The influence of surface roughness on surfactant adsorption was studied using a quartz crystal microbalance with dissipation (QCM-D). The sensors employed had root-mean-square (R) roughness values of 2.3, 3.1, and 5.8 nm, corresponding to fractal-calculated surface area ratios (actual/nominal) of 1.13, 1.73, and 2.53, respectively. Adsorption isotherms measured at 25 °C showed that adsorbed mass of cetyltrimethylammonium bromide per unit of actual surface area below 0.8 cmc, or above 1.2 cmc, decreases as the surface roughness increases. At the cmc, both the measured adsorbed amount and the measured dissipation increased dramatically on the rougher surfaces. These results are consistent with the presence of impurities, suggesting that roughness exacerbates well-known phenomena reported in the literature of peak impurity-related adsorption at the cmc. The magnitude of the increase, especially in dissipation, suggests that changes in adsorbed amount may not be the only reason for the observed results, as aggregates at the cmc on rougher surfaces are more flexible and likely contain larger amounts of solvent. Differences in adsorption kinetics were also found as a function of surface roughness, with data showing a second, slower adsorption rate after rapid initial adsorption. A two-rate Langmuir model was used to further examine this effect. Although adsorption completes faster on the smoother surfaces, initial adsorption at zero surface coverage is faster on the rougher surfaces, suggesting the presence of more high-energy sites on the rougher surfaces.

Negatively charged colloidal poly(methyl methacrylate-co-butyl acrylate) (P(MMA-BA)) particles and positively charged dissolved poly(ethyleneimine) (PEI) were adsorbed onto a cement block using a layer-by-layer (LBL) assembly technique. The block was fashioned so as to have a cylindrical hole running from one face to another along the long axis of the rectangular block, and a fluid containing either of the two charged materials was pumped through the block. The result was a film tens of micrometers thick, and the pressure required to crack the cement block was measured after one end of the hole was sealed. Latex particles with a T(g) near the use temperature showed the maximum improvement in the cracking stress of the blocks. In a multilayer coating with identically sized particles, the cracking stress of the blocks increased to an improvement of 25% and then dropped off with increasing number of layers, even though the relationship between film thickness and the number of layers was linear. An improvement of about 30% in the cracking stress of the coated blocks was obtained when using multiple layers with different particle sizes. The effects of the number of layers and particle size on the cracking stress suggest that both the morphology and the thickness of the film play a role in performance. Tests done under confinement, e.g., with an external stress applied to the outside of the blocks, suggest that not only does a film-forming mechanism contribute to performance but that filling of microcracks in the rock may also play a role.