Nulling interferometry for the DARWIN space mission

C. R. Acad. Sci. Paris, t. 2, Série IV, p. 149–156, 2001Techniques astronomiques/Astronomical techniques

DO

SS

IER

INTERFÉROMÉTRIE OPTIQUE AU SOL ET DANS L’ESPACE

OPTICAL INTERFEROMETRY AT GROUND LEVEL AND IN SPACE

Nulling interferometry for the DARWIN spacemissionMarc OLLIVIER a, Jean-Marie MARIOTTI †b, Alain LÉGER c, Predrag SÉKULIC c,Jacqueline BRUNAUDc, Guy MICHEL c

a Max Planck Institut für Astronomie, Heidelberg, Germanyb DESPA, Observatoire de Paris, Meudon, Francec Institut d’astrophysique spatiale, Orsay, France

(Reçu le 20 décembre 2000, accepté le 20 décembre 2000)

Abstract. The DARWIN observatory will include an interferometric instrument dedicated to thesearch for and spectral analysis of extrasolar planets. This instrument is based on theprinciple of Bracewell’s interferometric coronograph or nulling interferometer, proposedin 1978, but at present never experimentally demonstrated in the thermal infrared with highextinctions.

We have shown that high interferometric extinctions can only be obtained with ‘classical’optical pieces if the beams are cleaned by optical filtering during the recombination.After a short reminder of the nulling interferometer, we present different techniques ofoptical filtering to perform it. Then, we present a laboratory experimental bench we haveconstructed, which allows us, at present, to exhibit a monochromatic stable interferometricextinction better than103. Finally, we discuss this method, applied to ground-basedtelescopes. 2001 Académie des sciences/Éditions scientifiques et médicales Elsevier SAS

interferometry / nulling interferometry / coronography / extrasolar planets / DARWINmission

Coronographie interférentielle pour la mission DARWIN

Résumé. L’observatoire DARWIN possèdera un instrument interférométrique dédié à la rechercheet à la spectroscopie de planètes extrasolaires. Cet instrument fonctionne suivant leprincipe du coronographe interférométrique de Bracewell, ou interféromètre en frangenoire, proposé dès 1978, mais jusqu’à présent, jamais démontré expérimentalement dansl’infrarouge thermique, avec de fortes extinctions.

Nous avons montré que de fortes extinctions interférométriques ne peuvent être obtenuesavec des optiques « classiques » qu’à condition de nettoyer les faisceaux par filtrage optiquelors de la recombinaison. Après avoir rappelé brièvement le principe de l’interféromètreen frange noire, nous présentons les différentes techniques de filtrage optique nécessairesà sa réalisation. Nous présentons également une expérience de laboratoire que nousavons montée et qui nous a permis jusqu’à présent de mettre en évidence une extinctioninterférométrique monochromatique stable de plus de103. Enfin, nous discutons cetteméthode, appliquées à des telescopes au sol, comme précurseurs des futurs observatoiresspatiaux. 2001 Académie des sciences/Éditions scientifiques et médicales Elsevier SAS

Note présentée par Pierre ENCRENAZ.

S1296-2147(00)01151-3/FLA 2001 Académie des sciences/Éditions scientifiques et médicales Elsevier SAS. Tous droits réservés. 149

M. Ollivier et al. OPTICAL INTERFEROMETRY AT GROUND LEVEL AND IN SPACE

interférométrie / interférométrie en frange noire / coronographie / planètes extra-solaires / mission DARWIN

1. Nulling interferometry: optical requirements and demonstration strategy

Five years ago, when the study of DARWIN [1] began at ESA, several key-points of the mission werealready identified [2,3]. Among these key-points, the question of an efficient rejection of stellar light wascrucial. We decided to study this aspect from the point of view of optical constraints. The question of thearray configuration would be studied in parallel.

The concept of nulling interferometry on which the starlight rejection is based, was proposed byBracewell in 1978 to face the problem of the detection of faint objects near very bright ones, typically aplanet a fraction of arcsec away from a star [4]. It consists in introducing aπ phase shift in one arm of a twotelescope-interferometer to create an on-axis destructive interference — a zero-transmission zone — andan off-axis constructive interference — a full transmission zone. The angular position of the constructiveinterference is set by the baseline length of the interferometer and the orientation of the array.

In order to be efficient, the interferometer must have an internal extinction comparable to the contrast,that is to say105 to 107 in theN band. In addition, this extinction must be stable.

High values of extinction mean a great similarity of all the incoming wavefronts to be combined,concerning their amplitude, phase and polarization distributions. However, the different optical paths ofeach beam introduce differential defects on the wavefronts. The consequences are an immediate decreasein the quality of the null at the interferometer output.

We have particularly focused on the intensity and phase constraints. For instance, to get a106 rejectionrate with a two beam-interferometer, the intensity and the phase on each point of the beam combiner mustnot differ from one wavefront to the other by more than 0.4% and2 · 10−3 radian, respectively [5]. Amongthese specifications, the phase limitation is the more stringent: translated into optical constraints, it leadsto a wavefront quality better thanλ/3000 at the working wavelength (10 µm). Even in the optical range,the constraint is stillλ/190. Such constraint can be achievable at reasonable costs on small optical pieces,but not on several 1.5 m class telescopes. This last point was the major difficulty we faced 4 years ago andcould have killed the project.

A solution to this problem was proposed by Mariotti [6] developing an idea of Shao [7] for a similarproblem and consists in the use of optical filtering of the beams to be combined. Two kinds of optical filtercan be considered.

• Spatial filters: pinholes placed in the focal plane of a lens are spatial filters. They eliminate the highspatial frequencies of the wavefront. Their main drawback is that they are efficient only if the defects aremuch smaller than the pupil size. Spatial filtering is thus inefficient to correct low order defects such asdifferential piston, tilt defocus or other geometrical aberrations [8].

• Modal filters: single-mode waveguides perform a modal filtering. At the input of the guide, the beamis reduced over the base of its modes. Since it is single-mode, only the fundamental mode propagates.Whatever the input beam can be, the output beam shape is similar to the fundamental mode. The intensityat the output is directly linked to the injection condition, and the similarity of the beam to the fundamentalmode of the guide. Thus, the amplitude and phase defects are theoretically transformed at the output into aglobal amplitude and phase variation of the guide mode. This kind of optical filtering allows us to correctoptical defects over the complete range of scales [9]. Single-mode waveguides are already available inthe visible and in the near-infrared as fibres [10] or integrated optics waveguides [11]. They should beavailable in the near future in the thermal infrared, also as fibres [12], hollow waveguides, or integratedoptics waveguides with IR glasses (chalcogenides) that are under development [13].

150

OPTICAL INTERFEROMETRY AT GROUND LEVEL AND IN SPACE Nulling interferometry for DARWIN

DO

SS

IER

Even if an optical filtering device is used, getting a stable, deep interferometric extinction in a broad bandin the thermal infrared (DARWIN and its USA analog should observe between 6 and 18 µm) is not trivial.The difficulties are linked to:• the nulling interferometry principle itself that requires servo-controlled systems to warrant the stability

of the extinction;• the working wavelengths, in the thermal infrared, where the science signal is contaminated by thermal

emission, and where optical components are not as developed as in the visible or near infrared spectraldomain, particularly single-mode waveguides;

• the large wave-band over which the null should be simultaneously (more than an octave). Thisspecification requires us to develop achromatic devices such as phase-shifters, intensity compensatorsor optical filters. In the thermal infrared, such devices have sometimes still to be invented.In light of this, we decided on a step by step approach. We wanted to demonstrate the efficiency of optical

filtering in the thermal infrared, and we elaborate first a monochromatic demonstration strategy [14].Several approaches have been made later. A group at JPL chose to work in the visible, with a modified

compact Michelson interferometer, including an achromatic phase-shifter working by reflection [15] anda single-mode fibre as optical filter. Using a laser source this group obtained first a monochromatic105

rejection at 0.680 µm [16], then a 104 rejection with a 18% bandpass source [17]. Their phase shiftingtechnique is transposable in the thermal infrared. Another group at the University of Arizona chose to workdirectly in the thermal infrared, in order to get a demonstration instrument on a ground based-telescope.Controlling the thickness of the beam combiner and compensator, they could get a rejection of 24 in theN band onα Tau, using 2 mirrors of the MMT [18]. At present, they are developing, in the context ofthe TPF mission definition, an achromatic beam combiner where the phase shifter is based on balanceddispersive materials [19].

We have built at the IAS a laboratory interferometer working at 10.6 µm to show that high and stablerejection could be obtained in the thermal infrared if optical filtering is used. When we began to designthe experiment, optical waveguides working at 10 µm did not exist. However, we considered they shouldbe available in the coming years. We finally decided to design an interferometer that could work with bothspatial or modal optical filters and adopted the following strategy:• control of the low order defects: piston, tilt, and defocus of the wavefront, global equalization of the

intensity of the beams;• optical filtering of the higher order amplitude and phase defects.

This strategy led to a concept of interferometer described infigure 1.

2. Experimental setup and first results

The interferometer is a monochromatic modified Mach–Zehnder interferometer where delay lines, tip-tiltcorrector and a IR flux balancing device have been added. The output signal is sent to an optical filteringdevice where the superimposed wavefronts are cleaned. The rejection rate is measured as the ratio ofconstructive to destructive interference intensities as the interference state is modulated by modifications ofthe optical path difference.

The interferometer is working in the infrared (CO2 laser at 10.59 µm), and is servo-controlled using atunable visible laser. All its optical elements areλ(vis)/20 gold coated mirrors for reflective optics andZnSe or BaF2 plates for refractive optics.

Each sub-system has been designed and realized to get a 106 final rejection. Thetablesummaries thecharacteristics of the sub-systems and the error budget of the setup. The path difference is controlled bya delay line that also allows aπ phase-shift by aλ/2 optical path difference. The tip-tilt corrector allowsthe perfect angular and lateral superposition of the two beams on the combiner. The IR flux balance deviceallows us to equalize the beam intensities. The optical filtering bench allows the positioning of spatial ormodal filters. A complete description is available in French [5]; a description in English should be publishedsoon [20].

151


Figure 1. Principle scheme of the demonstration experiment. It is a classical Mach–Zehnder interferometer, modifiedby the implementation of intensity and wavefront correctors such as delay lines, orientation and intensity correctors,

or an optical filtering bench. All these devices are servo-controlled.

The spatial filters we use are simple metallic pinholes, whose diameter varies from a few hundreds toa few tens of microns. The use of single-mode waveguides is also planned but the efficiency of this methodis still limited by the 10 µm fibre technology. The absorption of the materials (at present 0.1 to 0.3 dB/cmfor IR single mode fibres) limits the length of the filter to a few centimetres while a 1.5 m long fibre appearsnecessary to reduce all the non-propagating modes by a factor of10−4 [21].

The experiment has been completed at the beginning of October 1999 and is now under development anduse.

The first measurements of contrast have been performed until now, using spatial filters only. Thefirst recordings of rejection rates allowed us to check the efficiency of the optical filtering. Using twobeams, whose intensity ratio had not been balanced(I1/I2 = 0.95), limiting theoretically the rejection rateachievable to 6000, we have measured rejection rates varying from 40 without spatial filtering to more than1000 using a 20 µm pinhole at the focus of a 50 mm ZnSe lens (around1/3 rd of the Airy disk).Figure 2shows the null we obtained using a 20 µm pinhole. It is interesting to notice how stable the null is. The pathdifference setting was performed manually before recording, and all the servo-control loops were then open.The stability of the null is obtained by the use of a protective enclosure where a top-oriented temperaturegradient had been imposed to limit the appearance of turbulence cells over the setup. As shown onfigure 2,the null was obtained during more than 3 minutes without any re-setting. This null duration was limitedby the number of samples recorded. A longer null can be obtained, providing the mechanical and thermalenvironment of the bench is stable.

152


DO

SS

IER

Table. Main characteristics of the interferometer sub-systems

Sub-system Parameter Required performance Measured performance

Laser: power 150 to 250 mW

stability ∆λ1/2/λ � 4 · 10−8 ∆λ1/2/λ = 10−8

Science beam: waist radius ω0 = 4.8 mm

divergence ∼ 0.07 mrad

Rayleigh length � 5 m 6.8 m

Flux balance device: range ±3% ±3%

resolution ∆I/I � 10−3 ∆I/I � 10−4

Delay lines: range ±2 cm

accuracy 3.5 nm (ptv) 3 to 5 nm (rms)

lateral drift 5 µm

angular drift 3 arcsec

bandpass 1.3 kHz

OPD resolution 3 nm 3 nm

Tip-tilt corrector: lateral superposition 10 µm � 2 µm

lateral dynamics ∼±100 µm ±80 µm

angular superposition ∼ 1 arcsec ∼ 1 arcsec

dynamics several arcsec 10 arcsec

bandpass � 1 Hz 3 Hz

HgCdTe detector: detectivity 2 to4 · 1010 cm·Hz1/2·W−1

NEP 1.5 to3 · 10−13 W·Hz−1/2

Pyroelectric detector: detectivity 3 · 108 cm·Hz1/2·W−1

NEP 4 · 10−10 W·Hz−1/2

Total transmission ∼ 1%

The experiment we have built is now ready to work as a test-bed for optical filtering devices. Therejection rates we obtained during the first measurements are very promising. The next steps will be toclose simultaneously every servo-control loop, and to get a perfect equalization of the intensity in everyarm of the interferometer. This operation requires us to check carefully all the settings, in order to avoida dramatic loss of photons through the system (the overall transmission of the interferometer is around 1%).

Finally, the use of optical fibres should allow us to enhance the quality of the optical filtering. This workis necessary before new considerations to increase the spectral range of the experiment by the introductionof achromatic devices.

3. Nulling interferometry development and future projects

The Nulling interferometry was first proposed as a method to look directly for extrasolar planets [4]. TheDARWIN mission [1] in Europe and the Terestrial Planet Finder project [22] in the USA, both using aninstrumental concept derived from the Bracewell’s interferometer, should achieve this goal around 2015–2020. The characteristics of these instruments are ambitious:• a104 to a fraction of106 rejection rate is required over the spectral range 6–18 µm;• the beam combiner should manage 4 to 6 telescope beams;• the system should be configurable, self-controlled in space, requiring free flying telescopes.

153


Figure 2. Evolution of the intensity at the output of the system. The signal extinction is obtained by aλ/2 pathdifference. This recording shows successive constructive and destructive interference. The extinction, better than103 ,

is stable over more than 3 minutes without any setting.

The development of such complex observatories requires the development of preliminary instruments.In that way, our laboratory test-bed can be considered as one of the first approaches that should lead to thisnew generation of instruments. Our setup cannot be used directly as an astronomical instrument. The twowavefronts are generated by amplitude splitting, and the beam-combiner compensates the beam-splitterphase-shift. In an astronomical interferometer, the two wavefronts are initially sepated and a differentstructure should be used to compensate the beam-combiner phase-shift. The setup proposed by Serabynet al. [16] is one of the possible solution.

A next step will be the realization of instruments adapted to ground-based interferometers, VLTI, LBTI orKeck-I. These instruments will observe within a spectral band reducing the range over which the extinctionshould be done. Their interest is double:• technological, as a tool for the development of the future space missions [23];• scientific, as a tool to study the exo-zodiacal clouds around the stars or stellar environments [23,24].

However, ground-based instruments have to face a strong supplementary difficulty: the effects ofatmospheric turbulence. Even if optical filtering is used and only the central part of the the Airy figureof the star is kept, the principle of interferometric extinction, as mentioned before, requires high qualitywavefronts, in terms of intensity, phase and polarization. For instance, without optical filtering, images witha Strehl ratio of 60% lead only to a rejection of about 3 of the star light. The observation of exo-zodiacaldisks, or dark companions requires much higher extinctions. A ground-based nulling interferometer shouldbe coupled to a performant adaptive optics system. Such systems have been proposed [25].

The choice of the observation band is the result of a trade-off between:• the type (and the temperature) of the astronomical object that should be observable;• the optical quality of the telescopes and beam combiner that increases with the wavelength;• the effects of atmospheric turbulence, that decrease as the wavelength increases;• the efficiency of adaptive optics correction, that is directly linked to the previous point;

154


DO

SS

IER

• the thermal emission of the instrument that increases with the wavelength to reach a maximum around10 µm. This last point limits the use of these instruments to the direct observation of giant planets neartheir parent star. In this case, the temperature of the object is higher, and near infrared spectral rangeappears more adapted.As an example, the trade-off adopted for the LBTI is to work in theN band [26], mainly to study

extra-solar zodiacal clouds. This choice is made easier by the small number of reflections (3 mirrors)before entering the beam combiner itself. The choice could be different for the VLTI, where the number ofreflections to reach the interferometric laboratory is five times bigger [27], increasing the level of thermalbackground.

Several concepts of nulling interferometer for ground-based telescopes have been proposed these re-cently, for single apertures, for instance, Gay and Rabbia’s Achromatic Coronographic Interferometer [28].Others are under study for multiple aperture interferometers [26,27].

Whatever the optical concept, the nulling interferometer appears to be one of the most powerfultechniques to study stellar environments, in the near future. As an interferometric method, it allowsobservations inside the Airy disk of a star which would be impossible with a single telescope and a classicalmask coronograph.

Acknowledgements.The authors are very grateful to their colleagues, who took place in the definition andrealization of the experiment. Special thanks to Patrick Bouchareine and Thierry Lépine from theInstitut d’optiquethéorique et appliquéein Orsay, Guy Artzner, Michel Decaudin, Patrick Boumier, Denis Barbet, Fernand Sarriau andBruno Crâne from theInstitut d’astrophysique spatialein Orsay, Jean Gay and Yves Rabbia from theObservatoire dela Côte d’Azur, Vincent Coudé du Foresto, Guy Perrin, Bertrand Mennesson, Pascal Bordé and Patrick Dierich from theDépartement d’études spatialesin Meudon, Pierre-Olivier Lagage from theService d’astrophysique du CEAin Saclay,Fabien Malbet and Pascal Puget from theLaboratoire d’astrophysique de l’observatoire de Grenoble. The authors wantto thank also Christian Chardonnet, Anne Amy-Klein and Alain Vanlerberghe from theLaboratoire de physique deslasersin Villetaneuse and Mr Courtois from theGroupe de spectrométrie moléculaire de l’université de Reimsfor theirhelp in the field of CO2 lasers.

This work was supported by theCentre national d’études spatiales(CNES), theCentre national de la recherchescientifique(CNRS), theUniversité de Paris-Sudand theEuropean Space Agency(ESA).

References

[1] DARWIN, The Infrared Space Interferometer, ESA-SP 451 (2000).[2] Bracewell R., Detecting nonsolar planets by spinning IR interferometer, Nature 274 (1978) 780.[3] Léger A. et al., Darwin Mission Concept (Proposal to ESA), 1993.[4] Léger A., Mariotti J.-M., Mennesson B., Ollivier M., Puget J.-L., Rouan D., Schneider J., Could we search for

primitive life on extrasolar planets in the near future? The DARWIN project, Icarus 123 (1996) 249.[5] Ollivier M., Contribution à la recherche d’exoplanètes — Coronographie interférentielle pour la mission

DARWIN, PhD thesis, Université de Paris-Sud (Paris XI), 1999.[6] Mariotti J.-M., Infrared space interferometry, in: C. Eiroa et al. (Eds.), Astrophysics & The Study of Earth-Like

Planets, Kluwer, Dordrecht, 1997, p. 219.[7] Shao M., Hubble extra-solar planet interferometer, in: Proc. SPIE, Vol. 1494, 1991, pp. 347–356.[8] Ollivier M., Mariotti J.-M., Improvement in the rejection rate of a nulling interferometer by spatial filtering, Appl.

Opt. 36 (1997) 5340.[9] Mennesson B., Ollivier M., Ruilier C., On the use of single-mode waveguides to correct the optical defects of a

nulling interferometer, J. Opt. Soc. Amer., submitted.[10] Monerie M., Alard F., Maze G., Fabrication and characterisation of fluoride-glass single-mode fibres, Electron.

Lett. 21 (1985) 1179–1181.[11] Malbet F., Kern P., Schanen-Duport I., Berger J.-P., Rousselet-Perraut K., Benech P., Integrated optics for

astronomical interferometry. I. Concept and astronomical applications, Astron. Astrophys. Suppl. Ser. 138 (1999)135–145.

[12] Perrin G., Ollivier M., Coudé du Foresto V., Spatial filtering with single-mode fibres for 10 µm interferometry, in:Proc. SPIE, Vol. 4006, 2000, pp. 1007–1013.

[13] Laurent E., Shanen-Duport I., Malbet F., Taillades G., Infrared waveguides for interferometry applications, in:Proc. SPIE, Vol. 4006, 2000, pp. 1090–1101.

155


[14] Ollivier M., Mariotti J.-M., Brunaud J., Michel G., Bouchareine P., Léger A., Artzner G., Malbet F., Puget P.,Coudé du Foresto V., Mennesson B., Coronographie interférentielle pour la mission DARWIN — Expérience devalidation en laboratoire, in: Proc. ICSO’97, CNES, Toulouse, France, 1997.

[15] Shao M., Colavita M.M., Long-baseline optical and infrared interferometry, Annu. Rev. Astron. Astrophys 30(1992) 457–489.

[16] Serabyn E., Experimental confirmation of deep nulling, in: Darwin and Astronomy, ESA-SP, Vol. 451, 1999,p. 101.

[17] Wallace K., Hardy G., Serabyn E., Deep and stable interferometric nulling of broadband light with implication forobserving planets around nearby stars, Nature 406 (2000) 700–702.

[18] Hinz P.M., Angel J.R.P., Hoffman W.F., McCarthy D.W., McGuire P.C., Cheselka M., Hora J.L., Woolf N.J.,Imaging circumstellar environments with a nulling interferometer, Nature 395 (1998) 251.

[19] Angel J.R.P., Woolf N.J., An imaging nulling interferometer to study extrasolar planets, Astrophys. J. 475 (1997)373–379.

[20] Ollivier M., Mariotti J.-M., Léger A., Sekulic P., Michel G., Interferometric coronography for the DARWIN spacemission — Laboratory demonstration experiment, Astron. Astrophys. Suppl. Ser., accepted.

[21] Leproux P., Pagnoux D., Reynaud F., How to determine the fibre length for an accurate spatial filtering in a spatialinterferometer, in preparation.

[22] Beichman C., Woolf N.J., Lindensmith C.A. (Eds.), Terrestrial Planet Finder, Origins of Stars, Planets and Life,JPL Publication 99-3, 1999.

[23] Ollivier M., Using LBT in preparation for DARWIN/TPF, in: T. Herbst (Ed.), Proc. Ringberg Conf. Science withthe LBT, Ringberg Castle, Germany, 24–28 July, 2000, in preparation.

[24] Woolf N.J., Terrestrial planet finder science with LBT, in: T. Herbst (Ed.), Proc. Ringberg Conf. Science with theLBT, Ringberg Castle, Germany, 24–28 July, 2000, in preparation.

[25] Angel J.R.P., Ground-based imaging of extrasolar planets using adaptive optics, Nature 368 (1994) 203–207.[26] Hinz P.M., Nulling Interferometry with the LBT, in: T. Herbst (Ed.), Proc. Ringberg Conf. Science with the LBT,

Ringberg Castle, Germany, 24–28 July, 2000, in preparation.[27] Glindemann A., Delplancke F., Kervella P., Paresce F., Richichi A., Schöller M., Ground interferometric searches,

in: Planetary Systems in the Universe, International Astronomical Union Symposium no. 202, Manchester,England, August 2000, in press.

[28] Gay J., Rabbia Y., Principe d’un coronographe interférentiel, C. R. Acad. Sci. Paris Sér. IIb 325 (1996) 51–56.

156

Documents

Nulling interferometry for the DARWIN space mission