Reference electrodes for use in the potentiometric determination of chloride. Part I. Assessment of mercury-mercury(I) sulphate electrodes

Abstract

Mercury-mercury(I) sulphate reference electrodes are used in the potentiometric determination of chloride. There have been various reports of their erratic behaviour, which is especially serious because chloride-selective electrodes are often working at the limits of their sensitivity in highly pure waters, such as condensed steam and boiler water. The loss of precision in e.m.f. in such circumstances has a much larger effect on the precision in concentration terms than when electrodes are working in their usual (Nernstian) sensitivity range. Commercial mercury-mercury(I) sulphate electrodes have been tested over several weeks of continuous operation so that their performance can be assessed when used under conditions similar to those of chloride monitoring, i.e., in pH 4.4 acetate buffer solution with silver chloride electrodes and 0.01 mol l^–1 nitric acid solution with mercury(I) chloride electrodes. Electrodes were obtained from two manufacturers and were of two types—one with ground-glass sleeve liquid junctions and one with a ceramic-frit junction. The latter was the best of those tested. The rate of change of its potential with respect to a calomel electrode was less than 0.5 mV over 100 h and the standard deviation of its potential over a week's operation was about 0.3–0.5 mV. Deviations of potential previously reported with this type of electrode could not be reproduced in this work. Some electrodes with ground-glass sleeve junctions became erratic after 2–4 d in the nitric acid solution normally used with mercury(I) chloride sensing electrodes. Their performance was better in the acetate buffer medium used with silver chloride sensing electrodes but inferior to that of similar electrodes with ceramic frit junctions.