In molecular dynamics (MD) simulations, interactions between water molecules and graphitic surfaces are often modeled as a simple Lennard-Jones potential between oxygen and carbon atoms. A possible method for tuning this parameter consists of simulating a water nanodroplet on a flat graphitic surface, measuring the equilibrium contact angle, extrapolating it to the limit of a macroscopic droplet, and finally matching this quantity to experimental results. Considering recent evidence demonstrating that the contact angle of water on a graphitic plane is much higher than what was previously reported, we estimate the oxygen-carbon interaction for the recent SPC/Fw water model. Results indicate a value of about 0.2 kJ/mol, much lower than previous estimations. We then perform simulations of cylindrical water filaments on graphitic surfaces, in order to compare and correlate contact angles resulting from these two different systems. Results suggest that a modified Young's equation does not describe the relation between contact angle and drop size in the case of extremely small systems and that contributions different from the one deriving from contact line tension should be taken into account.

No abstract

Polyhedral open cell lattice catalyst substrates are proposed based on results of numerical simulations and recent advances in Additive Manufacturing (AM) techniques. Detailed simulations have compared different polyhedral structures in terms of mass transfer (aiming at optimal reactivity in the mass transfer limited domain) and flow through resistance. The simulations have taken into account dimensional limits given by the possibilities of AM techniques. Comparisons with state of art honeycombs have been also used in order to identify the optimal shape. Substrates with these optimal characteristics have been manufactured out of Al2O3 with Stereolithography. Subsequently, these substrates have been coated and used for measurements of C3H6 oxidation in a model gas reactor. Measurements have focused in determining oxidation efficiency at different gas hourly space velocities as well as light-off behaviour. Simulation results show that the optimal open cell structures are comprised by a cubic elementary cell rotated by 45° so that one spatial diagonal of the cube is aligned to the main gas flow. Higher porosities and smaller strut diameters improve the reactivity to pressure drop trade off. However, given the current manufacturing limitations, it is not possible to produce structures with strut diameters lower than 0.5 mm. This results in high porosity but low specific surface area (i.e ε=0.95 and Sv=400m 2 /m 3 ). Thus, reaching a target conversion requires higher overall catalyst volume. The simulations show that for a series of geometrical parameters the open cell structures can reach identical conversion in respect to the honeycombs with only a fraction of the overall surface area and thus a fraction of the noble metals, while the overall dimensions are in the same order of magnitude and the pressure drop can reach lower levels. Measurements in the model gas reactor confirm the mass transfer advantages of the polyhedral structures as predicted by the simulations. Measurements also show that the polyhedral lattices have very similar light-off behaviour in spite the four times lower surface area. NomenclatureA: Cross section of catalyst AM:Additive Manufacturing CPSI:Cells Per Square Inch, commercial characterization of honeycomb catalyst substrates Cubic:Additive Manufactured (AM) catalyst substrate consisting of cubes as elementary cells aligned with the main flow Cubic45:AM catalyst substrate consisting of cubes as elementary cells rotated by 45° so that one spatial diagonal of the cube is aligned to the main gas flow Dij:Diffusivity of specie i in a gas j dc:Wetted width of a (square) honeycomb channel Dc:Inner width of a (square) honeycomb channel, difference to dc is the coating thickness ds:Strut diameter ghsv:Gas hourly space velocity through the catalyst, it corresponds to the ratio between the gas volume flow rate and the catalyst volume HC:Honeycomb catalyst substrate (conventional) K:Mass transfer coefficient Kelvin:AM catalyst substrate consisting of Kelvin cells (tetrakaidekahedral) as elementary cells

Wetting is a widespread phenomenon, most prominent in a number of cases, both in nature and technology. Droplets of pure water with initial radius ranging from 20 to 80 [Å] spreading on graphitic surfaces are studied by molecular dynamics simulations. The equilibrium contact angle is determined and the transition to the macroscopic limit is discussed using Young equation in its modified form. While the largest droplets are almost perfectly spherical, the profiles of the smallest ones are no more properly described by a circle. For the sake of accuracy, we employ a more general fitting procedure based on local averages. Furthermore, our results reveal that there is a possible transition to the macroscopic limit. The modified Young equation is particularly precise for characteristic lengths (radii and contact-line curvatures) around 40 [Å].

We report on a molecular dynamics investigation of the wetting properties of graphitic surfaces by various solutions at concentrations 1 − 8 wt% of commercially available non-ionic surfactants with long hydrophilic chains, linear or T-shaped. These are surfactants of length up to 160 [Å]. It turns out that molecular dynamics simulations of such systems ask for a number of solvent particles that can be reached without seriously compromising computational efficiency only by employing a coarse-grained model. The MARTINI force field with polarizable water offers a framework particularly suited for the parameterization of our systems. In general, its advantages over other coarse-grained models are the possibility to explore faster long time scales and the wider range of applicability. Although the accuracy is sometimes put under question, the results for the wetting properties by pure water are in good agreement with those for the corresponding atomistic systems and theoretical predictions. On the other hand, the bulk properties of various aqueous surfactant solutions indicate that the micellar formation process is too strong. For this reason, a typical experimental configuration is better approached by preparing the droplets with the surfactants arranged in the initial state in the vicinity of contact line. Cross-comparisons are possible and illuminating, but equilibrium contanct angles as obtained from simulations overestimate the experimental results. Nevertheless, our findings can provide guidelines for the preliminary assessment and screening of surfactants. Most importantly, it is found that the wetting properties mainly depend on the length and apolarity of the hydrophobic tail, for linear surfactants, and the length of the hydrophilic headgroup for T-shaped surfactants. Moreover, the T-shaped topology appears to favor the adsorption of surfactants onto the graphitic surface and faster spreading.