From Opto & Laser Europe March 2001
The European Space Agency is about to embark on its most ambitious project ever. Called Darwin, the
mission will search for planets that are outside our solar system and study them for any signs of life.
While this has been the aim of astronomers for many years, only now have advances in optical
technologies turned what once seemed like an impossible dream into an achievable reality. What makes
Darwin unique is the optical technique that will be used to find and study planets.
Anders Karlsson, study manager at the European Space Research and Technology Centre (ESTEC),
based in the Netherlands, explained: "Darwin is not an acronym, but it could stand for detection and analysis
of remote worlds by interferometric nulling.
"Numerous planets have been detected by indirect methods - for example, from spectral shifts in the
light from the parent star. However, these techniques cannot be used to detect Earth-like exoplanets, nor do
they allow spectroscopic studies. In contrast, Darwin will allow scientists to find planets and search for life."
In this project six telescopes will fly in a hexagonal formation around a central spacecraft. The craft will
combine the light from each telescope using nulling interferometry, which cancels out the light from the
bright parent star, leaving only the faint light from the planets around the star.
The light is then analysed in the thermal infrared spectrum, centred at 10 µm. This wavelength range is
chosen for two reasons: it provides optimal contrast between the planetary and stellar signals; and the spectral
features for life (CO2, H2O, O3, CH4) lie in the 6 to 18 µm
band.
Scientists hope to detect more than 100 Earth-like exoplanets, of which about one-third could be
observed in spectroscopic mode. However, astronomers will have to be patient: Darwin's provisional launch
date is not until 2014.
There are optical engineering challenges in every stage of the project - from controlling the optical path
difference between the arms of the interferometer with nanometre precision, to the spectroscopic analysis of
the planetary photons. And every stage involves using components or systems that have not yet been
developed.
"We need to make many technological improvements before the mission can take place," said
Karlsson. "We do not have materials that work in Darwin's spectral range of 5 to 20 µm, for example,
integrated optics. Another major challenge is achromatic phase shifting." A research team from the
Observatoire de la Côte d'Azur, France, is reviewing several techniques, but a decision has yet to be made as
to which is the best one to use.
Other technologies that still need to be developed are: detectors that work in the 6 to 18 µm range and
have active cooling that does not disturb the fragile set-up of the interferometer; techniques for controlling,
with nanometre precision, the optical path difference between two telescopes flying at up to 250 m apart; and
an effective filtering system that can clean the planetary photons before they are combined.
Thierry Viard, an optical engineer at Alcatel Space Industries in France, has been working on a feasibility
study for the Darwin project and believes that the filtering system is particularly important.
He told OLE:"The photons that are emitted by a planet and detected by the six telescopes are
weak compared with the parent star. The combination of the six beams has to be perfect in terms of
amplitude, wavefront quality and direction, or the instrument will not work. Every optical device in the hub is
involved in this procedure and any defect in the components can destroy the combined beams. This places
unrealistic requirements on the quality of the optical components. The project needs a more efficient filtering
system to reduce the quality requirements of the optical devices."
A modal optical filter could be the answer. However, to make one, researchers need a singlemode
optical fibre that works at 10 µm. Unfortunately, according to Marc Ollivier, a scientist at the Max Planck
Institute for Astronomy in Germany, this type of fibre does not exist. He said: "We have built a laboratory
interferometer that works at 10.6 µm to show that high and stable rejection can be obtained in the thermal
infrared if optical filtering is used.
"When we began to design the experiment, optical fibre that works at 10 µm did not exist. However,
we assume that it will be available in the next few years."
Researchers are hopeful that such fibre will soon be available in metre lengths. There is a demand, not
only from astronomers but also from the medical industry.
Another emerging technology that could be used on Darwin is the quantum-well infrared
photodetector (QWIP). Karlsson of ESTEC said: "We are concerned about finding a technology that will
meet our requirements. The detector must function over the 6 to 18 µm range and such devices do exist.
However, the main problem is cooling. While the temperature in space is sufficient to cool the optics to their
working temperature of 40 K, the detectors need active cooling methods to achieve a lower temperature to
decrease background noise.
"Most active cooling techniques induce vibrations that could disturb the interferometer. We are
looking at QWIP technology to help us to solve this problem."
Before any light can be detected, the central hub must ensure that all of the telescopes are in their correct
positions. Darwin will use three methods to achieve this: a generalized microwave ranging system, which can
determine distances with centimetre accuracy; a laser ranging device, which can only be used when the
spacecraft have been aligned to within 1 cm of their nominal array locations; and science object fringe
tracking, which uses the light collected by the telescopes to measure the optical path differences between the
six arms of the interferometer to provide pathlength stabilization with nanometre accuracy.
Laser interferometric metrology systems will be used to monitor optical path differences of up to 250
m with an accuracy of a few nanometres. Yves Salvade and colleagues at the University of Neuchâtel in
Switzerland commented: "Although classical high-resolution laser interferometers using a single wavelength
are well developed, this type of incremental interferometer has a drawback. Any interruption of the
interferometer signal results in the loss of the zero reference, so a new calibration would have to be carried
out, starting at zero optical path difference."
Consequently, the group has reviewed the different approaches to an absolute metrology system that is
based on multiple-wavelength interferometry. It believes that the availability of highly coherent lasers that
emit at approximately 1 µm should enable the generation of a chain of synthetic wavelengths of light that are
sufficient for the accurate control of the position of the six telescopes.
However, just because something is believed to be possible, it doesn't mean it is going to happen. For
Darwin to succeed, Europe's space and optics industries must collaborate more closely than ever before.
"Once we have overcome all of the challenges," said Alcatel's Viard, "all we have to do is to build a
system to communicate with the planets that we have found!"
The Darwin telescope array will operate in a nulling mode to enable it to search for planets around a star.
This technique uses a multiaperture stellar interferometer in which broadband destructive interference is
applied to the light of the parent star while the light from the planet escapes the destructive process and
remains detectable. The destructive process is performed by reversing the phase shift of one of the interfering
waves before it is combined.
For stellar and planetary observations this is more difficult because of the high contrast ratios between the
planets and their parent stars and owing to their very small angular separations.
The European Space Agency has asked Yves Rabbia and his colleagues at the Observatoire de la Côte d'Azur,
France, to review different methods of achromatic phase shifting for nulling interferometry. Rabbia told OLE:
"Achromatic phase shifting in this application is challenging because the process needs to be controlled over
every wavelength and the dephasing has to be precise enough to make the darkness very dark. Noise is also a
problem because it can mask the planet that you want to detect."
The team reviewed five methods: using dispersive materials; birefringence; focus-crossing phase-shift
properties; reversing the electric field vector; and phase changing at total reflection. However, each one had its
advantages and disadvantages.
One solution could be to design a system that combines the merits of two of the techniques, but, as Rabbia
pointed out: "A combined system could be too heavy and result in an increased payload." He believes that the
use of dispersive plates is the most promising solution because it gives an optical path difference that varies
linearly with wavelength over a given spectral range, which results in an achromatic phase shift. He also
points out that, compared with experiments in the visible range, the choice of materials for the infrared is not
rich; refractive indices are not as precisely known; and the machining of surfaces is not as well mastered.
Visit the Darwin website.
