“First Fringes” in Mid-Infrared Spectral Region with Two Giant Telescopes

Summary

Following several weeks of around-the-clock work, a team of
astronomers and engineers from Germany, the Netherlands, France and
ESO [2] has successfully performed the first observations with the
MID-Infrared interferometric instrument (MIDI), a new, extremely
powerful instrument just installed in the underground laboratory of
the VLT Interferometer (VLTI) at the Paranal Observatory (Chile).

In the early morning of December 15, 2002, two of the 8.2 m VLT unit
telescopes (ANTU and MELIPAL) were pointed towards the southern star
Epsilon Carinae and the two light beams were directed via the complex
intervening optics system towards MIDI. After a few hours of tuning
and optimization, strong and stable interferometric fringes were
obtained, indicating that all VLTI components – from telescopes to the
new instrument – were working together perfectly. Two more stars were
observed before sunrise, further proving the stability of the entire
system.

The first observations with MIDI mark one more important step towards
full and regular operation of the VLT Interferometer [3]. They are a
result of five years of determined efforts within a concerted
technology project, based on a close collaboration between ESO and
several European research institutes (see below). Now opening great
research vistas, they also represent several “firsts” in observational
astrophysics, together amounting to a real breakthrough in the field
of astronomical interferometry.

New views at mid-infrared wavelengths: MIDI is sensitive to light of a
wavelength near 10 micron, i.e., in the mid-infrared spectral region
(“thermal infrared”). This provides rich opportunities to study a wide
range of otherwise inaccessible, crucial astrophysical phenomena,
e.g., the formation of planets in dusty disks around newborn stars and
the innermost regions around black holes.

However, it is a great technical challenge to perform mid-IR
observations. This is first of all because the terrestrial
atmosphere, the telescopes, their mounts and, not least, the
complicated optics system needed to guide the beams the long way from
the telescopes to the MIDI instrument all glow bright at mid-IR
wavelengths. Thus, even the most luminous mid-IR stellar sources
“drown” in this bright background, calling for highly refined
observational methods and data reduction procedures.

Fainter objects with large telescopes: This is the first time
telescopes with mirrors as large as these have been used for mid-IR
interferometry. The use of the VLT giants at Paranal now allows
observing much fainter objects than before.

Sharper images with Interferometry: The distance between ANTU and
MELIPAL during these observations, 102 metres, is a new world record
for interferometry at this wavelength. The achieved angular resolution
is indeed the one theoretically possible with this instrumental
configuration, about 0.01 arcsec, better than what has ever been
achieved before from ground or space at this wavelength.

MIDI is the first of two instruments that will be placed at the focus of
the VLT Interferometer. It is a collaborative project between several
European research institutes:

  • European Southern Observatory (ESO)

  • Max Planck Institut fuer Astronomie (MPIA) (Heidelberg, Germany)

  • Netherlands Graduate School for Astronomy (NOVA) (Leiden, The Netherlands)

  • Department of Astronomy – Leiden Observatory (The Netherlands)

  • Kapteyn Astronomical Institute (Groningen, The Netherlands)

  • Astronomical Institute, Utrecht University (The Netherlands)

  • Netherlands Foundation for Research in Astronomy (NFRA) (Dwingeloo, The Netherlands)

  • Space Research Organization Netherlands (SRON) (Utrecht, Groningen; The Netherlands)

  • Thueringer Landessternwarte Tautenburg (TLS) (Germany)

  • Kiepenheuer-Institut fuer Sonnenphysik (KIS) (Freiburg, Germany)

  • Observatoire de Paris (OBSPM) (Paris, Meudon, Nancay; France)

  • Observatoire de la Cote d’Azur (OCA) (Nice, France)

The first observations with MIDI will now be followed up by thorough
tests of the new instrument before it enters into regular service. It
is planned that the first community observations will be performed at
the VLTI in mid-2003. Great efforts have gone into making observations
with this complex science machine as user-friendly as possible and,
contrary to what is normally the case in this technically demanding
branch of astronomy, scientists will find interferometric work at the
VLTI quite similar to that of using the many other, more conventional
VLT instruments.

A wonderful moment

[PR Photo 30a/02 – PR Video Clip 03/02 – PR Photo 30b/02]

Captions: PR Photo 30a/02 shows the “first fringes” of the star
Epsilon Carinae, as obtained at VLTI with the new MIDI instrument at
the mid-infrared wavelength of 8.7 micron. PR Video Clip 03/02 shows
the signature of the interference fringes, as they appear on the
computer screen when the optical path difference between the two arms
of the interferometer is slowly varied. From left to right, the
complex system of the VLTI delay lines changes this difference over a
length of about 1 mm (or 1/200,000 of the total path length of about
200 metres!). At each intermediate position, the MIDI instrument scans
the signal for the presence of fringes, by means of a digital
filtering algorithm. At the scan position in which fringes are
present, the signal increases dramatically: this is therefore the
position at which the light beams from ANTU and MELIPAL travel exactly
the same distance. PR Photo 30b/02 is a photo of the group responsible
for the MIDI installation and first tests, taken inside the VLT
Control Building, right after the successful “First Fringes” in the
early morning of December 15. From left to right: Front row
(sitting/kneeling) Julio Navarette, Lorena Faundez, Markus Schoeller,
Andrea Richichi – Back row (standing) Francesco Paresce, Andres Pino,
Uwe Gaser, Olivier Chesneau, Christoph Leinert, Andreas Glindeman,
Walter Jaffe, Sebastian Morel, Richard Mathar, Pierre Kervella, Eric
Bakker.

Another vital step has been accomplished as planned towards full
operation of the ESO Very Large Telescope (VLT) and the associated VLT
Interferometer (VLTI) at the Paranal Observatory in Chile, one of the
world’s foremost astronomical facilities. Indeed, plans had been made
more than one year ago for this milestone event to take place at the
end of 2002.

In the early morning of December 15, 2002, at 02:45 local time (05:45
UT), a team of astronomers and engineers from Germany, Netherlands,
France and ESO celebrated the first successful combination of
mid-infrared “light” beams from ANTU and MELIPAL, two of the four
8.2-m VLT Unit Telescopes.

This special moment, referred to as the “First Fringes”, occurred when
infrared radiation at a wavelength of 8.7 micron from the bright star
Epsilon Carinae was captured simultaneously by the two telescopes
(situated 102 metres apart) and then directed via a complex optics
system towards the MID-Infrared interferometric instrument (MIDI), a
new, extremely sensitive and versatile instrument just installed in
the underground VLT Interferometric Laboratory. Strong interferometric
fringes, well visible on the computer screen to the delighted team,
cf. PR Photo 30a-b/02 and PR Video Clip 03/02, were obtained
repeatedly by the MIDI instrument and the recorded data were of
excellent quality.

A great achievement

This is the first time ever interferometry in the near-infrared 8.7
micron-band (technically: the “N”-band”) with large telescopes has
been accomplished and the first time at 100-m baselines.

For this to happen, it was necessary to keep the difference in the
length of the light paths from the two telescopes to the focus of the
MIDI instrument stable and equal to within a small fraction of this
wavelength during the observations, in practice to about 1 micron
(0.001 mm). The team spent the first few hours of the night tuning the
system, positioning the many optical components and optimizing the
various feed-back mechanisms that involve precision-guided mirrors
below the two telescopes and the so-called “delay lines” in the
underground Interferometric Tunnel [3].

After a few attempts and successive on-line optimization, modulated
“fringes” – the typical signature of interferometric measurements –
became visible on the screens of the instrument computers,
demonstrating conclusively the validity of the overall concept, cf. PR
Video Clip 03/02. The rest of the night was used to further trim the
VLTI and MIDI. The team also observed two other objects before
sunrise, the young binary star Z Canis Majoris and the enigmatic Eta
Carinae – for both, interferometric fringes were convincingly
obtained.

The perfection of all of the 32 optical elements needed to guide the
starlight towards MIDI for these observations contributed to this, as
did the availability of advanced user-friendly control software,
specially developed for the VLTI and its instruments in order to
facilitate the future observations, also by non-specialists.

Advantages of MIDI

With its high sensitivity to thermal radiation, MIDI is ideally suited
to study cosmic material (dust and gas) near a central hot object and
heated by its radiation.

In the case of astronomical observations in the visible spectral
region, such material is usually hidden from view because of a strong
obscuring effect that is caused by the dust it contains. Most optical
observations of star-forming clouds only show the dark contours of the
cloud and nothing about the complex processes that happen
inside. Contrarily, this obscuring effect of the dust is often
entirely insignificant at the longer mid-infrared wavelengths around
10 micron (0.01 mm) at which MIDI observes, allowing direct studies of
what is going on inside.

MIDI science targets

Thanks to interferometry and the large collecting surface of the VLT
telescopes, MIDI achieves unsurpassed image sharpness (about 0.01
arcsec) and sensitivity at these “revealing” wavelengths, promising
extremely detailed views, also of faint and distant objects. Clearly,
the associated opportunities for exciting research are almost
unlimited.

Some of the first targets for the fully operational MIDI instrument
will thus include the enigmatic dust rings now believed to be located
around giant black holes at the centers of quasars and strong radio
galaxies.

Equally interesting will be in-depth studies of those disks of matter
that are known to accompany the creation of new stars and from which
exoplanets are forming. And with MIDI, it will now be possible to
investigate the outer zones of the extended atmospheres of giant stars
where the dust grains form in the first place – those complex
particles that, loaded with water ice, minerals and simple organic
molecules, eventually move into interstellar space and later play a
crucial role in the formation of stars and planets.

MIDI – a new and powerful instrument for the VLT Interferometer

[PR Photo 30c/02]

Caption: PR Photo 30c/02 shows the MIDI instrument installed in the
VLTI Laboratory at Paranal. Easily recognizable are the massive
liquid-Helium cryostat in the background (which keeps the MIDI
detector at a temperature of less than 7 K), the optical table with
injection optics (that brings the two VLTI beams from the left of the
picture into the cryostat), and two small alignment telescopes
(yellow).

The MIDI instrument has been developed by a European consortium of
astronomical institutes, under the leadership of the
Max-Planck-Institut fuer Astronomie (MPIA) in Heidelberg
(Germany). Following the installation in 2001 by ESO of the VLTI test
instrument, VINCI, to verify and tune the exceedingly complex optical
system [3], MIDI is the first of two scientific instruments that will
be devoted to interferometric observations with the VLT Interferometer
during the coming decade. The other is AMBER which will combine three
beams from different telescopes and will be sensitive in the
wavelength region of 1-2.5 micron.

The MIDI instrument weighs about 1.5 tons and is mounted on a 1.5 x
2.1 m precision optical table, placed at the centre of the underground
VLT Interferometric Laboratory at the top of the Paranal mountain,
cf. PR Photo 30c/02. The large cube at the back of the table is a
vacuum vessel that allows cooling of the infrared detector and the
surrounding optics to temperatures of -270 to -240 degC (4K to 35K on
the absolute temperature scale), which is necessary for observations
at these infrared wavelengths.

Despite its large dimensions, MIDI has to be very carefully adjusted
to the light beams arriving from the telescopes, with initial
precision exceeding 0.01 deg (angles) and 0.1 mm (position). The
electronic equipment necessary to run the instrument is installed in a
separate room in order to reduce any disturbances from heat, noise and
vibrations to the lowest possible level. During the observations, the
astronomers operate the entire instrument, as well as the VLT
Interferometer, from a building below the mountain top, more than one
hundred metres away.

This state-of-the-art instrument is the outcome of a close
collaboration between several European research institutes [1],
greatly profiting from their combined expertise in many different
technological areas. This involves the construction of large
astronomical instruments for infrared observations, involving
operation in vacuum and at low temperatures (MPIA in Heidelberg,
Germany), designing and manufacturing optics for the extreme cryogenic
environment (ASTRON in Dwingeloo, The Netherlands), designing and
creating the complex software needed to run the instrument in a
user-friendly way (NEVEC in Leiden, The Netherlands, and MPIA), as
well as other specialised contributions from the Kiepenheuer-Institut
fuer Sonnenphysik in Freiburg (Germany), Observatoire de Paris-Meudon
and Observatoire de la Cote d’Azur in Nice (France), and Thueringer
Landessternwarte in Tautenburg (Germany). This wide collaboration was
carried out in close cooperation with and profiting from the
professional experience of ESO that has built and now operates the
Paranal Observatory, ensuring the proper interfacing between MIDI and
the VLTI needed for high-performance interferometric measurements.

Brief history of the MIDI project

Work on the mid-infraredinterferometric instrument MIDI started in
1997 when MPIA proposed to ESO to build such a facility that would
conform with ESO’s plans for interferometric observations with the VLT
telescopes and which would most probably become the first of its kind
worldwide.

Soon thereafter, the Netherlands Science Organization NOVA with ASTRON
and NEVEC and the other partner institutes in France, the Netherlands
and Germany joined the project. With Christoph Leinert and Uwe Graser
from MPIA teaming up to lead the project, more than two dozen
engineers, astronomers and students worked intensively for three and a
half years on the planning, design and production, before the
integration of this highly complex instrument could start at the
Max-Planck-Institut fuer Astronomie in Heidelberg. This took place in
September 2001 and was followed by a period of extensive instrumental
tests.

Much preparatory work had to be done at Paranal in parallel, to be
ready for a smooth installation of MIDI [3]. After a positive,
concluding status review of MIDI by ESO in September 2002, the many
parts of the complex instrument were packed into 32 big wooden boxes,
with a total weight of 8 tons, and sent from Heidelberg to Paranal by
air freight.

The installation of MIDI in the VLT Interferometric Laboratory began
as scheduled in early November. The first test measurements were
carried out during the first days of December with two 40-cm
siderostats, the same that were used to obtain “first fringes” with
the VINCI test instrument in March 2001, cf. ESO PR 06/01. These
initial measurements led to stable, good-quality fringes on the bright
stars Alpha Orionis (Betelgeuse) and Omicron Ceti (Mira).

The total cost of MIDI is of the order of 6 million Euros. Of this,
1.8 million Euros are for equipment, materials and optical parts, with
the remaining for salaries during the extensive planning, construction
and testing of this front-line instrument.

Some related technical achievements

Astronomical observations of electromagnetic radiation at mid-infrared
wavelengths near 10 micron are difficult, because this is the spectral
region of thermal radiation from our environment.

If our eyes were sensitive to that radiation, everything around us
would be brilliantly bright, including the sky at night, and no stars
would then be visible to the naked eye. Sensitive imaging detectors
for these wavelengths have become available during the past years, but
to work satisfactorily, they must be cooled to a very low temperature
around -265 degC (4K – 10K) during operation. Also the optics in front
of the detector must be cooled to about -240 degC – otherwise all
images would be immediately overexposed, due to the added thermal
radiation from those lenses and mirrors.

In practice, the technical solution to this fundamental problem is a
so-called closed-cycle cooler that works with high-pressure helium gas
and achieves the required low temperatures on several “cold fingers”
inside the instrument. However, the associated moving pistons cause
vibrations which must be reduced to a minimum by means of special
damping materials and connections for the cooler and the
instrument. Otherwise this motion would be detrimental to the
sensitive measurements, which require near-perfect mechanical
stability, to within a fraction of the infrared wavelength, i.e., to
0.001 mm (1 micron) or better.

Similarly, slight bending effects of the instrument parts during
cool-down from room temperature would also compromise the
measurements. This has been avoided by manufacturing the support of
all optical parts near the detector from one single, carefully
selected block of special aluminium.

Still, as the light from the star being observed falls on the detector
inside MIDI, it will be surrounded by strong thermal radiation from
the terrestrial atmosphere in this direction and all uncooled (“warm”)
mirrors in the light path. The transfer of the digitally recorded
images from the detector to the computer data storage must therefore
occur at very high speed, one image per 0.001 sec, and always be
strictly synchronized with a modulation inherent in the measurement
process.

This requires powerful, highly specialized and yet flexible
electronics – this crucial part of the new instrument was developed
over the past years at MPIA. With this and many other technical
innovations successfully completed, and with the first on-the-sky
observations just accomplished to the full satisfaction of the MIDI
team, this new, powerful instrument will soon be ready to enter into
new and unknown research territory. Hundreds of astronomers in the ESO
members countries and their colleagues all over the world are now
eagerly waiting to get their hands on this new facility.

More information

Information about MIDI and its many components is available at several
dedicated websites, including those at MPIA, NOVA, NFRA and
FLUOR. Photos are available in the related ESO press releases [3] and
in the VLT Photo Gallery.

Notes

[1]: This press release is issued in coordination between ESO and the
research institutes participating in the MIDI project in Germany (Max
Planck Institut fuer Astronomie (MPIA), Thueringer Landessternwarte
Tautenburg (TLS) and Kiepenheuer-Institut fuer Sonnenphysik (KIS)), in
the Netherlands (Netherlands Graduate School for Astronomy (NOVA),
Department of Astronomy – Leiden Observatory, Kapteyn Astronomical
Institute, Astronomical Institute, Utrecht University, Netherlands
Foundation for Research in Astronomy (NFRA) and Space Research
Organization Netherlands (SRON)) and in France (Observatoire de Paris
(OBSPM) and Observatoire de la Cote d’Azur (OCA)). German-language
versions are available from MPIA. A Dutch-language version is
available from NOVA.

[2]: The members of the MIDI team are listed at the corresponding MIDI
webpage. Key personnel: Christoph Leinert (MPIA – PI Project
Scientist) and Uwe Graser (MPIA – PI, Project Manager), Andrea
Richichi (ESO Instrument Scientist for MIDI) and Francesco Paresce
(ESO VLTI Project Scientist).

[3]: The progress of the VLT Interferometer (VLTI) and its many parts
has been described at the VLTI website, as well as in various ESO
press releases: ESO PR 06/01 (“First Light” in March 2001 and
explanation of the interferometric measurements), ESO PR 23/01
(observations with two 8.2-m telescopes in October 2001), ESO PR 16/02
(observations with four 8.2-m telescopes in September 2002) and ESO PR
22/02 (observations of several small stars, in November 2002).

Contacts

Christoph Leinert (Principal Investigator)
Max-Planck Institut fuer Astronomie
Heidelberg, Germany
Phone: +49-49 6221 528 264
email: leinert@mpia.de

Andrea Richichi (ESO VLTI)
European Southern Observatory
Garching, Germany
Phone: +49-89-3200-6803
email: arichich@eso.org

Jakob Staude (MPIA PR Dept.)
Max-Planck Institut fuer Astronomie
Heidelberg, Germany
Phone: +49 6221 528 229
email: staude@mpia.de

Richard West (ESO EPR Dept.)
European Southern Observatory
Garching, Germany
Phone: +49-89-3200-6276
email: rwest@eso.org