QUANTITATIVE STUDIES OF DIFFERENT INJECTION SYSTEMS IN CAPILLARY ELECTROPHORESIS

For a rigorous assessment of the precise amount of sample loaded, for quantitation purposes, different sample injection systems were evaluat ed with two commercially available units, Waters Quanta 4000 and Beckm an P/ACE 2100. In the first system, sample introduction by hydrostatic means (i.e., placing the sample vial at some height, usually 10.1 cm, above the other capillary end) was evaluated. It was found that in th is system there is a constant positive bias, i.e. the amount of sample loaded lies on a curve parallel and above the theoretically predicted loading curve. However, the excess of mass loaded was constant along the injection times explored (covering from 5 to 35 s) and, for a 75 m u m capillary, was found to be of the order of +6 nL (above the expect ed injected value). Thus it is easy to correct for this sample bias. I n the electrokinetic mode, a very good correlation between expected an d predicted sample loads was obtained for both units. In the pressure system (by positive pressure from a nitrogen tank, Beekman unit) a sub stantial discrepancy was found between experimental and predicted valu es (13.5% overload). Since the manufacturer claims a constant pressure of 0.5 psi, i.e. 3447.5 Pa, it would appear that the injection pressu re is higher than the given value. Other causes for variation in sampl e load (e.g., as caused by diffusion of sample in the capillary just p rior to injection, by the sudden insertion of the capillary tip into t he sample vial, etc., lumped together in a general term as ''extraneou s injection'') have been evaluated and found to contribute to an addit ional sample volume injection of the order of 1-2 nL, i.e. quite negli gible. In conclusion, it is felt that both the hydrostatic and electro kinetic injection modes are highly reliable. The positive pressure mod e can fail due to leakage of O-rings, valve malfunctioning and to the typical problems of mechanical systems. Thus, the latter injection sys tem should be checked frequently for potential mechanical failures.