A blend of first-principles and kinetic lattice Monte Carlo computation to optimize samarium-doped ceria

Solid oxide fuel cells (SOFCs) have been acknowledged as a possible future source for clean and efficient electric power generation. One of the most important goals in the SOFCs research is to decrease the operating temperature, which in turn will improve the stability and decrease the cost of various components enabling its widespread utilization. For realizing the aforementioned goal, it is imperative to identify suitable electrolyte materials that show enhanced conductivity in the intermediate temperature range (773–1,073 K). Sm-doped ceria (SDC) is considered a promising candidate for use as an electrolyte material for SOFC operation in intermediate temperature range due to the high oxygen ion conductivity. In this article, we present a theoretical investigation using first-principles and kinetic lattice Monte Carlo (KLMC) computations to highlight the trends in oxygen ion conductivity as a function of dopant content and temperature in SDC. Using first-principles calculations, oxygen vacancy formation and migration were examined at first, second, and third nearest neighbor positions to a Sm ion. The activation energies for oxygen vacancy migration along various pathways in SDC computed using first-principles were used as input to the KLMC model to study vacancy mediated diffusion. SDC with 20 % mole fraction of dopant content yields the maximum conductivity, which is in very good agreement with experimentally identified compositions. Rationale for increase in conductivity as a function of increase in dopant content and subsequent decrease in conductivity at higher dopant fractions in SDC is presented. This combined methodology of first-principles and KLMC computations is a useful tool for the design and identification of various ceria-based electrolyte materials used in SOFCs.     

Comparative analysis and robustness study of logic styles

Microsystem Technologies - This paper explores design of various components that are extensively used in digital circuits namely AND gate, OR gate, NAND gate, NOR gate and Full Adder using GDI...     

Modular-structured resonant converter for multilevel converters in micro grids

In this work, an isolated modular resonant circuit with effective voltage regulation is presented as front-end for multilevel inverter system to interface renewable sources of energy with micro grid. Inputs of each module of the converter are connected to low-voltage DC sources. Modular-structured module outputs may be linked in sequence or parallel to attain the desired DC link capacitor voltage for multilevel inverter. The proposed converter overcomes imbalance in the capacitor voltage of diode-clamped inverter structure. Interleaved phase shift of 90° is supplied among the gate pulses of all four modules that minimizes the input-side current ripples. An effective zero voltage switching (ZVS) is achieved over wide load range using duty cycle and frequency-duty cycle double control method to improve system performance at light load condition. Adopting powder core magnetics and low forward voltage insulated-gate bipolar transistors (IGBTs) reduces magnetic and conduction loss. A 12-kW prototype of the designed modular resonant converter was verified by experimentation with a different-level diode-clamped inverter structures.     

EchoFakeD: improving fake news detection in social media with an efficient deep neural network

Neural Computing and Applications - The increasing popularity of social media platforms has simplified the sharing of news articles that have led to the explosion in fake news. With the emergence...     

Using gestures to convey internal mental models and index multimedia content

Gestures can serve as external representations of abstract concepts which may be otherwise difficult to illustrate. Gestures often accompany verbal statement as an embodiment of mental models that augment the communication of ideas, concepts or envisioned shapes of products. A gesture is also an indicator of the subject and context of the issue under discussion. We argue that if gestures can be identified and formalized they can be used as a knowledge indexing and retrieval tool and can prove to be useful access point into unstructured digital video data. We present a methodology and a prototype, called I-Gesture that allows users to (1) define a vocabulary of gestures for a specific domain, (2) build a digital library of the gesture vocabulary, and (3) mark up entire video streams based on the predefined vocabulary for future search and retrieval of digital content from the archive. I-Gesture methodology and prototype are illustrated through scenarios where it can be utilized. The paper concludes with results of evaluation experiments with I-Gesture using a test bed of design-construction projects.

---

Tables to LaTeX: structure and content extraction from scientific tables

International Journal on Document Analysis and Recognition (IJDAR) - Scientific documents contain tables that list important information in a concise fashion. Structure and content extraction from...     

FakeBERT: Fake news detection in social media with a BERT-based deep learning approach

In the modern era of computing, the news ecosystem has transformed from old traditional print media to social media outlets. Social media platforms allow us to consume news much faster, with less restricted editing results in the spread of fake news at an incredible pace and scale. In recent researches, many useful methods for fake news detection employ sequential neural networks to encode news content and social context-level information where the text sequence was analyzed in a unidirectional way. Therefore, a bidirectional training approach is a priority for modelling the relevant information of fake news that is capable of improving the classification performance with the ability to capture semantic and long-distance dependencies in sentences. In this paper, we propose a BERT-based (Bidirectional Encoder Representations from Transformers) deep learning approach (FakeBERT) by combining different parallel blocks of the single-layer deep Convolutional Neural Network (CNN) having different kernel sizes and filters with the BERT. Such a combination is useful to handle ambiguity, which is the greatest challenge to natural language understanding. Classification results demonstrate that our proposed model (FakeBERT) outperforms the existing models with an accuracy of 98.90%.

     

DAI based wireless sensor network for multimedia applications

In Wireless Multimedia Sensor Network (WMSN), the time critical and delay sensitive applications like video, audio, image demands high bandwidth and transmission resources. The provision of Cognitive Radio (CR) can effectively utilize the available spectrum in the most appropriate way to provide high bandwidth in the Wireless Sensor Network (WSN) environment as Cognitive Radio sensor network (CRSN). The CR features are applicable in WMSN paradigm with required changes in transmission parameter for bandwidth hungry multimedia applications. In this paper, we propose an approach for setting up a cost-efficient and higher data rates communication in Wireless Multimedia Cognitive Radio Sensor Network (WMCRSN). The process analyses power allocation for sensor nodes by dynamic channel modelling and allocates power using multi-agent based Distributed Artificial Intelligence (DAI) in WMCRSN applications. The novelty in the approach lies in analyzing the real-time spectrum sensing outputs system for high data rate wireless multimedia applications. The DAI makes the process of power allocation in a smart way for having low latency based intra and inter cluster communication between sensor nodes. The performance parameters of the network, i.e. probability of detection and false alarm with the modelled error rates are presented. The mathematical analysis and simulation results justifies the feasibility and merits of the proposed method over conventional methods.

     

Electrodeposited MnO2 as electrocatalyst for carbohydrate oxidation

The electrochemical oxidation of dextrose, fructose and sorbitol under galvanostatic conditions was carried out using both anodically and cathodically deposited MnO₂ layers on platinum and carbon. The coated electrodes showed better electrocatalytic activity for the oxidation of carbohydrates than the bare substrates. Catalytic participation of some higher valent states of manganese in electron transfer relay is speculated, which finds support in the chronopotentiogram and the observed pH-dependence of the electrochemical parameters. The current potential plots showed Tafel behaviour for the MnO₂/Pt electrode. The Tafel slopes were found to be relatively high, indicating kinetic complications. The observed unusual negative values of electrochemical reaction order on MnO₂/Pt electrode were accounted for by considering slow desorption of oxidation products from the electrode surface. The adsorption isotherms of dextrose and fructose on MnO₂ were determined. The current efficiencies of oxidation of carbonyl and hydroxyl groups were found to be and % respectively. SEM pictures showed the cathodically deposited MnO₂ on carbon to be more fine-grained and smoother than the corresponding anodic deposits.     

Linking Interfacial Step Structure and Chemistry with Locally Enhanced Radiation‐Induced Amorphization at Oxide Heterointerfaces

Nanostructured materials and their interfaces have attracted recent interest for their functionality in a wide variety of different applications. However, the origins of these properties in several instances remain unknown. One promising aspect of nanomaterials is their role in materials design for mitigating radiation damage. In particular, engineered radiation tolerant materials would exploit the presence of internal interfaces to act as recombination centers and suppress damage accumulation. Realizing this promise, however, requires a fundamental understanding of how radiation‐induced defects interact with interfaces. Thus, studying the interfacial atomic structure and chemistry before and after irradiation is critical. In this study, we have performed transmission electron microscopy on a series of pristine and ion‐irradiated oxide interfaces to probe radiation‐induced effects. The CeO₂/SrTiO₃ interface, chosen as a model system for these studies, is characterized by differences in SrTiO₃ terminations or steps. Our salient result is that steps are centers for preferential amorphization in SrTiO₃, which we attribute to defect flow across the interface induced by non‐stoichiometry in CeO₂. The study concludes the interfacial atomic ordering in the form of steps thereby modifies the response to ion irradiation and suggests interface patterning as another parameter to functionalize radiation tolerant materials.