Silicon最新文献

With the advancement of packaging technologies, wafer bonding has become increasingly critical in the fabrication of microelectronic devices, and stress concentration during this process has been recognized as a key factor affecting device reliability. To elucidate the stress response characteristics of wafers with different edge profiles and thicknesses during bonding, finite element simulations were conducted on R-type (rounded) and T-type (beveled) wafers with varying thinning thicknesses and chamfer angles. The simulations were performed under vertically applied uniform loads ranging from 15 to 60 MPa. The results indicate that while the stress distribution remains relatively uniform across the central wafer region, pronounced stress concentrations emerge in the edge regions, particularly near geometric discontinuities such as notches. T-type edges showed relatively better stress mitigation performance when thinned to an intermediate thickness (~ 562 μm), and larger chamfer angles were found to reduce stress concentrations associated with thinning. Further analysis of the stress concentration factor K revealed that R-type wafers exhibited a marked increase in K following thinning, whereas T-type structures demonstrated enhanced tolerance to thinning as chamfer angles increased. These findings indicate notable differences in stress adaptation between edge geometries under varying thicknesses and clarify the coupled influence of edge design and thinning on wafer stress behavior. The results may offer useful guidance for optimizing wafer edge structures to improve bonding performance and enhance the mechanical reliability of chip packages.

A novel, accurate charge-based MOSFET long-channel computational model is presented, which is portable and can be used across the electrical engineering domains ranging from sensing to power electronics, both under sub-threshold as well as super-threshold regime of MOSFET operation. The proposed physics-based model can be universally used to any long-channel MOS-transistor as it does not depend on any empirical factor and features extremely good computational efficiency. The model uses a novel two-step charge linearization, resulting into accurate drain current and charge model – valid for both the subthreshold and super-threshold regime of long-channel MOSFET operation. Another salient feature of the proposed model is a novel SPICE-compatible numerical solution strategy for the principal branch of the Lambert W function ((varvec{W}_{varvec{0}}varvec{(x)}) for (varvec{{x in {mathbb {R}} | x ge 0 }})). The algorithm is faster than present industry standard implementations, computationally efficient, accurate with maximum percentage error (varvec{approx O(10^{-14}%)}) and therefore may be incorporated in a SPICE engine for electrical design and optimization. The proposed computationally efficient long channel MOSFET model is validated against thorough TCAD simulations upto the fourth derivative and has been found to have fast convergence along with much higher degree of accuracy (maximum error (varvec{approx O{(10^{-14}%)}})) compared to existing MOSFET models.

This study elucidates the early-age reaction mechanisms governing strength development in ceramic waste powder (CWP) concrete activated with 6% potassium silicate (K₂SiO₃) solution. Microstructural evolution at three cement replacement levels (0%, 15%, 30%) was investigated through SEM–EDS analysis coupled with 7-day mechanical testing. Progressive pozzolanic activation was confirmed by decreasing Ca/Si ratios from 4.51 in control specimens to 2.61 at 15% CWP and 0.92 at 30% CWP replacement. Optimal strength development occurred at 15% CWP, achieving 7-day compressive strength of 42.05 MPa compared to 38.86 MPa for control specimens, with flexural strength reaching 5.90 MPa versus 5.70 MPa and split tensile strength of 5.33 MPa versus 4.20 MPa. At this replacement level, Ca/Si ratios approached ideal C-S–H stoichiometry (1.5–2.0) with balanced microstructural heterogeneity. EDS mapping revealed potassium silicate activation enhanced aluminosilicate dissolution, promoting C-A-S–H gel formation through localized Si enrichment and K-ion distribution at reaction interfaces. The 30% replacement level exhibited excessive microstructural heterogeneity, compromising mechanical performance despite continued pozzolanic activity. These findings demonstrate that controlled potassium silicate activation enables effective partial cement replacement with CWP, providing mechanistic insights for developing sustainable alternatives to conventional sodium-based activation systems while maintaining early strength requirements critical for construction applications.

Adopting sustainable and cleaner fuels has become imperative in light of the rapid depletion of petroleum reserves and escalating vehicular emissions. Biodiesel produced from agricultural and food-waste oils presents a promising alternative; however, its application in contemporary CRDI diesel engines is hindered by intrinsic drawbacks such as elevated viscosity, diminished volatility, and lower energy density. This study explores the synergistic potential of silicon dioxide (SiO₂) nanoparticles as energy catalytic additives to enhance the performance, combustion behavior, and emission profile of an Agri & Food-Waste Mixed Biodiesel (MME20) blend. Nanofuel formulations MME 20, MME20+SIO40, MME20+SIO80, and MME20+SIO120 were developed using CTAB-assisted ultrasonication for superior stability and uniformity. Experimental assessments were conducted on a single-cylinder CRDI diesel engine (3.7 kW, 1500 rpm, 18:1 compression ratio). The addition of SiO₂ nanoparticles led to a reduction in brake specific fuel consumption (BSFC) from 0.355 to 0.31 kg/kWh and an improvement in brake thermal efficiency (BTE) from 24.2% to 25.8% compared to neat biodiesel. Combustion analysis indicated elevated peak in-cylinder pressure (72 bar) and heat release rate (63 J/°CA), which are attributed to enhanced fuel atomization, micro-explosion phenomena, and catalytic oxidation mechanisms induced by SiO₂. Emission studies revealed marked decreases in hydrocarbon (20%), carbon monoxide (~18%), and smoke opacity (~25%), with a slight increase in nitrogen oxides (10%) likely resulting from higher in-cylinder flame temperatures. The findings underscore the effectiveness of SiO₂ nanoparticles in ameliorating the limitations of biodiesel, enabling improved engine performance, superior combustion, and cleaner exhaust emissions in CRDI diesel engines. Alongside the experimental evaluation, a comprehensive Artificial Neural Network (ANN) model was developed using MATLAB to predict engine performance, combustion, and emission parameters. The ANN was configured as a feedforward backpropagation network with input neurons representing engine operational variables and SiO₂ nanoparticle concentrations. The model was trained and validated using experimental data from varying nanoparticle blend levels (0, 40, 80, and 120 ppm) under different engine loads. The ANN predicted key outputs including brake thermal efficiency, brake-specific fuel consumption, hydrocarbon emissions, carbon monoxide emissions, nitric oxide emissions, and smoke opacity with high accuracy, achieving regression coefficients (R2R2) exceeding 0.98 and low mean squared error values. This predictive modeling complements the experimental results by enabling rapid parameter estimation and optimization without exhaustive engine testing, demonstrating the ANN's potential as an effective tool in biodiesel fuel engine studies.

Fiber Metal Laminates (FMLs) have emerged as advanced hybrid materials that combine the lightweight advantages of fiber composites with the structural stability of metal layers, making them attractive for automotive, aerospace, defense, and construction applications. However, achieving strong interfacial adhesion between natural fibers, fillers, and metal layers remains a significant challenge. This study focuses on enhancing the performance of natural fiber–based FMLs reinforced with prestressed areca nut fibers, SS316 foil, and nanosilica fillers through silane surface treatment. The 3-aminopropyltriethoxysilane (3-APTES) coupling agent was employed to promote Si–O–Si bonding and improve compatibility within the epoxy matrix. Laminates were fabricated using vacuum bagging and post-cured at 130 °C to achieve improved stability. Among all compositions, the laminate containing 40 vol.% silane-treated prestressed areca fiber, SS316 foil, and 2.0 vol.% nanosilica exhibited superior performance, achieving tensile and flexural strengths of 162 MPa and 182 MPa, along with enhanced shear properties and reduced drilling-induced damage. SEM analysis confirmed uniform nanosilica dispersion, reduced voids, and stronger fibre–matrix interaction. Overall, the combined use of surface treatment, nanosilica reinforcement, and natural fibers effectively improves the mechanical, shear, and machining performance of FMLs.

Purpose

Deficit irrigation presents significant potential for water savings, making it increasingly popular worldwide as a method to reduce freshwater consumption over time. The low water productivity of strawberries is often attributed to excessive water use and the limited ability of cultivars to optimize fruit set and yield in hydroponic systems. This study aimed to evaluate three management strategies cultivar selection, irrigation frequency, and potassium silicate application frequency on the growth, water conservation, and production efficiency of hydroponically grown strawberries.

Methods

The experiment followed a split-plot design, with 'Albion' and 'Chandler' cultivars as the main plot treatments, and a factorial arrangement of irrigation frequency (once/day vs. twice/day) and potassium silicate (AgSil16H) application frequency (6, 9, 12, 15 weeks) randomly assigned to the subplots.

Results

Results indicated that foliar application of potassium silicate enhanced plant vigor and contributed to water conservation (34%) in hydroponically grown strawberries compared to the control, where no potassium silicate was applied. Notably, a 12-week potassium silicate application boosted photosynthetic rates and improved water conservation, thereby enhancing plant productivity and water use efficiency. For 'Chandler' cultivar, potassium silicate treatment led to a 23% increase in net assimilation rate, a 29% rise in stomatal conductance, and a 33% reduction in transpiration loss. Additionally, electrolyte leakage decreased by 25%, while maintaining steady intercellular CO₂ concentrations. Strawberry plants treated with potassium silicate and irrigated once daily reduced water usage by 34% compared to untreated plants. Furthermore, flowering occurred 4 days earlier in treated plants, while fruit set increased by 16% and flower drop decreased by 13% compared to controls. Among all treatments, the 'Chandler' cultivar, irrigated once per day and treated with potassium silicate for 12 weeks, showed superior growth and significant water savings over the control group. Potassium silicate treatment for 12 weeks also resulted in a 28% higher marketable fruit yield compared to the control.

Conclusion

Therefore, potassium silicate (AgSil16H) demonstrated its potential as a promising fertilizer under deficit irrigation conditions, effectively conserving water and improving productivity in hydroponically grown strawberries.

Graphical Abstract

In pursuit of devices that mirror real neural characteristics, this paper illustrates a leaky-integrate-fire (LIF) neuron implementated using Si doped HfO(_2) ferroelectric junction-less Tunnel FET (FE-JLTFET). The leaky-integration process is thought to be mimicked using BTBT phenomena. Furthermore, back gate biasing is not necessary to cause the floating body to restrict the excess charge carrier. Avoiding metallurgical doping, the problems and the manufacturing complexity can be solved. Using the ferroelectric insulator with normal oxide in the gate stack acts as a step-up voltage transformer. Thus, it only needs drain biased at 0.3V, which is significantly less than its corresponding DGJL FET, PD-SOI MOSFET, LBIMOS, bulk FinFETs and Si NIPIN based silicon neurons. The proposed device requires only 7fJ of energy per spike, exhibiting markedly superior energy efficiency compared to existing neuromorphic devices such as Phase Change CMOS, PD-SOI MOSFET, PCMO RRAM, FBFET, LBIMOS, DGJL FET respectively. Additionally, it exhibits spiking frequencies in the GHz range (8 times higher than biological neurons) at threshold voltage of 0.42V and threshold current of 0.5mA/(mu )m. As a result, it is assumed that the proposed device based leaky-integrated fire neuron is better suitable for implementing the SNN at large-scale due to less energy consumption, area efficient and CMOS adoptable.

The extraction of critical elements from coal gangue has resulted in rapid accumulation of silica-rich byproducts, imposing a significant burden on the environment. In this study, a novel and environmentally sustainable approach was proposed for synthesizing high-purity white carbon black from coal gangue acid leaching residue via ammonium fluoride (NH₄F)-assisted activation. The F/Si molar ratio, roasting temperature, and roasting time were investigated in detail to elucidate their impacts on fluorination roasting process. Systematic characterization mechanistically revealed that siloxane (Si-O-Si) bonds in amorphous silica were cleaved through F⁻ ions nucleophilic attack. This selective bond cleavage fragments the tetrahedral framework into discrete [SiO₄] units, which subsequently react with NH₄F to form water-soluble (NH₄)₂SiF₆ intermediates. Through the integrated process of fluorination roasting, water leaching, and ammonia precipitation, the leaching efficiency of (NH₄)₂SiF₆ reached 99.76%, while the white carbon black product with purity and recovery rate of 98.16% and 95.20%, respectively. The closed-loop design achieved 81.45% recovery of high-purity NH₄F (99.57%) from fluoride-rich filtrate via evaporative crystallization, coupled with capture of 74.34% roasting-derived ammonia gas as 0.2313 mol/L aqueous ammonia (200 mL DI water) for precipitation reuse, thereby largely eliminating fluoride and ammonia discharge as environmental pollutants. This work offers a unique approach for the further study on the facile synthesis of white carbon black with similar mineral composition.

Graphical Abstract

In order to develop sustainable construction methods, it is necessary to incorporate waste materials into the concrete production process. This study evaluates the mechanical, permeability, and microstructural performance of green concrete incorporating waste glass powder (WGP) as a partial cement replacement under elevated temperature conditions. Concrete mixes with 0%, 10%, and 15% WGP were tested for compressive strength, flexural strength, elastic modulus, porosity, sorptivity, and microstructure. Results showed that incorporating 15% WGP increased compressive strength by up to 54% at ambient temperature and retained about 55% of its strength at 600 °C, compared to 40% for the control mix. Porosity and sorptivity were reduced by approximately 20–25%, indicating improved impermeability. The Si/Ca ratio increased by up to 38% as a result of improved pozzolanic reactivity and denser C-S–H formation, which were verified by SEM test. The results demonstrated that WGP not only reduces cement use but also improves durability and thermal resistance. This supports its application in structural concretes that are eco-efficient and fire-resistant. The findings demonstrate WGP’s potential for widespread use in sustainable infrastructure, which might lower CO₂ emissions and increase recycling of waste in the building sector.

Synthetic dye discharge into water sources offers substantial environmental and health problems due to its toxicity, probable carcinogenic effects, and resistance to natural degradation. These qualities make dye removal from wastewater more difficult, and traditional treatment procedures frequently fail. Recent developments have highlighted the potential of silica-based nanoparticles in addressing this issue. Functionalized silica and biogenic silica are two major groups. Biogenic silica generated from agricultural leftovers provides a long-term and cost-effective option, whereas functionalized silica is known for its high surface area and ability to be tuned for specific dye interactions. This paper looks at different dye kinds, removal processes, and the use of silica nanoparticles in wastewater treatment. It also investigates present limitations, environmental concerns, and the need for additional research to improve the usability of these materials in industrial applications.

Graphical Abstract