Many metals form alloys with the materials that are used for capsules or crucibles, e.g., Pt. This is particularly problematic for Sn, where not only is Sn lost from the sample, but the Sn alloys are liquids at relatively low temperatures. The formation of liquid alloys precludes the use of methods such as wire loop or presaturation. Therefore a technique has been developed that uses sintered SnO2 rods as a source of Sn and as a sample holder. The rods are roughly 1x1x30 mm3 that are cut such that a ~4 mm long 'T' remains on the top. Granitic glass is glued onto a SnO2 rod with polyvinyl alcohol binder and up to six rods are loaded onto an Al2O3 disk, 3.5 cm in diameter that contains slots to hold the T-shaped rods. The disk has a hole in the middle, by which it is attached to an Al2O3 rod that can be mounted into a 1 atm gas mixing furnace. SnO2 solubility studies in granitic melts are currently being undertaken with this method, initially for comparison with cassiterite solubilities determined by the diffusion profile method in hydrous granites. The latter studies predicted changes in tin valence and coordination, which owing to the method could not be tested by spectroscopic means. In the current work homogeneous samples can be produced, which opens the possibility to determine tin valence and coordination by spectroscopic or wet chemical techniques.