- The first galaxies are the frontier of extragalactic astronomy. Hubble and Spitzer have measured the buildup of stellar mass and specific star formation rates from z ~ 8 to 4. Extending this to z > 9 with the handful of candidates discovered in the Hubble Ultra Deep Field has not been possible due to their faint observed fluxes (~10 nJy). We expect the upcoming Frontier Fields to deliver another ~35 - 120 candidates at z ~ 9 - 12. It is imperative that we learn more about their physical properties, but many will still be too faint for follow-up. Fortunately, CLASH, a shallower survey of gravitational lensing clusters, has yielded 4 much brighter z > 9 candidates, including the most distant: MACS0647-JD at z ~ 10.8 (observed 420 Myr after the Big Bang). Our primary target, lensed image MACS0647-JD2, is magnified by a factor of 7 to 170 nJy in the rest-frame UV. With an intrinsic (unlensed) flux of ~25 nJy, it is a near-analog of z > 9 candidates discovered in 'blank' (unlensed) Hubble deep fields. The existing Spitzer IRAC imaging (5 hours in each of ch1 3.6um and ch2 4.5um, PI Egami) yields no detection in ch1 but a possible (3-sigma) intriguing detection of a Balmer break / [OII] 3727A emission in ch2. Here we propose imaging 10 times deeper (50 hours in each band) to detect the object significantly in both channels and obtain the first constraints on the age, stellar mass, and rest-frame UV slope of a galaxy observed in the heart of the epoch of reionization. This imaging will also enable study of a large sample of ~30 lensed z ~ 6 - 8 candidates (far more than revealed by most other cluster lenses) with delensed intrinsic fluxes down to ~1 nJy. This study, now possible before JWST, will contribute to measurements of scatter about the 'main sequence of star formation,' constraining SFR duty cycles at these early times.