Preparative Chromatography - Sample Mass Overload > Page 16
Examination of figure 6 shows that the descending steps of the curves, which correspond to the tails of the normal elution curves, are very diffuse compared with the ascending steps, which correspond to the fronts of the normal elution curves. If the sample is placed on the column by allowing the mobile phase to flow through the loop for a given time (determined from the flow rate and the volume of sample selected for injection) and the valve then rotated to allow the mobile phase to pass directly to the column, the dispersion effect of the sample tube is virtually eliminated. The improvement resulting from this technique (often called back-cutting injection) is shown in the upper curve in figure 6. It is seen that the descending steps of the curve are very similar to the ascending steps showing that the tube dispersion has been significantly reduced. It should be pointed out that this injection procedure does result in some loss of sample, due to that retained at the walls of the ample tube, but this can be easily recovered and recycled if necessary. Thisn technique is strongly recommended for preparative sampling and should be employed wherever possible.
Sample Mass Overload
The effect of excess mass of sample (mass overload) on the chromatographic process can be far more complex than volume overload. The theory of mass overload is, as one might expect, also complicated (5–7) and requires a considerable amount of basic physical chemical data, such as the adsorption isotherms of each solute measured over a wide range of concentration, before it can be applied to a practical problem. Only if the separation problem demands an extremely high through-put, and the process must be as economic as possible, will it be worthwhile to gather the necessary basic data. The problem of mass overload is more conveniently and economically solved by a simple experimental approach.
Depending on the ultimate concentration of solute at the point of injection, a number of effects can take place when a large sample mass is placed on the column. In the first instance, there will be the effect resulting from the limited capacity of the stationary phase. On injection, the sample will spread along the column, carried by the mobile phase, until it contacts sufficient stationary phase surface to allow it to be held on the surface under equilibrium conditions. This will result in a spreading process similar to sample volume overload and, if this were the sole contribution to mass overload, could be treated in a similar manner. The peaks would be square topped and similar in shape to those shown in figure 5. However, superimposed on this band dispersion process, is that arising from the deactivation of the adsorbing surface and the change in polarity of the mobile phase due to the presence of the solute. If the charge is substantial, the sample will occupy a significant slice of the column immediately after injection and the adsorbent (stationary phase) will become partially deactivated causing all the solutes in the mixture to be accelerated through the column with consequent reduced retention times. The increased migration rate is further aggravated by the higher polarity of the mobile phase, which results from the high concentration of the overloaded solute that it contains. This increase in mobile phase polarity causes the solutes to be further accelerated through the column reducing their retention time still more.