[Key factors in the control of electroosmosis with external radial electric field in CE].

Direct control of electroosmosis flow (EOF) by external radial electric field was performed at room temperature using a home-made field-modulated capillary electrophoresis (CE) system. The EOF was monitored at 206 nm by using DMSO as a probe. To apply a radial electric field across the CE capillary wall, the capillary was cased with a wide column. Both of the concentric space and the capillary bore were then filled with an identical running buffer and applied with an axial electric field of 150 V/cm but starting from different levels. All of the tubes used were made of fused silica with polyimide over-coating (from the Yongnian Optical Fiber Work, Hebei, P. R. China). The size of the CE capillaries adopted was 25-100 microns i.d. (375 microns o.d.) x 28.5/45 cm (effective/total length), and that of the casing column 400 microns i.d. x 32 cm. To investigate the fundamentals of the external EOF control when using the flexible fused silica capillaries, various parameters have been inspected such as pH, buffer composition, additives and capillary wall feature etc.. As expected, to well control both of the magnitude and direction of the electroosmosis, the buffer pH should be kept below 4 and the buffer concentration below 50 mmol/L. However, buffers below 1 mmol/L should be avoided because such a diluted running buffer may result in poor CE separation. Weak electrolytes like citric acid, tartaric acid and acetic acid were found to be capable of generating better EOF control than the strong electrolytes such as phosphate and chlorides. This is possibly due to the formation of looser electric double layer with the weak rather than the strong electrolytes. Some wall coatings like calix arene and its derivatives can evidently improve the EOF control even at pH 5. This reveals an exciting way to expend the controllable pH range. In addition, narrow-bore capillaries were demonstrated to be better than wide-bore tubes. Other conditions such as buffer additives and capillary rinse procedure were shown to have only negligible influence on the control.