What is the difference between extractive distillation and azeotropic distillation?

Answer 1

Extractive and azeotropic distillation have the common feature that a substance

not normally present in the mixture to be separated is deliberately introduced into the

system in order to increase the difference in volatility of the most hard to separate

components. Extractive distillation can be defined as distillation in the presence of a

substance which is relatively non-volatile compared to the components to be separa-

ted, and which, therefore, is charged continuously near the top of the fractionating

tower, so that an appreciable concentration is maintained on all plates in the tower

below its entry. Azeotropic distillation can be defined as distillation in which the add-

ed substance forms an azeotrope with one or more of the components in the feed, and

by virtue of this are present on most of the plates in the tower above its entry at an

These separation methods find their principal applications in the separation of

mixtures whose components boil too close together for the economical use of simple

fractionating equipment. These separation methods are particularly applicable when

the components to be separated differ in chemical type. The theoretical principles in-

volved are well documented, and will not be further considered here. The processes

differ in the means used to maintain be desired solvent concentration on the plates of

the tower. In extractive distillation the high concentration of solvent is maintained by

virtue of its non-volatility, and by the fact that it is charged at a high point in the

tower. The solvent is, necessarily, removed from the base of the principal tower.

In azeotropic distillation, most of the solvent is taken off from overhead, with rela-

tively small amounts (ideally, none) drawn off with the bottoms.

Extractive distillation is generally more flexible than azeotropic distillation, a

greater variety of solvents and a wider range of operation conditions are available; and

the concentration of solvent may be controlled by heat and material balances rather

than by the accident of azeotrope composition. Furthermore, since vaporization of

the solvent is not required beat loads are usually considerably less. It has been mainly

used for the separation of toluene, not benzene. But it is mentioned here for the sake

of completeness.

The use of azeotroptc distillation as a means of separation of BTX components

from other non-aromatic hydrocarbons has been known and employed for some thirty

years. Acetone is used as an entrainer to purify benzene from similar-boiling non-aro-

matic hydrocarbons. Toluene can be separated by the use of either methanol, or of

methyl ethyl ketone. Ethyl benzene may be separated from styrene either by isobuta-

nol, of by l-nitro-propane.

In a 1966 review paper, further information was made available. Fifty-eight pos-

sible entraining agents for separating ethyl benzene (bp. 136.20C) and para-xylene

(bp. 138.40C) have been examined. It would appear that 2-methyl butanol is the

most suceessful agent, requiring a column with only 48 percent of the number of the-

oretical plates required if no entrainet were used.

The separation of para-and meta-xylenes (bp. 138. 20C and 139. 2aC respectively ) is

much more difficult. Of thirty five entrainers examined the best appears to be 2-met-

bylbutanol, but the change of relative volatility is only from l. 020 t0 1. 029, and

hence it can be safely concluded that azeotropic distillation for the separation of the

para-and meta-xylenes is not an economic proposition.

Last, and easiest of the C8 aromatics is the separation of meta-from ortho-xylene

( bp. 139. 20C and 144. 50C respectively ). Twenty-eight entrainers were examined,

the best being formic acid, required a column containing only 70 percent of the theo-

retical plates required of no entrainer is used.

Two commercial processes have been developed for the separation of toluene

using azeotropic distillation. One using an aqueous solution of methyl ethyl ketone and wa-

ter and the other using methanol. Both processes operate on a narrow boiling range concen-