Strategies for low detection limit measurements with cyclic voltammetry.

Cyclic voltammetry of Nafion-coated, carbon-fiber electrodes is used to detect trace concentrations of dopamine, both in a flow injection apparatus and in the brain of an anaesthetized rat. To improve signal-to-noise ratios, the sources of noise during cyclic voltammetry have been determined and strategies have been developed to decrease the noise. With the potentiostat employed, the measured noise is comparable to that expected for Johnson noise from the feedback resistor of the current transducer. Additional noise arises from the waveform generator employed and, in some cases, line noise. Line noise is discriminated against by starting each cyclic voltammogram either in phase or 180 degrees out of phase with the line frequency. When used in vivo, additional noise also arises from the physiological activity of the animal. Detection limits are found to closely correspond to those predicted on the basis of simulation of the voltammetric shape and the measured noise. Detection limits are improved by the use of appropriate analog and digital filtering, ensemble averaging, and appropriate timing of repetitive cyclic voltammograms. The combined use of these techniques enables the in vivo detection of approximately 100 nM of dopamine with a signal-to-noise ratio of 25.