Organization of size-selected platinum and indium clusters soft-landed on surfaces

Metal cluster cations are produced in a laser vaporization source, size selected in the gas phase by an electrostatic quadrupole deflector, and deposited on transmission electron microscopy grids coated with various substrates. The morphology of cluster assemblies is very dependent on the occurrence of diffusion, collision, and coalescence of incident particles on the surface. In this respect, indium and platinum clusters show very different behaviors: the former easily coagulate to form large entities when the latter do not coalesce and tend to form structured arrays with a uniform distance between first neighbors.

Figure 1

TEM micrographs of size-selected indium clusters $(d_1=5.79\mathrm{nm})$ deposited on amorphous carbon. The coverage rate increases from left to right. The leftmost image (a) shows “as deposited” primary particles that have not coalesced on the surface.

Figure 2

Size histograms of indium clusters measured from the micrographs in Fig. 1. The horizontal axis $d^{*}$ is the effective particle diameter deduced from the projected area [hypothesis of perfect sphericity (Ref. 13)]. The arrows in (c) indicate the size sequence expected if the in-flight size-selected clusters [$d_1^{*}=5.79\mathrm{nm}$ at the first peak labeled by a star (*)] were the primary building blocks of the supported particles. The measured peak structure is shifted to larger sizes (maxima marked by dotted lines: $d_2^{*}=7.6\mathrm{nm}$, $d_3^{*}=8.85\mathrm{nm}$, $d_4^{*}=9.55\mathrm{nm}$, $d_5^{*}=10.2\mathrm{nm}$, $d_6^{*}=10.75\mathrm{nm}$, etc.) because of cluster flattening on the surface.

Figure 3

Plot of the cluster volumes $V_n^{*}$ [hypothesis of perfect sphericity (Ref. 13)] corresponding to the peak maxima in the histogram of Fig. 2 as a function of the number of incident clusters they are expected to be made of. The dotted line represents the function given in the inset and passes through the origin ($V_1^{*}=V_1$ where $V_1$ is the volume of the incident particles). The drawing on the right shows the geometry of a spherical particle that wets the substrate with a contact angle $\theta$ ($r^{*}$ is the radius of the truncated sphere).

Figure 4

TEM micrograph of size-selected indium clusters $(d_1=5.79\mathrm{nm})$ deposited on graphite (a). The histogram (b) gives the distribution of the cluster islands as a function of their equivalent diameter. The curves in (c) represent the normalized volume distribution of these islands in the spherical approximation $(\text{line}+\text{dots})$ and the predictions of the DLA model assuming pure juxtaposition (no coalescence of primary clusters) and the absence of evaporation (dotted line, only monomers are mobile; thin full line, the dimers are also mobile) (Ref. 5).

Figure 5

TEM micrographs of size-selected platinum clusters deposited on amorphous carbon for increasing coverage rates. Noncoalesced primary particles are $2.3\mathrm{nm}$ in diameter (leftmost image). The degree of particle attachment and coalescence is weak for any coverage rate. An enlarged view (inset in the middle micrograph) shows how particles tend to assemble in close-packed arrays without contact between them. The cluster size distribution of noncoalesced clusters is given in the histogram on the right.

Figure 6

TEM micrographs of size-selected platinum clusters (mean diameter $2.3\mathrm{nm}$) deposited on HOPG with a high coverage rate: (a) large scale view; (b) zoom of the island marked in (a); (c) zoom of a part of island (b) where nontouching primary particles can be identified.

Figure 7

TEM micrographs of sequential deposits of size-selected platinum (mean diameter $2.3\mathrm{nm}$) and indium (mean diameter $5.8\mathrm{nm}$) clusters on HOPG: (a) deposition of indium clusters followed by platinum clusters; (b) deposition of platinum clusters followed by indium clusters. Enlarged parts of each micrograph are given at the bottom.