Abstract: With the consumption of fossil fuels and the deterioration of the ecological environment, the need for developing new, efficient, and sustainable sources of clean energy is urgent. The importance of “green hydrogen” in electrolytic water splitting has attracted worldwide attention not only from the scientific community but also from governments and industries. Hydrogen energy is considered an ideal alternative to fossil fuels because of its high energy density, environmental friendliness, and low pollution level. Hydrogen production from renewable energy sources using the electrolysis of water is the lowest carbon emission process of the many current hydrogen source options. The electrolytic water reaction is subdivided into two half-reactions, namely, the hydrogen evolution reaction (HER) at the cathode and the oxygen evolution reaction (OER) at the anode. The HER is a relatively simple two-electron reaction. Compared to the HER at the cathode, the OER at the anode is a four-electron transfer process with slower kinetics and higher energy barriers. It is the decisive step in the electrolytic water reaction, receiving considerable attention from scholars. Recently, considerable developments in the research of high-performance electrolytic water catalysts have been reported as successful; however, the catalysts have been tested on a very small scale, usually under laboratory conditions, and can rarely operate continuously for hundreds of hours, far from meeting the needs of practical applications. Industrial-level electrocatalytic hydrogen production requires catalysts that are highly active, cost-effective, and stable at high current densities; thus, a great deal of work has explored efficient and highly durable active electrocatalysts to overcome the kinetic barriers that inhibit the reaction, particularly for the complex four-electron reaction of the OER. In summary, catalysts for oxygen precipitation reactions at high current densities will be the focus of future research. This paper reviews the current status of hydrogen energy development and various hydrogen production methods at home and abroad, focusing on an analysis of electrolytic water hydrogen production technology and proposing the requirements under large-scale industrial applications. Studying the OER mechanism has revealed that the activity of catalysts at high current densities can be enhanced by the following strategies: heteroatom doping, defect engineering, interface engineering, in situ self-growth, etc. Finally, the challenges in the field of high-current oxygen analysis at this stage of industrial development and the future direction of development are presented.
Abstract: As concerns over environmental contamination and rapid consumption of fossil fuels continue to grow, it is important for energy storage technology to reduce the intermittency of clean and renewable energy sources. So far, lithium-ion batteries (LIBs), commercialized by SONY corporation in 1991, have been the most widely used rechargeable batteries for various energy storage devices. Due to the ever-increasing demand for lithium employment in mobile electronic devices and electric vehicles (EVs), the price of Li resources is rising year by year. It is well known that worthwhile lithium resources are only found in a few countries (mainly in South America). Recently, sodium-ion batteries (SIBs) have been regarded as promising alternatives to LIBs for future large-scale energy storage systems (ESSs) owing to their low cost, abundant reservoirs of Na resources and similar characteristics to LIBs. Developing high-performance cathode materials is crucial to realize the commercialization of the SIB technology. Sodium transition metal oxides (NaxTMO2), especially for Ni–Mn-based compounds, have received significant attention thanks to their high specific capacity and operating voltage. Normally, layered NaxTMO2 materials have two types of crystal structures: P2 and O3, according to the surrounding Na environment and the number of unique oxygen layers occupied within the lattice. Compared with the O3 phase, the P2-type structure has open diffusion channels for the transport of Na+ and relatively rare phase transitions, which make P2-type Na0.67[Ni, Mn]O2 (NNMO) one of the most promising cathodes for SIBs. However, NNMO materials generally suffer from irreversible P2–O2 phase transformations, Na+/vacancy ordering transitions and Jahn–Teller distortion of Mn(III)O6 octahedra, leading to structural deterioration and performance degradation during the charge and discharge processes. In detail, the P2–O2 phase transition inevitably causes significant lattice volume change (~20%) and even the formation of cracks, resulting in the stripping of active substances from the collector and serious capacity decay during cycling. The Na+/vacancy ordering in NNMO causes the multi-step two-phase reactions, which may increase the activation energy barrier for Na+ hops between adjacent prismatic sites, consequently hindering Na+ diffusion. Additionally, the lattice distortion and P2-P’2 phase transition induced by the Jahn–Teller effect also impede Na+ migration, leading to the sluggish kinetics of Na+ (de)intercalation. In this review, the recent progress on NNMO cathodes is summarized, including ion-doping, surface modification and composite structure. The comprehensive and integrated explanation of the structure–function–performance relationship of these optimized cathodes is further presented. Moreover, the existing challenges of NNMO and possible remedies are also discussed. It is expected that this review can provide new insights into the commercialization of NNMO for SIBs.
Abstract: Under the background of peak carbon dioxide emissions and achieving carbon neutrality, clean energy technologies such as water electrolysis, metal–air batteries, and fuel cells have attracted extensive attention due to the advantages of high efficiency, good safety, a simple structure, low cost, and eco-friendliness. However, the key reactions on oxygen catalytic electrodes, the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER), are kinetically sluggish, which considerably hinders their commercial applications. The traditional oxygen catalytic electrodes with the use of binders have the disadvantages of a cumbersome synthesis process, low controllability, poor uniformity, high cost, and easy carrier catalyst agglomeration, which limit their catalytic performance. Recently, self-supported oxygen catalytic electrodes have attracted extensive attention due to their advantages of high catalytic active sites and a stabilized spatial framework, which can solve the problems faced by traditional oxygen catalytic electrodes and further improve the catalytic performance of the electrode. As the catalyst material carrier, the substrate materials play an important role in the catalytic performance of self-supporting oxygen electrodes. The high interaction forces between the substrate and the catalyst material lead to a single-direction growth orientation and uniform dispersion. Reportedly, the substrate materials for self-supporting oxygen catalytic electrodes have not been fully discussed in comprehensive reviews. Therefore, timely updates in this potential field must be provided. This paper summarizes the research progress and synthesis methods of commonly self-supporting oxygen catalytic electrodes based on different substrate materials, including two- and three-dimensional metal materials and carbon materials. In addition, this paper introduces the outstanding ORR/OER catalytic properties of common self-supporting oxygen catalytic electrodes, which are not only due to the intrinsic catalytic activity of the supported catalytic active materials but also related to the high specific surface area and high electron transfer rate caused by the structure of the self-supported electrode substrate. Finally, the future research and the development trend of self-supporting oxygen catalytic electrodes are addressed from the four aspects of density general function theory, improving electrode energy density, constructing an efficient gas–liquid–solid three-phase interface of an electrode, and establishing a standard evaluation protocol of self-supported oxygen catalytic electrodes. This review should provide new research insights for developing renewable energy storage and conversion systems.
Abstract: Developing new energy, reducing fossil energy consumption, and building a green and low-carbon energy system are important strategies to achieve carbon neutrality. To realize the grid connection of new energy generation, rechargeable potassium-ion batteries (KIBs) are expected to be used in large-scale energy storage, considering sufficient potassium resources and potential high-energy density. Anode, an essential battery component, directly determines the battery safety, cycle life, and energy density. Among various anodes, alloys can provide high theoretical specific capacity based on the multi-electron reaction mechanism, which is promising in terms of improving the energy density of a full battery. Their K-storage voltages also stay away from the deposition/stripping potentials of metallic K, thereby enhancing battery safety. However, the dramatic volume variation upon alloying and dealloying leads to electrode pulverization and capacity fading in traditional carbonate-based electrolytes. An effective method for stabilizing an alloy-based anode structure is the construction of a stable electrode–electrolyte interface by electrolyte optimization, which has the advantages of a simple process and low cost. Accordingly, the utilization of interfacial engineering to achieve stable alloy anodes has been frequently reported in the past few years. This topic includes the following: (1) regulation of the components of solid electrolyte interphase (SEI) layers to improve mechanical strength and ionic conductivity, buffer volume fluctuation, and reduce electrode corrosion; (2) adjustment of the solvated structure of K+ to enhance the diffusion rate and inhibit the electrolyte decomposition; and (3) utilization of solvent molecular chemisorption on the electrode to induce its microstructure change, improve the electrolyte wetting ability, and relieve volume change. The SEI is a passivation film generated on the electrode surface at the initial battery operation stage. Research on its structure, component, and formation mechanism is still basic due to its instability and complexity and limited research methods. Whether the solvated structure of cation and electrolyte adsorption affect the SEI structure and composition remains unclear. How the solvated structure and the electrolyte adsorption improve the electrode stability must also be further studied. This review covers recent research progress on the interfacial interaction between alloy anodes and electrolytes in KIBs, summarizes the electrolyte optimization strategies, analyzes the potassium storage mechanisms and electrochemical performance of alloy anodes, and highlights the interfacial interaction mechanisms. More importantly, this paper provides new insights for the future development of KIB electrolytes.
Abstract: Sodium is considered an ideal anode material for high-energy batteries because of its low cost, high natural abundance, low redox potential (?2.71 V vs SHE), and high theoretical specific capacity (1166 mA·h·g?1). However, due to the high reactivity, sodium rapidly reacts with the electrolyte to form an unstable solid electrolyte interface (SEI) layer during stripping/plating cycling. In addition, due to the large size change of sodium, the SEI layer repeatedly breaks and reassembles, resulting in the continuous consumption of sodium and electrolyte, as well as low coulombic efficiency and rapid capacity loss. Simultaneously, due to an uneven electric field distribution on sodium, numerous sodium dendrites generate during the repeated plating/stripping cycles. The growing Na dendrites easily pierce the separator, causing a short circuit and a series of safety issues. The above issues lead to the deterioration of battery performance and safety risks, thus considerably hindering the application of sodium metal batteries. Various studies have been conducted to solve these issues, including electrolyte engineering, artificial SEI layers, current collector and interlayer engineering, solid-state electrolyte engineering, and three-dimensional (3D) frameworks for sodium metal. Among various improvement strategies, the construction of a 3D conductive framework can effectively reduce the local current density, decrease nuclear energy, inhibit Na dendrite growth, and impede volume expansion, thus having a great potential in future applications. In this study, the current research progress in using various 3D conductive frameworks to improve the cycling stability of a sodium metal battery is reviewed, including carbon-based, metal-based, and MXene-based frameworks. Simultaneously, the pros and cons of different 3D conductive framework technologies in recent years are summarized and classified, and the electrochemical performance parameters of different 3D conductive frameworks for sodium metal batteries are compared. Finally, the development prospect and direction of 3D conductive frameworks in sodium metal anodes are discussed from basic research and practical applications. This review provides deeper insights into building more comprehensive and efficient sodium metal anodes. The 3D conductive framework technology can remarkably improve the cycle life and safety of a sodium metal battery. Multistrategy joint research methods will facilitate the practical applications of a sodium metal battery. Further exploration of the deposition behavior of sodium metal is required in the future, and we believe that it can definitely achieve commercial applications with continuous efforts.
Abstract: Development and utilization of renewable energy sources have gain great progress in recent years, which lead to increasing demands for large scale energy storage systems. Lithium-ion batteries have been widely used in portable electronic devices and electric vehicles. However, with the exploitation of the Earth’s lithium resources, the cost of lithium-ion batteries is gradually increasing. In contrast, the higher terrestrial potassium content promises inexpensive potassium-ion batteries, and the chemical properties of potassium and lithium ions are similar. Meanwhile, the low redox potential of K promises a high working voltage of potassium ion batteries. Thus, potassium-ion batteries have attracted considerable attention as a capable battery technology. However, the large radius of the potassium ion leads to unsatisfactory ion intercalation and extraction behavior during charging and discharging processes, resulting in poor cycling performance, unsatisfactory rate ability, and low capacity. The challenge remains to explore capable electrode materials for potassium-ion batteries to achieve a high energy density and power density. This review summarizes the anode and cathode materials of potassium-ion batteries in recent reports, including the research progress of graphite and other carbon materials, transition metal oxides/sulfides, alloys, and other anode materials, as well as Prussian blue, layered metal oxides, and polyanionic compound cathode materials, which will provide new ideas for developing high-performance potassium-ion batteries. We also discuss the potassium ion storage mechanism in these electrode materials. This review also demonstrates the approaches (nanotechnology, heteroatom doping, carbon coating, composite fabrication) to further improve the electrochemical performance of the cathode and anode. In addition, we point out the key factors for potassium ion batteries performance, such as the design of anode materials, exploitation of novel cathode materials, and optimization of full potassium ion cells fabrication, which would provide new thought for the development of potassium ion batteries with high performances.
Abstract: In the context of rapid social development, it is urgent to address the increasingly prominent issues of fossil energy depletion and environmental pollution. As a result, research has focused on the creation of new clean energy sources such as solar, wind, biological, geothermal, and hydrogen. Hydrogen energy is one of these new energy sources that has drawn a lot of attention because of its low weight, good thermal conductivity, high heating value, rich utilization forms, and diverse storage states. Nowadays, one of the most significant methods for producing clean energy is photoelectrochemical (PEC) water splitting for hydrogen. However, the intrinsic drawbacks of commonly used semiconductors, such as the low light absorption efficiency, high carrier recombination rate, and slow oxygen evolution kinetics, have become the main barriers preventing their advancement in PEC water splitting. This study used anodic oxidation to create N-doped 4H-SiC nanowire arrays (NWAs) from N-doped 4H-SiC single crystalline wafers. It can be verified that the highly oriented N-doped 4H-SiC NWAs are fully exposed by removing the cap layer. Additionally, the single bamboo-shaped nanowire that was produced has a diameter of ~30–50 nm. Focusing on the optimization of the oxygen evolution reaction (OER) conditions, the NWAs were used as an integrated photoanode in a typical three-electrode system to achieve effective PEC water splitting for hydrogen production under illumination and electric field. Notably, the N-doped 4H-SiC NWAs show better water splitting performance compared with the bulk; that is, the onset potential is decreased from 1.224 V to ?0.021 V versus the Ag/AgCl electrode, and the current density is increased from 2.64 mA?cm?2 to 3.61 mA?cm?2 at 1.4 V. Particularly, the N-doped 4H-SiC NWAs exhibit an extremely sensitive response to light. The improved optical absorption capacity and efficient charge transfer of N-doped 4H-SiC NWAs are responsible for the improvement in PEC water splitting performance. On the one hand, when the N-doped 4H-SiC NWAs are exposed to light, a significant amount of light shines into the gap between the nanowires. N-doped 4H-SiC obtains additional light absorption pathways with the constant reflection of the light, significantly enhancing the light absorption efficiency. On the other hand, the NWAs can considerably reduce the hole travel distance and avoid the recombination of the photogenerated electron-hole pairs, making more charges participate in the redox reaction to enhance the PEC water splitting performance of N-doped 4H-SiC. By building semiconductor photoanode nanostructures, it is possible to efficiently absorb light and transfer charge, significantly enhancing the PEC water splitting efficiency.
Abstract: Because of the global fossil energy crisis and environmental pollution problems, the efficient use of green, renewable, and clean energy has become a major trend. Mechanical energy is considered an ideal alternative energy source because of its abundance, accessibility, and non-polluting characteristics. A piezoelectric nanogenerator (PENG) can convert environmental mechanical energy into electrical energy to power electronic devices. However, conventional piezoelectric materials must induce dipole alignment by electrical polarization to obtain piezoelectric properties, which substantially increases the cost and energy consumption of device preparation. Meanwhile, depolarization occurs when the external electric field is removed, which severely affects the performance of the piezoelectric material. In this study, PVDF nanofiber film is prepared using the electrospinning method. The PVDF dipole is rearranged to achieve in situ polarization by a strong electric field and stretching force generated by the electrospinning process. The PVDF nanofiber film process has a high electroactive β-phase content of 78.7%, which is the main contributor to the piezoelectric properties. The PENG constructed based on this film achieves direct conversion of mechanical energy to electrical energy, greatly improving energy use. The open-circuit output voltage of the thin film PENG prepared based on the electrostatic spinning method is 1.6 V, and the short-circuit output current is 0.14 μA, which are 4.5- and 2.6-fold higher than those prepared using the spin-coating method, respectively. The PVDF–PENG can charge a 1-μF capacitor to 2 V through a bridge rectifier after 60 s of human finger tapping. The power density of the PVDF–PENG is analyzed by measuring the electrical parameters at both ends of the resistor. The maximum output power is 0.03 μW at an applied load of 200 MΩ. More electrical energy can be obtained based on the PVDF–PENG, which further illustrates its possibilities and reliability in practical applications. Further, the PVDF–PENG maintains approximately 100% output capacity after 2000 consecutive cycles of compression, verifying its long-term stable service capability. Finally, the energy collected from the mechanical energy of human motion by the PVDF–PENG is explored to drive low-power consumer electronics. Six commercial LEDs are lit by using a large PENG without using any storage device. In addition, a bridge rectifier is used to charge a 2.2-μF capacitor, which successfully lights up a commercial electronic watch.
Abstract: Organic carbonyl compounds have received great attention as electrode materials because of their fast reduction–oxidation kinetics, environment friendliness, and high theoretical capacity. Especially, the small molecular quinones, such as anthraquinone (AQ), can possess high theoretical (257 mA·h·g?1) and a discharge–charge voltage of 2.2–2.3 V, implying that it has the potential of up to 565 W·h·kg?1 energy density. However, it suffers from high solubility in organic electrolytes and low conductivity, leading to rapid capacity fading and inferior rate performance. Herein, we report 2,6-diaminoanthraquinone (2,6-AAQ) uniform self-assembly into a three-dimensional (3D) porous structure graphene foam, which was successfully fabricated through a gentle hydrothermal synthesis reaction with simultaneous in situ condensation of 2,6-AAQ on the reduced graphene surface, as a high-performance cathode for Lithium-organic batteries. Benefiting from the formation of a covalent bond (—CO—NH—) between the amino group (—NH2) of 2,6-AAQ and the carboxyl group (—COOH) of oxidized graphene, the molecular structure of AQ is uniformly anchored into a 3D graphene foam architecture. The strategy simultaneously solved the high dissolution and low conductivity of AQ. The as-obtained hybrid composites were characterized by various techniques. SEM and EDS mapping images demonstrated that the 2,6-AAQ within the hybrid architecture was not only uniformly anchored on the surface but also tightly wrapped in the interior of graphene foam. This unique architectural structure can improve the electronic conductivity of 2,6-AAQ in the lithiation process and effectively inhibit the dissolution of 2,6-AAQ in electrolytes, which is beneficial to hoist the electrochemical performance of the composite materials. XPS, XRD, FTIR, and Raman results indicated that hydrothermally assisted chemical bonding occurred between 2,6-AAQ and rGO, significantly facilitating the mass electron transformation and ion diffusion from graphene substrate to 2,6-AAQ for the fast reduction–oxidation reaction. Combined with the above results, UV–Vis spectroscopy tests also further disclosed that the 2,6-AAQ and rGO linked by covalent bonds significantly decrease solubility compared with 2,6-AAQ, indicating the greatly increased cycling stability of the hybrid material. Additionally, ex situ FTIR characterization results verified that the composite cathode material with two carbonyls (C=O) active sites has good lithium storage performance. By optimizing the 2,6-AAQ concentration, the 25% 2,6-AAQ in the as-prepared composite was used as the high-performance cathode for the lithium-ion battery. The composite material can display a high initial discharge capacity of 212.2 mA·h·g?1 at 100 mA·g?1 (based on the 2,6-AAQ mass) and a reversible capacity of 184 mA·h·g?1 with a capacity retention of 86.7% after 100 cycles at 500 mA·g?1 current density. This excellent electrochemical performance is attributed to fast lithium-ion diffusion and electric transport between the 2,6-AAQ and the 3D porous structure hybrid architecture, which also proposes a facile strategy for the immobilization of the small molecular quinones to construct advanced organic lithium batteries.
Abstract: With the increasing shortage of petroleum resources and serious environmental pollution, the demand for green technology development is growing stronger. Electrical energy storage is an excellent way to store intermittent clean energy and transport clean energy from one place to another. The lithium-ion battery (LIB) is broadly recognized as the first choice for electrical energy storage due to its high energy density, especially in smart electronics and electric cars. Nevertheless, the application of LIB in large-scale energy storage has been limited by various factors, including the limited and uneven distribution of lithium resources, safety issues and toxic organic electrolytes. The aqueous zinc-ion battery (AZIB) has been regarded as a potential substitute for LIB in large-scale energy storage devices because of the competitive theoretical volumetric capacity (5855 mA·h·cm?3) and gravimetric capacity (820 mA·h·g?1) of the Zn anode, the low electrochemical potential of Zn2+ (?0.76 V vs SHE), and the high ionic conductivity of the aqueous electrolyte, the ease of manufacturing (e.g., manufacture in an open-air environment), and the merits of rich resources, low cost and high safety. Finding a cathode material with high energy density and power density is proposed as a strategy to accelerate the progress of AZIB because the cathode material largely dominates the electrochemical properties and the cost of the battery. However, the strong electrostatic interaction between Zn2+ and the host material results in sluggish reaction kinetics, leading to inferior cycling performance and rate property. Some cathode materials are dissolved in aqueous electrolytes, which restrict the development of AZIB. In comparison to the reported AZIB cathodes, including vanadium-based materials, manganese-based materials, Prussian blue analogs, and organic materials, vanadium phosphates have received a lot of attention as cathodes due to their stable structures, high voltage plateaus, and high power densities. This review presents an overview of various vanadium phosphates such as Li3V2(PO4)3, Na3V2(PO4)3, VOPO4, Na3V2(PO4)2F3, NaVPO4F and their derivatives that are applied as cathodes for AZIB. The summary includes their phase structures, synthetic methods, electrochemical performance, electrochemical Zn2+ storage mechanisms and existing problems. The two major challenges in using vanadium phosphates as cathode materials for AZIB are low electronic conductivity and material dissolution problems, both of which result in inferior cycling performance and rate capacity. The resolution strategies for the mentioned challenges include designing the nanostructure, adjusting the electronic structure, coating with conductive materials, and regulating electrolytes to enhance electrochemical properties. Experimental techniques for studying electrochemical mechanisms are also proposed. Finally, the prospects for the future development of these cathodes in AZIB are advanced. It can be expected that this review has some significance for the development of new vanadium phosphates as cathode materials.
Abstract: In the steel manufacturing process, an accurate prediction of end sulfur content in KR is crucial for steadily controlling sulfur content in molten iron and improving steel properties. Regarding the end sulfur content prediction in the KR process, an integrated modeling method based on Kmeans clustering analysis and the BP neural network (BPNN) is proposed in this paper. As an unsupervised learning method, Kmeans clustering analysis can complete data classification according to the similarity of influencing factors instead of depending on target values. The BPNN, as a supervised learning method, can effectively explore the correlation between influencing factors and target values. The integration of these two methods can realize information exploration of data from different dimensions. Based on this understanding and the actual production data in one steel plant, the prediction model of end sulfur content in KR based on Kmeans–BPNN is studied. First, datasets of different operating conditions are constructed according to the pattern recognition and classification of production data in the KR process through Kmeans clustering. By establishing the relation curve between the number of clustering centers and the mean error of clustering results and selecting the adjacent positions to 10% of the maximum mean error difference, the number of Kmeans clustering centers is confirmed as five. Then, the prediction model is trained by different datasets based on the BPNN. The input layer and hidden layer have five nodes, and the output layer has one node in the BPNN-based prediction model of end sulfur content in KR. A piecewise linear function is selected as the activation function, and the maximum number of training is fixed at 1,000. Finally, the prediction models of different datasets are integrated and formulated in the final prediction model of end sulfur content in molten iron, realizing the prediction of different molten iron conditions and operating conditions. To test and verify the effectiveness and accuracy of the prediction model based on the Kmeans–BPNN method, the end sulfur content prediction of molten iron in KR is performed by applying prediction models based on desulfurization reaction kinetics, routine BPNN, and Kmeans–BPNN using the same training and testing datasets. The prediction results indicate that the end sulfur content prediction in KR based on the Kmeans–BPNN method is significantly more accurate than that of the prediction model based on the desulfurization reaction kinetics and the routine BPNN model.
Abstract: As a short-process hydrometallurgical technology, slurry electrolysis (SE) collects the stirring that improves the suspension of ore, the membrane bag that acts as purifying, and the cathodic and anodic plates that promote ion migration in one tank. The stirring helps to maintain the ore suspended. As the SE tank is stirred, the membrane bag will deform and become damaged, severely limiting production efficiency. In this research, the one-way fluid-structure interaction (FSI) was used to examine the impact of the solid–liquid suspension on membrane deformation, which was based on the computational fluid dynamics (CFD) and solid finite element method (FEM). Through the full 3D quantitative analysis, the database of membrane deformation under various conditions was established. The membrane was extruded to the center during the initial stirring conditions, and the greatest deformation measured 891.66 mm. Primarily, membrane deformation was brought on by the pressure differential brought on by liquid velocity, solid concentration distribution, and liquid level. The maximum deformation of the membrane first decreased and then increased with the increased liquid level difference between the cathode and anode. With the upper fixed constraint, the maximum deformation of the membrane appears at y = 1.2 m. The larger the stirring speed is, the smaller the optimal liquid level difference required to minimize the membrane deformation. The stirring speed changes the overall pressure distribution by changing the dynamic pressure in the anode domain. The maximum deformation of the membrane decreases first and then increases with the increase of electrolyte density in the cathode domain. The membrane bag is extruded to the cathode domain when the pressure in the cathode region is insufficient because of the low electrolyte density in the cathode domain. When the cathode pressure increases, the membrane bag bulges to both sides, and the inner bulge is greater than the outer. With an increase in solid volume concentration (SL) in the anode domain, the maximum membrane deformation first reduces and subsequently increases. When SL = 15%, the membrane deformation reaches the minimum value of 226.7 mm. The closer to the bottom of the tank, the greater the influence of solid content on absolute pressure. The maximum membrane deformation is drastically decreased to 0.664 mm when the frame restrictions are considered. It can support the industrial control process via visual analysis.
Abstract: To determine the dynamic matching of a mine ventilation system to onsite demands of automatic adjustment, we analyze the principle of air volume supply and demand matching and a linkage control method. Subsequently, we establish a mathematical model of main ventilator frequency adjustment, associate branch resistance adjustment, and joint adjustment with multi-feature fusion. We also propose a matching regulation model and a stability determination method for a ventilation network’s branch supply and demand. Based on the monitoring of harmful gases, intelligent emergency control software is developed by a mine ventilation supply and demand model. We realize the automatic calculation of the best working frequency of a ventilator when an unbalanced supply and ventilation demand is selected for frequency conversion adjustment. When selecting the associated branch wind resistance adjustment, we use a cellular automata model to calculate the optimal adjustment roadway. We obtain the adjusted wind resistance value using a winding network inversion calculation model. When a single adjustment method fails, a joint control scheme of fan frequency conversion and branch resistance adjustment is generated. A reliable adjustment of air volume supply and demand matching is realized through an advanced simulation analysis of the air network. A typical mine ventilation system is used to establish an experimental model for the automatic adjustment of the air demand of a branch of a winding network. The air demand adjustment and dilution experiment are carried out with the statistical law of onsite gas overrun as the guidance model of branch air demand control. The following results are obtained. The branch air volume changes according to the adjustment theory model under three adjustment methods. Further, the CO2 concentration change is evidently delayed in the air adjustment process. In the process of fan frequency conversion regulation, the air volume of each branch of the air network changes according to the ventilation network sensitivity, and the fluctuation of the air network is minimal. When a single associated branch resistance adjustment method is used to regulate the wind, the local wind network has great influence on air volume and thus fluctuates greatly. When the fan frequency and associated branch wind resistance are combined, the fluctuation of the branch air volume of the entire air network is the largest, and the system stability and security are the lowest. Therefore, the fan frequency and combined regulation methods of multiple associated branches are recommended to use in practical applications of mines. The experiment verified the practicability and feasibility of the deviation of mine ventilation supply and demand from intelligent control systems. It also provided theoretical and application guidance for mine ventilation linkage control.
Abstract: Material data are prepared in batches and stages, and data distribution in different batches varies. However, the average accuracy of neural networks declines when learning material data by batch, resulting in great challenges to the application of artificial intelligence in the materials field. Therefore, an incremental learning framework based on parameter penalty and experience replay was applied to learn streaming data. The average accuracy decline is due to two reasons: sudden variations of model parameters and a quite homogeneous sample feature space. By analyzing the model parameter variation, a mechanism of parameter penalty was established to limit the phenomenon of model parameters fitting toward new data when the model learns new data. The penalty strength of the parameters can be dynamically adjusted according to the speed of parameter change. The faster the speed of parameter changes, the higher the penalty strength, and vice versa, the lower the penalty strength. To enhance sample diversity, experience replay methods were proposed, which train the new and old data obtained by sampling from the cache pool. At the end of each incremental task, the incremental data were sampled and used for the update of the cache pool. Specifically, random sampling was adopted for the joint training, whereas reservoir sampling was used for the update of the cache pool. Further, the proposed methods (i.e., experience replay and parameter penalty) were applied to the material absorption coefficient regression and image classification tasks, respectively. The experimental results indicate that experience replay was more effective than parameter penalty, but the best results were obtained when both methods were used. Specifically, when both methods were used, the average accuracy of the benchmark increased by 45.93% and 2.62% and reduced the average forgetting rate by 86.60% and 67.20%, respectively. A comparison with existing methods reveals that our approach is more competitive. Additionally, the effects of specific parameters on the average accuracy were analyzed for both methods. The results indicate that the average accuracy increases with the proportion of experience replay and increases and then decreases when the penalty factor increases. In general, our approach is not limited by data modalities and learning tasks and can perform incremental learning on tabular or image data, regression, or classification tasks. Further, owing to the quite flexible parameter settings, it can be adapted to different environments and tasks.
Abstract: To date, artificial intelligence has been successfully applied in various fields of material science, but these applications require a large amount of high-quality data. In practical applications, many unlabeled data points but few labeled data points can be obtained directly. The reason is that data annotations require fine and expensive experiments, and the cost of time and money cannot be ignored. Active learning can select a few high-quality samples from many unlabeled data points for labeling and use as little labeling cost as possible to optimize task model performance. However, active learning methods suitable for material attribute regression are poorly understood, and the general active learning method cannot easily avoid the negative effects of noise data, resulting in decreased costs. Therefore, we propose a new active regression learning method that includes the following features: (1) outlier detection module: using the labeled data prediction from a task model trained to fit and the labeled dataset to train the auxiliary classification model for classifying outliers and then excluding the samples that are most likely to be outliers in the unlabeled dataset; (2) greedy sampling: an iterative method is adopted to select the data farthest from the data in the labeled dataset and the selected data in the geometric space to fully consider sample diversity; and (3) minimum change sampling: selecting the unlabeled data with minimum change before and after the task model, which is trained on the labeled dataset. This part of the data is relatively lacking in the feature space of the labeled dataset. We performed experiments on the concrete slump test dataset and the negative coefficient of thermal expansion dataset and compared our method with the latest active regression learning methods. The results show that other methods do not necessarily improve task model performance after labeling data in each active learning circle on noisy datasets, and the final performance cannot reach the level of the task model trained by all data. Under the same amount of data, the performance index of the task model trained by our method is improved by 15% on average compared with other models. Because of the addition of an outlier detection mechanism, our method can effectively avoid sampling outliers when selecting high-quality samples. The task model trained using only 30%–40% of the data can achieve or even exceed the accuracy of the task model trained by all data.
Abstract: Preparation of cerium oxide by conventional liquid phase method has the disadvantages of complex technological process and effluent discharge. Spray pyrolysis for making CeO2 has the disadvantages of nozzle plugging, and this traditional heating method produces a significant temperature gradient that results in unevenly heated reactants. To prevent the above issues, this study proposed an effective and environmental experimental scheme. Cerium chloride heptahydrate and deionized water were utilized for the raw material. High-purity nano cerium oxide particles were prepared by jet-flow pyrolysis technology via microwave heating. Combining the technology of microwave heating with jet-flow pyrolysis, whose Venturi reactor served as the primary piece of equipment, can improve the mixing of gas and liquid, increase chemical reaction efficiency, and reduce carbon emissions. It was a new effort in the area of pyrolysis. To visually analyze the distribution of each physical field and substance, numerical simulation was combined with x-ray diffraction, scanning electron microscopy, and energy dispersive spectroscopy to characterize the products. Effects of the various technological conditions (pyrolysis temperature, gas velocity, and adding citric acid) on the content of residual chloride element and microstructure of the product were examined. Results demonstrated that the temperature error between the experiment and simulation was below 20 ℃ with the condition of the same microwave power. When the pyrolysis temperature was at 500 ℃, CeO2 could be produced in a single phase, but the particle profile was unclear. The particles had sharp profiles when the temperature was 600 ℃. Nanoscale spherical CeO2 particles appeared when the average temperature reached 700 ℃. The results of the study’s simulations and experiments indicated that higher temperatures were associated with more regular microcosmic morphology and a lower content of residual chloride element. Increasing the gas velocity caused an obvious decrease in the average temperature, which led to more content of residual chloride elements. However, the gas collided with the solution more fiercely, which improved the mixing of the two phases. Experimental and simulated results showed that when gas velocity reached 1.2 m?s?1, better dispersity and less agglomeration of the product were obtained. Additionally, the residual chloride content was less than 1%. Because a significant amount of CO2 was produced during the burning of the citric acid, the spherical cerium oxide particles broke into irregular particles. Porous structures also appeared when citric acid was added. The residual chloride content decreased with the increase of citric acid concentration when citric acid concentration was greater than 0.05 mol?L?1.
Abstract: Data-driven material informatics is considered the fourth paradigm of materials research and development (R&D), which can greatly reduce R&D costs and shorten the R&D cycle. However, the data-driven method increases the risk of privacy disclosure when sharing and using materials data and sensitive information such as key processes in materials R&D. Therefore, privacy-preserving machine learning is a key issue in material informatics. The mainstream privacy protection methods in the current times include differential privacy, secure multi-party computation, federated learning, etc. The differential privacy model proposes strict definitions and metrics for quantitative evaluation of privacy protection, and the noise added by differential privacy is independent of the data scale. Only a small amount of noise is required to achieve a high level of protection, which considerably improves data usability. A novel differential privacy preserving random forest algorithm (DPRF) is proposed based on the fact that random forest is one of the most widely used models in material informatics. DPRF introduces the Laplace mechanism and exponential mechanism of differential privacy during the decision process tree building. First, the total privacy budget for the DPRF algorithm is set and then equally divided into each decision tree. During the tree-building process, the splitting features are randomly selected in the decision tree by the exponential mechanism and noise is added to the number of nodes by the Laplace mechanism, which is effective for differential privacy protection for the random forest. In experiments such as steel fatigue prediction experiments, the efficacies of DPRF under centralized or distributed data storage are verified. By setting different privacy budgets, the R2 of the predicted results of the DPRF algorithm can reach more than 0.8 for each target feature after adding differential privacy, which is not much different from the original random forest algorithm. A distributed data storage scenario shows that with the increase of privacy budget, the R2 of each target property prediction gradually increases. Comparing the effect of different tree depths in DPRF, it is shown that the overall R2 of the target prediction tends to increase and then later decrease .as the maximum depth of the tree increases. Overall, the best prediction accuracy is achieved when the maximum depth of the tree is set at 5. In summary, DPRF has good prediction accuracy in terms of achieving differential privacy protection of random forests. Specifically, in a distributed and decentralized data environment, DPRF can strike a balance between privacy-preserving strength and prediction accuracy by setting privacy budgets, tree depth, etc., which shows a wide range of application prospects of our algorithm.
Monthly, started in 1955 Supervising institution：Ministry of Education Sponsoring Institution：University of Science and Technology Beijing Editorial office：Editorial Department of Chinese Journal of Engineering Publisher：Science Press Chairperson：Ren-shu Yang Editor-in-Chief：Ai-xiang Wu ISSN 2095-9389CN 2095-9389
