Thermal effects and temperature profiles in capillary electrophoresis

Whilst temperature control is usually employed in capillary electrophoresis (CE) to aid heat dissipation and provide acceptable precision, internal electrolyte temperatures are almost never measured. In principle this limits the accuracy, repeatability and method robustness. This work presents a fundamental study that combines the development of new equations characterising temperature profiles in CE with a new method of temperature determination. New equations were derived from first principles relating the mean, axial and inner wall electrolyte temperatures (TMean, TAxis and TWall). TMean was shown to occur at a distance 1/‚Äöv†v¿2 times the internal radius of the capillary from the centre of the capillary and to be the average of TAxis and TWall. Conductance (G) and electroosmotic mobility (˜í¬¿EOF) were used to determine TMean and TWall, respectively. Extrapolation of curves of ˜í¬¿EOF versus power per unit length (P/L) at different temperatures was used to calibrate the variation of ˜í¬¿EOF with temperature (T), free from Joule heating effects. ˜í¬¿EOF increased at 2.22 % per ¬¨‚à´C. The experimentally-determined temperatures using ˜í¬¿EOF agreed to within 0.2 oC with those determined using G. The accuracy of G measurements was confirmed independently by measuring the electrical conductivity (˜í‚à´) of the bulk electrolyte over a range of temperatures and by calculating the variation of G with T from the Debye-H‚àö¬¿ckel-Onsager equation. TMean was found to be up to 20 oC higher than the external temperature under typical conditions using active air-cooling and a 74.0 ˜í¬¿m internal diameter (di) fused-silica capillary. A combination of experimentally-determined and calculated temperatures enabled a complete temperature profile for a fused-silica capillary to be drawn, the thickness of the stationary air layer to be determined and the heat transfer coefficient (h) for the capillary to be determined.