Engineering topics
Wang, Song
Publications and source records attributed to Wang, Song.
Prediction by Convolutional Neural Networks of CO 2 /N 2 Selectivity in Porous Carbons from N 2 Adsorption Isotherm at 77 K
Abstract Porous carbons are an important class of porous materials with many applications, including gas separation. An N 2 adsorption isotherm at 77 K is the most widely used approach to characterize porosity. Conventionally, textual properties such as surface area and pore volumes are derived from the N 2 adsorption isotherm at 77 K by fitting it to adsorption theory and then correlating it to gas separation performance (uptake and selectivity). Here the N 2 isotherm at 77 K was used directly as input (representing feature descriptors for the porosity) to train convolutional neural networks to predict gas separation performance (using CO 2 /N 2 as a test case) for porous carbons. The porosity space for porous carbons was explored for higher CO 2 /N 2 selectivity. Porous carbons with a bimodal pore‐size distribution of well‐separated mesopores (3–7 nm) and micropores (<2 nm) were found to be most promising. This work will be useful in guiding experimental research of porous carbons with the desired porosity for gas separation and other applications.
Prediction by Convolutional Neural Networks of CO 2 /N 2 Selectivity in Porous Carbons from N 2 Adsorption Isotherm at 77 K
Abstract Porous carbons are an important class of porous materials with many applications, including gas separation. An N 2 adsorption isotherm at 77 K is the most widely used approach to characterize porosity. Conventionally, textual properties such as surface area and pore volumes are derived from the N 2 adsorption isotherm at 77 K by fitting it to adsorption theory and then correlating it to gas separation performance (uptake and selectivity). Here the N 2 isotherm at 77 K was used directly as input (representing feature descriptors for the porosity) to train convolutional neural networks to predict gas separation performance (using CO 2 /N 2 as a test case) for porous carbons. The porosity space for porous carbons was explored for higher CO 2 /N 2 selectivity. Porous carbons with a bimodal pore‐size distribution of well‐separated mesopores (3–7 nm) and micropores (<2 nm) were found to be most promising. This work will be useful in guiding experimental research of porous carbons with the desired porosity for gas separation and other applications.
Preparation, surface functionalization and application of Fe3O4 magnetic nanoparticles
Not Available
Ion-gated carbon molecular sieve gas separation membranes
Membrane technology lies at the heart of many industrial gas separation processes and applications. Molecular sieving membranes that break the Robeson limit are desirable for energy-efficient gas separation. In this paper, we report a facile strategy of directly integrating ionic liquids (ILs) into porous membranes. Particularly, the ILs form an ultra-thin layer on the carbon molecular sieve (CMS) membranes rather than penetrating into the pores, acting as a smart gate for gas entry to boost the selectivity. The hybrid membrane exhibits CO 2 permeability >600 barrer and enhanced CO 2 /N 2 selectivity >50, which surpasses the Robeson limit and shows potential in CO 2 /N 2 separation process. Molecular dynamics simulations confirm the gating effect of the IL layer of molecular thickness. In conclusion, this work demonstrates a universal strategy to improve CMS membrane performance by creating an IL-membrane interface and tuning the ion-pore interaction.
Surpassing Robeson Upper Limit for CO2/N2 Separation with Fluorinated Carbon Molecular Sieve Membranes
Not Available
Conductivity of PEO/PLA Doped Liquid Crystal Ionomer Solid Polymer Electrolyte in Mesomorphic Range
A novel polyethylene oxide (PEO)/polylactic acid (PLA) solid polymer electrolyte (SPE) doped with Liquid crystal ionomer (LCI) was prepared by automatic scraping membrane technology and the liquid crystal property of LCI was proposed as an effective strategy to improve the conductivity. The highest conductivity of the SPE with 0.5 wt% LCI achieve a value of 2.19 × 10{sup −4} S/cm at 17 °C which is approximately four orders of magnitude higher than that of the original PEO solid polymer electrolyte, conductivity at room temperature is 10{sup −7}–10{sup −8} S/cm. This is because the crystallinity is reduced by 25.3% compared to the pure PEO host. Differential Scanning Calorimetry (DSC) and polarizing microscope (POM) characterize the structure and properties of LCI. Electrochemical Impedance Spectroscopy (EIS) shows that the conductivity of the polymer electrolyte with 0.5 wt% LCI increased abruptly at 35 °C as the temperature is higher than T{sub g} of LCI, reaching 6.37 × 10{sup −4} S/cm which is 191% higher than that at 17 °C. There ionic conductivity has been improved extremely by continuous channel for efficient ion transportation, especially at the microtherm. And this change is greater at the elevated temperature regions, in line with the VTF equation. In summary, LCI’s liquid crystal performance can be used as an effective strategy to improve the low temperature conductivity of SPE.
Cation-Exchange Synthesis of Highly Monodisperse PbS Quantum Dots from ZnS Nanorods for Efficient Infrared Solar Cells
Infrared solar cells that utilize low-bandgap colloidal quantum dots (QDs) are promising devices to enhance the utilization of solar energy by expanding the harvested photons of common photovoltaics into the infrared region. However, the present synthesis of PbS QDs cannot produce highly efficient infrared solar cells. Here in this paper, a general synthesis is developed for low-bandgap PbS QDs (0.65-1 eV) via cation exchange from ZnS nanorods (NRs). First, ZnS NRs are converted to superlattices with segregated PbS domains within each rod. Then, sulfur precursors are released via the dissolution of the ZnS NRs during the cation exchange, which promotes size focusing of PbS QDs. PbS QDs synthesized through this new method have the advantages of high monodispersity, ease-of-size control, in situ passivation of chloride, high stability, and a 'clean' surface. Infrared solar cells based on these PbS QDs with different bandgaps are fabricated, using conventional ligand exchange and device structure. All of the devices produced in this manner show excellent performance, showcasing the high quality of the PbS QDs. The highest performance of infrared solar cells is achieved using ≈0.95 eV PbS QDs, exhibiting an efficiency of 10.0% under AM 1.5 solar illumination, a perovskite-filtered efficiency of 4.2%, and a silicon-filtered efficiency of 1.1%.
From Highly Purified Boron Nitride to Boron Nitride-Based Heterostructures: An Inorganic Precursor-Based Strategy
The current approaches used to fabricate hexagonal boron nitrides (h-BN) from boron trioxide and urea always results in contamination of the h-BN product with carbon/oxygen. Thus, discovering a facile way of mass producing high-purity h-BN remains a challenge. A simple yet highly efficient thermal treatment approach to large-scale fabrication of nanoporous h-BN with high yield, high purity, and high crystallinity is described using NaNH 2 and NaBH 4 as the oxygen- and carbon-free precursors. The unique properties of inorganic metal salts, i.e., high melting point and strong electrostatic interaction with carbon substrates, render this strategy suitable for the production of homogeneous h-BN/mesoporous carbon and h-BN/carbon nanotube heterostructures of high crystallinity, high h-BN dispersity, and with a strong interfacial effect. Therefore, these unique features make them promising candidates for supercapacitor applications, resulting a significantly enhanced specific capacitance. This study provides new insight into the fabrication of high-purity h-BN and h-BN-based heterostructures thus expanding their application in the field of energy storage and transformation.
Synthesis and Characterization of Hydrophobic Fe{sub 3}O{sub 4} Magnetic Nanoparticles with High Saturation Magnetization
A simple one-step solvothermal method is proposed to prepare hydrophobic Fe{sub 3}O{sub 4} nanoparticles (MNPs) using iron acetylacetonate (Fe(acac){sub 3}), oleylamine (OAm), and ethylene glycol (EG). X-ray powder diffraction, scanning electron microscope, infrared spectroscopy, transmission electron microscopy, X-ray photoelectron spectrometry, and vibrating sample magnetometer are used to characterize the structure, morphology, and properties of products. The testing results indicate that the as-synthesized products are spherical-like hydrophobic MNPs, superparamagnetic at room temperature, with a saturation magnetization up to 76.8 emu/g. The good hydrophobic property of MNPs is attributed to the coverage of oleylamine, without causing a remarkable loss of magnetic property. In addition, the influences of material ratio, aging time, oleic acid, and stearic acid used as additives are studied in our research.
Influence of fluorination on CO 2 adsorption in materials derived from fluorinated covalent triazine framework precursors
Ultra-nanoporous materials derived from fluorinated covalent triazine frameworks (CTFs) have been developed for highly efficient CO 2 capture. A CO 2 uptake capacity of 6.58 mmol g –1 at 273 K, 1 bar (2.45 mmol g –1 at 0.15 bar) is achieved. The excellent performance is due to the presence of ultra-micropores (0.6–0.7 nm) that tightly fit CO 2 and strong electrostatic interactions from the residual fluorine atoms within the framework. Here, both molecular simulation and deep learning study predict that CTFs with a F content of ~4.8 wt% and pore size distribution around ~0.7 nm can give rise to the highest CO 2 uptake capacity.
Entropic selectivity in air separation via a bilayer nanoporous graphene membrane
Membranes represent an energy-efficient technology for air separation, but it is difficult to control the pore size to separate N 2 and O 2 due to their similar kinetic diameters. In this paper we demonstrate by molecular dynamics simulations that a bilayer nanoporous graphene membrane with continuously tunable pore sizes by the offset between the two graphene layers can achieve O 2 /N 2 selectivity of up to 26 with a permeance of over 10 5 GPU (gas permeation unit). We find that entropic selectivity is the main reason behind the high selectivity via the tumbling movement of the skinnier and shorter O 2 molecules entering and passing through the elliptic-cylinder-shaped nanopores of the bilayer membrane. Such motion is absent in the single-layer graphene membrane with pores of similar size and shape which yields an O 2 /N 2 selectivity of only 6 via molecular sieving alone. Hence the bilayer nanoporous graphene membrane provides a novel way to enhance the entropic selectivity for gas separation by controlling both the pore size and the 3D pore shape.
Construction of a Nanoporous Highly Crystalline Hexagonal Boron Nitride from an Amorphous Precursor for Catalytic Dehydrogenation
Abstract Hexagonal boron nitride (h‐BN) is regarded as a graphene analogue and exhibits important characteristics and vast application potentials. However, discovering a facile method for the preparation of nanoporous crystalline h‐BN nanosheets (h‐BNNS) is still a challenge. Herein, a novel and simple route for the conversion of amorphous h‐BN precursors into highly crystalline h‐BNNS was achieved through a successive dissolution–precipitation/crystallization process in the presence of magnesium. The h‐BNNS has high crystallinity, high porosity with a surface area of 347 m 2 g −1 , high purity, and enhanced thermal stability. Improved catalytic performance of crystalline h‐BNNS was evidenced by its much higher catalytic efficiency in the dehydrogenation of dodecahydro‐N‐ethylcarbazole, compared with its amorphous h‐BN precursor, as well as other precious‐metal‐loaded heterogeneous catalysts.
Construction of a Nanoporous Highly Crystalline Hexagonal Boron Nitride from an Amorphous Precursor for Catalytic Dehydrogenation
Abstract Hexagonal boron nitride (h‐BN) is regarded as a graphene analogue and exhibits important characteristics and vast application potentials. However, discovering a facile method for the preparation of nanoporous crystalline h‐BN nanosheets (h‐BNNS) is still a challenge. Herein, a novel and simple route for the conversion of amorphous h‐BN precursors into highly crystalline h‐BNNS was achieved through a successive dissolution–precipitation/crystallization process in the presence of magnesium. The h‐BNNS has high crystallinity, high porosity with a surface area of 347 m 2 g −1 , high purity, and enhanced thermal stability. Improved catalytic performance of crystalline h‐BNNS was evidenced by its much higher catalytic efficiency in the dehydrogenation of dodecahydro‐N‐ethylcarbazole, compared with its amorphous h‐BN precursor, as well as other precious‐metal‐loaded heterogeneous catalysts.
The Role of a Curved Electrode with Controllable Direction in the Radiofrequency Ablation of Liver Tumors Behind Large Vessels
Purpose: To investigate the role of a novel curved radiofrequency ablation (RFA) electrode with controllable direction in the ablation of tumors behind large hepatic vessels in ex vivo bovine and in vivo canine liver experiments. Materials and Methods: Approval from the institutional animal care and use committee was obtained. In ex vivo experiments, conventional multi-tines expandable electrodes, conventional monopolar straight electrodes and novel curved electrodes were used in the ablation of the bovine liver (n = 90). The ablated area, parallel axis, vertical axis and shape of different electrodes were compared. Then, 24 beagle dogs (10 months old, female) were used for in vivo experiments. Visual tumor targets deeply located in the portal vein were established, and ultrasound-guided liver ablation was performed with different electrodes. The ablation range, target coverage rate, percentage of normal tissue injury and damage to adjacent vessels were evaluated. The Kruskal–Wallis test and the Chi-squared test were used for statistical analysis. Results: For the ex vivo study with a 3-cm electrode, the ablation area of the multi-tines expandable electrode group (7.14 ± 0.16 cm{sup 2}) was significantly larger than that of the novel curved electrode group (5.01 ± 0.30 cm{sup 2}, P < 0.001) and the monopolar straight electrode group (5.43 ± 0.15 cm{sup 2}, P < 0.001). The results obtained with the 4-cm electrode in the three groups were in accordance with those of the 3-cm electrode. In vivo, the normal tissue damage area of the novel curved electrode group was smaller than that of the multi-tines expandable electrode group (1.10 ± 0.18 cm{sup 2} vs. 4.00 ± 0.18 cm{sup 2}, P < 0.001). The target coverage rate of the novel curved electrode group was better than that of the monopolar straight electrode group (100% vs. 80.86 ± 1.68%, P < 0.001). The hematoxylin and eosin (H&E) and TUNEL staining results showed that the ablation necrosis area was adjacent to large vessels, but the vascular wall was not significantly damaged in the novel curved electrode group. Conclusion: Our preliminary results showed that the novel curved RFA electrode with controllable direction could achieve accurate ablation for tumors behind large hepatic vessels, with a better target coverage rate and less damage to normal tissue, than conventional multi-tines expandable electrodes and monopolar straight electrodes.
Data-driven Feature Analysis in Control Design for Series-Compensated Transmission Systems
One challenge in power-system control designs is the gap between numerical model-based analysis and complex real-world power systems. With increased data and measurements being collected from power systems, data-driven analysis (e.g., machine learning) may provide an alternative approach to reveal hidden information through learning from the real system data, and provide insights for better control scheme design during the utility planning process. Data-driven feature analysis was performed to evaluate the relationships between series compensation, power generation, and path flows in a real transmission system, as well as temporal patterns. The main data-driven analysis methods, including statistical cross-correlation, multinomial logistical regression, and classification and regression trees, were integrated for feature selection and developing predictive models of series compensation. Analysis results demonstrated the effectiveness of the proposed methodology in feature analysis and the potential to help improve power-system control scheme design.
Domain Adaptation for Convolutional Neural Networks Based Remote Sensing Scene Classification
Remote sensing (RS) scene classification plays an important role in the field of earth observation. With the rapid development of the remote sensing techniques, a large number of RS scene images are available. Since manually labeling largescale RS scene images is both labor and time consuming, when a new unlabeled dataset is obtained, how to use the existing labeled datasets to classify the new unlabeled images is an important research direction. Different RS scene image datasets may be taken from different type of sensors, and the images may vary from imaging modalities, spatial resolutions and image scales, so the distribution discrepancy exists among different image datasets. As a result, simply applying Convolutional Neural Networks (CNN) trained on source domain cannot accurately classify the images on target domain. Domain Adaptation (DA) can be helpful to solve this problem. In this paper, we design a Subspace Alignment (SA) and CNN based framework to solve the DA problem in RS scene image classification. A new SA layer is proposed and added into CNN models for DA, which could align the source and target domains in some feature subspace. Finetuning the modified CNN model with the added SA layer is able to make the CNN model adapt to the aligned feature subspace and helps to relieve the domain distribution discrepancy. The experiments conducted on two public datasets show that adding the SA layer into CNN can improve the scene classification on target domain.