56016 research outputs found
Sort by
Designing Two-Dimensional Based Materials for Boosting Electrochemical Nitrogen Reduction Reaction
Get PDFDepartment of ChemistryAmmonia (NH3) is a carbon-free hydrogen carrier as a promising renewable fuel and a pivotal chemical for fertilizers in agriculture. Ammonia is feeding the world???s population which is essential for the subsistence of life on Earth. The Haber???Bosch method is mainly used for industrial-mass production of ammonia under harsh condition at drastic temperature and pressure, accounting for 1-2% of the global energy supply. The extraordinarily inert stable N???N molecules dissociate on iron-based catalysts during which huge amounts of fossil fuels are consumed. In this process, the CO2 emission has a highly-negative impact on the environment by producing huge amounts of CO2 per year (3% of total carbon dioxide emissions). Electrochemical nitrogen reduction reaction (NRR) has been proposed to mitigate the aforementioned issues under mild condition by utilization of green and renewable electrical energy. In this regard, many strategies have been explored, but the Faradaic efficiencies (FEs) are unsatisfactory and over-potentials are still very high. This thesis aims to design and employ some 2-dimensional (2D) based materials for the enhancing catalytic activity of NRR by a combined approach of density fictional theory (DFT) and artificial intelligence.
In the first chapter, we introduce some basic NRR mechanisms, reaction pathways, methods for ammonia production and evaluating efficiency. In the second chapter, 2D monolayer structures of BP, BAs, and BSb were constructed by immobilized some transition metals (TMs) on support aimed at designing single-atom catalysts (SACs) with both high activity and high selectivity for NRR. W-BAs shows high catalytic activity and excellent selectivity with a barrier of only 0.05 eV (practically no significant barrier along the distal pathway in the implicit solvation model) and a surmountable kinetic barrier of 0.34 eV. In addition, W-BSb and Mo-BSb exhibit high performances with limiting potentials of -0.19 and -0.34 V, respectively. The Bader charge descriptor reveals that in the first protonation step the charge transfers from substrate to *NNH, while in the last pronation step it transfers from *NH3 to the substrate. In fact, the electron transfer determines the binding strength of *NNH species and promotes NRR. Our machine learning (ML) models unfold that the N-TM (nitrogen to transition metal) and N-N bond lengths are the most important local features, determining the catalytic activity of a catalyst.
We present some new strategies in the third chapter to boost N2 reduction to NH3, while inhibiting the hydrogen evolution reaction (HER) using novel 2D transition metal borides (MBenes), defect-engineered 2D-materials, and 2D ??-conjugated polymer (2DCP)-supported SACs. DFT calculations show that nitrogen molecules can be captured in the hollow sites of MBenes, with significant increases in the adsorption strength and N???N bond length. Also, defective 2D-materials formed by the vacancy sites of Te, Se and S expose N2 molecules to a specific environment adjacent to three transition metals, which drastically improves the catalytic activity and selectivity (by a dramatic increase in the N???N bond length up to 1.38 ??). We report a new mechanism for NRR as a combination of dissociative and associative mechanisms. A ML based fast-screening strategy to predict efficient NRR electrocatalysts is described. Overall, TaB, NbTe2, NbB, HfTe2, MoB, MnB, HfSe2, TaSe2 and Nb@SAC exhibit impressive selectivities over the HER with overpotentials of 0.44 V, 0.40 V, 0.24 V, 0.60 V, 0.17 V, 0.17 V, 0.64 V, 0.37 V and 0.58 V, respectively.
In the fourth chapter, we designed a deep neural network (DNN) to predict efficient electrocatalysts for NRR among boron(B)-doped graphene SACs. This model can noticeably reduce the time of computation by removing non-efficient catalysts from screening. Also, the adsorption energy and free energy can be predicted by the feature-based light gradient boosting machine (LGBM) model. These features represent the geometrical structure as well as bonding characteristics. Among the catalysts evaluated, three candidates show very promising activity, offering excellent selectivity over the HER. CrB3C1 exhibits a minimal overpotential of 0.13 V for NRR.ope
Design Strategies Considering Degradation Mechanism by Measuring Current Distribution: Cathode with Two Characteristic Particle Sizes
Get PDFSchool of Energy and Chemical Engineering (Energy Engineering(Battery Science and Engineering))The performance of a lithium-ion battery can be improved by blending two or more kinds of cathode active materials in one electrode. In particular, energy density can be increased by blending active materials having different particle sizes because the electrode can be filled with more active materials. As the performance of a battery is determined by the combination of the properties of the material and electrode design conditions, it is of utmost importance to understand each impact independently. However, since the potential in a blended electrode is measured as the mixed potential of each active material, it is not easy to identify how each active material behaves electrochemically under the current load condition.
In this experiment, the effects of active materials are separated while minimizing the influence of the mass transport in the porous electrode by adjusting the electrode density. For this purpose, two electrodes consisting of Large-sized LiNi0.8Co0.1Mn0.1O2 (LNCM) and small-sized LiNi0.8Co0.1Mn0.1O2 (SNCM) are placed in one pouch cell and connected in parallel with Li metal as a counter electrode. By measuring the current flowing to each electrode individually, it is observed how LNCM and SNCM works in the blended electrode.
Herein, it is revealed that charging/discharging behavior of LNCM and SNCM is different by comparing the rate of solid diffusion and charge transfer using Potentiostatic Intermittent Titration Technique. It resulted in a difference in degradation mechanism which is actually implemented in the blended electrode. Through this study, it is proposed that to increase the energy density of blended electrode at early stage of cycle life, the ratio of SNCM should be decreased. In addition, considering degradation mechanism of each active material, LNCM should be designed with less crack by doping. In order to reduce charge transfer resistance generated at the grain boundary of SNCM, it must be made into a single crystal.ope
Strategies for boosting the activity of single-atom catalysts for future energy applications
No full textSingle-atom catalysts (SACs) containing "single-site" centers have become a distinct category in catalytic science. With positive achievements for various applications, SACs blur the line between heterogeneous and homogeneous catalysis, maximize metal efficiency, and provide ultrahigh activity. However, challenges remain to achieve the scalable production of SACs with atomically dispersed high-density and uniform sites, electronic structures, and fast reaction dynamics. This review summarizes the cutting-edge achievements of SACs for future energy applications. Specifically, we focus on recent efforts to boost the catalytic activity of SACs. Several key directions in creating local active surfaces, synthesis methodology, reaction mechanisms, and surface science from the viewpoint of coordination chemistry are introduced. Finally, we outlook on achieving star SACs up to the level of "precise catalysis" in the future
Observation of variations in cosmic ray single count rates during thunderstorms and implications for large-scale electric field changes
No full textWe present the first observation by the Telescope Array Surface Detector (TASD) of the effect of thunderstorms on the development of cosmic ray single count rate intensity over a 700 km(2) area. Observations of variations in the secondary low-energy cosmic ray counting rate, using the TASD, allow us to study the electric field inside thunderstorms, on a large scale, as it progresses on top of the 700 km(2) detector, without dealing with the limitation of narrow exposure in time and space using balloons and aircraft detectors. In this work, variations in the cosmic ray intensity (single count rate) using the TASD, were studied and found to be on average at the similar to(0.5-1)% and up to 2% level. These observations were found to be both in excess and in deficit. They were also found to be correlated with lightning in addition to thunderstorms. These variations lasted for tens of minutes; their footprint on the ground ranged from 6 km to 24 km in diameter and moved in the same direction as the thunderstorm. With the use of simple electric field models inside the cloud and between cloud to ground, the observed variations in the cosmic ray single count rate were recreated using CORSIKA simulations. Depending on the electric field model used and the direction of the electric field in that model, the electric field magnitude that reproduces the observed low-energy cosmic ray single count rate variations was found to be approximately between 0.2 GV-0.4 GV. This in turn allows us to get a reasonable insight on the electric field and its effect on cosmic ray air showers inside thunderstorms
Machine learning models for predicting maximum displacement of triple pendulum isolation systems
No full textMaximum displacement is an important engineering demand of an isolation system, including systems using triple friction pendulum bearings, during earthquakes. This response can be accurately predicted by time-history dynamic analysis of the nonlinear model of the system. However, this analysis approach is time-consuming and requires skillful analysts. To remedy the cumbersomeness, this study developed four machine learning models to confidently predict the important demand using limited number of parameters defining isolation system and earthquake event. Specifically, random forest, gradient boosting regression tree, adaptive boosting, and extreme gradient boosting approaches were employed to develop the machine learning models. The input features to the models include eight constitutive parameters of the triple pendulum bearings in the isolation system and five spectral accelerations at control periods of the average spectrum of the site. The database for constructing the machine learning models was obtained from time-history analysis of lumped-mass nonlinear model of isolation systems subjected to earthquake ground motions. The performance investigation showed that all proposed machine learning models can confidently predict the maximum displacement from the time-history analysis procedure. Among the four models, extreme gradient boosting model possesses the highest accuracy with an average ratio between analysis and predicted values of 0.9999 and a coefficient of variation of 0.017. A graphical user interface module based on this machine learning model was developed for practical uses. The module was written in Python and is free for download at GitHub
