Ng. Khlebtsov et al., Light absorption by the clusters of colloidal gold and silver particles formed during slow and fast aggregation, COLL J, 62(6), 2000, pp. 765-779
Spectra of absorption (400-800 nm) by the aggregates of colloidal gold (5,
15, and 30 nm in diameter) and silver (20 nm in diameter) particles were st
udied experimentally and theoretically. It was revealed that, during fast a
ggregation corresponding to the diffusion-limited cluster aggregation (DLCA
), the pattern of spectra is dependent on the size of primary particles. Sp
ectra with the additional absorption maximum in the red region are observed
for 15 and 30 nm gold and 20 nm silver particles, while the absorption spe
ctrum for 5 nm particles is characterized by only one maximum shifted to th
e red region. The slow aggregation resulted in a decrease in plasmon absorp
tion peak with no marked shift to the red region and to the broadening of l
ong-wave absorption wing. From data on electron microscopy, typical branche
d DLCA-clusters were formed during fast aggregation, whereas small compact
aggregates and a noticeable number of single particles were observed in a s
ystem during slow aggregation. The computer model of the diffusion-limited
cluster-cluster aggregation was used to explain these results. Optical prop
erties of aggregates were calculated using coupled dipole method (CDM or DD
A) and the exact method of a multipole expansion. Corrections for the size
effect were introduced into the bulk optical constants of metals for nanosi
zed particles. It was shown that a modified version of DDA (Markel et al.,
Phys. Rev: B, 1996, vol. 53, no. 5, p. 2425) allows us to explain the patte
rn of experimental spectra of DLCA-aggregates and their dependence on a mon
omer size. The exact method was applied to calculate: the extinction cross
sections of metallic aggregates demonstrating strong electrodynamic interac
tion between particles. The number of higher multipoles that are required t
o adequately describe this interaction is much larger than the number of te
rms of an ordinary Mie series and is the main obstacle to the exact calcula
tion of the spectra of metallic aggregates with a large number of particles
.