Hi gb

Might get rid of that ice-shell by have a multi-layer refelctor system, and some very low heat-transfer material between payload and shield. If 99% of the incoming light can be reflected then the first layer of the shield is at 1827 K. Add reflectors and insulation to get it down to something you can handle.

Masses less than an ablative ice-barrier.

And easier system might be the beryllium balloon sail that Matloff et.al. posted to the arXiv. Only have to dip into ~0.1 AU or so, so somewhat easier thermal environment, and you leave at ~400 km/s.

BTW- The sublimation energy of ice in a vacuum is ~0.45 eV per molecule, or about 2.4 MJ/kg. At 63.08 MW/m^2 it's about 26 kg/s sublimed away, or roughly an inch a second. Ice/water (under pressure) absorbs about 3 MJ/kg before it becomes steam and is then really difficult to handle. But it absorbs different frequencies in different ways. To IR liquid water is a blackbody, but semi-transparent to visible light... etc.