Free-standing polyaniline films are studied as pervaporation membranes for separating carboxylic acid/water mixtures. Changing polyaniline membranes from an undoped to a fully doped state has a dramatic effect on their selectivity. For example, with a 52 wt % acetic acid/48 wt % water feed, the undoped, base form of polyaniline leads to only a relatively modest separation of water (68 wt % water in permeate; alpha(H2O) = 2.4), while the doped, acid form of polyaniline shows a remarkably enhanced separation ability (>99 wt % water in permeate; alpha(H2O) > 1300). The improved selectivity of doped polyaniline is due in part to the hydrophilicity induced in the polymer by the acid dopants. Fully doped polyaniline appears to restrict the permeation pf carboxylic acids mainly based on their effective sizes. Whereas formic acid (solvated diameter = 3.5 Angstrom) can permeate through a doped polyaniline membrane, the larger acetic acid (4.5 Angstrom) has an exceedingly low rate of permeation, and the even larger propionic acid (5.5 Angstrom) is virtually excluded. Pervaporation of a mixture of all three carboxylic acids and water confirms that acetic and propionic acid are effectively blocked by doped polyaniline. The compositions of all feed and permeate mixtures were determined by H-1 NMR, which was found to be a very accurate method for quantitative analysis. Although the permeation rates of acetic acid/water feeds through dense, as-cast polyaniline membranes (similar to 80 mu m thick) are relatively low (2-13 g/(m(2) h)), a 2-fold or more increase in the permeation rate can be obtained by using asymmetric membranes of approximately the same overall thickness. Increasing the temperature of the pervaporation experiment can also lead to large increases in permeation rate. For example, the permeation rates of 50 wt % acetic acid/50 wt % water feeds through doped polyaniline membranes double on going from room temperature to 85 degrees C.