics will be cryogenically cooled to maintain the very low thermal background from only 3 warm reflections in total (primary, adaptive secondary, tertiary). For wide field imaging, the beams will be combined and stabilized so that in a long exposure every source across a ~ 1 arcminute field is crossed by interference fringes. From a set of such exposures the resultant deep image will have a resolution 0.02 arcsec in the 2.2µm K band. For high contrast studies of exo-planetary systems, a Bracewell nulling system will be used with superposition by division of amplitude, for 99.99% suppression of the stellar radiation.

KEYWORDS: interferometer, adaptive optics, telescope

1. INTRODUCTION

Adaptive optics and interferometry are techniques to overcome the blurring effects of the atmosphere and aperture diffraction, to recover the sharpest possible images. Their use at optical and infrared wavelengths has until now been restricted to observations of relatively bright objects with moderate-sized telescopes, and the two techniques have not been used together. Large telescopes are potentially very powerful when operated with adaptive correction, because image sharpness increases with aperture, along with light grasp. Correction of even the faintest objects will be possible with sodium laser guide star techniques. Furthermore, once atmospheric aberrations are removed, large telescopes have the potential for even higher resolution through interferometry.

The Large Binocular Telescope is being built to take maximum advantage of this potential. Its two 8.4 m telescopes are mounted side by side in a single alt-azimuth mount, with a 14.4 m center-to-center spacing. This configuration was chosen because of its profound advantages for studying faint objects at high resolution, and was felt to be valuable enough to sacrifice the possibility of independent telescope operation. Adaptive correction will be incorporated in the telescopes' secondary mirrors. The LBT is distinct from other ground based interferometers under construction that will use a combination of large and small telescopes as widely spaced, separately mounted units, with underground relays to bring the light to an interferometric focus when desired.

The LBT is an international project, with Italy and Germany as major partners in Europe, and the University of Arizona,Ohio State University and the Research Corporation in the USA. Figure 1 shows the telescope pier which was erected last year on Mt. Graham, Arizona. The first 8.4 m glass honeycomb mirror blank has been cast at the University of Arizona Mirror Lab, with its surface preformed by spinning to the steep f/1.14 curvature of the finished mirror. Contracts for the telescope and enclosure have been let, and completion with both mirrors in interferometric mode is expected in 2003.¹ A rendition of the telescope with the rotating enclosure removed is shown in figure 2.

Figure 1. The pier for the LBT, rising above the trees at 3200 m elevation on Mt. Graham, Arizona.

Figure 2. Computer rendition of the LBT on the pier. The interferometric beam combiners will be housed in the central dewar, shown as a dark cylinder linking the two mirrors. Tertiary mirrors located above the vertices of the primary mirrors direct the Gregorian foci into the dewar.

The advantages for interferometry of the common mount and close spacing of the LBT are:

1) Accurate images with resolution 2.7 times sharper than for a single 8.4 m telescope. This will be possible because the primary mirrors are set 6 m apart, slightly less than their diameter, so that the combined aperture covers all baselines from zero to the maximum spacing of 22.8 m. The u-v plane coverage of a single snapshot exposure yields elongated images. Because of field rotation of the alt-azimuth mount, several snapshots taken during the night will in general yield full u-v plane coverage to 22.8 m baseline, and round, high resolution images. Other interferometers with longer baselines from large separate telescopes will provide information about spatial structure at higher resolution than the LBT, but because of limited u-v plane coverage, the information needed to recover true images will be incomplete. Ancillary small telescopes planned for both the VLTI and Keck interferometers are aimed at remedying this deficiency, however, faint objects will remain inaccessible (see below). The baselines covered with big aperture telescopes by the LBT and the VLTI are complementary (0-23 m and 30-100 m respectively), so for brighter, complex objects the best images will come by combining data from both telescopes.

2) Because of its closely spaced large elements, the LBT will have the sensitivity to detect objects of low surface brightness, such as galaxies. The sensitivity of interferometers to objects of low surface brightness depends strongly on the ratio of baseline to diameter, and becomes very poor when this ratio is large (see sectiion 4.3 below).

3) A wide field of view for interferometric imaging is provided by the direct beam combining optics enabled by the common mount of the LBT. For example, it will image in the K band over a field of 1 arcminute, compared to a field of 2 arcsec for the separately mounted telescopes of the VLT, limited by the long, variable delay lines in the combining optics.

4) Thermal background emission at the interferometric focus is kept to an absolute minimum. The use of a common mount and adaptive correction at the secondary mirror provides a simple and direct path for beam combination. Only 3 reflections at the primary, secondary and flat tertiary mirrors are needed to bring the beams to the cryogenic beam combiner. For separately mounted telescopes, the large number of reflections needed to combine the beams results in high emissivity - estimated for the Keck Interferometer to be at 5 times that of the LBT system². The simplicity of the beam combination in the LBT is important not only for low background and high throughput, but also for practical reasons of reliability and cost.

5) The relatively short LBT baseline makes for deep star nulling The LBT is ideally suited for operation as a Bracewell nulling interferometer³, to detect planets and dust in extra-solar systems at high contrast ratio. Its short baseline makes for strong nulling of even the nearest solar type stars, whose disc is partly resolved and not well suppressed by long baseline systems. Further, especially high sensitivity to the thermal emission of dust and young or massive planets is achieved because of the low emissivity of the LBT.

Wide field imaging in the infrared with the LBT for has been reviewed earlier by Salinari.^.4 The LBT's potential for optical speckle interferometry has been discussed by Reinheimer et al.⁵ and for Bracewell interferometry to study extra-solar systems by Woolf and Angel.^6,7,8

2. SCIENTIFIC PROGRAMS

Based on these strengths, the LBT is to be equipped with two interferometric instruments, for wide field imaging and for Bracewell nulling. We list here examples of scientific programs for each instrument.

Compact galaxies in the HDF with z ~ 3 are about 0.3 arcsec in size, barely resolved by WFPC pixels of 0.1 arcsec. Already a single 8.4 m telescopes operating at the diffraction limit with adaptive optics will double this resolution. Diffraction limited images of 0.054 arcsec FWHM will be obtained in the K band, where the integrated magnitude of these galaxies is K ~ 21. Averaged over a typical galaxy size of 0.3 arcsec, their surface brightness at K is 18 - 19.5/ arcsec², and an integration of ~ 4 hours will be required to obtain a signal-to-noise ratio 5 per diffraction limited resolution element.

The resolution of the synthesized LBT images will be nearly 3 times sharper still, with a resolution size of 0.020 arcsec. The improvement is illustrated by the computer synthesis of reconstructed images⁹ shown in figure 5. As we discuss below, the price paid for the higher resolution is a 50 times increase in integration time. The LBT will be used to obtain deep 200 hour exposures as a key project. These will be unique in showing the morphology to 0.02 arcsec of the galaxies in a field of ~ 60 arcsec of the HDF (3000 x 3000 resolution elements).

At optical wavelengths, focus anisoplanatism prevents the sodium laser guide star method from recovering the coherent wavefronts needed for direct interferometry. But the full interferometric resolution of the LBT will still be recoverable for brighter objects, by the method of speckle masking. Galactic nuclei, especially those of the nearby Seyfert galaxies, are example of objects that will be accessible to speckle imaging with ~ 0.006 arcsec resolution in the V band (0.55 µm). Supermassive black holes are found at the center of galaxies, and show much structure and activity especially when matter is being actively accreted. The central infrared core of NGC 1068 has already been partially resolved by speckle masking imaging at the Russian 6 m SAO telescope, consistent with a nuclear torus or a circum-nuclear

scattering halo of 0.030 arcsec or 2 pc (Wittkowski et al.¹⁰). Similar structure are likely to be detectable at visible wavelengths in less dusty systems.

A detailed analysis of optical speckle masking interferometry applied to the LBT has been made by Reinheimer et al.⁵ In a simulation of excellent seeing conditions, they found that a well resolved image of a cluster of unresolved objects of V magnitude 15.6 - 17.1 and as close as 0.013 arcsec could be recovered from 200,000 specklegrams taken in an hour of integration spread through the night. With the addition of partial correction by laser guide star adaptive optics and longer exposures made with interferograms in many spectral bands simultaneously, diffraction-limited optical images of many galactic nuclei, as faint as 18th magnitude, will be accessible with LBT.

Dust debris around newly formed stars is observable in the mid and far infrared. For example, the disc recently imaged at 20 um wavelength around HR 4796 is 3.6 arcsec across (250 AU at distance 70 pc)5. The LBT will be able to image the "hole" in the disc, believed to be the result of planet formation, with the highest resolution at 20 um of 0.18 arcsec, or 13 AU. Newly formed massive planets radiating internal heat will be detectable at distances of up to 100 pc. The low emissivity of the LBT will give it especially high sensitivity to exo-planets in the window of low atmospheric emissivity at 3.8 µm where they are projected to be anomolously bright. When the star is nulled by the Bracewell method the first peak of high sensitivity is at angular separationλ/2d, i.e. 0.028 arcsec at 3.8 µm. From the models of Burrows et al.²² we find that exo-planets of mass 10 M_J will be observable at age 100 million years to 100 pc distance in orbits 3 AU, and at age 5 Gy to 10 pc at orbital radii 0.3 AU.

The limiting factor in the search and study of other stars from space for potentially habitable, Earth-like planets is likely to be the strength of zodiacal emission at around 10 µm wavelength. The dust in our own solar system, although far fainter than the emission from HR 4796 and β Pic, is nevertheless some 3 orders of magnitude brighter than any of the planets. If the zodiacal emission around the nearby solar type star candidates is even an order of magnitude brighter than in the solar system, its photon noise and spurious signals could potentially mask the detection of terrestrial planets by the Terrestrial Planet Finder mission.¹³

The LBT will be used in the Bracewell nulling mode to measure the strength of 10 µm dust emission down to solar system level. It is particularly well suited to this task because its response to zodiacal emission set by the nulling fringe pattern of the 14.4 m baseline reflects that of the TPF quite closely. It will detect the same outer component of zodiacal emission as will be seen by TPF. Its suppression of the star disc is also strong, because of the broad fringe spacing. Further, the very low thermal emission of the interferometer optics in the good 10 µm atmospheric window make for especially high sensitivity in the spectral region to be observed by TPF.

3. ADAPTIVE OPTICS AT THE LBT

Adaptive correction of the LBT is being built-in to its secondary mirrors. Following the original concept of Salinari et al.,¹⁴ the secondaries will be made of thin (2 mm) glass, and will be deformed by hundreds of force actuators to compensate for atmospheric wavefront distortion. The absolute shape of the deformation will be known, by reference to a thick, static plate of ULE glass directly behind the flexible glass membrane. Diffraction-limited images will thus be available to any infrared instrument, including the interferometric beam combiners, without the increase in thermal emissivity of conventional adaptive optics systems which use optical relays involving flat deformable and tip-tilt mirrors. Sodium laser guide star beams will be projected from above each secondary, so wavefront correction can be made for arbitrarily faint objects. Image motion and phase difference between the two telescopes will be obtained from measurements of either the program object or a field star. These measurements will be made in the infrared, where the image sharpening of the laser allows the use of faint field stars that can be found with good probability within the isoplanatic patch, even out of the galactic plane.¹⁵ Wavefront sensing will be with Shack Hartmann sensors, which measure slopes across the 8.4 m apertures. Because both the primary and adaptive secondary mirrors are continuous, numerical integration of the slopes yields accurately the wavefront displacement to be applied to the secondary mirror actuators.

4. WIDE FIELD APERTURE SYNTHESIS

4.1 Optical and detector configuration

In order to obtain high resolution imaging across a wide field by interferometry, an optical system is required that gives equal optical path lengths across both telescopes for any given star in the field. Suppose such an optical system lies between the wavy lines in figure 3a. The condition for an on-axis object to show interference is that all optical paths to the focus, through both telescopes, be equal in length. If objects in the nearby field are also to show interference, we require in addition the sine condition be satisfied, i.e. the rays incoming at distance h from the optical axis and converging to the focus at angle q to the axis must satisfy the condition sin qµ h. Another way to express this condition is to require that the exit and entrance pupils of the interferometer be similar.

If these conditions are satisfied, then the image plane of the interferometer has the character of that formed by a single large aperture telescope masked with circular openings corresponding to the two telescopes. In the absence of atmosperic aberration, the images of each point source in the field will be the Airy pattern of the individual aperture crossed by Young's fringes. The implementation for wide field imaging with the separate but co-mounted telescopes of the LBT is shown schematically in figure 3b. The individual afocal telescopes are represented by the Galilean telescopes with lenses L1 and L2. The beams are translated by mirrors M1 and M2 to bring their exit pupils side by side with the geometry of the entrance pupil. The camera lens L3 then forms the common image plane. The equivalent masked single telescope is shown in figure 3c.

Practical designs for the wide field beam-combiner at the folded Gregorian foci of the LBT covering near and mid-infrared wavelengths have been worked out by Byard and Bonaccini.¹⁶ A design concept for thecentral cryostat with optics to pick off reference stars has been developed by Salinari.⁴

a) the sine condition.

b) principle of beam combination for separate telescopes for a common wide field focus obeying the sine condition.

c) equivalent single telescope with 2 aperture stop.

This concept for wide field interferometric imaging with multiple telescopes was developed at the Multiple Mirror Telescope¹⁷, and was illustrated by the following experiment.¹⁸ A beam combining arrangement similar to 2b was used to combine the outputs of two 1.8 m telescopes separated by 2.5 m center to center, forming interference fringes with spacing 0.18 arcsec at 2.2 µm wavelength. Atmospheric fluctuations in path length and individual wavefront slope were corrected with piezoelectric actuators in a servo based on measurements of fringe motion in a reference star at the middle of the field. The resulting wide field of view for interference is illustrated by images of the star system g Andromeda, which consists of a bright star separated by 10 arcsec from a close binary (0.5 arcsec separation). A servo to correct phase fluctuations was closed about fringe motion measured in the brightest star with a fast scanning InSb detector, while a long exposure image of the fainter second and third stars was recorded with a separate integrating HgCdTe camera, as shown in figure 4. The Young's fringes crossing both star images are clearly visible

a) single telescope

b) double aperture

At the LBT, reference stars for the tilt and phase correction servo will be similarly selected by pick off mirrors in the image plane, while the fringe structure of all remaining objects over a field of 30 arcsec will be relayed to an integrating infrared detector mosaic of 5000 x 5000 pixels. The pixel size will be about 6 milliarcsec so as to adequately sample the 0.024 arcsec fringe spacing at 1.65 µm.

4.2 Image reconstruction

Figure 5. Simulation of a 0.64 arcsec square part of a 30 arcsec field obtained with the LBT at 1.65 µm wavelength⁹.

• Lower Left Single telesope image, corrected to the 8.4 m diffraction limit with adaptive optics.
• Lower Right Single exposure at the combined focus, showing horizontal Young's fringes crossing each feature.
• Upper Right Result of deconvulution of 3 images like (b), with the fringes at 45 degree intervals.
• Upper Left for comparison, simulation of the same galaxy images at the diffraction limit of a 22.8 m filled aperture telescope

Images taken at different times during the night show Young's fringes at different orientation, because of field rotation in the alt-azimuth telescope. From a set of such images a single high resolution image may be synthesized, as shown by Hege et al. ⁹ for K band images of a distant galaxy modeled for the LBT. For simplicity only three images were used, as would be obtained from exposures taken during the night with paralactic angles separated by intervals of 45 degrees. The u-v plane cover in this set of exposures are virtually complete, with all baselines from 0 - 23 m at nearly all position angles. We see that an image of a complex and diffuse galaxy is recovered with very significantly higher resolution than for a single 8.4 m aperture. For comparison, the whole galaxy image would be only 6 pixels across at the sampling of the WFPC.

We see that the LBT can recover the resolution of a filled aperture telescope of the same overall dimension, but we must consider the length of exposure needed to obtain good signal-to-noise ratio. As a starting point, we determine the exposure needed to image galaxies with a filled aperture telescope corrected by adaptive optics. The determining factor is the photon noise of the sky background, from OH emission and (in the K band) thermal emission of the telescope. The product of the solid angle W of the diffraction limited resolution element and the telescope collecting area A, given by AW = l², is independent of diameter. It follows that for a diffuse source like a galaxy, the photon rate for both the source and the background is independent of telescope aperture. and hence so is the signal/noise ratio. The background flux in the K band is such that detection at the 5 σ level a resolved galaxy of surface brightness K = 19/arcsec² will require an integration time of about 4 hours. In deriving this estimate, we have assumed an overall quantum efficiency of 0.2, a Strehl ratio for the adaptive optics system of 0.5 and a background brightness of K = 13.8 mag/arcsec². This estimate is in good agreement with Gillett and Mountain¹⁹ who find a limiting (10 σ) magnitude of K = 25.1 (18.9 mag/arcsec²) for a 2.8 hour integration with an 8 m telescope.

The sensitivity of an interferometer relative to a filled aperture telescope with the same resolution can be understood by realizing that it operates as a multiplex device to synthesize pixels smaller than the individual beam profile. If there is good sampling of the u-v plane, different pixels appear as orthogonal signals. So, like a Fourier transform interferometer, all the noise corresponding to the background flux in a the single aperture resolution element appears superposed on the signal for each high resolution pixel. The resolution element of the reconstructed image is dW = (l/b)² where b is the aperture of the corresponding filled aperture telescope. Assuming a source of approximately uniform surface brightness, the source photon flux in this resolution element received by the interferometer is reduced by a factor (2d/b)² compared to the filled aperture, the ratio of the receiving areas . The background photon flux, however, remains essentially the same. It follows that the signal-to-noise ratio depends as (d/b)², and the integration to reach given surface brightness increases as (baseline/aperture)⁴. Thus while the interferometer can increase resolution, the price paid is that integration time increases as the 4^th power of the resolution increase. For the LBT, the resolution increase is in the ratio 22.8/8.4, ie 2.7 times, and the exposure time is increased by 50 times, to 200 hours. Such long observations are still acceptable, given the unique scientific returns, but deep imaging of galaxies to higher resolution will not be practical for other, less compact interferometers.

5. BRACEWELL NULLING

A different optical configuration will be used at the LBT to implement the Bracewell's star nulling method. Its purpose is to image the circumstellar region at very high contrast by suppressing interferometrically the starlight. The wavefronts from two telescopes are exactly superposed with no relative tilt at a semitransparent beamsplitter, and the optical paths are configured so that at one of the two outputs the path difference is exactly half a wave, so cancellation takes place in detail across the whole pupil. This causes the entire star image, including its diffracted halo, to be suppressed. At the same time, if the two apertures are appropriately spaced, the wavefronts from the planet may add constructively, so its entire incident photon flux is transmitted to the focal plane. The second output of the beamsplitter is complementary, with all the star energy and none of the planet. For the 14.4 m separation of the LBT and at l = 10 µm, the first transmission peak adjacent to a nulled star is at an angular separation 0.07 arcsec. Thus the interferometer allows the isolation and spectral analysis of light from a faint companion that in the image of a single 8.4 m telescope is obliterated within the 0.25 arcsec diffraction limited image of the star. In the case of a faint, diffuse dust nebula around a star, we see a true thermal image at 0.25 arcsec resolution, with the star removed. This will be of particular value in the study of faint proto-planetary nebulae.

The Bracewell method was recently demonstrated at 10 µm wavelength by Hinz et al.,²⁰ using two 1.8 m telescopes of the MMT, separated 5 m center to center. An image of the dust nebula around Betelgeuse was recorded, with the star removed. The complete optical diagram, drawn out of scale for clarity, is shown in figure 6. A similar arrangement will be used at the LBT, with all the optics after the tertiary mirrors in a dewar. Note the geometry is quite different from the wide field interferometer of figure 3b. For nulling there is a single circular output pupil. No interference fringes are seen in the focal plane, they are located on the sky, and the star simply blinks on and off as the phase difference between the two beams is altered.

Figure 6. Schematic of the nulling interferometer at the MMT. The beam path-lengths are equalized by translating the beamsplitter vertically.

Two effects limit the extent to which the star image can be suppressed. The first is residual wavefront distortion. In the experiment of Hinz et al. only the phase and tilt differences caused by atmospheric turbulence were corrected, limiting the highest observed ratio of constructive to destructive image peaks to 25:1. But with the addition of adaptive optics and phase control at the LBT, residual aberrations will be reduced to the level where the peak reduction should be by a factor of ~ 50,000 at 10 µm for unresolved stars. This limit will not be reached for nearby stars because of the second effect. If the star's disc of diameter qs is partially resolved, a fraction (psqs/4l)2 is transmitted, where s is the element spacing. For a solar type star at 10 pc with diameter ~ 0.001 arcsec, the peak suppression from this effect alone would be limited to 33,000 for the LBT at 10 um, assuming ideal wavefront correction. For longer baselines, disc resolution will dominate over residual atmospheric errors. For example in the 3.8 µm band important for warm planet detection, the stellar disc suppression will be by a factor of only 140 for the 85 m baseline of the Keck interferometer, while still 4800 for the LBT. A detailed analysis of the sensitivity of different nulling interferometers, including the Terrestrial Planet Finder, to stars and zodiacal emission is given by Angel.⁸

6. IMAGING AT 350 µm

The LBT is located at a high site already used for sub-millimeter astronomy. Measurements taken at the nearby Heinrich Hertz 10 m telescope show the very low water vapor absorption needed for 350 µm observations occurs > 10% of the time. The LBT will be used at this wavelength for a few hours after dawn on the good mornings. Sources of low surface brightness will be targeted, taking advantage of the sensitivity of a well filled u-v plane that holds at sub-millimeter as well as shorter wavelengths. The resolution of the synthesized 22.8 m aperture will be 3 arcsec, and the telescope's optical quality and pointing accuracy (monitored optically) will be ideal. The morning switch-over will require simply a rotation of the tertiary mirrors, to bring the two beams from the two 8.4 m telescopes into a combiner dedicated for sub- millimeter work. Water column density measurements are made at the site every 30 minutes, 24 hours a day, so there will be some advance warning to bring observers to their computer terminals.

7. CONCLUSION

Radio astronomers have long chided their wavelength-challenged colleagues for not exploiting interferometry more effectively. This situation is now changing. In space, interferometry in the thermal infrared is the key to detecting Earth-like planets of other stars and searching them spectroscopically for signs of life.²¹ On the ground, the technique will come into its own at the new generation of very large telescopes, corrected for atmospheric aberrations so their strong signals will indeed interfere. Long baselines available at the separately mounted telescopes will be preferred for accurate astrometry and for imaging compact bright objects. But the LBT, with its unique, compact geometry that gives sensitivity to the faintest objects over a wide field, and its very low thermal background, will be most likely to move interferometry from a specialist area into the mainstream. The different interferometric methods to be used at the LBT for wide field and high contrast imaging have been developed and tested at the Multiple Mirror Telescope.

8. ACKNOWLEDGMENTS

Work described here has been supported by the Air Force Office of Scientific Research under grant F49620-96-1-0366 and by JPL under contract 961286.

9. REFERENCES

1. J. M. Hill and P. Salinari, "Large Binocular Telescope Project," SPIE Proc. Advanced Technology Optical/IR Telescopes VI," 3352, Kona, HI, 1998.
2. M Shao, private communication, 1997.

3. . R.N. Bracewell, "Detecting nonsolar planets by spinning infrared interferometer," Nature, 274, 780, 1978.

4. P. Salinari, "The Large Binocular Telescope Interferometer," Proc SPIE Optical Telescopes of Today and Tomorrow, ed. A. Ardeberg, 2871, pp. 564-574, Landskrona/Hven, Sweden.
5. T. Reinheimer, K.-H. Hofmann, M. Scholler and G. Weigelt, "Speckle masking interferometry with the Large Binocular Telescope," Astronomy and Astrophysics Supplement Series, 121, pp. 191-199, 1997.
6. . N. Woolf and J. R. P. Angel, "Use of the LBT to study zodiacal emission around other stars," Adaptive Optics, OSA Technical Digest Series 23, p. 44, Washington D.C., 1995.

7. J.R.P. Angel and N. Woolf, "The LBT: a unique scientific and technology precursor to planet finder," Planets Beyond the Solar System and the Next Generation of Space Missions, ed. D. Soderblom, 119, pp. 207-222, Astronomical Society of the Pacific Conference Series, 1997.
8. R. Angel, "Sensitivity of nulling interferometers to extra-solar zodiacal emission (EZE)," Exozodiacal Dust Workshop, eds. D. E. Backman, L.J. Caroff, S.A. Sandford, and D.H. Wooden, NASA/CP 1998-10155, 209-218, 1998.
9. E.K. Hege, J.R.P. Angel, M. Cheselka, and M. Lloyd-Hart, "Simulation of aperture synthesis with the Large Binocular Telescope," Proc. SPIE Conference on Optical Aperture Synthesis, ed. J. D. Gonglewski, 2566B, pp. 144-155, 1995.
10. M. Wittkowski, Y. Balega, T. Beckert, W.J. Duschl, K.-H. Hofmann and G. Weigelt: ''Diffraction-limited 76 mas speckle masking observations of the core of NGC 1068 with the SAO 6m telescope'', Astronomy and Astrophysics 329, pp. L45-L48, 1998.
11. D. Koerner, M. Ressler, M. Werner, and D. Backman, ApJ Letters, in press 1998.
12. R. Jayawardhana, L. Hartmann, G. Fazio, S. Fisher, Charles Telesco, and R. Pina, ApJ Letters, in press 1998.
13. J.R.P. Angel and N.J. Woolf, "An imaging nulling interferometer to study exta-solar planets," ApJ, 475, pp. 373-377, 1997.
14. P. Salinari, C. Del Vecchio, and V. Biliotti, 1993 "A study of an adaptive secondary mirror," Proc. ICO-16 Conference on Active and Adaptive Optics, ed. F. Merkle, Garching, 1993.
15. D. Sandler, S. Stahl, J.R.P. Angel, M. Lloyd-Hart, and D. McCarthy, "Adaptive optics for diffraction-limited infrared imaging with 8m telescopes," Journal of the Optical Society of America A, 11, pp. 925-945, 1994.
16. P. Byard and D. Bonaccini, "Optical design for interferometry with the Large Binocular Telescope," Proc. SPIE, 2200, p. 446, 1994.
17. E.K. Hege, J.M. Beckers, P.A. Strittmatter and D. W. McCarthy, "Multiple Mirror Telescope as a phased array telescope," Applied Optics, 24,#16, pp. 2565-2576, 1985.
18. M. Lloyd-Hart, P. Wizinowich, D. Wittman, D. Colucci, B. McLeod, R. Dekany, R. Angel, D. McCarthy, M. Rieke, and M. McCaughrean, "High resolution imaging at the Multiple Mirror Telescope using adaptive optics," Proc. SPIE Conference on Atmospheric Propagation and Remote Sensing, eds. A. Kohnle and W. B. Miller, 1688, pp. 527-535, Orlando, 1992.
19. F.C. Gillett and M. Mountain, "On the comparative performance of an 8m NGST and a ground-based 8m optical/IR telescope," in Science with the NGST, ASP Conference Series 133, p. 42, 1998.
20. P. M. Hinz, J.R.P. Angel, W.F. Hoffmann, D.W. McCarthy, Jr., P.C. McGuire, M. Cheselka, N.J. Woolf, and J.L. Hora, "Imaging circumstellar environments with a nulling interferometer," submitted to Nature, 1998.
21. N. Woolf and J. R. Angel, "Astronomical seraches for earth-like planets and signs of life," To be published in Annual Review of Astronomy and Astrophysics, 36, Annual Reviews, Inc., Palo Alto, CA, 1998.
22. A. Burrows, M. Marley, W.B. Hubbard, J.I. Lunine, T. Guillot, D. Saumon, R. Freedman, D. Sudarsky, and C. Sharp, "A nongray theory of extrasolar giant planets and brown dwarfs," ApJ, 491, pp. 856-875, 1997.

	http://medusa.as.arizona.edu/lbtwww/tech/interf98.htm