Micromorphological Studies of the Corrosion of Gold Alloys
0
A. J. Forty Department of Physics, University of Warwiek
, Coventry,
U.K
Considerable insight into the detailed mechanisms by which metals are corroded can be derived from direct microscopic observations. This article describes those micromorphological changes occurring near the surface that can be observed by transmission electron microscopy after gold alloys have been subjected to anodic dissolution in strong acids. These observations are used to discuss the important problem of corrosion by selective dissolution.
-
Direct observation of the surface of a metal after
exposure to a corrosive environment has frequently
been used as a basis for deductions concerning the
structural processes by which chemical reactions
inodify the surface and near-surface layers. Although
considerable information can be obtained in this way
using simple optical microscopy, as, for example, in
the case of the aqueous corrosion of gold-copper
alloys studied by Graf (1), Bakish and Robertson (2)
and Pickering (3), a much more detailed picture can
be obtained with the electron microscope,
particularly when this is coupled with the powerful analytical
techniques of selected area electron diffraction and
X-ray microanalysis. The availability of very high
resolution microscopes now makes it possible to
follow changes in the interaal microstructure, the
surface morphology and the composition of metals and
alloys on a scale approaching atomic dimensions.
The use of the transmission electron microscope
necessitates the study of very thin specimens. Great
care must be taken to ensure that the
micromorphological changes arising from corrosion can be
distinguished from those produced during the
preparation of such thin films from bulk samples. For
this reason, there may be doubts concerning, the
validity of some of the observations by Pickering
and Swann (4) and by others who have studied the
corrosion of alloys prepared as thin foils by
electropolishing techniques. Ion-sputtering methods for
thinning alloys from bulk specimens may also be
suspect, because different sputtering rates for the
various constituents can lead to compositional
changes. These difficulties have been largely
overcome recently by Durkin and Forty (5) who have
developed techniques for preparing thin films by
vapour deposition of an alloy from its individual
components. It. will be shown later how this has
contributed to a very detailed understanding of the
corrosion micromorphology in the special case of
gold-silver alloys.
Selective Dissolution
The most widely studied and possibly the most
important phenomenon involved in the aqueous
corrosion of gld alloys is that of selective dissolution,
whereby the less noble element is preferentially
removed from the alloy, leaving a gold-rich residue (6).
This is the basis of various practical methods for the
parting gold from its alloys. It is also thought to be an
important step in the stress corrosion of gold alloys
since rupture of the gold-rich surface layer by an
applied stress can lead to the initiation of localized,
deeper corrosion and subsequently of a stress
corrosion crack (1, 2, 3). As we shall discuss later, selective
dissolution might also be an important precursor of
other corrosion reactions, such as oxidation.
Fig. 1 Schematic representation on an atomic scale
of the surface of an alloy composed of dissolvable A
atoms and noble B atoms.
K is a kink site on a surface step
N is a non-kink site on a step
T is a terrace site
The fundamental question to be answered, as far
as the understanding of selective dissolution is
concerned, is why a gold-based alloy should continue to
dissolve in this way beyond the stage where the
surface should be passivated by the gold residue. Such
passivation might be expected to develop at a very
early stage, as can be seen from a consideration of the
atomic processes that might be occurring on the metal
surface during dissolution. These are depicted in
their simplest form in Figure 1, where we ignore
molecular adsorption, oxidation and complexing
effects associated with the electrolyte, and assume that
dissolution involves only ionization and solvation of
the metal atoms. Dissolution is expected to occur
preferentially from kink sites (K) in the surface steps
where the atoms are least firmly bound and, at
sufficiently low potentials, the dissolution current will
involve predominantly A atoms the less noble
species. As dissolution proceeds, however, this
current will be diminished as more and more kink sites
become occupied by more noble B atoms. Thereafter,
dissolution can proceed only by the removal of A
atoms from non-kink sites (N) on steps or from terrace
sites (T), which requires a greater activation energy or
overpotential. Eventually, the alloy becomes
completely passivated when all the surface sites are
occupied by B atoms only. For most alloys, and Cu 3Au
in particular, this passivation stage should be reached
after the removal of A atoms from only a few atomic
l (...truncated)