Cathodic Pulse Stripping Analysis of Iodine at the Parts-per-Billion Level R. C. Propst Savannah River Laboratory, E.
I. du Pont de Nemours & Co., Aiken, South Carolina 29801
The deposition-stripping behavior of iodide films at the hanglng mercury drop electrode Is investigated. Surface monolayers of Hg-HgI and Hg,12 are postulated based on the percentage occupancy of the available sites at the surface of the HMDE. Submonolayer films of Hg-HgI, as stripped from the eiectrode, are used to determine iodine concentrations from 1 to 50 ppb with a standard deviation of 0.3 ppb.
Sensitive analytical methods are required for environmental studies of the amounts of trace substances released and their ultimate fate. Stripping methods are frequently selected for this application because the capital investment in equipment is low and because the high sensitivity of the method largely eliminates the need for preconcentration of the sample. Existing stripping methods for iodide did not exhibit the 0.5-ppb sensitivity desired for the determination of iodine in surface water. Perone et al. (1) have described the stripping analysis of iodide a t the silver electrode. The deposition-stripping processes were uncomplicated; however, a 30-min deposition was required to achieve a detection limit of 5 ppb. The nature of our samples precluded incorporation of organic solvent (2) into the electrolyte system to improve sensitivity. Methods for the stripping analysis of iodide at the hanging mercury drop electrode (HMDE) have been plagued by the anomalous deposition-stripping behavior of mercurous iodide films. Kemula et al. ( 2 )report a detection limit of about 65 ppb for iodide in 0.1 M KNOB. They observed deviations from linear stripping behavior with both more dilute and more concentrated solutions of iodide, especially after prolonged electrolysis. Also, the appearance of two peaks of different heights complicated estimation of the height of the stripping peak. Perchard et al. (3) report that, for concentrations of less than about 13 ppb I-, the relationships governing the analytical applications no longer hold. The anomalous depositionstripping behavior was attributed to the special properties of a monolayer deposit of mercurous iodide. Boult et al. ( 4 ) investigated the formation of halide films on mercury. They report that iodide films do not exhibit interfacial orientation but form loosely organized deposits from dilute solution in which the crystals are mixtures of mercurous iodide, mercuric iodide, and mercuric oxide. Also, the presence of 0.01 M "OB apparently prevented formation of mercuric oxide and, to a greater extent, mercuric iodide. Because of the unknown deposition-stripping behavior of iodide at the HMDE from dilute (0.023 cm2 and for an accumulation time of 16 min because the drop detached before the required accumulation time had elapsed. The results expressed in terms of the rate of change of charge with surface area were 87.8 &/cm2 a t 2.63 X M I- and 170.7 pC/cm2 at 2.63 X M I-. Thus, the limiting slope must be associated with a surface phenomenon; and in the limiting slope region, the ratio of mercury to iodide must be changing from 2:l to 1:1,i.e., from HgHgI to Hg212. The development of the stripping peaks as a function of surface coverage is readily ascertained by comparing the stripping curves (Figure 3) with the graph of the integrated stripping current (Figure 2). The correspondence between
the curves (Figure 3) and the points on graph (Figure 2) is indicated by the symbols a-h. The stripping curves are typical of those recorded in this study. For short accumulation times at low concentrations, a small peak near -0.6 V appears on the stripping scan (Curve a, Figure 3). This peak is quite diffuse but shows increased resolution and an anodic shift in potential (towards 4 . 4 5 V) as the amount of deposit increases. As the coverage approaches 50%, a shoulder appears near -0.38 V; as the coverage nears loo%, the peak at -0.38 V becomes equal in height to the peak at -0.45 V. When the coverage exceeds loo%, the character of the peak near -0.38 V begins to change. At multilayer coverages, this peak assumes the form of a stripping peak where the activity of the deposit is constant. This behavior is illustrated in the curves of Figure 4. Graphs of the integrated stripping current vs. concentration or deposition time for the region from 0 to 50% coverage are linear (3) by visual inspection; however, the RMS error is reduced by a factor of 20 when linear regression techniques are used in going from a linear to a cubic fit. The RMS error is not significantly reduced on going to a polynomial of higher degree. This nonlinear behavior is apparently related to an observed change in the accumulation rate with coverage a t low (
