34254 research outputs found
Sort by
2D Vanadium Oxide Nanosheets by Hydrothermal Treatment Assisted Wet Chemical Method for Photothermal and Biosensing Applications
No full textIn the first study, vanadium dioxide has excellent absorbance in near-infrared (NIR) wavelengths making it a perfect candidate for photothermal applications based on solar energy. However, up to date, it was mainly applied in a lithium-ion battery. This study is for the second time, to utilize VO2-based nanomaterials as a photothermal material for solar steam generation to purify seawater and wastewater. The study has proposed a straightforward and promising room temperature approach to synthesize a mixture of two vanadium oxide nanosheets with a major fraction of VO2 (B) nanosheets with the major part for VO2.nH2O and a minor fraction of V2O5.nH2O which have demonstrated excellent photothermal capability for water purification. The highest steady state temperature of ~87 \uc2\ub0C was reported by the synthesized nanosheets that were spray coated on cellulose fabric for the solar absorption purpose under one sun illumination for water purification. The as-prepared photo absorber was capable of reaching the highest steady temperature at high speed (within 10 minutes). The synthesized 2D VO2/V2O5.nH2O nanosheets and cellulose fabric combination of solar steam generator gained a competitive evaporation rate of 1.31 kg m-2 h-1 and a percentage efficiency of 89.7% under one sun illumination. The current approach was successfully applied to three real-world samples to obtain clean or drinking water. All three samples showed excellent improvements in their water quality compared with their initial states in a single distillation. Two samples can even reach the quality of WHO-recommended standards for drinking water. The current approach has opened a new platform for the utilization of VO2-based nanosheets for obtaining clean water from seawater and wastewater by solar energy utilization.
In the second study, for the first time, we report a Dual Optical Sensor Method (DOSM) using novel 2D VO2 nanosheets to act as fluorometric and colorimetric sensors to perform quantitative analysis for epinephrine and dopamine. The wide color spectrum of the 2D vanadium oxidation series and specifically metastable blue 2D VO2 nanosheets were used to develop a DOSM biosensor. Dopamine and Epinephrine are the major catecholamines in the human body that play vital roles as neurotransmitters and stress-responsive hormones of the endocrine system, respectively. Accurate and selective detection of these bio-molecules can assist in the diagnosis of many neuroendocrine system-related diseases. The newly synthesized 2D VO2 nanosheet sensor showed bright greenish fluorescence as the first-ever fluorescence from 2D VO2 nanosheets. This sensor showed a dual function sensing towards epinephrine by color change dominantly and fluorescence quenching. It is capable of individually detecting and quantifying both epinephrine and dopamine with high selectivity and sensitivity by using both colorimetry and fluorometry simultaneously with the detection limits of 1.07 \uc2\ub5M and 5.54 \uc2\ub5M for colorimetric analysis, respectively, and 48.07 \uc2\ub5M and 3.98 \uc2\ub5M for fluorescence analysis, respectively. The DOSM sensor was directly applied to real urine samples and gained satisfactory recovery above 90% in the means of spiked concentrations. This study has opened a new platform using the DOSM and the vanadium oxidation spectrum in a much more effective way for bio-sensing. The fluorescence capabilities of this metal oxide can be further applied to many sensor applications based on both fluorescent and colorimetric detection
Feasibility of TikTok Promoting Taiwanese Dramas
No full textDigital advertising in Taiwan totaled NT$54,428,000,000 in 2021, an increase of 12.8% more than 2020, of which social media accounted for 38%, a growth rate of 11.9%, of which 85.7% came from mobile users. As of December 2021, there are approximately 4,200,000TikTok users in Taiwan, with approximately 91% of users aged 18 to 34 and 45 and older. In early 2022, about 21,350,000 social media users in Taiwan, or 20.6% of the Taiwanese population, were affected by TikTok ads. At present, Taiwan dramas have not formed a complete marketing system on TikTok. In other words, if TikTok can establish a sound marketing model for Taiwan dramas, it will improve the TV ratings and viewing rate.
This study explores the feasibility of TikTok's promotion of Taiwan drama from the perspective of market analysis and user analysis; and discusses the overall environment for TikTok's promotion in Taiwan, including political, economic, social and technological influences, as well as the competitive environment and competitors for the promotion of Taiwan drama in Taiwan. User analysis uses TikTok recommended algorithm to understand TikTok user appearance. The method of action research was adopted, and the data collection methods were observation and analysis of secondary data. The main research questions are as follows: 1. What is the market competition of TikTok's promotion of TV series? What are the competitors? 2. what is the user's willingness to watch and search? What is the user's recommended satisfaction and willingness to share/forward?
Research shows that TikTok has advantages in promoting Taiwanese drama in Taiwan. At present, the usage rate of TikTok in Taiwan is becoming more and more popular, and the Taiwanese are gradually accustomed to and addicted to TikTok. The marketing promotion of TikTok is gradually popular, but the security problem of TikTok is still the main reason for the rejection of the use of TikTok in Taiwan. TikTok has a competitive advantage in promoting Taiwanese dramas in Taiwan; TikTok users are more interested in stars and celebrities, film and television clips, popularize knowledge, law, media related videos with large fluctuations and positive and negative emotional confrontation. TikTok users' willingness to watch TV series and recommended satisfaction is high, while their willingness to share and search is low.
The researcher put forward two suggestions for TikTok: First, it is expected that TikTok can release data analysis and audience analysis reports focusing on Taiwan; Second, it is expected that TikTok officials can establish convenient advertising channels for Taiwan audiences. The following suggestions are put forward for enterprises and creators who use TikTok as a promotion tool in Taiwan: 1. remember to "raise the account" after registering a new account; 2. Keywords can match TikTok's recent activities; 3. Active account, you can recommend videos by browsing, liking and commenting; 4. Positioning accounts and Posting videos are relevant; 5. Keep the identity and content of the account unique; 6. the content has various emotions can obtain large clicks
Research on Electrical Characteristics and Reliability Mechanism of GaN High Electron Mobility Transistors and Fin-Field Effect Transistors
No full textIn recent years, the growing demand for higher data bandwidth, driven by digitalization across various industries and the widespread use of smartphones, has fueled the rapid development of 5G communication. Additionally, the increasing awareness of environmental concerns and the rising demand for electric vehicles have led to a continuous growth in the market's demand for power devices. These shifts in market demand have prompted significant attention toward Gallium Nitride (GaN). GaN is highly regarded for its advantages, including high electron mobility, excellent thermal stability, and a high breakdown voltage. Therefore, GaN High Electron Mobility Transistors (HEMTs) are particularly anticipated for high-frequency and high-voltage applications. However, power conversion systems always generate heat during operation, which can affect other components, such as central processing units. Silicon-based Fin Field-Effect Transistors (FinFETs), as the most widely used devices in logic operations, exhibit degradation mechanisms closely related to environmental temperature. Therefore, this dissertation primarily focuses on analyzing the electrical characteristics and reliability issues of GaN HEMTs and silicon-based FinFETs through electrical properties, reliability testing, and electrical simulations.
In Chapter 3, this dissertation explores the three-stage leakage mechanisms in p-GaN HEMTs. Through the investigation of leakage current contributions at various endpoints, repeated experiments on leakage measurements, low/high voltage off-state stress tests, and temperature-dependent off-state stress tests, the mechanisms at each stage are clarified. These stages are primarily dominated by punch-through leakage, gate electron injection, and defect-assisted thermal field emission. Once the leakage mechanisms are elucidated, the introduction of variables like un-doped GaN (UGaN) thickness and process temperature provides a feasible pathway for adjusting leakage characteristics through process improvements. Experimental results ultimately reveal that a thin UGaN layer and lower process temperatures offer the more effective suppression of punch-through leakage. Conversely, a thick UGaN layer is more effective in curbing current generation.
In Chapter 4, the dissertation examines the anomalous saturation current trends between p-GaN HEMTs with low and high carbon doping concentration buffer layers. The p-GaN HEMTs with high carbon doping concentration buffer layers exhibit higher current levels in the saturation region. This result contradicts the common understanding that carbon doping reduces 2DEG concentration. The reasons behind this anomalous trend are clarified through temperature experiments and saturation region stress tests. This anomaly is due to the lower energy barrier in the GaN layer of p-GaN HEMTs with low carbon doping concentration buffer layers. Under saturation region conditions, hot electron injection into the GaN layer occurs more easily, affecting current characteristics. Lastly, the impact of carbon doping concentration on the energy barrier is verified through Silvaco TCAD simulations.
In Chapter 5, this dissertation investigates the Drain-Induced barrier lowering (DIBL) effect saturation phenomenon in Schottky-gate GaN HEMTs. Utilizing Silvaco TCAD simulations involving electric fields, energy band diagrams, and 2DEG concentrations at various drain voltage (Vd), the analysis reveals that the saturation of the DIBL effect is rooted in the T- gate structure. This gate structure has the ability to deplete additional 2DEG under large Vd, thus dispersing the electric field that would normally concentrate in the channel, consequently suppressing the DIBL effect. Finally, the thesis examines the relationship between the ability to suppress DIBL and the geometric aspect of the T-gate structure from the perspective of parasitic capacitance.
In Chapter 6, this dissertation explores the influence of temperature on the degradation mechanisms in 60nm and 14nm FinFETs. Through fitting the mechanisms, the impact of different voltage conditions on degradation mechanisms is clarified. With increasing gate voltage (Vg), the Hot Carrier Stress (HCS) degradation mechanism transitions from Single Vibrational Excitation (SVE) or Electron-Electron Scattering (EES) mechanisms with relatively higher field dependence to the Multiple Vibrational Excitation (MVE) mechanism with lower field dependence. In the 60nm sample, due to the impact of phonon scattering, the part predominantly governed by SVE transitions to the EES mechanism as the temperature rises. In the 14nm sample at higher temperatures, under higher Vg, the transition shifts from EES to MVE. Finally, the relationship between lifetime (\ucf) and temperature validates this argument
Taiwan Stock Futures Price Prediction with Machine Learning
No full textFinancial markets are characterized by price uncertainty and volatility. For traders, predicting market trends is crucial for maximizing profits and minimizing risks. However, traditional reliance on various technical and fundamental analysis tools to predict market movements often has limitations. With the increasing application of machine learning in finance, a new approach is offered to process and analyze large volumes of market data. This involves learning from historical data, identifying potential patterns in price movements, and providing predictions for future market trends.
This study aims to use Support Vector Regression (SVR), XGBoost Regression, and Random Forest Regression to predict the price behavior of Taiwan stock futures. It utilizes daily price data from January 2020 to June 2023 of six listed stock futures on the Taiwan Futures Exchange: Taiwan Semiconductor (TSMC), United Microelectronics (UMC), AU Optronics (AUO), Cathay Financial, Fubon Financial, and Shin Kong Financial. The data includes opening price, highest price, lowest price, closing price, trading volume, open interest, and positions not yet offset. The study also incorporates Simple Moving Average (SMA), Relative Strength Index (RSI), and Exponential Moving Average Convergence Divergence (MACD) for empirical analysis. The results show that Support Vector Regression performs best for predicting Shin Kong Financial stock futures, XGBoost Regression is most effective for Taiwan Semiconductor stock futures, and Random Forest Regression provides the best predictions for United Microelectronics, AU Optronics, Fubon Financial, and Cathay Financial futures. Furthermore, XGBoost Regression and Random Forest Regression demonstrate superior predictive performance over Support Vector Regression, as evidenced by their lower MAPE, higher R2 values, and more accurate prediction of price trends
Investigating Fish Biota and Seasonal Dynamics in a Tropical River using eDNA: Gaoping River, Taiwan, as a case study
No full textFreshwater ecosystems are increasingly threatened by habitat fragmentation, destruction, and biological invasion. To enhance the efficiency of biodiversity monitoring, environmental DNA (eDNA) metabarcoding has emerged as a powerful method for ecological assessments. Given the limited understanding of the efficacy of eDNA in tropical rivers, this study utilized eDNA metabarcoding to evaluate its effectiveness in revealing the fish biota and seasonal distribution patterns in Taiwan's Gaoping River basin. Over six seasons spanning from 2020 to 2022, 75 three-liter stream water samples were collected and analyzed. Totally, 122 fish species were detected, validating over half of the historical freshwater fish records from conventional surveys and identifying four previously undocumented invasive fish species. Fish distribution patterns exhibited relative consistency throughout the year, with specific seasons (such as summer vs. spring and summer vs. winter) displaying distinct patterns, which are likely associated with changes in food sources within the environment and reproductive behaviors triggered by high and low flow periods. The co-utilization of co-occurrence networks facilitated the summarization and visualization of cross-species seasonal distribution patterns, revealing potential ecological relationships. Furthermore, comparative analyses with on-site electrofishing surveys confirmed the reliability of eDNA read abundance for macro-level assessments, e.g., the comparison of overall abundance changing patterns. While refinements are still necessary, this study underscores the significance of the eDNA metabarcoding method in providing ecological insights into tropical freshwater fish ecosystems, contributing to biodiversity assessment and conservation management in the face of rapidly changing environments
A Two-Stage Offline Test and Protection Method for Undervolted SRAM
No full textThe surge in powerful mobile device usage has led to a significant increase in the need for multimedia applications on these devices. These complex applications require frequent access to embedded memory due to their intensive computational demands and heavy reliance on memory. Consequently, the substantial power consumption of embedded Static Random-Access Memories (SRAMs) restricts the battery life of mobile devices. Voltage scaling stands as the prevailing method to achieve low-power memory, substantially reducing SRAM power consumption by leveraging the square relationship between voltage and power. However, as voltage decreases, the probability of undervolting faults grows exponentially. Employing conventional offline memory repair techniques leads to a significant increase in repair costs due to a large number of undervolting faults. While past literature has developed offline protection techniques based on undervolting faults, they haven't integrated these protective measures with offline testing processes nor discussed the feasibility of in-field testing and protection.
This thesis proposes a two-stage offline testing and protection technique based on undervolted SRAM memory. The proposed technique utilizes bit-shuffling and redundancy repair approaches as the first and second-stage protection methods. In the first-stage testing phase, a single test sequence \ue2w1r1w0r0 is employed targeting on undervolting faults, resolving the issue of generating shift values that cannot be feasibly implemented on-chip due to testing costs. Simultaneously, a shift value generation algorithm is developed, minimizing the time required to generate shift values during the testing phase using a Look-Up Table. Finally, a cost-effective hardware architecture is proposed that, compared to exhaustive approach, reduces an average of only 2% deviation while reducing permutations by a factor of 2^16. In the second stage, it employs the March testing algorithm to test manufacturing defects and uses redundancy repair techniques coupled with the concept of non-uniform protection to repair remaining high-bit errors, achieving further error mitigation.
Comparing the protective effects and implementation costs with past literature, our experimental results indicate that at an error rate of 10^(-3), compared to bit-shuffling, there is an 81% reduction in Mean Absolute Error (MAE) with a 1.6% increase in total area cost. Compared to redundancy repair techniques, there is a 46% reduction in MAE with a 5.1% decrease in total area cost. Finally, calculating the performance of each method in terms of implementation cost and MAE across all possible configurations, a Pareto Optimality analysis is conducted. The experimental results indicate that in the Pareto frontier of optimal solutions at error probabilities of 10^(-3) and 10^(-2), respectively, 83% and 73% of the solutions are attributed to our proposed technique, demonstrating that the proposed technique offers more configuration choices based on user constraints of area cost and MAE across different error probabilities
Improving the Mechanical Properties of CuZr-Based Metallic Glass with the Addition of Nb Element
No full textith the advancement of technology, the devastation of resources, and the pollution of the air environment becoming unavoidable issues, the extensive focus on ecological sustainability has driven the critical research into the development of lightweight materials. The lightweighting of public transportation vehicles can enhance fuel efficiency and reduce carbon emissions. In the process of developing lightweight materials, the improvement of specific strength becomes crucial, especially when applied to transportation as a key factor in passenger safety.
Metallic glasses, due to their high strength and excellent corrosion resistance, are considered promising structural materials. However, their lack of room temperature plastic deformation capability severely limits their application in structural materials. To address this crucial issue, this experiment utilized the addition of trace elements to induce atomic-level microalloying within the internal amorphous microstructure. The goal of this experiment was to maintain the non-crystalline structure of the parent phase and thereby enhance the glass-forming ability. To achieve this target, CuZrAlTiNb was selected as the composition basis to enhance the strain capacity of the metallic glass. The optimal component obtained through composition refinement were determined to be (CuZr)85(AlTiNb)15. Experimental results revealed that changes in internal impurity atom concentrations led to alterations in the structure. Through the investigation of Nb addition, it was found that samples without Nb exhibited lower yield strength and ductility compared to those with Nb. The sensitivity of the mechanical properties of metallic glass to compositional changes was explored in the study, and the optimized composition (CuZr)85(AlTiNb)15 exhibited both plastic and non-plastic compressive behaviors.
Specifically, the yield strength was 1747.3 \uc2\ub1 67.3 MPa with a strain of 7.2 \uc2\ub1 0.8 % for the non-plastic samples, and 1761.8 \uc2\ub1 13.1 MPa with a strain of 13.9 \uc2\ub1 3.2 % for the plastic samples. According to the abovementioned, we can infer that during the rapid cooling process, differences in internal pores result in the phenomenon of mechanical properties varying for the same element at the macro level
Attribute-Aware Flow Management Strategy in SDN-Based Data Center Networks
No full textThe recent escalation in network activity triggered by advancements in cloud computing, big data, the Internet of Things, and artificial intelligence has presented significant challenges to data centers. Traditional network architectures, characterized by manual configuration and hardware control, need to help handle the diverse and expanding traffic, leading to issues like uneven distribution and reduced performance. The advent of software-defined networking (SDN) can effectively address the above challenges by automatically adjusting path and bandwidth allocation based on traffic conditions, thereby enhancing network performance.
This paper introduces an SDN-based traffic management policy tailored to address various link conditions. When all available links have ample bandwidth, centralized control optimally schedules transmission paths to alleviate congestion. Conversely, when all links are saturated, the rate is scaled down based on current traffic size, duration, and value, prioritizing the protection of high-value traffic. The proposed dynamic scaling monitoring mechanism continuously assesses link status, facilitating the reallocation of optimal bandwidth resources. Compared with the link status-based mechanisms, our strategy has been empirically proven to enhance transmission efficiency and reduce packet loss
Theoretical and experimental investigations of Janus transition metal dichalcogenides
No full textJanus transition metal dichalcogenides (TMDs) are new kinds of two-dimensional (2D) materials that possess novel properties due to their asymmetric structure. The breaking of the out-of-plane mirror symmetry in Janus TMDs results to interesting phenomena such as Rashba splitting, vertical piezoelectric effect, and a second harmonic generation (SHG) performance, among others. In this work, we performed theoretical and experimental investigations of Janus TMDs. First, we explored the anisotropic Rashba splitting on Pt-based Janus monolayers (ML) PtXY (X,Y = S, Se, Te). We found that all three Pt-based Janus TMD monolayers are thermodynamically stable based on the Phonon dispersion plot. Furthermore, all the Pt-based Janus TMD monolayers retained the 1T-phase as the stable structure, like their classical counterpart (e.g., PtS2, PtSe2, and PtTe2). In terms of electronic properties, PtSSe, PtSTe, and PtSeTe are insulating materials with an indirect bandgap of 2.108, 1.335, and 1.221 eV, respectively. Due to centrosymmetry breaking in the Pt-based Janus TMD monolayers, spin-orbit coupling (SOC)-induced anisotropic Rashba splitting can be observed. Among the Janus PtXY monolayers, PtSSe has the greatest Rashba strength at 1.654 and 1.333 eV\ue2\ua2\uc3-1 from M to \uce (\uce\ub1_R^(M-\uce)) and M to K (\uce\ub1_R^(M-K)), respectively. Meanwhile, for PtSTe and PtSeTe, the Rashba strengths for \uce\ub1_R^(M-\uce)are 1.103 and 1.244 eV\ue2\ua2\uc3-1, while the values for \uce\ub1_R^(M-K) are 0.435 and 0.746 eV\ue2\ua2\uc3-1, respectively. Also, based on the spin texture results, Dresselhaus effect contributed to the anisotropic spin-splitting. Furthermore, the Rashba splitting and bandgap can be manipulated by applying biaxial strain. Next, we discuss our successful synthesis of ML Janus TMDs through plasma-assisted selenization process (PASP). By carefully controlling the kinetic parameters such as temperature, plasma power, and selenization time, the top sulfur (S) atom can be replaced with Se, creating the Janus TMDs. Based on the Raman results, PASP is suitable for high-yield production as 135 individual MoS2 flakes were all converted to Janus MoSSe. Furthermore, Janus MoSSe and WSSe monolayers can be created through PASP, indicating the universality of the process for synthesizing Janus TMDs. By changing the selenization temperature, phase-transition can be observed. At 200 \uc2\ub0C, Janus MoSSe at 2H can be obtained while increasing the temperature to 400 \uc2\ub0C or 600 \uc2\ub0C can lead to 1T/1T\ue2 phase of MoSSe. Finally, wafer-scale Janus MoSSe was synthesized by PASP from 2-inch continuous MoS2. Our findings provide significant contributions towards full exploration of Janus TMDs for future novel electronics and optoelectronics devices application
The Role of Adjuvant Radiotherapy for Non-small Cell Lung Cancer in Modern Era and the Evaluation of Associated Biogenetic and Pathophysiologic Predicting Factors.
No full text
With the improvement of radiotherapy technique and well-established knowledge regarding clinicopathological factors of non-small cell lung cancer(NSCLC), the aims of this study were to evaluate treatment effect of modern radiotherapy for completely-resected pN2-stage III(R0-resected pN2-stage III) NSCLC and to search for possible predicting factors for post-operative radiotherapy(PORT) through systematic review and meta-analysis and cohort study based on local real world data.
First, we conducted a systematic review of literature regarding possible predicting factors of PORT for pN2-stage III NSCLC. Then, local real world data cohort study was carried out by retrospectively collecting clinicopathological factors, treatment courses, and clinical outcomes of R0-resected pN2-stage III NSCLC from the Cancer Registry Database of a medical center. Survival outcomes were analyzed with Kaplan-Meier Method. Impacts of clinicopathological factors were evaluated by Cox regression model. The primary endpoint was disease-free survival(DFS). Other endpoints included overall survival(OS), distal metastasis(DM) and locoregional recurrence(LRR).
We found 15 relevant literature published during 2009-2023, including 1 randomized control trial and 14 retrospective cohort studies, half of which were based on large cancer registry database of the USA. Considering possible repeated sampling and high heterogeneity regarding clinicopathological factors and reported outcomes among literatures, meta-analysis was not performed.
As for the cohort study, there were 82 R0-resected pN2-stage III NSCLC included from 2010 to 2021. Among them, 70.1% were test positive for epidermal growth factor receptor(EGFR) mutation. There were 73.2% received PORT, with the medium dose of 54 Gy. After medium follow up of 42 months, the 3-year DFS and OS were 40.6% and 77.3%, respectively. PORT was associated with better 3-year DFS (44.9% vs 29.8%; HR: 0.552\uef\ubcp=0.045). The benefit was more profound in subgroups with EGFR mutation and absent extranodal extension, with a statistically significant high heterogeneity.
For R0-resected pN2-stage III NSCLC, PORT with modern technique might improve DFS. The DFS benefit could be predicted by some clinicopathological factors and might even translate into OS improvement in some subgroups. Further evaluation of possible predicting factors and development of practical nomogram or polygenic risk score might help guide the personalized and precision treatment of NSCLC