Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userSimulating diffusion properties of solid-state electrolytes via a neural network potential: Performance and training scheme
87
0
0.0
(
0
)
Added by
Aris Marcolongo Dr.
Publication date
2019
fields
Physics
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
The recently published DeePMD model (https://github.com/deepmodeling/deepmd-kit), based on a deep neural network architecture, brings the hope of solving the time-scale issue which often prevents the application of first principle molecular dynamics to physical systems. With this contribution we assess the performance of the DeePMD potential on a real-life application and model diffusion of ions in solid-state electrolytes. We consider as test cases the well known Li10GeP2S12, Li7La3Zr2O12 and Na3Zr2Si2PO12. We develop and test a training protocol suitable for the computation of diffusion coefficients, which is one of the key properties to be optimized for battery applications, and we find good agreement with previous computations. Our results show that the DeePMD model may be a successful component of a framework to identify novel solid-state electrolytes.
rate research
Read More
One hidden yet important issue for developing neural network potentials (NNPs) is the choice of training algorithm. Here we compare the performance of two popular training algorithms, the adaptive moment estimation algorithm (Adam) and the extended Kalman filter algorithm (EKF), using the Behler-Parrinello neural network (BPNN) and two publicly accessible datasets of liquid water. It is found that NNPs trained with EKF are more transferable and less sensitive to the value of the learning rate, as compared to Adam. In both cases, error metrics of the test set do not always serve as a good indicator for the actual performance of NNPs. Instead, we show that their performance correlates well with a Fisher information based similarity measure.
Prediction of material properties from first principles is often a computationally expensive task. Recently, artificial neural networks and other machine learning approaches have been successfully employed to obtain accurate models at a low computational cost by leveraging existing example data. Here, we present a software package Properties from Artificial Neural Network Architectures (PANNA) that provides a comprehensive toolkit for creating neural network models for atomistic systems. Besides the core routines for neural network training, it includes data parser, descriptor builder and force-field generator suitable for integration within molecular dynamics packages. PANNA offers a variety of activation and cost functions, regularization methods, as well as the possibility of using fully-connected networks with custom size for each atomic species. PANNA benefits from the optimization and hardware-flexibility of the underlying TensorFlow engine which allows it to be used on multiple CPU/GPU/TPU systems, making it possible to develop and optimize neural network models based on large datasets.
Finding new ionic conductors that enable significant advancements in the development of energy-storage devices is a challenging goal of current material science. Aside of material classes as ionic liquids or amorphous ion conductors, the so-called plastic crystals (PCs) have been shown to be good candidates combining high conductivity and favourable mechanical properties. PCs are formed by molecules whose orientational degrees of freedom still fluctuate despite the material exhibits a well-defined crystalline lattice. Here we show that the conductivity of Li+ ions in succinonitrile, the most prominent molecular PC electrolyte, can be enhanced by several decades when replacing part of the molecules in the crystalline lattice by larger ones. Dielectric spectroscopy reveals that this is accompanied by a stronger coupling of ionic and reorientational motions. These findings, which can be understood in terms of an optimised revolving door mechanism, open a new path towards the development of better solid-state electrolytes.
For machine learning of interatomic potentials a scalable sparse Gaussian process regression formalism is introduced with a data-efficient on-the-fly adaptive sampling algorithm. With this approach, the computational cost is effectively reduced to those of the Bayesian linear regression methods whilst maintaining the appealing characteristics of the exact Gaussian process regression. As a showcase, experimental melting and glass-crystallization temperatures are reproduced for Li7P3S11, Li diffusivity is simulated, and an unchartered phase is revealed with much lower Li diffusivity which should be circumvented.
Solid-state lithium-ion batteries (SSLIBs) are considered to be the new generation of devices for energy storage due to better performance and safety. Poly (ethylene oxide) (PEO) based material becomes one of the best candidate of solid electrolytes, while its thermal conductivity is crucial to heat dissipation inside batteries. In this work, we study the thermal conductivity of PEO by molecular dynamics simulation. By enhancing the structure order, thermal conductivity of aligned crystalline PEO is obtained as high as 60 W/m-K at room temperature, which is two orders higher than the value (0.37 W/m-K) of amorphous structure. Interestingly, thermal conductivity of ordered structure shows a significant stepwise negative temperature dependence, which is attributed to the temperature-induced morphology change. Our study offers useful insights into the fundamental mechanisms that govern the thermal conductivity of PEO but not hinder the ionic transport, which can be used for the thermal management and further optimization of high-performance SSLIBs.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
