A disadvantage of the electrokinetic or electromigration injection method in capillary electrophoresis is the sample bias caused by differences in the electrophoretic mobilities of the sample ions. This sample bias problem can be eliminated by isolating the sample-introducing end of the capillary from the injection electric field, since, in this case, the sample solution is pulled into the capillary by the electroosmotic flow independently of the electrophoretic migrations of the sample ions. This paper reports an injection method using the pulling action of the electroosmotic flow. The sample-introducing end of the capillary is isolated from the injection electric field by utilizing an on-column Nafion joint. In order to demonstrate that the sample bias problem is eliminated in the method,
p-hydroxylbenzoic acid (1) and nicotinic amide (2) were selected as model samples since they have different electrophoretic mobilities. The model sample is dissolved into the Na
2B
4O
7-NaOH buffer solution (pH 10.5). After the sample is introduced into the capillary, electrophoresis is carried out under an electric voltage of 20 kV. The ratio of peak areas S
1/S
2 of the model samples (1) to (2) are calculated from the electropherogram obtained with a UV detector. The values of S
1/S
2 for the conventional electrokinetic injection method, the gravity flow injection method, and the present method are compared. Experimental results show that the value of S
1/S
2 for the conventional electrokinetic method is much larger than that for the gravity flow method; moreover, the value of S
1/S
2 increases with the injection voltage. On the other hand, the value of S
1/S
2 for the present method is the same as that for the gravity flow method, and does not vary with change in the injection voltage.
View full abstract