High-entropy engineering with regulated defect structure and electron interaction tuning active sites for trifunctional electrocatalysis.

High-entropy alloy nanoparticles (HEANs) possessing regulated defect structure and electron interaction exhibit a guideline for constructing multifunctional catalysts. However, the microstructure-activity relationship between active sites of HEANs for multifunctional electrocatalysts is rarely reported. In this work, HEANs distributed on multi-walled carbon nanotubes (HEAN/CNT) are prepared by Joule heating as an example to explain the mechanism of trifunctional electrocatalysis for oxygen reduction, oxygen evolution, and hydrogen evolution reaction. HEAN/CNT excels with unmatched stability, maintaining a 0.8V voltage window for 220 h in zinc-air batteries. Even after 20 h of water electrolysis, its performance remains undiminished, highlighting exceptional endurance and reliability. Moreover, the intrinsic characteristics of the defect structure and electron interaction for HEAN/CNT are investigated in detail. The electrocatalytic mechanism of trifunctional electrocatalysis of HEAN/CNT under different conditions is identified by in situ monitoring and theoretical calculation. Meanwhile, the electron interaction and adaptive regulation of active sites in the trifunctional electrocatalysis of HEANs were further verified by density functional theory. These findings could provide unique ideas for designing inexpensive multifunctional high-entropy electrocatalysts.

Scalable Production of Highly Conductive 2D NbSe2 Monolayers with Superior Electromagnetic Interference Shielding Performance.

Thin, flexible, and electrically conductive films are in demand for electromagnetic interference (EMI) shielding. Two-dimensional NbSe2 monolayers have an electrical conductivity comparable to those of metals (106-107 S m-1) but are challenging for high-quality and scalable production. Here, we show that electrochemical exfoliation of flake NbSe2 powder produces monolayers on a large scale (tens of grams), at a high yield (>75%, monolayer), and with a large average lateral size (>20 µm). The as-exfoliated NbSe2 monolayer flakes are easily dispersed in diverse organic solvents and solution-processed into various macroscopic structures (e.g., free-standing films, coatings, patterns, etc.). Thermal annealing of the free-standing NbSe2 films reduces the interlayer distance of restacked NbSe2 from 1.18 to 0.65 nm and consequently enhances the electrical conductivity to 1.16 × 106 S m-1, which is superior to those of MXenes and reduced graphene oxide. The optimized NbSe2 film shows an EMI shielding effectiveness (SE) of 65 dB at a thickness of 5 µm (>110 dB for a 48-µm-thick film), among the highest in materials of similar thicknesses. Moreover, a laminate of two layers of the NbSe2 film (2 µm thick) with an insulating interlayer shows a high SE of 85 dB, surpassing that of the 20-µm-thick NbSe2 film (83 dB). A two-layer theoretical model is proposed, and it agrees with the experimental EMI SE of the laminated NbSe2 films. The ability to produce NbSe2 monolayers on a tens of grams scale will enable their diverse applications beyond EMI shielding.

Dynamically Cross-Linked, Self-Healable, and Stretchable All-Hydrogel Supercapacitor with Extraordinary Energy Density and Real-Time Pressure Sensing.

Wearable electronics with flexible, integrated, and self-powered multi-functions are becoming increasingly attractive, but their basic energy storage units are challenged in simultaneously high energy density, self-healing, and real-time sensing capability. To achieve this, a fully flexible and omni-healable all-hydrogel, that is dynamically crosslinked PVA@PANI hydrogel, is rationally designed and constructed via aniline/DMSO-emulsion-templated in situ freezing-polymerization strategy. The PVA@PANI sheet, not only possesses a honeycombed porous conductive mesh configuration with superior flexibility that provides numerous channels for unimpeded ions/electron transport and maximizes the utilization efficiency of pseudocapacitive PANI, but also can conform to complicated body surface, enabling effective detection and discrimination of body movements. As a consequence, the fabricated flexible PVA@PANI sheet electrode demonstrates an unprecedented specific capacitance (936.8 F g-1 ) and the assembled symmetric flexible all-solid-state supercapacitor delivers an extraordinary energy density of 40.98 Wh kg-1 , outperforming the previously highest-reported values of stretchable PVA@PANI hydrogel-based supercapacitors. What is more, such a flexible supercapacitor electrode enables precisely monitoring the full-range human activities in real-time, and fulfilling a quick response and excellent self-recovery. These outstanding flexible sensing and energy storage performances render this emerging PVA@PANI hydrogel highly promising for the next-generation wearable self-powered sensing electronics.

Sketch-Segformer: Transformer-Based Segmentation for Figurative and Creative Sketches.

Sketch is a well-researched topic in the vision community by now. Sketch semantic segmentation in particular, serves as a fundamental step towards finer-level sketch interpretation. Recent works use various means of extracting discriminative features from sketches and have achieved considerable improvements on segmentation accuracy. Common approaches for this include attending to the sketch-image as a whole, its stroke-level representation or the sequence information embedded in it. However, they mostly focus on only a part of such multi-facet information. In this paper, we for the first time demonstrate that there is complementary information to be explored across all these three facets of sketch data, and that segmentation performance consequently benefits as a result of such exploration of sketch-specific information. Specifically, we propose the Sketch-Segformer, a transformer-based framework for sketch semantic segmentation that inherently treats sketches as stroke sequences other than pixel-maps. In particular, Sketch-Segformer introduces two types of self-attention modules having similar structures that work with different receptive fields (i.e., whole sketch or individual stroke). The order embedding is then further synergized with spatial embeddings learned from the entire sketch as well as localized stroke-level information. Extensive experiments show that our sketch-specific design is not only able to obtain state-of-the-art performance on traditional figurative sketches (such as SPG, SketchSeg-150K datasets), but also performs well on creative sketches that do not conform to conventional object semantics (CreativeSketch dataset) thanks for our usage of multi-facet sketch information. Ablation studies, visualizations, and invariance tests further justifies our design choice and the effectiveness of Sketch-Segformer. Codes are available at https://github.com/PRIS-CV/Sketch-SF.

Self-sacrifice-template epitaxial growth of hierarchical MnO2@NiCo2O4 heterojunction electrode for high-performance asymmetric supercapacitor.

Constructing three-dimensional (3D) hierarchical bimetallic pseudocapacitive materials with abundant opening channel and heterojunction structures is rather promising but still challenging for high-performance supercapacitors. Herein, a self-sacrifice-template epitaxial growth strategy was proposed for the first time to construct 3D hierarchical bimetallic pseudocapacitive material. By using this strategy, NiCo2O4 nanowires (NiCo2O4NW) arrayed randomly to form a porous shell via in-situ epitaxial growth fully enclosing a MnO2 tube core, forming multiple transport channels and nano-heterojunctions between MnO2 and NiCo2O4NW, which facilitates electron transfer, i.e. exhibiting high electronic conductivity than any single component. As a result of the self-sacrifice-template epitaxial growth method, special hollow tectorum-like 3D hierarchical structure with considerable inter-nanowire space and hollow interior space enables easy access of electrolyte to NiCo2O4NW surface and MnO2 core, thereby resulting in highly exposed redox active sites of MnO2 core and NiCo2O4NW shell for energy storage. Comprehensive evaluations confirmed MnO2@NiCo2O4NW was a supercapacitor electrode candidate, delivering a superior energy density of 106.37 Wh kg-1. Such performance can be ascribed to the synergistic coupling effect of 3D hierarchical tube and nano-heterojunction structures. The proposed self-sacrifice-template epitaxial growth strategy provides important guidance for designing high-performance energy storage materials.

Regulating the Lithium Ions' Local Coordination Environment through Designing a COF with Single Atomic Co Site to Achieve Dendrite-Free Lithium-Metal Batteries.

The detrimental growth of lithium dendrites and unstable solid electrolyte interphase (SEI) inhibit the practical application of lithium-metal batteries. Herein, atomically dispersed cobalt coordinate conjugated bipyridine-rich covalent organic framework (sp2 c-COF) is explored as an artificial SEI on the surface of the Li-metal anode to resolve these issues. The single Co atoms confined in the structure of COF enhance the number of active sites and promote electron transfer to the COF. The synergistic effects of the Co─N coordination and strong electron-withdrawing cyano-group can adsorb the electron from the donor (Co) at a maximum and create an electron-rich environment, hence further regulating the Li+ local coordination environment and achieving uniform Li-nucleation behavior. Furthermore, in situ technology and density functional theory calculations reveal the mechanism of the sp2 c-COF-Co inducing Li uniform deposition and promoting Li+ rapid migration. Based on these advantages, the sp2 c-COF-Co modified Li anode exhibits a low Li-nucleation barrier of 8 mV, and excellent cycling stability of 6000 h.

Searching for Network Width With Bilaterally Coupled Network.

Searching for a more compact network width recently serves as an effective way of channel pruning for the deployment of convolutional neural networks (CNNs) under hardware constraints. To fulfil the searching, a one-shot supernet is usually leveraged to efficiently evaluate the performance w.r.t. different network widths. However, current methods mainly follow a unilaterally augmented (UA) principle for the evaluation of each width, which induces the training unfairness of channels in supernet. In this article, we introduce a new supernet called Bilaterally Coupled Network (BCNet) to address this issue. In BCNet, each channel is fairly trained and responsible for the same amount of network widths, thus each network width can be evaluated more accurately. Besides, we propose to reduce the redundant search space and present the BCNetV2 as the enhanced supernet to ensure rigorous training fairness over channels. Furthermore, we leverage a stochastic complementary strategy for training the BCNet, and propose a prior initial population sampling method to boost the performance of the evolutionary search. We also propose a new open-source width search benchmark on macro structures named Channel-Bench-Macro for the better comparisons of the width search algorithms with MobileNet- and ResNet-like architectures. Extensive experiments on the benchmark datasets demonstrate that our method can achieve state-of-the-art performance.

On the Comparisons of Decorrelation Approaches for Non-Gaussian Neutral Vector Variables.

As a typical non-Gaussian vector variable, a neutral vector variable contains nonnegative elements only, and its l1 -norm equals one. In addition, its neutral properties make it significantly different from the commonly studied vector variables (e.g., the Gaussian vector variables). Due to the aforementioned properties, the conventionally applied linear transformation approaches [e.g., principal component analysis (PCA) and independent component analysis (ICA)] are not suitable for neutral vector variables, as PCA cannot transform a neutral vector variable, which is highly negatively correlated, into a set of mutually independent scalar variables and ICA cannot preserve the bounded property after transformation. In recent work, we proposed an efficient nonlinear transformation approach, i.e., the parallel nonlinear transformation (PNT), for decorrelating neutral vector variables. In this article, we extensively compare PNT with PCA and ICA through both theoretical analysis and experimental evaluations. The results of our investigations demonstrate the superiority of PNT for decorrelating the neutral vector variables.

Learning Calibrated Class Centers for Few-Shot Classification by Pair-Wise Similarity.

Metric-based methods achieve promising performance on few-shot classification by learning clusters on support samples and generating shared decision boundaries for query samples. However, existing methods ignore the inaccurate class center approximation introduced by the limited number of support samples, which consequently leads to biased inference. Therefore, in this paper, we propose to reduce the approximation error by class center calibration. Specifically, we introduce the so-called Pair-wise Similarity Module (PSM) to generate calibrated class centers adapted to the query sample by capturing the semantic correlations between the support and the query samples, as well as enhancing the discriminative regions on support representation. It is worth noting that the proposed PSM is a simple plug-and-play module and can be inserted into most metric-based few-shot learning models. Through extensive experiments in metric-based models, we demonstrate that the module significantly improves the performance of conventional few-shot classification methods on four few-shot image classification benchmark datasets. Codes are available at: https://github.com/PRIS-CV/Pair-wise-Similarity-module.

Distributed interval state estimation with l∞-gain optimization for cyber-physical systems subject to bounded disturbance and random stealthy attacks.

This paper studies the distributed interval state estimation problem for cyber-physical systems with bounded disturbance and random stealthy attacks. Since conventional interval observers cannot complete the task of real-time monitoring system under random attacks, an attack-resistant distributed interval observer is designed by using attack frequency and interval attack estimation. Using the designed observer, upper- and lower-bounding estimation error systems are modeled by positive interconnected systems with hybrid deterministic and random bounded inputs. To explicitly attenuate the effect of disturbance and attacks, the resulting deterministic positive error system between upper- and lower-bounding estimates is formulated. By linear programming, the results of interval observer design and l∞-gain optimization are proposed. The remote monitoring of vehicle lateral dynamic is given for numerical verification of the results.

GPCA: A Probabilistic Framework for Gaussian Process Embedded Channel Attention.

Channel attention mechanisms have been commonly applied in many visual tasks for effective performance improvement. It is able to reinforce the informative channels as well as to suppress the useless channels. Recently, different channel attention modules have been proposed and implemented in various ways. Generally speaking, they are mainly based on convolution and pooling operations. In this paper, we propose Gaussian process embedded channel attention (GPCA) module and further interpret the channel attention schemes in a probabilistic way. The GPCA module intends to model the correlations among the channels, which are assumed to be captured by beta distributed variables. As the beta distribution cannot be integrated into the end-to-end training of convolutional neural networks (CNNs) with a mathematically tractable solution, we utilize an approximation of the beta distribution to solve this problem. To specify, we adapt a Sigmoid-Gaussian approximation, in which the Gaussian distributed variables are transferred into the interval [0,1]. The Gaussian process is then utilized to model the correlations among different channels. In this case, a mathematically tractable solution is derived. The GPCA module can be efficiently implemented and integrated into the end-to-end training of the CNNs. Experimental results demonstrate the promising performance of the proposed GPCA module. Codes are available at https://github.com/PRIS-CV/GPCA.

Dirichlet Process Mixture of Generalized Inverted Dirichlet Distributions for Positive Vector Data With Extended Variational Inference.

A Bayesian nonparametric approach for estimation of a Dirichlet process (DP) mixture of generalized inverted Dirichlet distributions [i.e., an infinite generalized inverted Dirichlet mixture model (InGIDMM)] has been proposed. The generalized inverted Dirichlet distribution has been proven to be efficient in modeling the vectors that contain only positive elements. Under the classical variational inference (VI) framework, the key challenge in the Bayesian estimation of InGIDMM is that the expectation of the joint distribution of data and variables cannot be explicitly calculated. Therefore, numerical methods are usually applied to simulate the optimal posterior distributions. With the recently proposed extended VI (EVI) framework, we introduce lower bound approximations to the original variational objective function in the VI framework such that an analytically tractable solution can be derived. Hence, the problem in numerical simulation has been overcome. By applying the DP mixture technique, an InGIDMM can automatically determine the number of mixture components from the observed data. Moreover, the DP mixture model with an infinite number of mixture components also avoids the problems of underfitting and overfitting. The performance of the proposed approach is demonstrated with both synthesized data and real-life data applications.

Advanced Dropout: A Model-Free Methodology for Bayesian Dropout Optimization.

Due to lack of data, overfitting ubiquitously exists in real-world applications of deep neural networks (DNNs). We propose advanced dropout, a model-free methodology, to mitigate overfitting and improve the performance of DNNs. The advanced dropout technique applies a model-free and easily implemented distribution with parametric prior, and adaptively adjusts dropout rate. Specifically, the distribution parameters are optimized by stochastic gradient variational Bayes in order to carry out an end-to-end training. We evaluate the effectiveness of the advanced dropout against nine dropout techniques on seven computer vision datasets (five small-scale datasets and two large-scale datasets) with various base models. The advanced dropout outperforms all the referred techniques on all the datasets. We further compare the effectiveness ratios and find that advanced dropout achieves the highest one on most cases. Next, we conduct a set of analysis of dropout rate characteristics, including convergence of the adaptive dropout rate, the learned distributions of dropout masks, and a comparison with dropout rate generation without an explicit distribution. In addition, the ability of overfitting prevention is evaluated and confirmed. Finally, we extend the application of the advanced dropout to uncertainty inference, network pruning, text classification, and regression. The proposed advanced dropout is also superior to the corresponding referred methods. Codes are available at https://github.com/PRIS-CV/AdvancedDropout.

Progressive Learning of Category-Consistent Multi-Granularity Features for Fine-Grained Visual Classification.

Fine-grained visual classification (FGVC) is much more challenging than traditional classification tasks due to the inherently subtle intra-class object variations. Recent works are mainly part-driven (either explicitly or implicitly), with the assumption that fine-grained information naturally rests within the parts. In this paper, we take a different stance, and show that part operations are not strictly necessary - the key lies with encouraging the network to learn at different granularities and progressively fusing multi-granularity features together. In particular, we propose: (i) a progressive training strategy that effectively fuses features from different granularities, and (ii) a consistent block convolution that encourages the network to learn the category-consistent features at specific granularities. We evaluate on several standard FGVC benchmark datasets, and demonstrate the proposed method consistently outperforms existing alternatives or delivers competitive results. Codes are available at https://github.com/PRIS-CV/PMG-V2.

DS-UI: Dual-Supervised Mixture of Gaussian Mixture Models for Uncertainty Inference in Image Recognition.

This paper proposes a dual-supervised uncertainty inference (DS-UI) framework for improving Bayesian estimation-based UI in DNN-based image recognition. In the DS-UI, we combine the classifier of a DNN, i.e., the last fully-connected (FC) layer, with a mixture of Gaussian mixture models (MoGMM) to obtain an MoGMM-FC layer. Unlike existing UI methods for DNNs, which only calculate the means or modes of the DNN outputs' distributions, the proposed MoGMM-FC layer acts as a probabilistic interpreter for the features that are inputs of the classifier to directly calculate the probabilities of them for the DS-UI. In addition, we propose a dual-supervised stochastic gradient-based variational Bayes (DS-SGVB) algorithm for the MoGMM-FC layer optimization. Unlike conventional SGVB and optimization algorithms in other UI methods, the DS-SGVB not only models the samples in the specific class for each Gaussian mixture model (GMM) in the MoGMM, but also considers the negative samples from other classes for the GMM to reduce the intra-class distances and enlarge the inter-class margins simultaneously for enhancing the learning ability of the MoGMM-FC layer in the DS-UI. Experimental results show the DS-UI outperforms the state-of-the-art UI methods in misclassification detection. We further evaluate the DS-UI in open-set out-of-domain/-distribution detection and find statistically significant improvements. Visualizations of the feature spaces demonstrate the superiority of the DS-UI. Codes are available at https://github.com/PRIS-CV/DS-UI.

A low-temperature solution-processed indium incorporated zinc oxide electron transport layer for high-efficiency lead sulfide colloidal quantum dot solar cells.

Colloidal quantum dot solar cells (CQDSCs) have achieved remarkable progress recently in terms of mainly surface passivation and composition-matching matrices on CQDs, while improving the overall photoelectric conversion efficiency (PCE) through electron transport layer (ETL) modifications is less explored. We report a low-temperature solution route to synthesize donor (Al3+/Ga3+/In3+) incorporated zinc oxide (AZO/GZO/IZO) ETL films for PbS CQDSCs. Spectroscopic characterization studies indicate that the IZO ETL fabricated with 150 °C annealing can increase the bandgap the most from 3.56 eV to 3.74 eV, possesses enhanced light transmission (∼94%) and finer particle sizes, and importantly shows the most suitable band alignment and charge transfer ability. Well-dispersed PbS CQDs of around 3 nm are synthesized by a N2-protected reflux method and are surface exchanged with 1-ethyl-3-methylimidazolium iodide (EMII) to allow I- grafting and ethanedithiol (EDT) for the active layer and hole transport layer, respectively. The IZO based PbS CQDSC, with a device architecture of ITO/IZO/PbS-EMII/PbS-EDT/Au, shows an enhanced PCE of 11.1% (comparatively 18% higher than that of the ZnO ETL), a VOC value of 0.64 V, and a JSC of 25.8 mA cm-2. The improved performances benefit from the higher recombination resistance and constrained photoluminescence emission with the utilization of the IZO ETL that provides a superior charge transfer property.

Lattice Defect Engineering Enables Performance-Enhanced MoS2 Photodetection through a Paraelectric BaTiO3 Dielectric.

Carrier mobility and density are intrinsically important in nanophoto/electronic devices. High-dielectric-constant coupled polarization-field gate ferroelectrics are frequently studied and partially capable in achieving large-scale tuning of photoresponse, but their light absorption and carrier density seem generally ineffective. This raises questions about whether a similarly high-dielectric-constant paraelectric gate dielectric could enable tuning and how the principles involved could be established. In this study, by deliberately introducing lattice defects in high-dielectric-constant paraelectric, cubic BaTiO3 (c-BTO) was explored to fabricate MoS2 photodetectors with ultrahigh detection ability and outstanding field-effect traits. An organic-metal-based spin-coating cum annealing method was used for the c-BTO synthesis, with an optimized thickness (300 nm), by introducing lattice defects properly but maintaining a large dielectric constant (55 at 1k Hz) and low dielectric loss (0.06 at 1k Hz), which renders the enhanced visible-light region absorption. As a result of the synergistically enhanced mobility and photoabsorption, the MoS2/BTO FET exhibits promising merits, for example, on/off ratio, subthreshold swing, and mobilities for high-performance photodetectors with excellent responsivity (600 AW-1) and detectivity (1.25 × 1012 Jones). Thus, this work facilitates the establishment of a lattice defect induced sub-bandgap absorption landmap for synergistically enhanced photoresponse for high-performance photodetector exploration.

Influence of Cu on the Improvement of Magnetic Properties and Structure of L10 FePt Nanoparticles.

L10 ordered FePt and FePtCu nanoparticles (NPs) with a good dispersion were successfully fabricated by a simple, green, one-step solid-phase reduction method. Fe (acac)3, Pt (acac)2, and CuO as the precursors were dispersed in NaCl and annealed at different temperatures with an H2-containing atmosphere. As the annealing temperature increased, the chemical order parameter (S), average particle size (D), coercivity (Hc), and saturation magnetization (Ms) of FePt and FePtCu NPs increased and the size distribution range of the particles became wider. The ordered degree, D, Hc, and Ms of FePt NPs were greatly improved by adding 5% Cu. The highest S, D, Hc, and Ms were obtained when FePtCu NPs annealed at 750 °C, which were 0.91, 4.87 nm, 12,200 Oe, and 23.38 emu/g, respectively. The structure and magnetic properties of FePt and FePtCu NPs at different annealing temperatures were investigated and the formation mechanism of FePt and FePtCu NPs were discussed in detail.

AP-CNN: Weakly Supervised Attention Pyramid Convolutional Neural Network for Fine-Grained Visual Classification.

Classifying the sub-categories of an object from the same super-category (e.g., bird species and cars) in fine-grained visual classification (FGVC) highly relies on discriminative feature representation and accurate region localization. Existing approaches mainly focus on distilling information from high-level features. In this article, by contrast, we show that by integrating low-level information (e.g., color, edge junctions, texture patterns), performance can be improved with enhanced feature representation and accurately located discriminative regions. Our solution, named Attention Pyramid Convolutional Neural Network (AP-CNN), consists of 1) a dual pathway hierarchy structure with a top-down feature pathway and a bottom-up attention pathway, hence learning both high-level semantic and low-level detailed feature representation, and 2) an ROI-guided refinement strategy with ROI-guided dropblock and ROI-guided zoom-in operation, which refines features with discriminative local regions enhanced and background noises eliminated. The proposed AP-CNN can be trained end-to-end, without the need of any additional bounding box/part annotation. Extensive experiments on three popularly tested FGVC datasets (CUB-200-2011, Stanford Cars, and FGVC-Aircraft) demonstrate that our approach achieves state-of-the-art performance. Models and code are available at https://github.com/PRIS-CV/AP-CNN_Pytorch-master.

Ultrafine Ru species within confined space: An efficient adsorbent for ultra-deep desulfurization of benzene.

The ultra-deep adsorptive desulfurization (ppb level) of benzene remains a challenging subject with the need to construct efficient adsorbent systems. Herein, a kind of ruthenium-based adsorbent functionalized with bimetallic Ru-Al was rationally designed using Al2O3 as support (denoted as 0.8%Ru-1.2%Al/Al2O3). It was found that the co-anchoring of Ru and Al species endows the Ru-based adsorbent unique adsorption capability, which is able to completely eliminate sulfur compounds in benzene, and exhibiting a much higher breakthrough sulfur capacity than that of the 0.8%Ru/Al2O3. Remarkably, under the industrial experiment conditions, 0.8%Ru-1.2%Al/Al2O3 exhibited excellent long-term stability for more than 1200 h, showing the potential for industrial application. Various characterization techniques, including BET, XRD, SEM, TEM, TPD-MS, TPR and XPS, were used to investigate the correlation between the adsorption performance and the microstructure of the adsorbents. Over 0.8%Ru-1.2%Al/Al2O3, the ultra-thin aluminum additive is beneficial to improve the dispersion of Ru species, which therefore exhibits desirable desulfurization efficiency. Moreover, the enhanced performance is also correlated to the presence of the suitable Ru active centers generated from the selective coverage by Al species. It leads to an optimal exposure of the Ru active centers, which would facilitate the interaction of S-Ru and the improvement of the desulfurization activity.