Abstract
Continuous separation of three carbohydrate mixtures [fructose‐dextran (M. W. ≈ 9,400), raffinose‐dextran (M. W. ≈ 6,000), and fructose‐raffinose] has been carried out using a simulated countercurrent process with silica gel as the sorbent and deionized water as the eluent. Experiments were conducted using the commercial four‐section arrangement and a modified three‐section flow scheme which does not recycle part of the eluent for reuse. Data for the four‐section flow scheme reflect a much higher product concentration in the raffinate and a substantially lower rate of fresh eluent consumption than for the three‐section scheme. It is concluded that when both the extract and raffinate products are desired, the four‐section scheme should be used for the separation of a mixture with a small separation factor which usually requires a large quantity of eluent to achieve an efficient separation. The behavior of the simulated countercurrent process was simulated theoretically by representing the process in terms of an equivalent true countercurrent system. A good fit to the experimental data was obtained when the process was configured in the four‐section mode. However, the same theoretical approach failed to provide a good representation of the behavior of the three‐section scheme in one respect: predicted concentrations of the raffinate product were about two times higher than experimental values. The discrepancy may be attributed to the periodic switching of fluid inlet and outlet points in the operation of the three‐section simulated countercurrent unit.