Page 17 - MODERN ASPECTS OF ELECTROCHEMISTRY

Michael SpiroA
4
mercury surface since the exchange current density of the hydrogen couple
is known to be larger on a zinc electrode than on one of mercury. 5
Several groups of woÀers in the literature have since tested the
additivityprinciple. In most cases it, or corollaries based on it, were found
to hold within the uncertainties of measurement.= Modern texts on corro -
6,7
sion t¸e the additivity principle for granted throughout, while a recent
8,9
compilation on electroless plating cites various literature tests in this
area, not all of them favorable. It is preciselyin corrosion and in electroless
plating that most doubt arises about the validityof the test. To be certain
of the significance of the results, it is essential that each process be carried
out on the same surface.= But as shown for the zinc corrosion example, this
cannot be done directly for one of the partial reactions: the reduction of
the oxidant in corrosion or the oxidation of the reducing agent in elec-
troless plating. Attempts to obtain the electrochemical information indi-
rectly by studying the polarization curves over a much lower or much
higher potential range, respectively, and extrapolating the current to the
corrosion or plating potential, are not valid if the rate-determining step
(and hence the Tafel slope) changes over the extrapolated section.= They
are also not valid if the catalytic properties of the surface change with
10
potential over the region concerned. Moreover, in both corrosion and
plating the surface is continuouslybeing regenerated, a point rarelyt¸en
11
into account. In spite of these limitations, however, a small number of
careful studies has shown clearlythat the two partial reactions in corrosion
or electroless plating are not always independent of each other. The main
evidence from two major studies is summarized below.

IV.A SELECTED CORROSION AND ELECTROLESSAPLATINGI
STUDIES
10
Andersen et al. carried out a detailed investigation of the corrosion of
o
copper in stirred oxygenated H SO solution at 25 C.=The stable corrosion
4
2
products, at times up to 40 hr, were found to be equimolar amounts of Cu 2+
ions and H O . The rates of copper oxidation at various potentials were
2
2
determined bytwo independent methods: weight loss and polarization
curves. The results of these two types of measurement agreed moderately
well with N -saturated solutions whereas in O 2 -saturated solutions the
2
copper weight loss was always substantially larger than that derived from
the current-potential curve and Faraday's law. Furthermore, the weight
loss at the corrosion potential E corr in oxygenated solutions was some ten