Nulling interferometrie: van exozodiacaal stof met de LBTI tot exoplaneten met Asgard/NOTT bij de VLTI

The search for exoplanets -- planets located outside the Solar system -- has become a major field of astronomy in the recent decades. In May 2025, the NASA Exoplanet Archive reported almost 5900 confirmed planets. Among these detections, planets orbiting around main-sequence stars are essential targets for understanding the formation of planetary systems, the chemistry that is taking place in their atmosphere, and their potential habitability. The characterization of these confirmed planets requires the combination of a mosaic of techniques. Among them, the transit, radial velocity, and astrometry are useful to constrain the mass, radius, semi-major axis, eccentricity, or the inclination of a planet. Transit spectroscopy can also be used to analyze the composition of its upper atmosphere, but it is limited by the photon noise generated by the stellar light. Direct imaging of exoplanets are techniques that disentangle the stellar light from the planet's signal, and enable exhaustive understanding of their properties. Direct imaging needs to achieve both high-contrast and high-angular resolution due to the proximity of the planets with their more luminous host star. Nulling interferometry is a technique of direct imaging imagined by Bracewell (1978). It is able to attenuate the light from the star by several orders of magnitude at angular separations down to a few milliarcseconds. The technique has already been implemented at the Multiple Mirror Telescope (MMT: Hinz et al. 1998), at the Keck Interferometer (Serabyn et al. 2012), or at the Large Binocular Telescope Interferometer (LBTI: Hinz et al. 2016), and will be implemented at the Very Large Telescope Interferometer (VLTI) with the upcoming Asgard/NOTT visitor instrument (Defrère et al. 2018b, 2022; Defrère et al. 2024). However, past and present nulling interferometers have been unable to achieve the contrast and sensitivity for exoplanet imaging. Their main science case has been the observation of exozodiacal dust, which are debris material located in the inner region of planetary systems. Two types of exozodiacal dust can be identified: the warm dust located mainly around the HZ with a temperature 200K, and the hot dust located in the vicinity of the host star with temperatures up to 1500K. These debris are typically studied to give insights on the architecture and dynamics of the system. The LBTI carried out the largest survey of warm dust with the Hunt for Observable Signatures of Terrestrial Systems (HOSTS: Ertel et al. 2018, 2020). One of the main results of the survey was the evidence for a strong correlation between the detection of warm dust and the presence of a Kuiper belt-like debris disk that could be the source of the inner dust. However, a few targets such as θ Boo show evidence of bright exozodiacal disks without the detection of an outer belt. The first goal of this thesis is to investigate the warm dust around θ Boo using the LBTI nulling observations obtained in the context of HOSTS (Chapter 2). This work was able to constrain three possible dust distributions that could explain the observations, the brightness increase in the HZ of the star, and the size of the dust grains. New L- and H-bands observations at the LBTI also constrained the presence of substellar companions in the system. The study highlights the current capabilities of nulling interferometry and its limitations. The results on the warm dust are indeed very dependent on the assumptions made on the system's inclination or the geometry of the exozodiacal disk. The enigmatic origin of the warm dust could not be confirmed, and the understanding of the θ Boo planetary system remains incomplete. Asgard/NOTT is the new generation of nulling interferometer, designed to operate in the L'-band (3.5-4µm), and the first one expected to image young giant planets (Dandumont et al. 2022a; Laugier et al. 2023). The instrument is also expected to observe exozodiacal dust and study the connection between warm and hot dust. Its development is led by KU Leuven, and its installation at the VLTI is planned at the beginning of 2026. The warm optical design of Asgard/NOTT and its test bench is the second goal of this thesis. Chapter 3 presents the warm optical design and its expected performance. The instrument takes advantage of the breakthrough in integrated optics for the L'-band to perform the beam recombination (Gretzinger et al. 2019), and proposes a novel strategy to inject the light beams inside the photonic chip without fibers. With this strategy, Asgard/NOTT is expected to reach a throughput of 40% after the injection in the chip. The instrument is also expected to reach a raw null ratio of 0.2%, which is compatible with the requirements for exoplanet imaging, and is shown to be dominated by the effect of atmospheric turbulences and vibrations. Finally, Chapter 4 presents the test bench of Asgard/NOTT implemented at KU Leuven, and its first experimental results. The warm optical design described in Chapter 3 was successfully implemented, with injection of L'-band light in the chip and the measurement of two-beams interferences. A throughput of 20% could be obtained on the test bench. Null measurements were also performed at ambient and cryogenic temperatures using a test cryostat, where the chip was cooled down to 138K. Null ratios of 1.4% and 2.4% could be obtained with a stability of a few 1e-4 at ambient and cryogenic temperatures, respectively. The test bench confirms that the chip and the strategy for beam injection and recombination is compatible with cryogenic nulling experiments, but more work is necessary to reach the Asgard/NOTT requirements for exoplanet imaging. The first implementation of nulling interferometry at the VLTI and its successful observation of exoplanets would add nulling into the mosaïc of techniques that can be used to detect and characterize them.