Soluble aqueous samples do not have to be digested. Samples and gravimetric standards are aspirated (sucked up via a pump) and sent to a plasma torch for analysis. If you wanted to know the aluminum content in polyethylene (for example), there is no physical way to introduce the polyethylene sample into the plasma torch. Therefore, the sample is usually digested with an oxidizing acid (nitric acid for example) to char and remove the organic portion and make the aluminum portion water soluble. The digestion serves a secondary purpose of making the density of the solvent carrier similar to the density of the solvent carrier for the standards. With oil, the primary complication I would be concerned about would include gumming up the tubing that draws the liquid to the flame. Also, most ICP standards are aqueous. The standards would have to be made in oil or a correction factor to account for the differing solvent densities would be needed. There might also possibilities of matrix interferences that would need to be addressed.