Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userLearning intermolecular forces at liquid-vapor interfaces
103
0
0.0
(
0
)
Added by
David T. Limmer PhD
Publication date
2021
fields
Physics
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
By adopting a perspective informed by contemporary liquid state theory, we consider how to train an artificial neural network potential to describe inhomogeneous, disordered systems. We find that neural network potentials based on local representations of atomic environments are capable of describing some properties of liquid-vapor interfaces, but typically fail for properties that depend on unbalanced long-ranged interactions which build up in the presence of broken translation symmetry. These same interactions cancel in the translationally invariant bulk, allowing local neural network potentials to describe bulk properties correctly. By incorporating explicit models of the slowly-varying long-ranged interactions and training neural networks only on the short ranged components, we can arrive at potentials that robustly recover interfacial properties. We find that local neural network models can sometimes approximate a local molecular field potential to correct for the truncated interactions, but this behavior is variable and hard to learn. Generally, we find that models with explicit electrostatics are easier to train and have higher accuracy. We demonstrate this perspective in a simple model of an asymmetric dipolar fluid where the exact long-ranged interaction is known, and in an ab initio water model where it is approximated.
rate research
Read More
The excited state dynamics of chromophores in complex environments determine a range of vital biological and energy capture processes. Time-resolved, multidimensional optical spectroscopies provide a key tool to investigate these processes. Although theory has the potential to decode these spectra in terms of the electronic and atomistic dynamics, the need for large numbers of excited state electronic structure calculations severely limits first principles predictions of multidimensional optical spectra for chromophores in the condensed phase. Here, we leverage the locality of chromophore excitations to develop machine learning models to predict the excited state energy gap of chromophores in complex environments for efficiently constructing linear and multidimensional optical spectra. By analyzing the performance of these models, which span a hierarchy of physical approximations, across a range of chromophore-environment interaction strengths, we provide strategies for the construction of ML models that greatly accelerate the calculation of multidimensional optical spectra from first principles.
We perform molecular dynamics simulations to understand the translational and rotational diffusion of Janus nanoparticles at the interface between two immiscible fluids. Considering spherical particles with different affinity to fluid phases, both their dynamics as well as the fluid structure around them are evaluated as a function of particle size, amphiphilicity, fluid density, and interfacial tension. We show that as the particle amphiphilicity increases due to enhanced wetting of each side with its favorite fluid, the rotational thermal motion decreases. Moreover, the in-plane diffusion of nanoparticles at the interface becomes slower for more amphiphilic particles, mainly due to formation of a denser adsorption layer. The particles induce an ordered structure in the surrounding fluid that becomes more pronounced for highly amphiphilic nanoparticles, leading to increased resistance against nanoparticle motion. A similar phenomenon is observed for homogeneous particles diffusing in bulk upon increasing their wettability. Our findings can provide fundamental insight into the dynamics of drugs and protein molecules with anisotropic surface properties at biological interfaces including cell membranes.
Many atomic liquids can form transient covalent bonds reminiscent of those in the corresponding solid states. These directional interactions dictate many important properties of the liquid state, necessitating a quantitative, atomic-scale understanding of bonding in these complex systems. A prototypical example is liquid silicon, wherein transient covalent bonds give rise to local tetrahedral order and consequent non-trivial effects on liquid state thermodynamics and dynamics. To further understand covalent bonding in liquid silicon, and similar liquids, we present an ab initio simulation-based approach for quantifying the structure and dynamics of covalent bonds in condensed phases. Through the examination of structural correlations among silicon nuclei and maximally localized Wannier function centers, we develop a geometric criterion for covalent bonds in liquid Si. We use this to monitor the dynamics of transient covalent bonding in the liquid state and estimate a covalent bond lifetime. We compare covalent bond dynamics to other processes in liquid Si and similar liquids and suggest experiments to measure the covalent bond lifetime.
Machine learning promises to deliver powerful new approaches to neutron scattering from magnetic materials. Large scale simulations provide the means to realise this with approaches including spin-wave, Landau Lifshitz, and Monte Carlo methods. These approaches are shown to be effective at simulating magnetic structures and dynamics in a wide range of materials. Using large numbers of simulations the effectiveness of machine learning approaches are assessed. Principal component analysis and nonlinear autoencoders are considered with the latter found to provide a high degree of compression and to be highly suited to neutron scattering problems. Agglomerative heirarchical clustering in the latent space is shown to be effective at extracting phase diagrams of behavior and features in an automated way that aid understanding and interpretation. The autoencoders are also well suited to optimizing model parameters and were found to be highly advantageous over conventional fitting approaches including being tolerant of artifacts in untreated data. The potential of machine learning to automate complex data analysis tasks including the inversion of neutron scattering data into models and the processing of large volumes of multidimensional data is assessed. Directions for future developments are considered and machine learning argued to have high potential for impact on neutron science generally.
Liquid phase exfoliation is a commonly used method to produce 2D nanosheets from a range of layered crystals. However, such nanosheets display broad size and thickness distributions and correlations between area and thickness, issues that limit nanosheet application potential. To understand the factors controlling the exfoliation process, we have liquid-exfoliated 11 different layered materials, size-selecting each into fractions before using AFM to measure the nanosheet length, width, and thickness distributions for each fraction. The resultant data show a clear power-law scaling of nanosheet area with thickness for each material. We have developed a simple nonequilibrium thermodynamics-based model predicting that the power-law prefactor is proportional to both the ratios of in-plane-tearing/out-of-plane-peeling energies and in-plane/out-of-plane moduli. By comparing the experimental data with the modulus ratio calculated from first-principles, we find close agreement between experiment and theory. This supports our hypothesis that energy equipartition holds between nanosheet tearing and peeling during sonication-assisted exfoliation.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
