Context. Wide binaries offer a unique opportunity to test gravity in the low-acceleration regime, where deviations from Newtonian dynamics may appear. Aims. We used high-resolution VLT–ESPRESSO archival spectra to study 26 wide binaries with projected separations > 13 000 AU. By combining precise radial velocities with Gaia proper motions and parallaxes, we tested whether these systems are consistent with Newtonian gravity in the low-acceleration regime. Methods. We used multiple radial-velocity measurements and stellar parameters to remove systems affected by unresolved triple or chance alignments as well as young systems. For the remaining binaries, we combined radial velocities (corrected for convective shift and gravitational redshift) with Gaia proper motions, parallaxes, and positions in a bid to find bound Newtonian orbital solutions. Results. Of the 26 initial systems, 14 were discarded: 12 due to radial-velocity variability indicating unresolved close binaries, 1 that hosts a faint Gaia companion, and 1 that is too young. Of the remaining 12, 9 can be fitted with a bound orbital solution, while the velocity differences of the other 3 are too large to be reconciled with any bound Newtonian orbit. Conclusions. For the three systems that cannot be fitted with a bound orbit, repeated radial-velocity observations allowed us to confidently exclude, with one possible exception, unresolved triple stellar companions or massive close-in planets as causes. Given their likely large 3D separations, these binaries may have been dynamically perturbed or disrupted by stellar encounters or Galactic tides, and may no longer be gravitationally bound. This highlights how utmost caution must be applied when studying wide binaries as isolated systems.
Pasquini et al. (Thu,) studied this question.