Managing Expectations and Imbalanced Training Data in Reactive Force Field Development: An Application to Water Adsorption on Alumina

ReaxFF is a computationally efficient model for reactive molecular dynamics simulations that has been applied to a wide variety of chemical systems. When ReaxFF parameters are not yet available for a chemistry of interest, they must be (re)optimized, for which one defines a set of training data that the new ReaxFF parameters should reproduce. ReaxFF training sets typically contain diverse properties with different units, some of which are more abundant (by orders of magnitude) than others. To find the best parameters, one conventionally minimizes a weighted sum of squared errors over all of the data in the training set. One of the challenges in such numerical optimizations is to assign weights so that the optimized parameters represent a good compromise among all the requirements defined in the training set. This work introduces a new loss function, called Balanced Loss, and a workflow that replaces weight assignment with a more manageable procedure. The training data are divided into categories with corresponding "tolerances", i.e., acceptable root-mean-square errors for the categories, which define the expectations for the optimized ReaxFF parameters. Through the Log-Sum-Exp form of Balanced Loss, the parameter optimization is also a validation of one's expectations, providing meaningful feedback that can be used to reconfigure the tolerances if needed. The new methodology is demonstrated with a nontrivial parametrization of ReaxFF for water adsorption on alumina. This results in a new force field that reproduces both the rare and frequent properties of a validation set not used for training. We also demonstrate the robustness of the new force field with a molecular dynamics simulation of water desorption from a γ-Al2O3 slab model.

Chemical and Structural Characterization of Amorphous and Crystalline Alumina Obtained by Alternative Sol-Gel Preparation Routes

Aluminum oxide is one of the most commonly used materials in the industry. It is used in the field of catalysis, refractories, and optics. Despite the fact that there are many techniques available, there is still a great challenge in obtaining a material with desired and designed properties. Nevertheless, there is a great flexibility in making customized alumina materials with desired physicochemical properties synthesized by sol–gel methods. This work consists in characterizing the physicochemical properties of sol–gel synthesized aluminum oxide using different sol–gel preparation routes. Three different sols were obtained by using organic precursors and underwent thermal treatment. The structure (Middle Infrared Spectroscopy, Diffused Reflectance Infrared Spectroscopy, X-ray Diffraction, Magic Angle Spinning Nuclear Magnetic Resonance) and microstructure (Scanning Electron Microscopy with Electron Dispersive Spectroscopy) tests of the materials were carried out. The specific surface area was determined by using the Brunauer–Emmett–Teller (BET) method. Thermal analysis was performed for all the powders, in order to analyze the specific temperature of materials transformation.

Molecular dynamics simulations of the mechanical behavior of alumina coated aluminum nanowires under tension and compression

For materials with high oxygen affinity, oxide layers will significantly change the material properties. This is of particular importance for aluminum nanowires which have many applications because of their ultrahigh strengths. Recent studies show that thin amorphous oxide shell layers on aluminum surfaces significantly change the responses of the material. However, the relations between the thickness of the oxidized layer, the strain rate and the mechanical response of nanowires to compression and tension have not been investigated intensively. In this study, we use a ReaxFF potential to analyze the influences of oxide shell layers on the material responses of the nanowires under uniaxial tension and compression at different strain rates. The Al–O interface leads to an increased defect nucleation rate at the oxide interface preventing localized deformation. During tension, we observe a reorganization of the structure of the oxide layer leading to bond healing and preventing fracture. While ductility is increasing with coating thickness during tension, the thickness of the coating is less decisive during compression.

Alumina coatings increase the ductility of aluminum nanowires by reorganization of the Al–O layer and stabilization of bonds.