The agreement between experimental data (solid line in Figure
4b) and mean free path corrected theory is excellent, especially
as the absorption was calculated quantitatively from size and
concentration of the particles without any adjustable parameter.
The calculated position of the surface plasmon peak as
displayed in Figure 5 (circles) is in perfect agreement with the
experimental data (triangles in Figure 5) for particle sizes from
25 to 120 nm. For sizes smaller than 25 nm the experimentally
observed peak position is somewhat lower than predicted by
theory. This may be attributed to the pronounced increase of the
ratio of surface atoms to bulk atoms for particle diameters smaller
than 20 nm. However, for sizes larger than 25 nm both the
theoretical and the experimental peak positions (ìspr) are both
fitted precisely by a simple exponential function (R2 ) 0.99, see
the dotted and the dashed lines in Figure 5)