Assessing the effects of stellar binarity on Neptunian exoplanets

Recent literature regarding exoplanets has focused increasingly on intermediate-sized planets, commonly referred to as Neptunians. An area of particular interest involves their architecture, specifically the inclination between the host star's spin axis and the planet's orbital axis. It is possible to infer the projected angle between the two, commonly referred to as the projected spin-orbit obliquity, by modeling the Rossiter-McLaughlin (RM) effect, which consists of an apparent shift in the radial velocity (RV) during a planetary transit. Knowing the rotation period of the star it is possible to deproject this quantity from the plane of the sky into a physical tridimensional angle. This quantity is known as the true obliquity, and it is regarded as a fundamental tracer of the dynamical history of the planet. During my B.Sc. thesis, I conducted a systematic analysis of the RM effect, producing the largest catalog of true obliquities to date, which we later published in A&A (Rossi et al. 2026). We highlighted that Neptunian planets exhibit a dichotomous distribution of true obliquities, sharply divided into a cluster of aligned planets (orbiting along the equator of the star) and a second cluster of perpendicular planets (orbiting the poles of their host stars, a peculiar but not uncommon occurrence). We also underscored how the small sample size could be a major factor hindering the characterization of the real true obliquity distribution. Another key factor that could perturb the architecture of a planetary system is the presence of binary companions, as numerous studies have highlighted. This thesis aims to investigate the effect of stellar binarity on the architecture of Neptunian exoplanets. Using a comprehensive multi-technique approach, I have conducted a systematic search for binary companions around Neptunian hosts to identify potential correlations with orbital misalignment. I updated the sample from Rossi et al. (2026) by adding six newly characterized targets, bringing the total to 38 Neptunian planets with a measured true obliquity, a 20% increase over the previous record. The methodological cornerstone of this work is the development of an original astrometric pipeline. By cross-calibrating the historical Tycho-2 catalog, the second reduction of the Hipparcos star mapper, with modern Gaia DR3 data, I exploited a 24.5 year temporal lever arm to compute the proper motion anomaly. This framework represents a significant scale-up in multiplicity studies, opening the possibility to infer binarity for 2.5 million targets, a potential that has remained largely untapped until now. To ensure a thorough vetting of the stellar environment, this astrometric search was integrated with high-resolution imaging analysis (speckle and adaptive optics) from ExoFOP and a search for linear RV drifts in the DACE database, covering the entire range of physical separations from a few au to hundreds of au. Furthermore, the analysis was bolstered by a dedicated literature review for every target and a comprehensive search across eleven multiplicity catalogs, providing a complete census of the stellar environment across all accessible scales. The results demonstrate that stellar binarity plays a neutral role in shaping Neptunian architecture. No statistically significant difference was found between the obliquity populations of planets in single and binary systems, suggesting that Neptunian misalignments are likely not driven by external stellar perturbations. Conversely, the presence of a compact companion seems to induce alignment, as suggested by recent literature. Most importantly, this work provides evidence that the previously claimed Neptunian dichotomy is starting to collapse with an increase in sample size. Bootstrap smoothed model selections strongly favor a unimodal distribution with a prominent alignment peak and an isotropic tail, yielding a combined winning rate of almost 90\% for single-peaked models.