We describe the formation process of thin films consisting of surface-attached polymer networks formed through C, H-insertion based cross-linking (CHic process). To this thin films of polymers containing benzophenone or sulfonyl azide groups are photochemically or thermally activated, which leads to simultaneous cross-linking and surface attachment of the deposited polymer. A simple percolation model is used to describe the formation of such polymer networks as a function of reaction time and reaction conditions. It is based on known parameters such as the kinetics of the activation of the cross-linker, cross-linker contents, molecular weight of the polymer, film thickness, temperature (respectively the light dose), and absorption coefficient of the polymer. The developed model allows, with only one adaptable parameter, namely the efficiency factor j, the prediction of the complete cross-linking behavior of CHic reactions. The kinetic model enables the identification of suitable conditions for network formation and thus facilitates to use this very simple and versatile method to generate tailor-made surfaces.

Deposition of fullerenes on the CaF(2)(111) surface yields peculiar island morphologies with close similarities to previous findings for (100) surfaces of other ionic crystals. By means of noncontact atomic force microscopy we find a smooth transition from compact, triangular islands to branched hexagonal islands upon lowering the temperature. While triangular islands are two monolayers high, hexagonal islands have a base of one monolayer and exhibit a complicated structure with a second-layer outer rim and trenches oriented towards the interior. By developing a kinetic growth model we unravel the microscopic mechanisms of the structure formation.

The capture numbers entering the rate equations (RE) for submonolayer film growth are determined from extensive kinetic Monte Carlo (KMC) simulations for simple representative growth models yielding point, compact, and fractal island morphologies. The full dependence of the capture numbers σs(Θ, Γ) on island size s, and on both the coverage Θ and the Γ = D/F ratio between the adatom diffusion coefficient D and deposition rate F is determined. Based on this information, the RE are solved to give the RE island size distribution (RE-ISD), as quantified by the number ns(Θ, Γ) of islands of size s per unit area. The RE-ISDs are shown to agree well with the corresponding KMC-ISDs for all island morphologies. For compact morphologies, however, this agreement is only present for coverages smaller than Θ ≃ 5% due to a significantly increased coalescence rate compared to fractal morphologies. As found earlier, the scaled KMC-ISDs nss 2 /Θ as a function of scaled island size x = s/s approach, for fixed Θ, a limiting curve f∞(x, Θ) for Γ → ∞. Our findings provide evidence that the limiting curve is independent of Θ for point islands, while the results for compact and fractal island morphologies indicate a dependence on Θ.

We show that mean-field rate equations for submonolayer growth can successfully predict island size distributions in the pre-coalescence regime if the full dependence of capture numbers on both the island size and the coverage is taken into account. This is demonstrated by extensive Kinetic Monte Carlo simulations for a growth kinetics with hit and stick aggregation. A detailed analysis of the capture numbers reveals a nonlinear dependence on the island size for small islands. This nonlinearity turns out to be crucial for the successful prediction of the island size distribution and renders an analytical treatment based on a continuum limit of the mean-field rate equations difficult.

We study the nonlinear hopping transport in one-dimensional rings and open channels. Analytical results are derived for the stationary current response to a constant bias without assuming any specific coupling of the rates to the external fields. It is shown that anomalous large effective jump lengths, as observed in recent experiments by taking the ratio of the third-order nonlinear and the linear conductivity, can occur already in ordered systems. Rectification effects due to site energy disorder in ring systems are expected to become irrelevant for large system sizes. In open channels, in contrast, rectification effects occur already for disorder in the jump barriers and do not vanish in the thermodynamic limit. Numerical solutions for a sinusoidal bias show that the ring system provides a good description for the transport behavior in the open channel for intermediate and high frequencies. For low frequencies temporal variations in the mean particle number have to be taken into account in the open channel, which cannot be captured in the more simple ring model.