Relation between optical non-contact profilometry and AFM roughness parameters on coated papers with oil-filled nanoparticles

In parallel with the development of nanoparticle coatings for protection of paper substrates, detailed descriptions of the surface topography with micro-to nanoscale roughness features are needed. In this work, papers have been coated with poly(styrene-co-maleic anhydride) nanoparticles including different types of vegetable oils and the surface roughness was evaluated at 2000 x 2000 mu m(2) to 0.2 x 0.2 mu m(2) length scales by combining non-contact optical profilometry (NCP) and atomic force microscopy (AFM). The relationships between roughness data were studied for statistical roughness parameters, spatial roughness parameters and in the frequency domain. In order to compare AFM roughness more accurately, the original images were flattened to remove effects of the underlaying fibrous substrate and highlight features of the nanoparticle coating. More detailed information on the coating topography could be obtained by considering bearing ratio curves and histograms, where it was concluded that the oil-filled coatings form a rather thin and continuous coating that closely follows the shape of the cellulose fibers. The relation between statistical roughness parameters from NCP and AFM follows an exponential trend with relatively low coefficient of determination. The increase in surface roughness with length scale showed a transition point attributed to short-and long-range surface features. Therefore, the correlation length was used as a spatial roughness parameter that provides a successful extrapolation of the average roughness over different length scales in a double logarithmic diagram with very high coefficient of determination. Based on the power spectral density, it was difficult to exactly distinguish between the different types of SMI/oil coatings, as they include similar nanoscale features. The frequency roughness parameters were better suited for extrapolation than statistical roughness parameters but little less efficient than the spatial roughness parameters. (C) 2015 Elsevier Ltd. All rights reserved.