arXiv subjects

Gun Park

Publications and source records attributed to Gun Park.

3 recordsLinked to original sources

arXiv2022

Correlating Nanoscale Structure with Electrochemical Property of Solid Electrolyte Interphases in Solid-State Battery Electrodes

Here, we correlate the nanoscale morphology and chemical composition of solid electrolyte interphases (SEI) with the electrochemical property of graphite-based composite electrodes. Using electrochemical strain microscopy (ESM) and X-ray photoelectron spectroscopy (XPS), changes of chemical composition and morphology (Li and F distribution) in SEI layers on the electrodes as a function of solid electrolyte contents are analyzed. As a result, we find a strong correlation between morphological variations on the electrode, Li and F distribution in SEI layer, and Coulomb efficiency. This correlation determines the optimum composition of the composite electrode surface that can maximize the physical and chemical uniformity of the solid electrolyte on the electrode, which is a key parameter to increase electrochemical performance in solid-state batteries.

cond-mat.mtrl-sci↗

arXiv2021

Task-Adaptive Neural Network Search with Meta-Contrastive Learning

Most conventional Neural Architecture Search (NAS) approaches are limited in that they only generate architectures without searching for the optimal parameters. While some NAS methods handle this issue by utilizing a supernet trained on a large-scale dataset such as ImageNet, they may be suboptimal if the target tasks are highly dissimilar from the dataset the supernet is trained on. To address such limitations, we introduce a novel problem of \emph{Neural Network Search} (NNS), whose goal is to search for the optimal pretrained network for a novel dataset and constraints (e.g. number of parameters), from a model zoo. Then, we propose a novel framework to tackle the problem, namely \emph{Task-Adaptive Neural Network Search} (TANS). Given a model-zoo that consists of network pretrained on diverse datasets, we use a novel amortized meta-learning framework to learn a cross-modal latent space with contrastive loss, to maximize the similarity between a dataset and a high-performing network on it, and minimize the similarity between irrelevant dataset-network pairs. We validate the effectiveness and efficiency of our method on ten real-world datasets, against existing NAS/AutoML baselines. The results show that our method instantly retrieves networks that outperform models obtained with the baselines with significantly fewer training steps to reach the target performance, thus minimizing the total cost of obtaining a task-optimal network. Our code and the model-zoo are available at https://github.com/wyjeong/TANS.

cs.LG↗

arXiv2019

Nanoscale characterization of the impact of beverages on the enamel surface of human teeth

Here we quantitatively evaluate the early stages of mechanical and morphological changes of polished human enamel surfaces induced by soft drinks using atomic force microscopy. With an increase of the immersion time in soft drinks, we found a significant increase of surface roughness (Ra) and a considerable decrease of elastic modulus (E) of the enamel. The prismatic structure of enamel was clearly observed after a one-hour immersion in Coca-Cola, which shows its strong erosion effect. A high surface roughness of enamel results in a high chance of cavities due to easier bacterial adhesion on rougher surface, while a drastic deterioration of the mechanical properties of the enamel weakens its protection property. Our findings show the variation of enamel surface at the very beginning stage of etching process by acidic drinks, which can also be applicable to the etching mechanism of enamel surface by other sources.

cond-mat.mtrl-sci↗