Structure of self-assembled layers on silicon: Combined use of spectroscopic variable angle ellipsometry, neutron reflection, and atomic force microscopy

Abstract

Identification of the atomic scale structures of the gold-thiol interfaces of molecular nanowires by inelastic tunneling spectroscopy J. Chem. Phys. 136, 014703 (2012) Simultaneous step meandering and bunching instabilities controlled by Ehrlich-Schwoebel barrier and elastic interaction Appl. Phys. Lett. 99, 263106 (2011) Density functional theory simulations of amorphous high-κ oxides on a compound semiconductor alloy: aAl2O3/InGaAs(100)-(4×2), a-HfO2/InGaAs(100)-(4×2), and a-ZrO2/InGaAs(100)-(4×2) J. Chem. Phys. 135, 244705 (2011) Additional information on J. Appl. Phys. Neutron reflection ͑NR͒, spectroscopic ellipsometry ͑SE͒, and atomic force microscopy ͑AFM͒ have been used to characterize the structure of self-assembled octadecyltrichlorosilane ͑OTS͒ layers on silicon. The first two of these techniques rely on modeling of the experimental data and may thus result in the unrealistic representation of the composition and structure at the interface. Ambiguities arise from model-dependent analysis complicated by the lack of sufficient external constraints to converge nonunique solutions to a unique one. We show in this work that AFM measurements provide extra constraints to allow us to obtain a physical description closer to the actual structure of the film. It was found that ''the simpler the better'' modeling strategy very often employed during the fitting of ellipsometric and neutron reflection data is, therefore, not necessarily the best way to obtain a reliable description of the interfacial structure. Our AFM findings necessitated the refit of both neutron and ellipsometric data that were previously described by a single-layer model. Interpretation of the structure of thin layers that is based only on indirect measurements such as SE, NR, and x-ray reflection techniques may be, therefore, misleading. A combined analysis of SE, NR, and AFM data suggests that the OTS film may comprise a rough layer, with pinholes down to bare silicon oxide surface, consisting at least of mono-, bi-and trilayers of OTS molecules.

Octadecyltrichlorosilane (OTS): a resist for OMCVD gold nanoparticle growth

“…Obviously, a uniform single layer is too simplistic a model and does not describe the surface structure well when tested against AFM results. [74] The roughness of the layers is of paramount importance to the final conclusions about the OTS layer structure. Therefore, the mean surface roughness (R a ) was measured for six randomly chosen 300 nm × 300 nm areas in each of the 1 µm × 1 µm AFM images and averaged (Table 2).…”

Comparative characterisation by atomic force microscopy and ellipsometry of soft and solid thin films

“…Fang et al 28 studied the relationship between the AFM roughness and the ellipsometric parameters for a chemically treated rough Si surface and showed that AFM roughness without the roughness spectral density information does not determine the ellipsometry parameters unambiguously. However, to our knowledge, all papers published to date deal with either soft organic films 3,20,27,29 or solid semiconductor/dielectric surfaces, 28,30,31 but not both together.…”

“…The features of an intermediate group of rough soft organic films may be reduced to the elements of the two main groups. 3 In the literature, there are at least two approaches that are used to interpret the experimental data obtained by ellipsometry for the characterisation of randomly rough surfaces: the effective medium approximation (EMA), which connects the microstructure of a heterogeneous thin film with its macroscopic dielectric response; 4,5 and the Rayleigh-Rice theory (RRT), which is based on the vector perturbation approach for the solution of Maxwell's equations on rough surfaces. 6 The RRT theory is more accurate, as, in contrast with EMA, it takes into account the light scattering from rough surfaces.…”

Applications of All-Atom Molecular Dynamics to Nanofluidics

“…Cottin-Bizonne et al (2008) and Joseph & Tabeling (2005) indicate the peak to peak distance between asperities in their OTS coatings in a few nanometers. Additionally, spectroscopic variable angle ellipsometry, neutron reflection and atomic force microscopy by Styrkas et al (1999) suggest that OTS films often consist of nearby multilayer domains of different thickness, ranging from 1 to 3 and even 4 times the monolayers thickness. Our conjecture is that apparent slip effects occurring on nanopatterned surfaces can, in principle, explain the difference observed between MD simulations and experiments.…”