In a two-fluid thorium reactor, thorium nuclei in the blanket will intercept neutrons and breed--first to protactinium-233, then to uranium-233. If the protactinium is isolated chemically after it is formed, then the U-233 in the protactinium decay tank will have little U-232 contamination.
The real question I've had is: how can we generate U-232 contaminated material, even in the Pa decay tank? Ionium is thorium-230, which is part of the natural decay chain of uranium-238, which is rather abundant. If the thorium in the blanket was "spiked" with ionium, it would be impossible to chemically separate the two forms of thorium (since they are chemically identical) but the ionium would preferentially absorb neutrons and form Pa-231.