The lifecycle of hollows on Mercury: An evaluation of candidate volatile phases and a novel model of formation

On Mercury, high-reflectance, flat-floored depressions called hollows are observed nearly globally within low-reflectance material, one of Mercury’s major color units. Hollows are thought to be young, or even currently active, features that form via sublimation, or a “sublimation-like” process. The apparent abundance of sulfides within LRM combined with spectral detections of sulfides associated with hollows suggests that sulfides may be the phase responsible for hollow formation. Despite the association of sulfides with hollows, it is still not clear whether sulfides are the hollow-forming phase. To better understand which phase(s) might be responsible for hollow formation, we calculated sublimation rates for 57 candidate hollow-forming volatile phases from the surface of Mercury and as a function of depth beneath regolith lag deposits of various thicknesses. We found that stearic acid (C18H36O2), fullerenes (C60, C70), and elemental sulfur (S) have the appropriate thermophysical properties to explain hollow formation. Stearic acid and fullerenes are implausible hollow-forming phases because they are unlikely to have been delivered to or generated on Mercury in high enough volume to account for hollows. We suggest that S is most likely the phase responsible for hollow formation based on its abundance on Mercury and its thermophysical properties. We discuss the possibility that S is the phase responsible for hollow formation within the hollow-formation model framework proposed by Blewett et al. (2013). However, several potential limitations with that model lead us to suggest an alternative hollow-formation model: a subsurface heat source (most often impact-induced) generates thermal systems that drive sulfur-rich fumaroles in which S and other phases accumulate on and within the surface at night and sublimate during the day to create hollows. We call this hollow-formation model “Sublimation Cycling Around Fumarole Systems” (SCArFS). We suggest that thermal decomposition of sulfides within LRM is a main contributor to S and S-bearing gases within the proposed fumarole systems and that (re-)precipitation of sulfides may occur at the surface along hollow floors and rims.