Low-density foam shells are currently being employed as direct drive targets on the Omega laser facility at the University of Rochester. For cryogenic shots, only a thin layer of glow discharge polymer (GDP) is required over these foam shells to hold the D2 (or DT) fill provided the capsules are re-filled after cooling. Room temperature surrogate experiments, however, require an additional permeation barrier of aluminum on GDP coated foam shells. This barrier should have a permeation time constant of at least 4 h for D2 at room temperature. To study this coating, 0.1 m layers of Al were deposited via magnetron sputtering onto the surface of GDP shells and GDP coated foam shells. The foam shells were 180 mg/cc resorcinol formaldehyde (RF) with a GDP thickness of 3-5 m; the GDP shells used for this study had a wall thickness of 25-30 m. Preliminary data shows that the permeation rate of D2 for smooth GDP shells is lower than for GDP coated RF shells with a similar thickness of Al. The main factor in this difference appears to be the surface roughness of the shells.