Apodization in high-contrast long-slit spectroscopy. II. Concept validation and first on-sky results with VLT/SPHERE

Spectral characterization of young, giant exoplanets detected by direct imaging is one of the tasks of the new generation of high-contrast imagers. For this purpose, the VLT/SPHERE instrument includes a unique long-slit spectroscopy (LSS) mode coupled with Lyot coronagraphy in its infrared dual-band imager and spectrograph (IRDIS). The performance of this mode is intrinsically limited by the use of a non-optimal coronagraph, but in a previous work we demonstrated that it could be significantly improved at small inner-working angles using the stop-less Lyot coronagraph (SLLC). We now present the development, testing, and validation of the first SLLC prototype for VLT/SPHERE. Based on the transmission profile previously proposed, the prototype was manufactured using microdots technology and was installed inside the instrument in 2014. The transmission measurements agree well with the specifications, except in the very low transmissions (<5% in amplitude). The performance of the SLLC is tested in both imaging and spectroscopy using data acquired on the internal source. In imaging, we obtain a raw contrast gain of a factor 10 at 0.3" and 5 at 0.5" with the SLLC. Using data acquired with a focal-plane mask, we also demonstrate that no Lyot stop is required to reach the full performance, which validates the SLLC concept. Comparison with a realistic simulation model shows that we are currently limited by the internal phase aberrations of SPHERE. In spectroscopy, we obtain a gain of ~1 mag in a limited range of angular separations. Simulations show that although the main limitation comes from phase errors, the performance in the non-SLLC case is very close to the ultimate limit of the LSS mode. Finally, we obtain the very first on-sky data with the SLLC, which appear extremely promising for the future scientific exploitation of an apodized LSS mode in SPHERE.