Laboratory characterization of FIRSTv2 photonic chip for the study of substellar companions
K. Barjot, E. Huby, S. Vievard, N. Cvetojevic, S. Lacour, G. Martin, V. Deo, V. Lapeyrere, D. Rouan, O. Guyon, J. Lozi, N. Jovanovic, C. Cassagenettes, G. Perrin, F. Marchis, G. Duchêne, T. Kotani
LLaboratory characterization of FIRSTv2 photonic chip for thestudy of substellar companions
K. Barjot a , E. Huby a , S. Vievard b,c , N. Cvetojevic d , S. Lacour a , G. Martin e , V. Deo b,a , V.Lapeyrere a , D. Rouan a , O. Guyon b,c,f , J. Lozi b , N. Jovanovic g , C. Cassagenettes h , G. Perrin a ,F. Marchis i,a , G. Duchˆene e,j , and T. Kotani c,ba LESIA, Observatoire de Paris, Universit´e PSL, CNRS, Sorbonne Universit´e, Universit´e deParis, 5 place Jules Janssen, 92195 Meudon, France b National Astronomical Observatory of Japan, Subaru Telescope, 650 North Aohoku Place,Hilo, HI 96720, U.S.A. c Astrobiology Center of NINS, 2-21-1, Osawa, Mitaka, Tokyo, 181-8588, Japan d Universit´e Cˆote d’Azur, Observatoire de la Cˆote d’Azur, CNRS, Laboratoire Lagrange, France e Universit´e Grenoble Alpes / CNRS, Institut de Plan´etologie et d’Astrophysique de Grenoble,38000 Grenoble, France f College of Optical Sciences, University of Arizona, Tucson, AZ 85721, U.S.A. g California Institute of Technology, 1200 E California Blvd, Pasadena, CA 91125, U.S.A. h Teem Photonics, F-38240, Meylan, France i Carl Sagan Center at the SETI Institute, 189 Bernardo Av., Mountain View, CA 94043, USA j Astronomy Department, University of California at Berkeley, Berkeley, CA 94720, USA
ABSTRACT
FIRST (Fibered Imager foR a Single Telescope instrument) is a post-AO instrument that enables high contrastimaging and spectroscopy at spatial scales below the diffraction limit. FIRST achieves sensitivity and accuracyby a unique combination of sparse aperture masking, spatial filtering by single-mode fibers and cross-dispersionin the visible. The telescope pupil is divided into sub-pupils by an array of microlenses, coupling the lightinto single-mode fibers. The output of the fibers are rearranged in a non redundant configuration, allowing themeasurement of the complex visibility for every baseline over the 600-900 nm spectral range. A first version ofthis instrument is currently integrated to the Subaru Extreme AO bench (SCExAO). This paper focuses on theon-going instrument upgrades and testings, which aim at increasing the instrument’s stability and sensitivity, thusimproving the dynamic range. FIRSTv2’s interferometric scheme is based on a photonic chip beam combiner.We report on the laboratory characterization of two different types of 5-input beam combiner with enhancedthroughput. The interferometric recombination of each pair of sub-pupils is encoded on a single output. Thus,to sample the fringes we implemented a temporal phase modulation by pistoning the segmented mirrors of aMicro-ElectroMechanical System (MEMS). By coupling high angular resolution and spectral resolution in thevisible, FIRST offers unique capabilities in the context of the detection and spectral characterization of closecompanions, especially on 30m-class telescopes.
Keywords:
Exoplanets, high contrast imaging, interferometry, pupil remapping, single-mode fiber filtering
1. INTRODUCTION
High contrast imaging at high angular resolution is crucial for the imaging and the spectroscopic study of faintstellar companions such as exoplanets. As the angular resolution is inversely proportional to the diameter of thetelescope, larger telescopes are needed to improve it. Thus the interferometric combination of several telescopes(e.g. the four 8-meter VLT) is currently used to increase the resolution power in images. It is a successful
Send correspondence to K. Barjot, E-mail: [email protected] a r X i v : . [ a s t r o - ph . I M ] F e b echnique but demands the use of several telescopes and the recombination of their light. As an alternative, thepupil masking technique proposes to interfere sub-divisions of one telescope pupil (via a mask with holes). Itsimplifies the scheme while improving the angular resolution up to twice the diffraction limit of the telescope.However, the sub-apertures have to be located non redundantly to minimize the degradation of the opticaltransfer function (OTF) of the telescope. As a consequence, it limits the dynamic range because only a smallfraction of the pupil can be used. To overcome this issue and improve the image contrast, the pupil remappingconcept has been proposed,
2, 3 where the entire surface of the pupil is divided into sub-apertures that are rear-ranged into a non-redundant pattern for the recombination. Single-mode optical fibers are used to perform thisrearrangement and to filter the wavefront from aberrations.The fibered imager for a single telescope (FIRST ) is an instrument installed on the Subaru CoronagraphicExtreme Adaptive Optics (SCExAO ) at the Subaru Telescope, aiming at high angular resolution and highcontrast imaging using pupil remapping with single-mode optical fibers. Prior to this, the instrument successfullyvalidated these concepts with its first on-sky results at the Lick Observatory, demonstrating that images canbe recovered at an angular resolution lower than the telescope diffraction limit at visible wavelengths.In this paper, we present the upgrade and laboratory characterization of the second version of the instrument,FIRSTv2, now including a photonic chip, in order to improve the stability and sensitivity. Integrated optics(IO) is now a key technology to perform the interferometric combination in a small volume. As an example,the GRAVITY instrument operating at the very large telescope interferometer (VLTI) is based on a photonicrecombination and allows the characterization of exoplanets.
2. INSTRUMENT DESCRIPTION2.1 FIRSTv2 testbed
FIRSTv2 is a laboratory testbed and is the upgrade of the FIRST instrument installed on the SCExAO benchat the Subaru Telescope since 2013. It aims at improving the contrast obtained on astrophysical data at highresolution.
Figure 1:
Schematic of the FIRSTv2 instrument. From left to right are shown: (1) the pupil sampling part,where the blue sub-pupils are the ones which light is injected into fibers, (2) the micro-lenses for theinjection into the optical fibers, (3) the optical delay lines, (4) the single-mode fibers, (5) the photonicchip for the recombination, (6) the prism and (7) the final image on the camera.Fig. 1 depicts a schematic of the FIRSTv2 instrument. Denoted by (1) is the telescope pupil (with only acentral obstruction here) represented on top of the 37-segment deformable mirror which sub-divides the pupilinto sub-pupils. The blue sub-pupils are the ones that are used in the FIRSTv2 experiment. Since optical delaylines (ODL) have been included to control the optical path length differences (3), the configuration of thesesub-pupils can actually be modified at will. Then (2) are the micro-lenses that inject the light of each sub-pupilinto single-mode optical fibers (4) which filter the atmospheric wavefront aberrations. Only the differential pistonemains. The photonic chip (5) performs the recombination between the sub-pupils (see section 2.2). Finally,the prism (6) cross-disperses the light before being focused onto the camera (7).The FIRSTv2 upgrade consists in improving the contrast of the instrument by: − performing the interferometric recombination in a photonic chip instead of at the focal plane (on thecamera), − improving the search for the fringes with the use of ODLs to accurately equalize the optical path length ofall optical fibers, − encoding each interference pattern for a given baseline on few pixels instead of hundreds (as shown in Fig. 1(7)) which increases the sensitivity of the instrument. The photonic chip is manufactured by Teem Photonics ∗ . It consists in a block of glass in which optical waveguides are engraved by photolithography. For this experiment, they have been optimized for 650 nm . The chipshave a number n inputs of inputs corresponding to the number of sub-pupils that interfere with each other. Asshown in Fig. 2, the waveguide corresponding to each input is divided into n inputs − X-coupler type(left schematic of Fig. 2), leading to two outputs per recombined baseline, i.e. n outputs = n inputs ( n inputs −
1) =5 × Y-coupler type (right schematic of Fig. 2), leading to one output per recombined baseline,i.e. n outputs = n inputs ( n inputs − / × / Figure 2:
Schematics of the two photonic chips we characterized for FIRSTv2. The inputs are located at thebottom and the outputs at the top. Left is a
X coupler type and right is a
Y coupler type.
3. PHOTONIC CHIP CHARACTERIZATION
To characterize the two photonic chips designed and manufactured for FIRSTv2, we use a broadband Halogenlight source
Ocean Optics, HL-2000-FHSA-HP † . By putting the source light through an optical fiber focusedon the camera, its spectrum is used to normalized the spectra obtained by injecting the light into the chips, inorder to estimate their throughput as a function of wavelength. Three features of the chips were assessed: (1)the flux cross-talk between the outputs, (2) the throughput and (3) the contrast performances. ∗ † .1 Cross-talk measurement This measurement aims at quantifying the unwanted light leaking from one waveguide to others, in particularat the location of the couplers. In the case where light is injected into one input only, only 4 or 8 outputs areexpected to show non-zero flux for the Y-coupler and X-coupler chip respectively. The unwanted flux measuredin the other outputs corresponds to the cross-talk leakage.Fig. 3 presents the cross-talk characterization obtained for the X-coupler chip on top and for the Y-couplerchip at the bottom. For each chip, there are five bar-plots (titled from input 1 to input 5) showing the fluxesmeasured on all outputs while the
Ocean Optics source is injected in one input. The blue bars are the outputswhere light is expected, while the red bars are the other outputs expected to show no light. There is at worst a10% cross-talk for the X-coupler chip and a 20% cross-talk for the Y-coupler chip.
Figure 3:
Cross-talk leakage characterization for the X-coupler (top) and the Y-coupler (bottom) chips. For eachchip, there are five bar-plots, showing the flux measured in all outputs, while light is injected in oneinput only (the illuminated input corresponds to the number in the title). The blue bars represent theoutputs where light is expected, while the red bars represent the other outputs expected to have nolight. .2 Throughput comparison
The measured throughput are presented on Fig. 4 for the two types of couplers. Between ∼ nm and ∼ nm (working bandwidth of the chips) the throughput of the X-coupler represented by the continuous lineis measured to ∼
30 % and the throughput of the Y-coupler represented by the doted line is measured to ∼
13 %.This throughput levels are satisfying in comparison with the throughput of our first test chips reaching no morethan 1 % throughput.
Figure 4:
Measured throughput as a function of wavelength for the X- and Y-coupler chips in continuous anddoted lines, respectively.
When illuminating two inputs with light intensities I and I , their interferometric combination is measured onone of the outputs with an intensity I given by: I = I + I + 2 (cid:112) I I cos (cid:18) π × OP Dλ (cid:19) (1)where
OP D is the optical path length difference between the two beams and λ the wavelength.From equation 1, the contrast of the fringes is obtained by the factor 2 √ I I / ( I + I ). Thus we can usethis definition to estimate the contrast performance expected for each baseline of the chips using the flux valuesof Fig. 3. These measurements were performed with only one input illuminated at a time. Hence, the expectedcontrast for baseline i is assessed with the combination of two fluxes represented by two blue bar-plots. Forinstance, the contrast of the fifth baseline of the X-coupler chip is obtained from the combination of the fluxmeasured on the fifth outputs of the bar-plot titled input 2 and of the bar-plot titled input 3 , normalised by thesum of the two intensities.Fig. 5 shows the expected fringe contrast estimated from the combination of every pair of output fluxes, shownin Fig. 3. Thus, the best contrast that can be obtained is measured to 2 √ I I = 2 √ . × × . × / (2 . × ) ≈ .
866 with the X-coupler and to 2 √ I I = 2 √ . × × . × / (3 . × ) ≈ .
998 with the Y-coupler.It appears from this study that the X-coupler type has twice the throughput ( ∼
30 %) of the Y-coupler type( ∼
13 %) but has lower contrast (contrast equals to ∼ .
87 at best for the first one and equals to ∼ igure 5: Interferometric contrasts obtained from the flux intensities of the Fig. 3 as a function of the baselinesfor the X-coupler in continuous line and for the Y-coupler in doted line.
4. CONCLUSION
Based on the pupil remapping technique with photonic technology, FIRSTv2 is currently under lab characteriza-tion. In this paper we have shown a comparison between two chips performing the interferometric combinationof pairs of beams by two different types of couplers: X-type and Y-type. Our study shows that the first oneis twice better in transmission but has a lower performance in contrast than the second. As a consequence thework to come on FIRSTv2 will focus on the improvement of the Y-type photonic chip throughput.
REFERENCES [1] Tuthill, P. G., Monnier, J. D., and Danchi, W. C., “Aperture masking interferometry on the keck i tele-scope: new results from the diffraction limit,” in [
Interferometry in Optical Astronomy ], , 491–498,International Society for Optics and Photonics (2000).[2] Perrin, G., Lacour, S., Woillez, J., and Thi´ebaut, E., “High dynamic range imaging by pupil single-modefiltering and remapping,” Monthly Notices of the Royal Astronomical Society (2), 747–751 (2006).[3] Lacour, S., Thi´ebaut, E., and Perrin, G., “High dynamic range imaging with a single-mode pupil remappingsystem: a self-calibration algorithm for redundant interferometric arrays,”
Monthly Notices of the RoyalAstronomical Society (3), 832–846 (2007).[4] Kotani, T., Perrin, G., Lacour, S., Thi´ebaut, E., Woillez, J., Fedou, P., Berger, J.-P., Bord´e, P., Chesneau,O., Kervella, P., et al., “The first project: a single-mode fiber-based very high-dynamic range diffraction-limited imaging instrument at visible to near-infrared wavelengths,” in [
Ground-based and Airborne Instru-mentation for Astronomy II ], , 70141P, International Society for Optics and Photonics (2008).[5] Huby, E., Perrin, G., Marchis, F., Lacour, S., Kotani, T., Duchˆene, G., Choquet, E., Gates, E., Woillez,J., Lai, O., et al., “First, a fibered aperture masking instrument-i. first on-sky test results,” Astronomy &Astrophysics , A55 (2012).[6] Vievard, S., Cvetojevic, N., Huby, E., Lacour, S., Martin, G., Guyon, O., Lozi, J., Kotani, T., Jovanovic,N., Perrin, G., et al., “Capabilities of a fibered imager on an extremely large telescope,” arXiv preprintarXiv:2010.10733 (2020).[7] Vievard, S., Huby, E., Lacour, S., Barjot, K., Martin, G., Cvetojevic, N., Deo, V., Guyon, O., Lozi,J., Kotani, T., Jovanovic, N., Perrin, G., Marchis, F. Lapeyrere, V., and Rouan, D., “First, a pupil-remapping fiber interferometer at the subaru telescope: on-sky results,” in [
SPIE Astronomical Telescopes+ Instrumentation (AS20) ], International Society for Optics and Photonics (2020).8] Jovanovic, N., Martinache, F., Guyon, O., Clergeon, C., Singh, G., Kudo, T., Garrel, V., Newman, K.,Doughty, D., Lozi, J., Males, J., Minowa, Y., Hayano, Y., Takato, N., Morino, J., Kuhn, J., Serabyn, E.,Norris, B., Tuthill, P., Schworer, G., Stewart, P., Close, L., Huby, E., Perrin, G., Lacour, S., Gauchet,L., Vievard, S., Murakami, N., Oshiyama, F., Baba, N., Matsuo, T., Nishikawa, J., Tamura, M., Lai, O.,Marchis, F., Duchene, G., Kotani, T., and Woillez, J., “The Subaru Coronagraphic Extreme AdaptiveOptics System: Enabling High-Contrast Imaging on Solar-System Scales,”
Publications of the AstronomicalSociety of the Pacific , 890 (Sept. 2015).[9] Huby, E., Duchˆene, G., Marchis, F., Lacour, S., Perrin, G., Kotani, T., Choquet, ´E., Gates, E., Lai, O.,and Allard, F., “First, a fibered aperture masking instrument-ii. spectroscopy of the capella binary systemat the diffraction limit,”
Astronomy & Astrophysics , A113 (2013).[10] Perraut, K., Jocou, L., Berger, J., Chabli, A., Cardin, V., Chamiot-Maitral, G., Delboulb´e, A., Eisenhauer,F., Gamb´erini, Y., Gillessen, S., et al., “Single-mode waveguides for gravity-i. the cryogenic 4-telescopeintegrated optics beam combiner,”
Astronomy & Astrophysics , A70 (2018).[11] Lacour, S., Nowak, M., Wang, J., Pfuhl, O., Eisenhauer, F., Abuter, R., Amorim, A., Anugu, N., Benisty,M., Berger, J., et al., “First direct detection of an exoplanet by optical interferometry-astrometry and k-bandspectroscopy of hr 8799 e,”
Astronomy & Astrophysics , L11 (2019).[12] Martin, G., Foin, M., Gardillou, F., Cassagnettes, C., Ulliac, G., Courjal, N., Cvetojevic, N., Vievard,S., Huby, E., and Lacour, S., “Recent results on electro-optic visible multi-telescope beam combiner fornext generation first/subaru instruments: Hybrid and 3d devices,” in [