Table 111

The value as will be seen from p. 290, is equal to and

1 Reilly and Hickinbottom, Proc. Royal Dublin Soc, 1921. if is assumed to be constant is inversely proportional to the initial volume of liquid in distilling vessel. This value varies also to a small extent with the concentration of the liquid distilled and, therefore, in specifying distillation constants of the above type it is necessary to specify the initial volume of liquid distilled and also its concentration. As will be seen from Table 111 the distillation constant (A) for ethyl alcohol is lower than the corresponding constants for the higher alcohols. In the case of fermented wash the higher alcohols present (fusel oil) will correspond mainly to a mixture of iso-amyl alcohol with smaller quantities of iso-butyl and other alcohols. The total quantity of fusel oil present is small compared with the ethyl alcohol content (1 : 20). It is evident that in the distillation of fermented wash the fusel oil will tend to concentrate in the first fractions of the distillate. From the value of the distillation constants a table can be made for any known mixture of ethyl alcohol and fusel oil, which will show the extent of the separation of the alcohols in the various fractions of the distillate. The presence of other impurities such as esters will influence these results.

Removal of By-products by Rectification, - As the means adopted for the removal of volatile by-products in discontinuous rectification differ from those employed in continuous rectification the two cases will be considered separately.

Discontinuous Rectification

The still is charged with raw spirit at about 38 per cent by weight, for it is more difficult to eliminate the impurities from very concentrated alcohol.

The raw spirit contains organic acids which react with the various alcohols present to produce esters. This action is more pronounced in discontinuous rectification, for in this case the acids remain in contact with concentrated alcohol for a longer time than when the rectification is continuous. It is therefore desirable to neutralise the acids by the addition of caustic soda.

The raw spirit must, however, not be completely neutralised for then amines may be evolved which will give an unpleasant smell to the alcohol.

Behaviour of various Impurities during the Rectification.

- Impurities for which K' is very high, such as ethyl formate, methyl acetate, ethyl acetate (see Fig. 102), will tend to pass over in the first runnings.

For a reason which will be explained later the concentration of the ethyl alcohol on the plates of the rectifying column is allowed to attain its maximum before the collection of the "first runnings" is started.

In these circumstances impurities such as those we are considering are easily eliminated from the still, but their separation from the concentrated alcohol on the upper plates of the column is much more difficult. The result is that the distillate will continue to be contaminated by such impurities for a considerable time. The same difficulties occur but in a more aggravated form with such substances as ethyl iso-butyrate. For this substance K' = 1 when the alcoholic strength is 89.7 per cent by weight, so that this impurity will tend to collect near the plate containing alcohol of this concentration. This point is dangerously near the level at which the finished product is run off, so that the continuous contamination of the distillate is very probable.

The contamination of the finished spirit by head products is prevented to a considerable extent by running off the ethyl alcohol from a point a few plates below the top of the column. The head products mixed with ethyl alcohol pass to the top of the column and enter the condenser. The major part of this mixture is condensed and returned to the top plate of the column, but a small portion rich in head products passes uncondensed through the condenser and is collected separately. By this process, which was first employed by Barbet, and called by him "Pasteurisation," the ethyl alcohol obtained is freed from head products to a very considerable extent.

Impurities such as amyl alcohol, for which K' is low, are retained in the still or lower parts of the column as long as possible. Completely to prevent amyl alcohol from passing over into the distillate, it is necessary to maintain a very high concentration of ethyl alcohol on the upper plates of the column, and the alcoholic content of the distillate should be not less than 93.8 per cent by weight. For this purpose a very large number of plates are required in the column.

As amyl alcohol is the principal volatile impurity it is important to retain it effectively from the start of the rectification. The alcoholic strength on the plates is therefore raised to a high point before the first runnings are collected, so as to prevent the amyl alcohol from mounting to the upper part of the column and contaminating the finished product.

The amyl alcohol will be present in greatest concentration near the point in the column at which the ethyl alcohol strength is 334 per cent by weight, and K' = l. Towards the end of the rectification, as the amount of ethyl alcohol remaining in the still decreases, the point at which the alcoholic strength is 334 per cent by weight will gradually rise in the column and the distillate will become more and more contaminated with amyl alcohol.

The danger of amyl alcohol passing over into the distillate increases as the rectification proceeds ; for as the alcoholic strength of the liquid in the still decreases the proportion of the total amyl alcohol contained in the column increases owing to the rise of K' with the fall of the alcoholic strength in the still.

Impurities such as iso-amyl acetate, iso-butyl alcohol, and propyl alcohol behave in a similar manner to amyl alcohol, but as they tend to collect at points in the column where the alcoholic strength is 62.5 to 73.5 per cent by weight, they are more difficult to retain than amyl alcohol.

"The distillation of a mixture of three constituents and the continuous separation of these in different phases have recently been investigated by Gay.1 Formulae have been worked out for calculating the composition of the different phases, and rules established concerning the minimum and maximum heat compatible with the correct operation of the fractionating and rectifying column.

1 Chim. etlnd., 1920, 4, 735-748; J. Chem. Soc, A. 1921, 120, ii. 85.