2025MNRAS.544.1981S

The recent discovery of 40 binary Jupiter-mass free-floating planets in the Trapezium cluster by the James Webb Space Telescope (S. G. Pearson & M. J. McCaughrean 2023) has opened new windows and questions in planets studies. These systems were not predicted beforehand by star and planet formation scenarios. In this work, we study characterizations of microlensing events due to binary free-floating planetary (BFFP) systems, and potentially detecting them in the Roman Galactic Exoplanet Survey (RGES). Detecting binary signatures in these light curves is challenging, because of finite-source sizes and short durations. In a 2D space made from the total mass of these systems ($M_{\rm {tot}}$) and their separation (s) in wide ranges, only massive and close systems with $M_{\rm {tot}}\gtrsim 0.2\,\mathrm{ M}_{\rm{\oplus }}$ and $s\lesssim 0.4\,\rm {au}$ can create close or intermediate caustic topologies. Earth (and Jupiter)-mass BFFP systems with separations $\lesssim 0.025~\rm {au}$ (and $2~\rm {au}$) create caustics larger than source sizes. We evaluate the Roman efficiency for detecting binary signatures in microlensing events due to BFFPs with $\log _{10}[s(\rm {au})]\in [-2.25,~1.5]$ and $\log _{10}[M_{\rm {tot}}(\mathrm{M}_{\rm{\oplus }})]\in [-1.75,~2]$. This efficiency is virtually zero for systems with $M_{\rm {tot}}\lesssim 0.02\,\mathrm{ M}_{\rm{\oplus }}$, but it reaches 10 per cent and 25 per cent for Earth and Saturn-mass systems by assuming a log-uniform distribution for binary separations. The detection efficiency drops to zero for systems with $s\gtrsim 1~\rm {au}$. Improving the Roman cadence to 10 min enhances this efficiency up to 7 per cent. Roman can discover $\lesssim 4$, $\lesssim 16$, and $2-50$ Earth, Super-Earth, and Saturn-mass BFFP systems during RGES. Some of these light curves exhibit two wide and flattened peaks that are degenerate with microlensing from binary sources and low-mass lens objects.