Testing the nature of a strongly lensed dual AGN candidate in the early Universe

In the hierarchical framework of galaxy evolution, galaxies are expected to assemble through mergers. As most massive galaxies are known to host supermassive black holes (SMBHs) in their central regions, mergers should naturally produce pairs of SMBHs. When one or both are in the process of accreting matter, they appear as active galactic nuclei (AGN) and this makes the pair directly observable. One of the critical populations of dual AGN to test the models of galaxy evolution is that of low-separation systems, on sub-galactic scales, especially when they are found at high redshift, when galaxies were more gas rich and mergers more frequent. However, confirming such systems is challenging because of their small angular separations and the sensitivity limits of observing facilities. In this thesis, we investigate the nature of the background source (at z>3) of the strong gravitational lensing system MG B2016+112, which had previously been identified as a candidate dual AGN. We analyzed simultaneous multi-frequency VLBI observations in order to compute spectral index measurements, to assess whether two flat-spectrum core structures are present, which would strongly support the dual AGN scenario. Furthermore, we performed the first grid-based lens modelling, for this source, on near-infrared adaptive optics observations to produce a robust lens mass model and a pixellated reconstruction of the background source. While more work on the lens mass model is needed in order to obtain conclusive results from the source reconstruction, the spectral index analysis revealed only one flat-spectrum component in the background source. This constitutes a strong piece of evidence that disfavors the dual AGN scenario for MG B2016+112.