Direct imaging is crucial to increase our knowledge on extrasolar planetary systems. It can detect long orbits planets that are inaccessible by other methods and it allows the spectroscopic characterization of exoplanet’s athmospheres. During the past fewyears, several giant planetswere detected by direct imaging methods. Yet, as exoplanets are 103 to 1010 fainter than their host star in visible and near-infrared wavelengths, direct imaging requires extremely high contrast imaging techniques, especially to detect low-mass and mature exoplanets. Coronagraphs are used to reject the diffracted light of an observed star and obtain images of its circumstellar environment. Nevertheless, coronagraphs are efficient only if the wavefront is flat because aberrated wavefronts induce speckles in the focal plane which mask exoplanet images. Thus, wavefront sensors associated to deformable mirrors are mandatory to correct speckles by reducing aberrations. To test coronagraph techniques and focal plane wavefront sensors at very high contrast level, we developed the THD2 bench in the optical wavelengths. On the THD2 bench, we routinely reach 108 raw contrast level inside the dark hole over broadbands but this level is not sufficient to detect low-mass exoplanets. At this level, it seems that many experimental factors can affect the contrast and understanding which one is limiting the final detection contrast will be useful to upgrade the THD2 bench and to develop the next generation of space-based instruments (LUVOIR, HabEx) aiming to reach 10<sup>-10</sup> contrast level. We started a complete study of the instrumental limitations of the THD2 bench, focusing on scattering which could add intensity on the detector or polarization effects and residual laboratory turbulences. In this paper, we present the methods used to estimate the amount of scattered light that reaches the final detector on the THD2 bench.
While radial velocity and transit techniques are efficient to probe exoplanets with short orbits, the study of long-orbit planets requires direct imaging and coronagraphic techniques. However, the coronagraph must deal with planets that are 10<sup>4</sup> to 10<sup>10</sup> fainter than their hosting star at a fraction of arcsecond, requiring efficient coronagraphs at short angular separation. Phase masks proved to be a good solution in monochromatic or limited spectral bandwidth but expansion to broadband requires complex phase achromatization. Solutions use photonic crystals, subwavelength grating or liquid crystal polymers but their manufacturing remains complex. An easier solution is to use photolithography and reactive ion etching and to optimize the azimuthal phase distribution like achieved in the six-level phase mask (SLPM) coronagraph (Hou et al. 2014). We present here the laboratory results of two SLPM coronagraphs enabling high-contrast imaging in wide-band. The SLPM is split in six sectors with three different depths producing three levels of optical path difference and yielding to uniform phase shifts of 0, π or 2π at the specified wavelength. Using six sectors instead of four sectors enables to mitigate the chromatic effects of the SLPM compared to the FQPM (Four-Quadrant Phase Mask) while keeping the manufacturing easy. Following theoretical developments achieved by University of Shanghai and based on our previous experience to fabricate FQPM components, we have manufactured SLPM components by reactive ion etching at Paris Observatory and we have tested it onto the THD2 facility at LESIA. The THD2 bench was built to study and compare high-contrast imaging techniques in the context of exoplanet imaging. The bench allows reducing the starlight below a 10<sup>−8</sup> contrast level in visible/near-infrared. In this paper, we show that the SLPM is easy to fabricate at low cost and is easy to implement with a unique focal plane mask and no need of pupil apodization. Detection of a planet can be achieved at small inner working angle down to 1 λ/D. The on-axis attenuation of the best SLPM component reaches 2 × 10<sup>−5</sup> at λ = 800 nm and is better than 10<sup>−4</sup> in intensity over a 10% spectral bandwidth. Along the diagonal transition, we show that the off-axis transmission is attenuated by less than 3% over a 10% bandwidth and will need to be calibrated. Any etching imperfections can affect the SLPM performance, by lowering the on-axis attenuation and by changing the optimal wavelength. Despite few nanometers of uncertainty for etching the depths, we show that this first component can provide a high-contrast attenuation in laboratory
High-contrast imaging (HCI) techniques appear like the best solutions to directly characterize the atmosphere of large orbit planets and planetary environments. In the last 20 years, different HCI solutions have been proposed based on coronagraphs. Some of them have been characterized in the laboratory or even on the sky. The optimized performance of these coronagraphs requires a perfect wavefront unreachable without active control of the complete electrical field (phase and amplitude) at the entrance of the instrument. While the correction of the phase aberrations is straight forward using deformable mirrors (DM), correcting amplitude defects is complex and still under study at the laboratory level. The next generation of HCI instrument either for ground-based (PCS instrument for ELT) or space-based (LUVOIR, HabEx) telescopes will require a practical and operational solution for amplitude corrections. The implementation of a DM located at a finite distance from the pupil is a simple solution that has been chosen by most of the projects. There have been only a few investigations on the optimization of the mirror positions for dedicated optical designs. In this paper, we give an intuitive approach that helps defining the best deformable mirror position in an instrument. Then, we describe its application to the THD2 and the performance in the laboratory that reaches a contrast level below 10<sup>-8</sup> at distance larger than 6 λ/D.