Search

A number of experimental approaches that should offer improved S/N in electrochemical detection were evaluated and compared. In addition, to evaluate and optimize the electrochemical response behavior of analytes under actual CE conditions, an on-line cyclic voltammetry (CV) system was developed. The experimental parameters examined included waveform shape, waveform frequency and various signal treatments, using a lock-in amplifier and Fourier analysis. A multiple-step pulse waveform provided the maximum S/N enhancement [up to 10-fold relative to pulse amperometric detection (PAD)], with detection limits in the range of 2·10−8 to 2·10−7 mol/l. The use of the second harmonic from Fourier analysis offered the best improvement in baseline stability. Other approaches such as high pulse frequency (100 to 200 Hz), data collection over selected time windows, digital filters and average smoothing also enhanced S/N 2- to 5-fold relative to PAD. On-line CV studies showed that the adsorption of organic electrolytes on electrode surfaces can inhibit O2 reactions, and thus give low and stable background currents without O2 removal. The CV studies also showed that detection of Pb2+, Cu2+ and Ag+ affected subsequent electrode response, and that H+ evolution contributed to the cathodic signals of the analytes Ni2+, Co2+ and Zn2+.

The electronics and construction for an end-column ultramicroelectrode (3–10 μm) detection system that permits the use of medium-sized capillaries (25 μm I.D.) without appreciable effects from the potential field at the end of the capillary. Normal peak-to-peak noise over 10 s was 0.01–0.1 pA. The background noise observed for a 200 × μm carbon-fiber electrode placed either 180 μm within a 25-μm capillary or at a point 500 μm away from the capillary was essentially the same. A study of detector response as a function of the position of the electrode has shown that accurate location of the electrode is important for sensitive and reproducible detection. These studies also showed that differences between the density of the electrolyte existing the capillary and the electrolyte in the detection cell could cause anomalous electrode response depending on the location of the electrode relative to the end of the capillary. Application of a carbon fiber or an Hg film electrode gave detection limits (twice the peak-to-peak noise over 10 s) of 2 · 10−8 mol/l for Pb2+, 1 · 10− 5 mol/l for NO2− and 5 · 10−10 mol/l for catechol.