No abstract

Waveforms used in pulsed electrochemical detection (PED) have frequencies of ca. 1 Hz when optimized to give a maximum signal-to-noise ratio (SIN) for carbohydrates. However, higher frequencies are desired for applications of PED to capillary electrophoresis (CE) and microcolumn liquid chromatography (MLC), which can produce narrow elution peaks. Minimization of the time periods for oxidative cleaning and reductive reactivation in PED waveforms generated by the Dionex Pulsed Electrochemical Detector allows the increase in waveform frequency to ca. 3 Hz without change in the traditional time period of 200 ms, prescribed for integration of the anodic current. However, further increase in frequenq requires a decrease in the integration period with a corresponding loss of signal strength.

The anodic charge (9) for oxide formation, obtained by potential-step chronocoulometry at a Au minidisk electrode, is nearly a linear function of log (t/ms> for t = ca. 3-30 ms with a slope proportional to the applied overpotential for oxide formation. It was demonstrated that, contrary to popular opinion based on voltammetric observation at slow scan rates ($1, the anodic current for oxide formation obtained using a linear potential-scan waveform is not a linear function of + at large values of + . It was also demonstrated that a reverse potential-step waveform can be used t o interrupt the rapid oxide growth and, perhaps, inhibit the conversion of the hydrous oxide (AuOH) to the inert oxide (AuO). This result has probable significance for decreasing the background current during anodic pulsed amperonietric detection of amines and sulfur compounds whose oxidation rneclranisrns are believed catalyzed by AuOH but not bj7AuO. The cathodic charge (&I for oxide dissolution is approximately a linear function of t with a slope that is proportional to the applied potential.

Results of a potential-step chronocoulometric study of oxide formation at a Pr-rotated minidisk electrode (0.00785 cm') indicate that the anodic charge (qcJ grows ca. as a linear function of the log time ( t ) for t = ca. 2-30 ms. Furthermore. the slope of the linear qo -log (rims} plot is proportional to the applied overpotential for oxide formation. The anodic peak current observed during linear potential-scan voltammetric experiments is nearly a linear function of scan rate (4) for small 4 (1000 < mi. 5-l) bur shous substantial negative deviation from lineari? for 4 > 1000 mi7 s-'. The peak potential for oxide formation shifts in a positive direction for increasing values of 4, suggesting that this process is kinetically slow relative to large 4 values. Reverse potential-step chronocoulometric measurements demonstrate that the background current in so-called "reverse-pulsed amperometric detection (WAD)" can be decreased by inhibiting the conversion of the hydrous oxide (PtOH) to the inert oxide (PtO). Data also indicate that the rate of cathodic dissolution of surface oxide is dependent on applied potential for t < 2 to 30 ms. Oxide reduction continues even for t > 1000 ms.

INTRODUCTION 102 EXPERIMENTAL 105 Reagents 105 Equipment 105 Sample preparation Ill Waveforms Ill RESULTS AND DISCUSSION System Performance Variation of oxidative cleaning time, tgjtd Variation of oxidative cleaning potential, Eoxd Variation of reductive cleaning time, t^ Variation of the integration period, ti^