引入特征交互的红外与可见光图像自适应融合北大核心CSCD

在图像融合领域,现有的基于卷积神经网络(CNN)或Transformer架构的方法存在两个局限性:首先,浅层纹理特征与深层语义特征之间无法有效聚合;其次,红外与可见光特征的权重比例无法自适应变化。本文提出一种引入特征交互的红外与可见光图像自适应融合方法。首先,构建一种基于Transformer的特征交互模块,聚合跨尺度特征信息,增强特征表达能力。其次,设计一种融合模块,自适应地调整特征权重比例。所提出的融合方法通过两阶段训练策略完成。第一个阶段,应用创新的特征交互概念训练编码器,增强特征表达,重建特征图像。第二个阶段,基于设计的权重自适应调整模块训练红外与可见光特征融合任务。公开数据集的实验结果表明,与现有方法相比,本方法在主观和客观的评价方面均优于其他典型方法。     

Room-Temperature Metal-Catalyzed Ultrafast Gasification of Ultrathin Boron Flakes

The trivalent outer shell of boron renders this element electron-poor but chemically rich, exhibiting more than one dozen allotropes. Its 2D polymorph has been recently synthesized on metal substrates under ultrahigh vacuum and has attracted intense interest. However, probing its properties ex situ has been challenging due to the quality degradation—surface oxidation—that occurs upon exposure to ambient environments. Herein, this surface chemistry is investigated in regard to the air stability of ultrathin boron flakes on metals prepared by atmospheric-pressure chemical vapor deposition. The characteristic Volmer–Weber growth is recognized by the stacking of polygon-shaped, thin flakes as isolated islands. Significantly, the metal-catalyzed, ultrafast gasification of boron flakes at room temperature, exemplified by the complete, spontaneous vanishment of 200 nm-thick boron islands in 3 h is observed. A two-step mechanism, first oxygen-involved surface oxidation and then subsequent reactions with water forming a highly volatile boric acid layer, is unambiguously revealed by combined surface characterizations. The catalysis by metal substrates, corroborated by theoretical calculations, is attributed as the crucial cause of the unprecedented gasification. The concept of oxygen-free growth is thereby proposed for air-sensitive material growth by introducing in situ oxygen scavengers. These findings significantly expand the fundamental understanding of the surface chemistry of boron and pave the way for the chemical vapor deposition growth of hydrophobic materials.     

Selenium Substitution in Bithiophene Imide Polymer Semiconductors Enables High-Performance n-Type Organic Thermoelectric

Designing n-type polymers with high electrical conductivity remains a major challenge for organic thermoelectrics (OTEs). Herein, by devising a novel selenophene-based electron-deficient building block, the pronounced advantages of selenium substitution in simultaneously enabling advanced n-type polymers is demonstrated with high mobility (≈2 orders of magnitude higher versus their sulfur-based analogues due to both intensified intra- and inter-chain interactions) and much improved n-doping efficiency (enabled by the largely lowered LUMO level with a ≈0.2 eV margin) of the resulting polymers. Via side chain optimization and donor engineering, the selenium-substituted polymer, f-BSeI2TEG-FT, achieves a highest conductivity of 103.5 S cm⁻¹ and power factor of 70.1 µW m⁻¹ K⁻², which are among the highest values reported in literature for n-type polymers, and f-BSeI2TEG-FT greatly outperformed the sulfur-based analogue polymer by 40% conductivity increase. These results demonstrate that selenium substitution is a very effective strategy for improving n-type performance and provide important structure-property correlations for developing high-performing n-type OTE materials.     

Light-Fueled Nonequilibrium and Adaptable Hydrogels for Highly Tunable Autonomous Self-Oscillating Functions

Nonequilibrium oscillation fueled by dissipating chemical energy is ubiquitous in living systems for realizing a broad range of complex functions. The design of synthetic materials that can mimic their biological counterparts in the production of dissipative structures and autonomous oscillations is of great interest but remains challenging. Here, a series of environmentally adaptable hydrogels functionalized with photoswitchable spiropyran derivatives that display a tunable equilibrium-shifting capability, thus endowing those hydrogels with a high degree of freedom and flexibility is reported. Such nonequilibrium hydrogels are able to responsively adapt their shapes under constant light illumination due to asymmetric deswelling, which in turn generates self-shadowing and consequently creates autonomous self-oscillating behaviors through a negative feedback process. Diverse oscillation modes including bending, twisting, and snap-through buckling with tunable frequency and amplitude are widely observed in three different molecular systems. Density functional theory calculations and finite element simulations further demonstrated the robustness of such a photoadaptable self-oscillation mechanism. This study provides a useful molecular design strategy for construction of highly adaptable hydrogels with potential applications in self-sustained soft robots and autonomous devices.     

Efficient Inverted Perovskite Solar Cells with a Fill Factor Over 86% via Surface Modification of the Nickel Oxide Hole Contact

The poor interface quality between nickel oxide (NiO_x) and halide perovskites limits the performance and stability of NiO_x-based perovskite solar cells (PSCs). Here a reactive surface modification approach based on the in situ decomposition of urea on the NiO_x surface is reported. The pyrolysis of urea can reduce the high-valence state of nickel and replace the adsorbed hydroxyl group with isocyanate. Combining theoretical and experimental analyses, the treated NiO_x films present suppressed surface states and improved transport energy level alignment with the halide perovskite absorber. With this strategy, NiO_x-based PSCs achieve a champion power conversion efficiency (PCE) of 23.61% and a fill factor of over 86%. The device's efficiency remains above 90% after 2000 h of thermal aging at 85 °C. Furthermore, perovskite solar modules achieve PCE values of 18.97% and 17.18% for areas of 16 and 196 cm², respectively.     

An Intrinsically Nonflammable Electrolyte for Prominent-Safety Lithium Metal Batteries with High Energy Density and Cycling Stability

Nex-generation high-energy-density storage battery, assembled with lithium (Li)-metal anode and nickel-rich cathode, puts forward urgent demand for advanced electrolytes that simultaneously possess high security, wide electrochemical window, and good compatibility with electrode materials. Herein an intrinsically nonflammable electrolyte is designed by using 1 M lithium difluoro(oxalato)borate (LiDFOB) in triethyl phosphate (TEP) and N-methyl-N-propyl-pyrrolidinium bis(trifluoromethylsulfonyl)imide [Pyr₁₃][TFSI] ionic liquid (IL) solvents. The introduction of IL can bring plentiful organic cations and anions, which provides a cation shielding effect and regulates the Li⁺ solvation structure with plentiful Li⁺-DFOB⁻ and Li⁺-TFSI⁻ complexes. The unique Li⁺ solvation structure can induce stable anion-derived electrolyte/electrode interphases, which effectively inhibit Li dendrite growth and suppress side reactions between TEP and electrodes. Therefore, the LiNi_0.9Co_0.05Mn_0.05O₂ (NCM90)/Li coin cell with this electrolyte can deliver stable cycling even under 4.5 V and 60 °C. Moreover, a Li-metal battery with thick NCM90 cathode (≈ 15 mg cm⁻²) and thin Li-metal anode (≈ 50 µm) (N/P ≈ 3), also reveals stable cycling performance under 4.4 V. And a 2.2 Ah NCM90/Li pouch cell can simultaneously possess prominent safety with stably passing the nail penetration test, and high gravimetric energy density of 470 Wh kg⁻¹ at 4.4 V.     

A Visible and Near-Infrared Light-Fueled Omnidirectional Twist-Bend Crawling Robot

In recent years, although light-driven soft actuators have attracted intense scientific attention and achieved remarkable progress, the design and construction of an intelligent robotic system with maneuverability, self-adaptability, untethered control, and greater freedom of action, in particular the omnidirectional motion capability on a plane, remains challenging. Herein, four types of photo-thermal fillers and an unprecedented twist-bend actuation mode is introduced into a liquid crystal elastomer-based soft robot. The obtained twist-bend crawling robot not only exhibits in situ rotation, four-way turning, and four-way linear motion under light irradiation with four wavelength bands (520, 655, 808, and 980 nm), but also demonstrates the ability to avoid obstacles in complex geographical environments. This work may bring a new perspective for fabrication and development of soft robots that can adapt to dynamic and complex environmental conditions.     

A Conductive 2D Conjugated Tetrathia[8]circulene-Based Nickel Metal–Organic Framework for Energy Storage

Herein, a novel D₄ symmetrical redox-active ligand tetrathia[8]circulene-2,3,5,6,8,9,11,12-octaol (8OH-TTC) is designed and synthesized, which coordinates with Ni²⁺ ions to construct a 2D conductive metal-organic framework (2D c-MOF) named Ni-TTC. Ni-TTC exhibits typical semiconducting properties with electrical conductivity up to ≈1.0 S m⁻¹ at 298 K. Furthermore, magnetism measurements show the paramagnetic property of Ni-TTC with strong antiferromagnetic coupling due to the presence of semiquinone ligand radicals and Ni²⁺ sites. In virtue of its decent electrical conductivity and good redox activity, the gravimetric capacitance of Ni-TTC is up to 249 F g⁻¹ at a discharge rate of 0.2 A g⁻¹, which demonstrates the potential of tetrathia[8]circulene-based redox-active 2D c-MOFs in energy storage applications.     

Improving the Reliability of MLC NAND Flash Memories Through Adaptive Data Refresh and Error Control Coding

NAND Flash memory has become the most widely used non-volatile memory technology. We focus on multi-level cell (MLC) NAND Flash memories because they have high storage density. Unfortunately MLC NAND Flash memory also has reliability problems due to narrower threshold voltage gap between logical states. Errors in these memories can be classified into data retention (DR) errors and program interference (PI) errors. DR errors are dominant if the data storage time is longer than 1 day and these errors can be reduced by refreshing the data. PI errors are dominant if the data storage time is less than 1 day and these errors can be handled by error control coding (ECC). In this paper we propose a combination of data refresh policies and low cost ECC schemes that are cognizant of application characteristics to address the errors in MLC NAND Flash memories. First, we use Gray code based encoding to reduce the error rates in the four subpages (MSB-even, LSB-even, MSB-odd, LSB-odd) of a 2-bit MLC NAND Flash memory. Next, we apply data refresh techniques where the refresh interval is a function of the program/erase (P/E) frequency of the application. We show that an appropriate choice of refresh interval and BCH based ECC scheme can minimize memory energy while satisfying the reliability constraint.     

Interface Reaction Between Electroless Ni–Sn–P Metallization and Lead-Free Sn–3.5Ag Solder with Suppressed Ni3P Formation

The voids formed in the Ni₃P layer during reaction between Sn-based solders and electroless Ni–P metallization is an important cause of rapid degradation of solder joint reliability. In this study, to suppress formation of the Ni₃P phase, an electrolessly plated Ni–Sn–P alloy (6–7 wt.% P and 19–21 wt.% Sn) was developed to replace Ni–P. The interfacial microstructure of electroless Ni–Sn–P/Sn–3.5Ag solder joints was investigated after reflow and solid-state aging. For comparison, the interfacial reaction in electroless Ni–P/Sn–3.5Ag solder joints under the same reflow and aging conditions was studied. It was found that the Ni–Sn–P metallization is consumed much more slowly than the Ni–P metallization during soldering. After prolonged reaction, no Ni₃P or voids are observed under SEM at the Ni–Sn–P/Sn–3.5Ag interface. Two main intermetallic compounds, Ni₃Sn₄ and Ni₁₃Sn₈P₃, are formed during the soldering reaction. The reason for Ni₃P phase suppression and the overall mechanisms of reaction at the Ni–Sn–P/Sn–3.5Ag interface are discussed.