Current theory considers two options for the formation of the Pluto-Charon binary (e.g., Canup 2005Canup , 2011Desch 2015). In the 'hit-and-run' model, a lower mass projectile barely hits the more massive Pluto, kicks up some debris, and remains bound to Pluto (see also Asphaug et al. 2006). In a 'graze-and-merge' scenario, the projectile ejects substantial debris as it merges with Pluto (see also Canup & Asphaug 2001). To investigate the graze-and-merge idea in more detail, we consider the growth of Charon-mass objects within a circum-Pluto ring of solids. Numerical calculations demonstrate that Charon analogs form rapidly within a swarm of planetesimals with initial radii r 0 ≈ 145-230 km. On time scales of ∼ 30-100 days, newly-formed Charon analogs have semimajor axes, a ≈ 5-6 r P , and orbital eccentricities, e ≈ 0.1-0.3, similar to Charon analogs that remain bound after hit-and-run collisions with Pluto. Although the early growth of Charon analogs generates rings of small particles at a ≈ 50-275 r P , ejection of several 145-230 km leftovers by the central Pluto-Charon binary removes these small solids in 10-100 yr. Simple estimates suggest small particles might survive the passage of 10-20 km objects ejected by the central binary. Our results indicate that the Pluto-Charon circumbinary satellite system was not formed by a graze-and-merge impact when the formation of Charon within a circum-Pluto disk leads to the ejection of several 100-200 km particles through the orbital plane of the Pluto-Charon binary. If a growing Charon ejects only much smaller particles, however, graze-and-merge impacts are a plausible formation channel for the Pluto-Charon binary and an ensemble of small, circumbinary satellites.