Pushing ALMA to the limit: 140 pc resolution observations of a z=6.6 quasar-galaxy merger resolve strikingly different morphologies of dust continuum and [CII] 158 um emission

Abstract: We present $0."026$ $(140\ \rm{pc})$ resolution ALMA observations of [C II] $158\ \mu\rm{m}$ and dust continuum emission of the $z=6.6$ quasar J0305--3150, resolved over $\sim 300-400$ independent resolution elements. The dust continuum emission is compact with $\sim 80\%$ recovered within $r<0."3$ $(1.6\ \rm{kpc})$, whereas the [C II] emission profile is composed of a central Gaussian ($r<0."4$, i.e. $<2.2\ \rm{kpc}$) and an extended component (detected up to $\sim 10\ \rm{kpc}$ at $>3\sigma$). We infer a direct contribution of the quasar to the observed 260\ \rm{GHz} continuum $S_{\nu,\rm{QSO}} / S_{\nu,\rm{QSO+Host}} \lesssim 1\%$. We report the detection of FIR-detected star-forming clumps with $r<200 \ \rm{pc}$ and properties similar to that of rest-frame UV-optical clumps reported in the literature. The $200\ \rm{pc}$ resolved [C II]/FIR ratio follows the global relation with the FIR surface brightness established in low- and high-redshift galaxies, even at the quasar location. We find that dust continuum is emitted in regions of $\sim0."02-0."04$ consistent with the size of photo-dissociation regions (PDR), whereas $50\%$ of the [C II] originates from larger physical scales ($\theta \gtrsim 2"$). The large-scale [C II] emission presents a velocity gradient aligned with a nearby companion with perturbed kinematics, and misaligned with the kinematics of the small-scale emission. The absence of significant [C II] emission by structures with physical scale $\lesssim 1\ \rm{kpc}$ implies that [C II] emission is not produced in dense PDR located at the boundary of Giant Molecular Clouds. We argue instead that [C II] is produced in low-density PDRs in the interstellar medium and diffuse HI gas tidally-stripped during the ongoing merger.