Nanoparticle networks on silicon: Self-organized or disorganized?

A structural motif that appears very frequently not only in a wide range of nanostructured systems but also on mesoscopic to macroscopic length scales is the "cellular network". We present a quantitative analysis of the morphology of cellular networks formed by thiol-passivated Au nanoparticles, and, for comparison, organometallic molecules, spin cast onto native oxide-terminated silicon substrates. The structural parameters determined from Voronoi tessellation and Minkowski functional analyses of the experimental data are compared to those extracted from Monte Carlo simulations of nanoparticle network formation. The key result of this comparative study is that although the cell positions are spatially correlated, i.e., they deviate strongly from those expected for a Poisson point set, this correlation arises simply from a coalescence of neighboring cells during network formation. Complex nonlinear processes such as spinodal decomposition or Marangoni convection are therefore not always a prerequisite for the formation of spatially correlated networks.