1. Introduction

The ability to detect and measure small amplitude signals electrochemistry can be very useful when there is a need minimize the input signal in order to perturb as little as ssible the electrode/electrolyte interface as well as to crease unwanted effects such as the ohmic drop error, d when the aim of the research is to study the spontanes fluctuations in current and potential, which are comonly referred to as electrochemical noise. This area of deavor is receiving increasing attention in the field of ectrochemistry [1]', and it is being applied for the detecon of corrosion [2]. One of the problems encountered in ese studies is caused by the need of controlling the potenl of the electrode without causing an unacceptably high wel of interference from the noise generated by the potenstat [3]. In order to improve its performance with respect commercially available instruments, a low-noise potenstat was designed and built, whose characteristics will be scribed in the following sections.

2. Instrument Characteristics

The low-noise potentiostat was designed and developed the Special Analytical Instrumentation Group of the enter for Analytical Chemistry at the National Bureau of andards.

Center for Analytical Chemistry, National Measurement Laboratory.
Center for Materials Science, National Measurement Laboratory.
Figures in brackets indicate literature references at the end of this paper.

The design has exceptional low noise and stable low frequency performance as well as moderately wide bandwidth. The characteristics of the potentiostat include:

a. Highly stable operation over a wide range of loads with up to 5 mA output.

b. A useful frequency range from dc to greater than 2 kHz.

c. A precision low noise voltage reference source variable from ± 5 V, having a stability of better than 10 μV for periods greater than 1 hr.

d. A fast recovery, low-noise, ac current signal amplifier with a frequency response of between 5 × 10-2 and 2 × 103 Hz.

Additional features include an external input to allow small signal modulation of the internal voltage reference and the ability to be used as a constant current source or galvanostat. The circuit scheme is shown in figure 1.

3. Design and Construction Criteria

In order to achieve the characteristics desired, the design was chosen so that the working electrode was grounded, facilitating the shielding of stray signals. To reduce capacitative currents across the reference electrode, the lead shield was maintained at the same potential, effecting a guarded input circuit. The high impedance guarded input amplifier is followed by the potentiostat loop amplifier and an output current buffer. (Total noise referred to input is less than 3 μV rms.) The frequency response of the circuit

crosses unit gain at approximately 6 kHz. Immunity to oscillation for all values of cell R and C was secured by tailoring the frequency characteristics of the feedback amplifiers. No provision to compensate for cell parameters at high frequencies is included. A differential amplifier is used to extract the dc current signal from the potentiostat current driver. Signal output comes from an ac coupled amplifier having a very low frequency cutoff (5 x 10-2 Hz) and exhibits extremely low noise and prompt overload recovery. It has three gain settings up to a maximum gain of 1000. For ac components within the frequency range, an ultimate resolution as small as 3 nA (3 mV output) can be measured. A fast recovery circuit brings the amplifier into its linear range within a few seconds for any overload.

Power is supplied by a pair of rechargeable 12 V gelled lead acid batteries to assure a flat and constant discharge as well as complete freedom from ac pickup. All amplifier circuits make use of high performance operational amplifiers. The amplifier chosen has excellent noise and drift performance with long-term drift stability approaching that of chopper stabilized amplifiers.

A double box construction was used to reduce ambient thermal effects and improve immunity to electrical noise. Resistors used were glass-tin oxide, and all capacitors used

in active circuit locations were of polystyrene foil constra tion. All loose wiring was mechanically secured by using dabs of silicone rubber sealant to reduce vibrations causing microphonics.

4. Performance Characteristics

The potentiostat is employed as a part of the circu shown in a block diagram in figure 2. For testing purposes the cell was substituted with a passive load, which was either a pure resistor or a resistor and a capacitor in para lel. With the switch in the upper position the fluctuations the current were analyzed, while with the switch down, the noise voltage applied to the load was recorded. If desired either a constant or a swept frequency signal of small amp tude was superimposed to the control voltage. All spectra shown in this work were recorded in four sections, over the 5, 50, 500, and 5,000 Hz range. The spectrum analyzer has a frequency resolution of 1/200 of the range, requiring a time of 40, 4, 0.4 and 0.04 s respectively for each spectrum. recordings are averages over 64 spectra for the lowes range, and 256 or 512 spectra for the higher frequency ranges. All data given are rms values.

When the resistive load R is sufficiently low (<500 Q), Icv predominates, as shown by the two upper curves in figure 3. When R is large, however, IN ~ IA, as shown by the lowest curve of figure 4. I, is also largely independent of frequency above 1 Hz and results to be about 3 x 10-11 A/√Hz.

The potentiostat was also tested on a circuit formed by a 10 resistor in series with a parallel combination of a 5 KQ resistor and a 5μF capacitor (as shown in the inset in figure 5). In this case, the load impedance is frequency dependent. The voltage and current spectra are given in figure 5. Taking the ratio between these values at the same frequency should give the absolute value of the impedance Z.

10-6

However, as shown before in the case of a resistive load, pedances greater than several hundred ohms are too la to give a current significantly above I, if one relies only the control voltage noise. Because of the low intrinsic no of the potentiostat, the addition of a very low amplit swept frequency signal to the control voltage was suffici for obtaining good impedance values. The voltage and rent spectra obtained are shown in figure 6, and the plot |Z❘ vs. frequency from which the value of the circuit com nents can be obtained, is shown in figure 7. These me urements were obtained applying no more than a few ten of a mV (in the lowest frequency range) and generating c rents less than 100 nA. The possibility of measuring at. amplitudes can be an important advantage when, otherwi distortions or irreversible changes in the electrochem: system could be produced.

5. Performance as a Galvanostat

As mentioned before, the instrument can be switched i configuration where it supplies a constant dc current, up 5 mA, provided the total load does not exceed 5 V. The