A feasibility study was made of the use of physical methods for separating SO2 from flue gas. The study included both direct removal techniques and physical techniques for converting SO2 to a more easily removable form such as SO3. It was found to be technically feasible to directly separate SO2 from flue gas by gaseous diffusion, permeation, centrifugation, mass spectrometry, vortex tube, and Joule-Thompson condensation. Condensation by surface heat exchangers or by the use of an expansion turbine does not appear to be technically feasible because of operating problems. The conversion of SO2 to SO3 by electrical excitation was demonstrated to be technically feasible and there are indications that such conversion could also be accomplished by electromagnetic excitation and nuclear radiation excitation. It was found that a magnetic field selectively affects the molecular diffusivity of different molecular species. However, at the pressures considered for this application, extremely strong magnetic fields would be required to get a slight effect. All of the processes for direct separation of SO2 which were found to be technically feasible were found not to be economically feasible by a substantial margin, and the physical processes for conversion of SO2 to SO3 were shown to have minimum (theoretical) energy requirements which were excessive. (Author)