Method of column calibration

The cleaning of plate columns is more difficult because traces of liquid tend to remain on the plates. The most satisfactory procedure is to distil a volatile solvent through the column and then to dry it with a current of hot air, the column being surrounded meanwhile by a heated jacket. 

The amount of test mixture should be 8 times the total column hold-up, and this figure should previously be determined, as described in section 4.10.5. The size of the still pot should be such that it is initially about two-thirds full. The composition of the test mixture should be adjusted to 30—40mol% of low-boding component. The influence of the composition on the separating effect has as yet not been fuDy established [208]. 

With the carefully purified components mixtures are prepared in the range of 5 to 95% (wt), their refractive indices are measured and a calibration curve ( -% wt) is plotted. If desired, the number of stages required with v = oo for a series of enrichments may be calculated by means of the Fenske equation (114), and a chart for these values (Fig. 71) against the refractive index constructed. From this chart the number of stages may be read off directly after taking samples and determining the refractive indexes. 

Destinorm sampling device for normal and reduced pressures 

If reliable data are to be obtained it is necessary to operate the column adiabati-cally. The stiU pot and all components up to the condenser should therefore be carefully insulated, for example with asbestos cord and glass wool, so as to minimize loss of heat (c/. section 7.7). Since, even at temperatures below 80°C, a certain amount of false reflux is formed, it is essential to compensate for the loss of heat from the 

---