Etching 10 min with a current of 100 mA (1.14 mA/mm2)
produced a surface with ~5 lm wide furrows of up to 5 lm depth around the grains (Rz = 9.8 lm) (Fig. 2). The
furrow depth was measured by SEM analysis of a polished
section. Applying a current of 200 mA (2.28 mA/mm2) for
10 min produced a very rough (Rz = 15.8 lm), granulated
surface (Fig. 3). In order to get detailed information about
the selectivity of the etching process in dependence of the
applied current, polished sections were made of stents
etched with different currents and etching times. The
parameters were chosen in a way that the furrow depth is
smaller than half the grain perimeter, so that detachment of
grains is avoided. From each stent strut the average length
of the five longest crevices were measured, as well as the
thickness loss of the strut. The ratio of the average of the
furrow depth plus the thickness loss to the thickness loss
was regarded as an index for the selectivity of the etching
process. Interestingly it was found that the selectivity does
not vary significantly (Fig. 4) within the measured range of
50 mA (0.57 mA/mm2) to 800 mA (9.12 mA/mm2). This
observation is in contrast to former grain boundary etching
models where the U/I curve of grain boundaries and matrix
is assumed as illustrated in Fig. 1. According to this model
a high selectivity could only be achieved in a small range
of voltage, where the current curve has a high slope.