We performed the resonance shear measurement (RSM) for evaluating the properties of water confined between silica surfaces with and without water vapor plasma treatment, which was used to increase the density of the silanol groups on the surfaces. We compared the properties of the confined water, such as viscosity and lubricity, by controlling the surface separation at a 0.1 nm resolution. The observed resonance curves for water between the plasma-treated and untreated silica surfaces showed the following results: (1) The viscosity of the water confined between the plasma-treated silica surfaces increased due to water structuring at separations less than 3 nm, while the value for the water between the untreated silica surfaces was 8 nm. (2) The water confined between the plasma-treated surfaces could maintain lubricity under the normal pressure of more than 1.7 MPa; however, the water confined between the untreated surfaces lost lubricity under the normal pressure of more than 0.4 MPa. To discuss these properties in terms of water structures on the silica surfaces, we performed sum frequency generation (SFG) vibrational spectroscopy for water on the plasma-treated and untreated silica surfaces. The main peak of SFG spectra for the water on the plasma-treated silica was around 3200 cm −1 , and that for water on the untreated silica was around 3400 cm −1 , indicating that the hydrogen bonding network of the water on the plasma-treated silica surface was stronger than that on the untreated one due to the higher silanol density. The strongly networked water could exhibit higher lubricity with the increased silanol density.

Material informatics (MI) is a promising approach to liberate us from the time-consuming Edisonian (trial and error) process for material discoveries, driven by machine-learning algorithms. Several descriptors, which are encoded material features to feed computers, were proposed in the last few decades. Especially to solid systems, however, their insufficient representations of three dimensionality of field quantities such as electron distributions and local potentials have critically hindered broad and practical successes of the solid-state MI. We develop a simple, generic 3D voxel descriptor that compacts any field quantities, in such a suitable way to implement convolutional neural networks (CNNs). We examine the 3D voxel descriptor encoded from the electron distribution by a regression test with 680 oxides data. The present scheme outperforms other existing descriptors in the prediction of Hartree energies that are significantly relevant to the long-wavelength distribution of the valence electrons. The results indicate that this scheme can forecast any functionals of field quantities just by learning sufficient amount of data, if there is an explicit correlation between the target properties and field quantities. This 3D descriptor opens a way to import prominent CNNs-based algorithms of supervised, semi-supervised and reinforcement learnings into the solid-state MI.

Functional materials, especially those that largely differ from known materials, are not easily discoverable because both human experts and supervised machine learning need prior knowledge and datasets. An autonomous system can evaluate various properties a priori, and thereby explore unknown extrapolation spaces in high-throughput simulations. However, high-throughput evaluations of molecular dynamics simulations are unrealistically demanding. Here, we show an autonomous search system for organic molecules implemented by a reinforcement learning algorithm, and apply it to molecular dynamics simulations of viscosity. The evaluation is dramatically accelerated (by three orders of magnitude) using a femto-second stress-tensor correlation, which underlies the glass-transition model. We experimentally examine one of 55,000 lubricant oil molecules found by the system. This study indicates that merging simulations and physical models can open a path for simulation-driven approaches to materials informatics.

Different hypotheses have been proposed to explain the mechanism for the extremely low friction coefficient of carbon coatings and its undesired dependence on air humidity. A decisive atomistic insight is still lacking because of the difficulties in monitoring what actually happens at the buried sliding interface. Here we perform large-scale ab initio molecular dynamics simulations of both undoped and silicon-doped carbon films sliding in the presence of water. We observe the tribologically-induced surface hydroxylation and subsequent formation of a thin film of water molecules bound to the OH-terminated surface by hydrogen bonds. The comparative analysis of silicon-incorporating and clean surfaces, suggests that this two-step process can be the key phenomenon to provide high slipperiness to the carbon coatings. The water layer is, in fact, expected to shelter the carbon surface from direct solid-on-solid contact and make any counter surface slide extremely easily on it. The present insight into the wettability of carbon-based films can be useful for designing new coatings for biomedical and energy-saving applications with environmental adaptability