Flash and relative volatility

Most separations in chemical plants use the difference in how easily things boil. Heat a mixture until part of it boils, let vapor and liquid reach equilibrium, and the vapor is richer in the light component. That's one equilibrium stage, and a flash drum does exactly this: hot liquid enters at lower pressure, part of it flashes to vapor, and vapor leaves the top, liquid the bottom.

How easy the split is depends on the relative volatility, α (alpha): roughly the ratio of the two pure vapor pressures, α = P*_light ÷ P*_heavy. With α = 1 the vapor and liquid are the same and boiling can't separate them. The bigger α, the easier the separation.

For a two-component mixture, the vapor in equilibrium with liquid of mole fraction x (of the light one) is y = αx ÷ (1 + (α − 1)x). With α = 2.5 and x = 0.5: y = 1.25 ÷ 1.75 = 0.714. One stage takes a 50/50 liquid to a 71% vapor, an improvement, but nowhere near pure.

A flash drum's two products come from a balance plus the equilibrium relation, and the fraction of feed that vaporizes is set by the temperature and pressure you choose. Flashes are cheap and used where a rough split will do, such as knocking liquid out of a gas or dropping gas out of a reactor product.

To get purer products you repeat the stage many times, with each stage's vapor rising to the next and its liquid falling to the one below. That stacked set of stages is a distillation column.