Search

A novel strategy was developed to fabricate FeNx-doped carbon quantum dots (Fe-N-CQDs) to detect Cu2+ ions selectively as a fluorescence probe. The Fe-N-CQDs were synthesized by an efficient electrolysis of a carbon cloth electrode, which was coated with monoatomic iron-anchored nitrogen-doped carbon (Fe-N-C). The obtained Fe-N-CQDs emitted blue fluorescence and possessed a quantum yield (QY) of 7.5%. An extremely wide linear relationship between the Cu2+ concentration and the fluorescence intensity was obtained in the range from 100 nM to 1000 nM (R2 = 0.997), and the detection limit was calculated as 59 nM. Moreover, the Fe-N-CQDs demonstrated wide range pH compatibility between 2 and 13 due to the coordination between pyridine nitrogen and Fe3+, which dramatically reduced the affection of the protonation and deprotonation process between H+ and Fe-N-CQDs. It is notable that the Fe-N-CQDs exhibited a rapid response in Cu2+ detection, where stable quenching can be completed in 7 s. The mechanism of excellent selective detection of Cu2+ was revealed by energy level simulation that the LUMO level of Fe-N-CQDs (−4.37 eV) was close to the redox potential of Cu2+, thus facilitating the electron transport from Fe-N-CQDs to Cu2+.

Rational design of advanced structure for transition metal oxides (TMOs) is attractive for achieving high-performance supercapacitors. However, it is hampered by sluggish reaction kinetics, low mass loading, and volume change upon cycling. Herein, hierarchical NiCo2O4 architectures with 2D-nanosheets-shell and 3D-nanocages-core (2D/3D h-NCO) are directly assembled on nickel foam via a facile one-step way. The 2D nanosheets are in-situ generated from the self-evolution of initial NCO nanospheres. This 2D/3D hierarchical structures ensure fast ion/electron transport and maintain the structural integrity to buffer the volume expansion. The 2D/3D h-NCO electrode with an ultrahigh mass loading (30 mg cm−2) achieves a high areal capacity of 4.65 C cm−2 (equivalent to 1.29 mAh cm−2) at a current density of 4 mA cm−2, and retains 3.7 C cm−2 even at 50 mA cm−2. Furthermore, the assembled solid-state hybrid supercapacitor yields a high volumetric energy density of 4.25 mWh cm−3 at a power density of 39.3 mW cm−3, with a high capacity retention of 92.4% after 5000 cycles. Therefore, this work provides a new insight to constuct hierarchical electrodes for energy storage application.

With the aggressive and amazing scaling of the feature size of semiconductors, hotspot detection has become a crucial and challenging problem in the generation of optimized mask design for better printability. Machine learning techniques, especially deep learning, have attained notable success on hotspot detection tasks. However, most existing hotspot detectors suffer from suboptimal performance due to two-stage flow and less efficient representations of layout features. What is more, most works can only solve simple benchmarks with apparent hotspot patterns like ICCAD 2012 Contest benchmarks. In this paper, we firstly develop a new end-to-end hotspot detection flow where layout feature embedding and hotspot detection are jointly performed. An attention mechanism-based deep convolutional neural network is exploited as the backbone to learn embeddings for layout features and classify the hotspots simultaneously. Experimental results demonstrate that our framework achieves accuracy improvement over prior arts with fewer false alarms and faster inference speed on much more challenging benchmarks.

As the circuit feature size continuously shrinks down, hotspot detection has become a more challenging problem in modern DFM flows. Developed deep learning techniques have recently shown their superiorities on hotspot detection tasks. However, existing hotspot detectors can only handle defect detection from one small layout clip each time, thus may be very time-consuming when dealing with a large full-chip layout. In this paper, we develop a new end-to-end framework that can detect multiple hotspots in a large region at a time and promise a better hotspot detection performance. We design a joint auto-encoder and inception module for efficient feature extraction. A two-stage classification and regression framework is designed to detect hotspot with progressive accurate localization, which provides a promising performance improvement. Experimental results show that our framework enables a significant speed improvement over existing methods with higher accuracy and fewer false alarms.

Photoelectrochemical (PEC) water-splitting using solar energy holds great promise for the renewable energy future, and a key challenge in the development of industry viable PEC devices is the unavailability of high-efficient photoanodes. Herein, we designed a TiO2 model photocatalyst with nano-groove pattern and different surface orientation using low-energy Ar+ irradiation and photoetching of TiO2, and significantly improved the intrinsic activity for PEC water oxidation. High-resolution transmission electron microscopy directly manifests that the grooves consist of highly stepped surface with steps and well-crystallized. Transient absorption spectroscopy reveals the groove surface that allows for increased recovery lifetime, which ensures promoted electron-hole separation efficiency. Surface photovoltage directly shows the carrier separation and transportation behaviors, verified by selective photodeposition, demonstrating the groove surface on TiO2 contributes to electron-hole separation. This work proposes an efficient and scalable photoanode strategy, which potentially can open new opportunities for achieving efficient PEC water oxidation performance.

Surface oxygen vacancy defects of mesoporous CeO2 nanosheets assembled microspheres (D-CeO2) are engineered by polymer precipitation, hydrothermal and surface hydrogenation strategies. The resultant D-CeO2 with a main pore diameter of 9.3 nm has a large specific surface area (~102.3 m2/g) and high thermal stability. The mesoporous nanosheets assembled microsphere structure prevents the nanosheets from aggregation, which is beneficial to effective mass transfer and shortens the migration distance of charge carriers. After surface hydrogenation, the photoresponse extends to long wavelength region, combing with the band gap from 2.63 eV reduced to 2.39 eV. Under AM 1.5 G radiation, the photocatalytic degradation rate of tetracycline (TC) can be up to 99.99%, which is three times as high as that of pristine CeO2 microspheres. The excellent solar-driven photocatalytic performance can be attributed to the efficient surface oxygen vacancy engineering and the mesoporous nanosheets assembled microsphere structure, which narrows the band gap, shortens the migration distance of carriers, promotes the spatial separation of photogenerated electron-hole pairs and favors mass transfer. The strategy provides new insights for fabricating other high-efficient oxide photocatalysts .

Potassium‐ion batteries (PIBs) are a promising candidate for next‐generation electric energy storage applications because of the abundance and low cost of potassium. However, the development of PIBs is limited by sluggish kinetics and huge volume expansion of anodes, leading to poor rate capability and cycling stability. Herein, an advanced superstructure anode, including Te-doped SnS2 nanosheets uniformly anchored on MXene surface (Te-SnS2/MXene), is rationally designed for the first time to boost K+ storage performance. Featuring with strong interface interaction and self-autoadjustable interlayer spacings, the Te-SnS2/MXene can efficiently accelerate electron/ion transfer, accommodate volume expansion, inhibit crack formation, and improve pseudocapacitive contribution during cycling. Thus, the novel Te-SnS2/MXene anode delivers a high reversible capacity (343.2 mAh g−1 after 50 cycles at 0.2 A g−1), outstanding rate capability (186.4 mAh g−1 at 20 A g−1), long cycle stability (165.8 mAh g−1 after 5000 cycles at 10 A g−1 with a low electrode swelling rate of only 15.4%), and reliable operation in flexible full battery. The present Te-SnS2/MXene becomes among the best transition metal-based anode materials for PIBs reported to date.

Although neoantigen-based cancer vaccines show great potential in cancer immunotherapy due to their ability to induce effective and long-lasting anti-tumor immunity, their development is hindered by the limitations of neoantigens identification, low immunogenicity, and weak immune response. Cyclophosphamide (CTX) not only directly kills tumors but also causes immunogenic cell death, providing a promising source of antigens for cancer vaccines. Herein, a combined immunotherapy strategy based on temperature-sensitive PLEL hydrogel is designed. First, CTX-loaded hydrogel is injected intratumorally into CT26 bearing mice to prime anti-tumor immunity, and then 3 days later, PLEL hydrogels loaded with CpG and tumor lysates are subcutaneously injected into both groins to further promote anti-tumor immune responses. The results confirm that this combined strategy reduces the toxicity of CTX, and produces the cytotoxic T lymphocyte response to effectively inhibit tumor growth, prolong survival, and significantly improve the tumor cure rate. Moreover, a long-lasting immune memory response is observed in the mice. About 90% of the cured mice survive for at least 60 days after being re-inoculated with tumors, and the distant tumor growth is also well inhibited. Hence, this PLEL-based combination therapy may provide a promising reference for the clinical promotion of chemotherapy combined with cancer vaccines., Graphical abstract Image 1, Highlights • PLEL based-CTX hydrogel system avoided the rapid clearance of CTX and reduced systemic toxicity. • PLEL-assisted tumor lysate vaccine was cheap, safe, and contained all tumor antigens. • This strategy promoted the maturation and activation of DCs, enhanced cancer-specific CD8+ T cell responses. • PLEL-assisted combination strategy achieved a good tumor inhibition effect and generate a lasting immune memory. . • This local administration strategy could kill tumors that could not be detected or removed surgically in the clinic.

Barium Strontium Titanate (Ba1-xSrxTiO3, BST) possesses the advantages of high permittivity, low remnant polarization and dielectric loss, suggesting great potential in the application of power capacitors. However, the low maximum polarization severely restricts the achievement of outstanding energy storage properties (ESP). Herein, we design a lead-free bulk ceramic: (1-x)Ba0.55Sr0·45TiO3-xBi0.5Na0.5TiO3 [(1-x)BST-xBNT] (x = 0.1–0.4), prepared by two-step solid-state reaction method. The introduction of BNT contributes to (1) the boost of maximum polarization (from 21 μC/cm2 up to 32.6 μC/cm2) due to Bi3+ lone pair covalent effect (2) strong relaxor behavior (1.71≪ γ 1.84) derived from weakly coupled nanopolar regions, which can be verified by high activation energy Ea = 0.341 eV at x = 0.3, improves the ESP and thermal stability of the ceramics. According to the above strategies, high Wrec = 1.73 J/cm3 and η = 84.4% are achieved in the 0.7BST-0.3BNT ceramics at 206 kV/cm, which is preferable to the previous results regarding BST-based bulk ceramics. The excellent thermal stability (20–120 °C), frequency stability (1–500 Hz), as well as anti-fatigue characteristic (cycling number: 105) were obtained in the 0.7BST-0.3BNT. Furthermore, 0.7BST-0.3BNT ceramics exhibit high-power density (PD = 44.9 MW/cm3) and rapid discharge rate (t0.9 = 99.2 ns) from 20 °C to 120 °C. This work provides a feasible way to achieve good energy storage and charge-discharge properties for pulse power energy storage systems.

The combination of multiple loss characteristics is an effective approach to achieve broadband microwave wave absorption performance. The Fe-doped SiOC ceramics were synthesized by polymer derived ceramics (PDCs) method at 1500 °C, and their dielectric and magnetic properties were investigated at 2–18 GHz. The results showed that adding Fe content effectively controlled the composition and content of multiphase products (such as Fe3Si, SiC, SiO2 and turbostratic carbon). Meanwhile, the Fe promoted the change of the grain size. The Fe3Si enhanced the magnetic loss, and the SiC and turbostratic carbon generated by PDCs process significantly increased the polarization and conductance loss. Besides, the magnetic particles Fe3Si and dielectric particles SiO2 improved the impedance matching, which was beneficial to EM wave absorption properties. Impressively, the Fe-doped SiOC ceramics (with Fe addition of 3 wt %) presented the minimum reflection coefficient (RCmin) of −20.5 dB at 10.8 GHz with 2.8 mm. The effective absorption bandwidth (EAB, RC

In this study, MgAl2O4-based ceramics with high quality factor (Qf) and low dielectric constant (er ≤ 10) were obtained by fabricating MgAl2-x(Zn0.5Ti0.5)xO4 (x = 0–0.5) ceramics via conventional solid-state reaction method. Excellent microwave dielectric properties were achieved for samples at x = 0.5 and sintered at 1550 °C, i.e., er = 9.86, Qf = 263 900 GHz (five times better than that for x = 0 sample) and τf = −92 ppm/°C. The X-ray diffraction (XRD) patterns displayed characteristic peaks of MgAl2O4 with spinel structure. MgTi2O5 and MgTiO3 were considered as secondary phases. Scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS) and relative density analysis indicated that ultra-high Qf values were dominated by dense microstructure, secondary phase and cation vacancies; whereas er values were mainly affected by secondary phase and ionic polarizability. MgAl2-x(Zn0.5Ti0.5)xO4 ceramics with excellent microwave dielectric properties have potential application in millimeter-wave communication, dielectric filters, dielectric antennas and resonators.

Nowadays, iron ions as a ubiquitous heavy metal pollutant are gradually concerned and the convenient and quick removal of excessive iron ions in groundwater has become a major challenge for the safety of drinking water. In this study, boron-doped biochar (B-BC) was successfully prepared at various preparation conditions with the addition of boric acid. The as-prepared material has a more developed pore structure and a larger specific surface area (up to 897.97 m²/g). A series of characterization results shows that boric acid effectively activates biochar, and boron atoms are successfully doped on biochar. Compared with the ratio of raw materials, the pyrolysis temperature has a greater influence on the amount of boron doping. Based on Langmuir model, the maximum adsorption capacity of 800B-BC1:2 at 25 °C, 40 °C, 55 °C are 50.02 mg/g, 95.09 mg/g, 132.78 mg/g, respectively. Pseudo-second-order kinetic model can better describe the adsorption process, the adsorption process is mainly chemical adsorption. Chemical complexation, ions exchange, and co-precipitation may be the main mechanisms for Fe2+ removal.

A dual-band (12.5/14.25 GHz) orthogonally polarized contour beam transmitarray has been designed to accomplish the contour beam pattern for a European coverage region in both transmit and receive frequency bands. A quad-layer transmitarray unit cell configuration is optimized theoretically for an efficient dual-band operation. An orthogonally polarized 3-dipole unit cell with 360° phase range is used to control transmission phases at both frequencies independently. A modified intersection approach with a fast Fourier transform algorithm is used to optimize the aperture phase distribution of transmitarray at each frequency independently. An antenna prototype has been fabricated and tested. The measured contour beam patterns are in good agreement with the calculated patterns in both frequency bands. The measured minimum gains in the required coverage region are 25 and 25.3 dBi at 12.5 and 14.25 GHz, respectively.

Batteries can undergo overdischarge in actual applications without triggering safety issues. The interesting one is whether the battery that is overdischarged can have a second life in the following normal charge and discharge process. Herein, the effects of overdischarge are investigated using the LiCoO2/mesocarbon microbeads (MCMB) batteries with different depth of discharge (DOD) (102% DOD, 105% DOD, and 115% DOD) for ten times; the performance and aging mechanisms of the overdischarged battery in the following normal cycling are studied mainly through the electrochemical methods, scanning electron microscopy (SEM), X-ray diffractometry (XRD), X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy (TEM). For the batteries subjected to multiple shallow overdischarge (e.g., 102% DOD and 105% DOD), the aging mechanism remains unchanged, and the battery can be further used for a second life. In contrast, the multiple deep overdischarge (e.g., 115% DOD) remarkably leads to battery bulges, a drastic decrease in capacity, and a changed aging mechanism. It is noteworthy that the performance decay of LiCoO2 is deteriorated owing to the thickened Cathode-Electrolyte-Interface (CEI) and the damaged microstructure, which is different from overdischarge for one time. At the same time, the thermostability of LiCoO2 is remarkably reduced after deep overdischarge for ten times.

Piezoelectric bioorganic thin films Piezoelectric materials enable a reversible conversion between mechanical pressure and electric charge and are useful for sensors, actuators, and high-precision motors. Yang et al. developed a method for making high-quality crystalline thin films of piezoelectric γ-glycine crystals that are grown and refined between layers of polyvinyl alcohol (PVA) (see the Perspective by Berger). The PVA layers are essential to promoting the crystallization of the preferred crystal phase with the polar axis oriented perpendicular to the film plan because of hydrogen bonding at the PVA-glycine interface. The thin films show a macroscopic piezoelectric response and high stability in aqueous environments. The films are water soluble and, when suitably packaged, could be implanted into a biodegradable energy-harvesting device. Science , abf2155, this issue p. 337 ; see also abj0424, p. 278

Layout hotspot detection is of great importance in the physical verification flow. Deep neural network models have been applied to hotspot detection and achieved great successes. The layouts can be viewed as binary images. The binarized neural network (BNN) can thus be suitable for the hotspot detection problem. In this article, we propose a new deep learning architecture based on BNNs to speed up the neural networks in hotspot detection. A new binarized residual neural network is carefully designed for hotspot detection. Experimental results on ICCAD 2012 and 2019 benchmarks show that our architecture outperforms previous hotspot detectors in detecting accuracy and has an $8\times $ speedup over the best deep learning-based solution. Since the BNN-based model is quite computationally efficient, a good tradeoff can be achieved between the efficiency and performance of the hotspot detector by applying ensemble learning approaches. Experimental results show that the ensemble models achieve better hotspot detection performance than the original with acceptable speed loss.

Rumor propagation on complex networks is rapidly affecting people’s life. As we all know, the regulatory control of rumors by regulators has a specific impact on the spread of rumors. Limited regulatory resources may saturate the regulatory level. Therefore, in this paper, we have introduced a saturation treatment function to model this phenomenon and further establish an SIS rumor propagation model with consideration of some comprehensive influence on rumor diffusion. First of all, we prove the boundedness of solutions, and the basic reproduction number R 0 is obtained by the method of the next generation matrix . Secondly, by constructing Lyapunov function and applying the linearization method of differential equations, the stability conditions of the equilibrium points are derived. Further, we determine the condition for the backward bifurcation of the rumor propagation model. Moreover, in order to control the spread of rumors, we propose targeted immunization control, acquaintance immunization control and optimal control strategies based on complex networks. Finally, the sensitivity analysis of the basic reproduction number is carried out, and the correctness of the theoretical results is verified by numerical simulations. Our results may provide us with useful insights into the dynamics of online rumor propagation. For example, the basic reproduction number R 0 gives the threshold of rumor propagation, which provides control conditions for suppressing the spread of rumors. Stability analysis indicates the likelihood of local or global outbreaks for rumors. In addition, the comparison of the effect for control strategies gives a variety of options for the control method.

Deep learning is increasingly being used in high-stake decision making applications that affect individual lives. However, deep learning models might exhibit algorithmic discrimination behaviors with respect to protected groups, potentially posing negative impacts on individuals and society. Therefore, fairness in deep learning has attracted tremendous attention recently. We provide a review covering recent progresses to tackle algorithmic fairness problems of deep learning from the computational perspective. Specifically, we show that interpretability can serve as a useful ingredient to diagnose the reasons that lead to algorithmic discrimination. We also discuss fairness mitigation approaches categorized according to three stages of deep learning life-cycle, aiming to push forward the area of fairness in deep learning and build genuinely fair and reliable deep learning systems.

The fast exploration of low energy configuration by surface atoms is believed to favor the formation of ultrastable metallic glasses, prepared by physical vapor deposition. Here, we find that the rearrangement of surface atoms is collective, rather than being dominated by individual atoms. Specifically, we experimentally observe the growth process of ultrastable metallic glasses at monolayer resolution, which follows a periodic island-layer-island pattern with morphology variation between islands and flat surfaces. The estimated surface diffusion coefficient is orders of magnitude higher than that for bulk diffusion. The fast surface dynamics allow the newly deposited clusters on the flat surface to form local islands with spherical shape, which substantially reduces the surface free energy in each island-layer-island growth cycle. Our findings are helpful for understanding the growth mechanisms of ultrastable metallic glasses and potentially for tailoring their properties. Ultrastable metallic glasses form by surface diffusion on a substrate during deposition. Here, this process is shown to involve the collective diffusion of atoms in a cyclic process, involving the formation of islands, their coalescence into a layer, followed by further island formation.

With the rapid development of laser sources and welding technologies in recent years, laser welding has been widely used in the production. In laser welding process, plasma plume often shields the input energy and defocuses the laser beam, which has significant influences on the welding efficiency and quality. It has been found that laser welding under subatmospheric environment could effectively suppress plasma plume, increase weld penetration depth, improve weld pool stability, reduce spatter and undercut, and obtain high-quality weld. After a comprehensive review of the latest research results, the plasma plume, molten pool dynamics, weld morphology, microstructure, and mechanical characteristics of various materials during laser welding under subatmospheric environment are analyzed. Finally, the future research works are discussed aiming at some remaining issues in this topic.

The memristor is a foundational device for an artificial synapse, which is essential to realize next-generation neuromorphic computing. Herein, an optoelectronic memristor based on a two-dimensional (2D) transitional-metal trichalcogenide (TMTC) is designed and demonstrated. Owing to the excellent optical and electrical characteristics of titanium trisulfide (TiS3), the memristor exhibits stable bipolar resistance switching (RS) as a result of the controllable formation and rupturing of the conductive aluminum filaments. Multilevel storage is realized with light of multiple wavelengths between 400 and 808 nm, and the synaptic properties such as conduction modulation and spiking timing-dependent plasticity (STDP) are achieved. On the basis of the photonic potentiation and electrical habitual ability, Pavlovian-associative learning is successfully established on this TiS3-based artificial synapse. All these results reveal the large potential of 2D TMTCs in artificial neuromorphic chips.

Low-cost and efficient bifunctional catalysts are urgently needed for overall water splitting used in large-scale energy storage. In this study, we develop a nickel and iron (di)sulfide (Ni-Fe-S) composite catalyst that is in situ synthesized and fixed within the intergranular nanopores inside high pure polycrystalline graphite. Two precursor solutions (reactants) may permeate the graphite intergranular pores to a depth of more than 3.5 mm. The nanoscale pores serve as an array of nanoreactors for the synthesis of the Ni-Fe-S nanoparticles under conditions much milder than usual. The prepared catalyst efficiently catalyzes both the hydrogen and oxygen evolution reactions (HER and OER) in 1.0 M KOH. It delivers a current density of 400 mA cm-2 at a full cell voltage of around 2.3 V without considerable activity decay over 24 h electrolysis. The active species of the catalyst are different for the HER and OER and discussed accordingly. The synthesis strategy based on the nanopores in a monolithic conductive substrate proves to be a simple, efficient, and promising way to prepare electrocatalysts that are cheap, abundant, and industrially attractive.

Although high-performance graphene-based micro/nano flexible electronic devices have shown promising applications in numerous fields, there are still many problems in converting graphene into practical applications. Heteroatom-doped graphene materials are of huge importance because heteroatom doping can significantly change the electronic structure and introduce the active site, which benefits the integration with a promising substrate and achieves nondestructive transfer of carbon materials. Herein, we analyze in detail the pyrolysis gas composition of heteroatom-enriched phosphazenes with different structures and prepare a series of high-quality in situ N, P-codoped carbon-based films from phosphazene solid sources on a low-cost glass substrate by a convenient one-step method. The N, P-codoped carbon film shows reflectivity, good conductivity, and transparency. In addition, with the help of in situ "molecular welding", we achieve nondestructive transfer of a conductive carbon-based film from a glass substrate to promising layer-polyimide (PI) and prepare a flexible free-standing carbon/PI hybrid film with an excellent binding interface. The flexible conductive hybrid film shows excellent durability under an extremely low temperature environment and superior bending stability after 800 bending cycles. The results suggest that a phosphazene precursor is an amazing choice for constructing high-quality heteroatom-doped conductive carbon films. Besides, this work provides a promising way for nondestructive transfer of the conductive carbon-based films and large-scale preparation of large-area patterned conductive thin films.

Replacing scarce and expensive platinum (Pt) with metal–nitrogen–carbon (M–N–C) catalysts for the oxygen reduction reaction in proton exchange membrane fuel cells has largely been impeded by the low oxygen reduction reaction activity of M–N–C due to low active site density and site utilization. Herein, we overcome these limits by implementing chemical vapour deposition to synthesize Fe–N–C by flowing iron chloride vapour over a Zn–N–C substrate at 750 °C, leading to high-temperature trans-metalation of Zn–N4 sites into Fe–N4 sites. Characterization by multiple techniques shows that all Fe–N4 sites formed via this approach are gas-phase and electrochemically accessible. As a result, the Fe–N–C catalyst has an active site density of 1.92 × 1020 sites per gram with 100% site utilization. This catalyst delivers an unprecedented oxygen reduction reaction activity of 33 mA cm−2 at 0.90 V (iR-corrected; i, current; R, resistance) in a H2–O2 proton exchange membrane fuel cell at 1.0 bar and 80 °C. Replacing platinum with metal–nitrogen–carbon catalysts for the oxygen reduction reaction in proton exchange membrane fuel cells has been impeded by low activity. These limitations have now been overcome by the trans-metalation of Zn–N4 sites into Fe–N4 sites.

Polyvinylidene fluoride (PVDF) thin films with high β-phase content exhibit excellent piezoelectric and dielectric properties. In this work, PVDF–NaCl composite film with high β-phase content was prepared by near-electric-field three-dimensional (3D) printing. The mechanism of the influence of NaCl on the nucleation and phase transition of the PVDF thin films was revealed based on the synergistic effect of the electric field and salt addition. The composite films were characterized by scanning electron microscopy, fourier transform infrared spectroscopy, x-ray diffraction analysis, and measurements of electrical properties. The results showed that NaCl can increase the density of the membrane, while the β-phase content and dielectric properties of the PVDF composite membrane are positively correlated with the amount of NaCl added. The β-phase content reached a maximum value of 52.68% when the amount of NaCl added was 0.05% by weight. The dielectric constant of the PVDF composite film increased with increasing NaCl content. The electrical properties also indicated that PVDF composite film with higher NaCl content showed higher electrical output voltage, while the response voltage of the composite film with 0.05 wt.% NaCl was higher than that of pure PVDF film.

An integrated approach combining scanning tunneling microscopy (STM) and X-ray photoelectron spectroscopy (XPS) is used to investigate the atomic structures and electronic properties of Cr-doped ZnO( ) surfaces. When deposited at 300 K, Cr at low surface coverage (l 0.1 ML) appeared either as isolated atoms on the surface terrace of ZnO( ) or substituting Zn atoms in the ZnO lattice. Their structural models could be identified from atomic-resolution STM images and their oxidation states were found as Cr3+ based on XPS measurements. Rectangular islands nucleated at step edges along the [0001] direction could also be observed during the initial growth of Cr at 300 K and were assigned as Cr islands. The density of Cr islands as well as their average size increased with the increasing of Cr surface loading. Thermal treatments at above 600 K could facilitate the decomposition of Cr islands and the re-dispersion of Cr atoms into the ZnO lattice, indicating a strong interaction between Cr and ZnO. The adsorption of CO at 78 K showed no preferential adsorption at Cr3+ sites embedded in the surface lattice of ZnO. However, the re-dispersion of Cr atoms into the ZnO bulk at above 600 K could induce a significant upward band bending, causing a negative shift of core level XPS peaks of Zn 2p and O 1s by ~0.5–0.7 eV. Our study has thus constructed a model catalyst for Cr-doped ZnO and provided atomic insight for understanding ZnO-based catalysts.

Biomedical imaging is a relevant noninvasive technique aimed at generating an image of the biological structure under analysis. The arising visual representation of the characteristics of the object is affected by both the measurement process and reconstruction algorithm. This procedure can be considered as a hybridization of data information, measurement physics, and prior information.

The Gaoshiti-Moxi Gas Field has been discovered recently in the central part of Sichuan Basin, which might be the biggest gas field in China. In this gas field, a large amount of gas is from Neoproterozoic Sinian (Ediacaran) karstic carbonate rocks. The gas was originated from cracking of oil, and the oil source rocks are the Sinian and Cambrian strata. In this study, the characteristics of these carbonate reservoir rocks have been studied based on subsurface data and outcrop observation. The carbonate rocks are at the Sinian Dengying Formation, and the reservoir properties are controlled by sedimentary facies and diagenesis. The Dengying Formation is composed of algal mounds and grainstone build-ups, which were developed in a restricted platform setting from glacial period to interglacial period. Due to tectonic movement and eustatic cycle in Sichuan Basin during Neoproterozoic Era, five sedimentary cycles of transgression and regression have been developed in the Dengying Formation. According to the sedimentary facies and carbon-oxygen isotope variation, five sedimentary cycles have been recognized. Due to the penecontemporaneous dolomitization, together with hypergene karstification and late burial dissolution, the reservoir properties of the carbonates rocks have been improved greatly.

An environment-friendly, water-soluble, and cellulose based binder (lithium carboxymethyl cellulose, CMC-Li) was successfully synthesized by using Li+ to replace Na+ in the commercial sodium carboxymethyl cellulose (CMC-Na). Li-O2 batteries based on the CMC-Li binder present enhanced discharge specific capacities (11151 mAh/g at 100 mA/g) and a superior cycling stability (100 cycles at 200 mA/g) compared with those based on the CMC-Na binder. The enhanced performance may originate from the electrochemical stability of the CMC-Li binder and the ion-conductive nature of CMC-Li, which promotes the diffusion of Li+ in the cathode and consequently retards the increase of charge transfer resistance of the cathode during cycling. The results show that the water-soluble CMC-Li binder can be a green substitute for poly(vinylidene fluoride) (PVDF) binder based on organic solvent in the lithium oxygen batteries (LOBs).

通过表面氧空位缺陷调控催化剂的表面电子结构, 实现催化剂表面易于富集参与反应的电子, 从而提高还原催化效率. 本文中, 将导电玻璃隔绝空气加热, 在其表面形成氧空位缺陷. 通过控制热处理时间调控表面氧空位浓度梯度. 氧空位能导致SnO2导带弯曲, 因此调控氧空位梯度可增强SnO2导电玻璃表面电子富集速率, 从而实现高效电催化还原反应. 在相同条件下, 使其催化性能可以和金属铂相媲美.

This paper presented an improved switching transient model of silicon carbide (SiC)-based asymmetric H-bridge (AHB) power converter for a switched reluctance generator (SRG), which takes the nonlinear phase inductance of the SRG into consideration. First, a systematic mathematic derivation is carried out and the switching transient model is established. Second, the impact of the nonlinear phase inductance of the SRG on switching transients is verified by PSpice & Simulink co-simulations. Third, a power loss model is established by the transient model through PSpice & Simulink co-simulations. The model is conducted to indicate the power loss characteristics of the SiC-based AHB power converter. Simulation results indicate that the nonlinear phase inductance of the SRG is able to accelerate the switching speed of SiC-MOSFETs, and that the SiC converter is advantageous in terms of power loss. Experimental results illustrate that the established power loss model experiences high accuracy. In addition, SiC devices are able to strengthen the power density and efficiency of a converter while reducing its heat dissipation requirements.

A wing is an important part of the aircraft to improve aerodynamic performance. The current study is focused on an adaptive surrogate algorithm for airfoil aerodynamic optimization, which is based on a multi-output Gaussian process model. The conventional design method seriously relies on wind tunnel experiments and expensive computational simulations. The metamodels can significantly improve design efficiency and hence reduce the overall design costs. An active learning algorithm is proposed to improve the effectiveness of the multi-output Gaussian process model. The NSGA-II algorithm is adopted to obtain the optimal Pareto set with the optimization objectives of lift and drag coefficients for adaptive airfoil shapes. Besides, the Bezier curve and radial basis function are utilized in this study for airfoil mesh deformation. The results show that the airfoil shape can be obtained effectively by integrating the metamodel, active learning algorithm, and multi-objective optimization algorithm. The optimized results are of great engineering applications.

Intelligent decision control is one key issue of automated vehicles. Complex dynamic traffic flow and multi-requirement of passengers including vehicle safety, comfort, vehicle efficiency bring about tremendous challenges to vehicle decision making. Overtaking maneuver is a demanding task due to its large potential of traffic collision. Therefore, this paper proposes a fuzzy-inference-based reinforcement learning (FIRL) approach of autonomous overtaking decision making. Firstly, the problem of overtaking is formulated as a multi-objective Markov decision process (MDP) considering vehicle safety, driving comfort, and vehicle efficiency. Secondly, a temporal difference learning based on dynamic fuzzy (DF-TDL) is presented to learn optimized policies for autonomous overtaking decision making. Fuzzy inference is introduced to deal with continuous states and boost learning process. The RL algorithm decides whether to overtake or not based on the learned policies. Then, the automated vehicle executes local path planning and tracking. Furthermore, a simulation platform based on simulation of urban mobility (SUMO) is established to generate the random training data, that is, various traffic flows for algorithm iterative learning and validate the proposed method, extensive test results demonstrate the effectiveness of the overtaking decision-making method.

The underbody of a vehicle system, either military or civil, is typically made of a relatively thin metallic plate, thus vulnerable to mine blast attacks. To improve the blast resistance, a multitude of protective structures have been proposed as attachments to the thin plate. In the present study, a novel ultralight all-metallic sandwich panel with three-dimensional (3D) tube cellular cores mounted to the vehicle underbody was envisioned as such a protective system. A metallic substrate (mimicking vehicle bottom) was placed above the proposed sandwich panel to construct a sandwich-substrate combinative structure. A series of sandwich panels having 3D tube cellular cores were fabricated via argon protected welding and laser welding. Mechanical responses of the combinative structure subjected to the denotation of 6 kg TNT explosives shallow-buried in dry sand were experimentally measured. Full numerical simulations with the method of finite elements (FE) were subsequently carried out to explore the physical mechanisms underlying the observed dynamic performance and quantify the effects of key geometrical parameters and connection conditions of the protective system. The performance of the proposed sandwich panel under shallow-buried explosives was also compared with competing sandwich constructions having equal mass. Finally, a preliminary optimal design of the 3D tube cellular core was carried out.

The combination of graphene oxide (GO) and iron oxide (Fe2O3) may induce property enforcement and application extension. Herein, GO/Fe2O3 nanocomposites were synthesized via the vacuum-freeze-drying method and used for the thermal decomposition of ammonium perchlorate (AP). A series of characterization techniques were applied to elucidate the as-obtained nanomaterial's physicochemical properties. These results show that the treated GO is consistent with the pristine GO after the freeze-drying treatment. Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) analyses show that iron oxide nanoparticles are anchored on and between the GO sheets. The catalytical effect investigation on AP with different Fe2O3: GO ratios indicates that the high-temperature decomposition temperature of AP could be decreased by a temperature as high as 77 °C compared to pure AP accompanied by 3 wt % GO/Fe2O3 nanocomposite which proves the high catalytic performance of the nanocomposites. The first-principles calculation was employed to elaborate the synergistic effect, and the findings demonstrate that the presence of graphene in the catalyst can enhance the catalytic effect via reducing the activation energy barrier by ∼17% in the reaction of AP thermal decomposition.

Alkylation of benzene to value-added, high octane number and low toxic toluene and xylenes provides a way to lower benzene content in gasoline pool, and is hence a method to promote fuel quality. On the other hand, CO2 accumulation in the atmosphere causes global warming and requires effective route for its valorization. Utilization of CO2 as a carbon source for benzene alkylation could achieve both goals. Herein, alkylation of benzene with CO2 and H2 was realized by a series of low-cost bifunctional catalysts containing zinc/titanium oxides (Zn/Ti oxides) and HZSM-5 molecular sieves in a fixed-bed reactor. By regulating and controlling oxygen vacancies of Zn/Ti oxides and the acidities of HZSM-5, benzene conversion and CO2 conversion reached 28.7% and 29.9% respectively, along with a total selectivity of toluene and xylene higher than 90%. In this process, more than 25% CO2 was effectively utilized and incorporated into the target products. Moreover, the mechanism of the reaction was analyzed and the course was simultaneously traced. CO2 was transformed into methanol firstly, and then methanol reacted with benzene generating toluene and xylene. The innovation provides a new method for upgrading of fuels and upcycling the emissions of CO2, which is of great environmental and economic benefits.

We review high baud rate operation of silicon photonic modulators with emphasis on short-reach and metro-haul applications. The current status of both devices and digital signal processing are discussed. The basic design of a typical carrier-depletion-based dual-drive traveling-wave Mach-Zehnder modulator (TW-MZM) is studied by simulation. Based on an all-silicon TW-MZM, we report the transmission experiments of high baud rate single sideband 4-ary pulse amplitude modulation (PAM-4) signals with linear equalization and artificial neural network based equalization, respectively. Our work indicates all-silicon modulators can address the requirements of future optical communication systems operating at 100 Gbaud.

Background The demand for home healthcare devices arises; however, many home healthcare devices on the market are not designed to reflect the needs and features of the end-users. This study explored the user knowledge factors that hindered the design of new home healthcare devices and the interrelationships between the factors. Methods The abovementioned factors were identified from analysing the project documents of thirty-eight carefully selected home healthcare devices produced by five manufacturers; followed by interviewing the thirty stakeholders playing key roles in developing the devices. Results The design of the home healthcare devices was influenced by (1) the user insights utilised in formulating project strategies; (2) the sources of user information; (3) the execution of user research; and (4) the formulation of the manufacturers’ principal innovation processes. Conclusions The users’ characteristics and needs were not sufficiently reflected in developing new home healthcare devices. One root cause was that the end-users were not perceived by the manufacturers as a key success factor in most cases, given that most of the devices were initiated following the public sector’s requests. Actual or potential applications of this study include the facilitation of the appropriate application of human factors methods in developing new home healthcare devices and the improvement of the user performance of the end-devices.

Time series classification and retrieval are two important tasks of time series analysis. Existing methods solve these two tasks separately, which ignores the sharable information among different tasks. In this paper, we propose a deep multi-task representation learning method (MTRL) for time series classification and retrieval, which exploits both related supervised and unsupervised information. Specifically, supervised representation learning for classification task tries to maximize the inter-class variations and minimize the intra-class variations. Unsupervised representation learning for retrieval task aims at preserving pairwise dynamic time warp (DTW) distances. These two tasks can benefit from each other via shared networks, which consist of deep wavelet decomposition networks and residual networks. These networks can extract the information hidden in different time and frequency domains, and can achieve easier information flow from the lowest level to the highest level than traditional convolutional neural networks. Furthermore, we propose a distance-weighted sampling strategy, which focuses on the more discriminative samples to achieve high convergence speed and accuracies. Extensive experiments on UCR datasets demonstrate that MTRL outperforms the state-of-the-art methods.

Novel degraded potassium channel-modulatory peptides were recently found in thermally processed scorpions, but their pharmacological properties remain unclear. Here, we identified a full-length scorpion toxin (i.e., BmKcug2) and its four truncated analogs (i.e., BmKcug2-P1, BmKcug2-P2, BmKcug2-P3 and BmKcug2-P4) with three conserved disulfide bonds in processed scorpion medicinal material by mass spectrometry. The pharmacological experiments revealed that the recombinant BmKcug2 and BmKcug2-P1 could selectively inhibit the human Kv1.2 and human Kv1.3 potassium channels, while the other three analogs showed a much weaker inhibitory effect on potassium channels. BmKcug2 inhibited hKv1.2 and hKv1.3 channels, with IC50 values of 45.6 ± 5.8 nM and 215.2 ± 39.7 nM, respectively, and BmKcug2-P1 inhibited hKv1.2 and hKv1.3, with IC50 values of 89.9 ± 9.6 nM and 1142.4 ± 64.5 nM, respectively. The chromatographic analysis and pharmacological properties of BmKcug2 and BmKcug2-P1 boiled in water for different times further strongly supported their good thermal stability. Structural and functional dissection indicated that one amino acid, i.e., Tyr36, determined the differential affinities of BmKcug2 and four BmKcug2 analogs. Altogether, this research investigated the different pharmacological properties of BmKcug2 and its truncated analogs, and the findings highlighted the diversity of K+ channel blockers from various scorpion species through thermal processing.

SiCN-based ceramics with broadband and strong microwave absorption properties are desired for the structural and functional integration of ceramic matrix composites. The elemental composition and thermal expansion coefficients of the ceramics matrix crucially affect its microstructure and electromagnetic wave (EMW) absorption properties. BN layer with lower electrical conductivity and higher specific area, exhibits both the impedance matching characteristic and EMW attenuation in the process of multiple reflections, electrical conductivity loss, dipole polarization and interfacial polarization. Therefore, Si3N4-BN-SiCN ceramics, which were synthesized using chemical vapor infiltration (CVI) method, construct unique hetero-interface of Si3N4-BN, Si3N4–SiCN and BN-SiCN. Therefore, the Si3N4-BN-SiCN ceramics have outstanding EMW absorption performance and realize an effective absorption bandwidth (EAB) that covers the whole X band and the minimum reflection coefficient (RC) reaches -18.43 dB at a thickness of 3.37 mm.

Sand production has become a common phenomenon in the exploitation of unconsolidated natural gas hydrate reservoirs, which will hinder the long-term production of natural gas hydrate reservoirs. However, there are few literatures reported on the influences in reservoir physical properties such as permeability and porosity, and production laws caused by sand production. This paper provides a numerical model, coupled with reservoir sand-gas-water multiphase flow processes, which is capable to simulate the process of sand production in natural gas hydrate reservoirs. The simulation results indicate that sand settlement is mainly concentrated near the wellbore due to the high concentration of migrated sand. The decrease in reservoir porosity and permeability caused by sand settlement has a significant impact on production. The impact of sand production on reservoir fluid fluidity shows that fluid flow is inhibited near the wellbore, while fluid flow performance increases far away from the wellbore. The numerical model and analysis presented here could provide useful insight into changes in reservoir physical properties and production laws caused by sand production in the natural gas hydrate-bearing marine sediments using depressurization method.

Multiple Object Tracking (MOT) in the wild has a wide range of applications in surveillance retrieval and autonomous driving. Tracking-by-Detection has become a mainstream solution in MOT, which is composed of feature extraction and data association. Most of the existing methods focus on extracting targets’ individual features and optimizing the association by hand-crafted algorithms. In this paper, we specially consider the interrelation cue between targets and we propose Human-Interaction Model (HIM) to extract interaction features between the tracked target and its surrounding. The interaction model has more discriminative features to distinguish objects, especially in crowded (dense) scene. Meanwhile we propose an efficient end-to-end model, Deep Association Network (DAN), to optimize the association with graph-based learning mechanism. Both HIM and DAN are constructed by three kinds of deep networks, which include Convolutional Neural Network (CNN), Recurrent Neural Network (RNN) and Graph Neural Network (GNN). The CNNs extract appearance features from bounding box images, the RNNs encoder motion features from historical positions of trajectory. And then the GNNs aim to extract interaction features and optimize graph structure to associate the objects in different frames. In addition, we present a novel end-to-end training strategy for Deep Association Network and Human-Interaction Model. Our experimental results demonstrate performance of our method reaches the state-of-the-art on MOT15, MOT16 and DukeMTMCT datasets.

For biomedical photoacoustic applications, an ongoing challenge in simultaneous volumetric imaging and spectroscopic analysis arises from ultrasonic detectors lacking high sensitivity to pressure transients over a broad spectral bandwidth. Photoacoustic impulses can be measured on the basis of the ultrafast temporal dynamics and highly sensitive response of surface plasmon polaritons to the refractive index changes. Taking advantage of the ultra-sensitive phase shift of surface plasmons caused by ultrasonic perturbations instead of the reflectivity change [as is the case for traditional surface plasmon resonance (SPR) sensors], a novel SPR sensor based on phase-shifted interrogation was developed for the broadband measurement of photoacoustically induced pressure transients with improved detection sensitivity. Specifically, by encoding the acoustically modulated phase change into time-varying interference intensity, our sensor achieved an almost five-fold sensitivity enhancement (∼98 Pa noise-equivalent pressure) compared with the reflectivity-mode SPR sensing technologies (∼470 Pa) while retaining a broadband acoustic response of ∼174 MHz. Incorporating our sensor into an optical-resolution photoacoustic microscope, we performed label-free imaging of a zebrafish eye in vivo, enabling simultaneous volumetric visualization and spectrally resolved discrimination of anatomical features. This novel sensing technology has potential for advancing biomedical ultrasonic and/or photoacoustic investigations.

How to identify the key nodes effectively in urban traffic networks to achieve the equitable resource allocation face to the complex traffic network? This issue needs to be solved in current traffic management. This study considered the urban traffic network topology and network traffic status, put forward an improved model based on the economics of the input-output method by introducing a virtual node to the selected network set up with the flow of urban traffic network, sensor nodes by Leontief inverse matrix calculation coefficient to determine node importance, according to the node importance to deliberate attack traffic network to analyze its robustness, to test the accuracy and practicability of the method. The results show that this improved method adopted to measure the importance of traffic nodes from the global scope has the advantages of fast calculation and simple process and provides a more reliable basis for rational allocation of transport resources.

Direct electricity generation from water flow/evaporation, coined hydrovoltaic effect, has recently attracted intense interest as a facile approach to harvest green energy from ubiquitous capillary water flow or evaporation. However, the current hydrovoltaic device is inferior in output power efficiency compared to other renewable energy devices. Slow water evaporation rate and inefficient charge collection at device electrodes are two fundamental drawbacks limiting energy output efficiency. Here, we report a bioinspired hierarchical porous fabric electrode that enables high water evaporation rate, efficient charge collection, and rapid charge transport in nanostructured silicon-based hydrovoltaic devices. Such an electrode can efficiently collect charges generated in nanostructured silicon as well as induce a prompt water evaporation rate. At room temperature, the device can generate an open-circuit voltage (Voc) of 550 mV and a short-current density (Jsc) of 22 μA·cm-2. It can output a power density over 10 μW·cm-2, which is 3 orders of magnitude larger than all those reported for analogous hydrovoltaic devices. Our results could supply an effective strategy for the development of high-performance hydrovoltaic devices through optimizing electrode structures.

Lead-free ceramic capacitors play an important role in electrical energy storage devices because of their ultrafast charge/discharge rates and high power density. However, simultaneously obtaining large energy storage capability, high efficiency and superior temperature stability has been a huge challenge for practical applications of ceramic capacitors. Here, the relaxor ferroelectric (1-x)[0.8Bi0.5Na0.5TiO3-0.2Ba(Zr0.3Ti0.7)O3]-xSr0.7La0.2TiO3 ((1-x)(BNT-BZT)-xSLT) ceramics are prepared through solid-state reaction method to obtain excellent comprehensive energy storage performances. Particularly, high recoverable energy density (Wrec ∼ 2.6 J/cm3) as well as superior efficiency (η ∼ 92.2 %) can be achieved simultaneously under 210 kV/cm with composition of x = 0.3. Meanwhile, the corresponding ceramic shows excellent temperature (20−140 °C), frequency (1−200 Hz) and cycle stabilities (106 st). Additionally, the 0.7(BNT-BZT)-0.3SLT ceramic also displays high power density (PD ∼ 38.8 MW/cm3) and extremely short discharge time (τ0.9 ∼ 0.11 μs). Therefore, this study provides a useful guideline for designing novel BNT-based ceramics with superior comprehensive energy storage performances.

Anaerobic digestion is promising for waste activated sludge (WAS) degradation. However, conventional processes were generally stuck with limited hydrolysis and poor pathogen destruction. Hyperthermophilic digestion at 70 °C has drawn attention in overcoming those issues at a relatively low energy requirement and operating difficulties. In order to illuminate its operation characteristics, a single-stage hyperthermophilic digester was controlled at 70 °C and operated continuously to degrade WAS. 88.7 mL/g VSadded of methane yield could be achieved in the hyperthermophilic system, fourfold higher than that in the mesophilic system. Kinetic analysis revealed that hyperthermophilic digestion was advantageous in converting the non-degradable fraction. Consequently, hydrolysis under the hyperthermophilic condition was able to be significantly improved. Above 10 d was necessary for the hyperthermophilic system to gain such a high methane production. In the case of stability, the organic loading of higher than 10.2 g VS/L/d resulted in increasing limitation from methanogenesis and accumulation of propionic, butyric and valeric acids. In addition to the dominant acetoclastic genus Methanothrix for methane production in the hyperthermophilic system, two hydrogenotrophic methanogens Methanospirillum and Methanothermobacter reached 18.84% and 8.31%, respectively. The genus Coprothermobacter, affiliated with the phylum Firmicutes, made more contribution to protein hydrolysis in the hyperthermophilic digester.

Objective: The absolute image reconstruction problem of electrical impedance tomography (EIT) is ill-posed. Traditional methods usually solve a nonlinear least squares problem with some kind of regularization. These methods suffer from low accuracy, poor anti-noise performance, and long computation time. Besides, the integration of a priori information is not very flexible. This work tries to solve EIT inverse problem using a machine learning algorithm for the application of thorax imaging. Methods: We developed the supervised descent learning EIT (SDL-EIT) inversion algorithm based on the idea of supervised descent method (SDM). The algorithm approximates the mapping from measured data to the conductivity image by a series of descent directions learned from training samples. We designed a training data set in which the thorax contour, and some general structure of lungs, and heart are embedded. The algorithm is implemented in both two-, and three-dimensional cases, and is evaluated using synthetic, and measured thoracic data. Results, and conclusion: For synthetic data, SDL-EIT shows better accuracy, and anti-noise performance compared with traditional Gauss–Newton inversion (GNI) method. For measured data, the result of SDL-EIT is reasonable compared with computed tomography (CT) scan image. Significance: Using SDL-EIT, prior information can be easily integrated through the specifically designed training data set, and the image reconstruction process can be accelerated. The algorithm is effective in inverting measured thoracic data. It is a potential algorithm for human thorax imaging.

Deep learning (DL) can effectively extract the features of attack behaviours and identify unknown attack behaviours. However, the current DL-based methods separately learn spatial feature and temporal features and fail to consider the spatiotemporal correlation of cyber events. To make up for the gap, this paper proposes an identification strategy for unknown attack behaviours through the joint learning of spatiotemporal features. First, a double-layer long short-term memory (LSTM) was adopted to learn the spatial features of data packet and the temporal feature of the network flow, which makes the attack behaviour recognition less dependent on prior knowledge. Next, the temporal attention was constructed to suppress the noises in the spatial features of the data packet; the spatial attention was designed to reduce the temporal features of low-density information; the spatial attention was fused with the temporal attention to establish the spatiotemporal dependence of cyber-attack behaviours and distinguish the importance of spatiotemporal features. Finally, our identification strategy was experimentally compared with the identification models solely based on spatial features or temporal features. The comparison shows that our strategy outperformed the contrastive models by 2% in recognition accuracy. Thus, the fusion between spatial and temporal features can effectively promote the identification accuracy of unknown attack behaviours.