Photonic time-delayed reservoir computing based on series-coupled microring resonators with high memory capacity.

On-chip microring resonators (MRRs) have been proposed to construct time-delayed reservoir computing (RC) systems, which offer promising configurations available for computation with high scalability, high-density computing, and easy fabrication. A single MRR, however, is inadequate to provide enough memory for the computation task with diverse memory requirements. Large memory requirements are satisfied by the RC system based on the MRR with optical feedback, but at the expense of its ultralong feedback waveguide. In this paper, a time-delayed RC is proposed by utilizing a silicon-based nonlinear MRR in conjunction with an array of linear MRRs. These linear MRRs possess a high quality factor, providing enough memory capacity for the RC system. We quantitatively analyze and assess the proposed RC structure's performance on three classical tasks with diverse memory requirements, i.e., the Narma 10, Mackey-Glass, and Santa Fe chaotic timeseries prediction tasks. The proposed system exhibits comparable performance to the system based on the MRR with optical feedback, when it comes to handling the Narma 10 task, which requires a significant memory capacity. Nevertheless, the dimension of the former is at least 350 times smaller than the latter. The proposed system lays a good foundation for the scalability and seamless integration of photonic RC.

Data augmentation using continuous conditional generative adversarial networks for regression and its application to improved spectral sensing.

Machine learning-assisted spectroscopy analysis faces a prominent constraint in the form of insufficient spectral samples, which hinders its effectiveness. Meanwhile, there is a lack of effective algorithms to simulate synthetic spectra from limited samples of real spectra for regression models in continuous scenarios. In this study, we introduced a continuous conditional generative adversarial network (CcGAN) to autonomously generate synthetic spectra. The labels employed for generating the spectral data can be arbitrarily selected from within the range of labels associated with the real spectral data. Our approach effectively produced spectra using a small spectral dataset obtained from a self-interference microring resonator (SIMRR)-based sensor. The generated synthetic spectra were subjected to evaluation using principal component analysis, revealing an inability to discern them from the real spectra. Finally, to enhance the DNN regression model, these synthetic spectra are incorporated into the original training dataset as an augmentation technique. The results demonstrate that the synthetic spectra generated by CcGAN exhibit exceptional quality and significantly enhance the predictive performance of the DNN model. In conclusion, CcGAN exhibits promising potential in generating high-quality synthetic spectra and delivers a superior data augmentation effect for regression tasks.

A Temperature and pH Dual-Sensitive Multifunctional Polyurethane with Bacteria-Triggered Antibacterial Activity.

An effective and practical antibacterial strategy is to design multifunctional and stimuli-responsive materials that exhibit antibacterial activity in response to bacterial triggers. In this study, because the metabolism of Staphylococcus aureus (S. aureus) can acidify the surrounding environment and pH level can affect the lower critical solution temperature of temperature/pH dual-sensitive polymers, a monomer containing a temperature-sensitive N-isopropyl amide derivative and pH-sensitive tertiary amine groups is first synthesized. Then, the monomer is copolymerized with a polyurethane chain, and partial tertiary amine groups are quaternized to obtain bactericidal activity. The modified polyurethane exhibits temperature/pH sensitivity, antibacterial adhesion activity, bactericidal activity, and good cytocompatibility. An in situ investigation of bacterial behavior and pH changes in the bacterial suspension during the process confirms that the temperature/pH dual-sensitive polyurethane successfully achieves antibacterial activity though the metabolic activity of S. aureus without external intervention. This design concept provides a new perspective for antibacterial material design.

Regulating the Crystallinity and Self-Aggregation of Fused Ring Electron Acceptors via Branched Side-Chain Engineering for Efficient Organic Solar Cells.

Fine-tuning the crystallinity and self-aggregation features of donors/acceptor materials toward high-efficiency organic solar cells (OSCs) is of crucial importance. Here, a convenient yet effective way to simultaneously control the crystallinity and self-aggregation of the fused ring electron acceptor (FREA) is demonstrated by altering the length of the first-position branched alkyl chain on the cyclic unit. Specifically, three carbazole-based FREAs, 4TC-4F-C6C6, 4TC-4F-C8C8, and 4TC-4F-C10C10, are synthesized by changing the length of the first-position branched alkyl chain on the carbazole unit. The crystallinity of the studied acceptors decreases as the branched alkyl chain is lengthened. The ability of the acceptors to undergo self-aggregation decreases in the order 4TC-4F-C10C10, 4TC-4F-C6C6, and 4TC-4F-C8C8. The medium crystallinity and lower self-aggregation properties of 4TC-4F-C8C8 result in favorable phase separation when blended with poly-[(2,6-(4,8-bis(5-(2-ethylhexyl-3-fluoro)thiophen-2-yl)-benzo[1,2-b:4,5-b']dithiophene))-alt-(5,5-(1',3'-di-2-thienyl-5',7'-bis(2-ethylhexyl)benzo[1',2'-c:4',5'-c']dithiophene-4,8-dione))] (PM6), which is conducive to effective exciton dissociation and charge transport. Consequently, the OSC device based on PM6:4TC-4F-C8C8 delivers the best power conversion efficiency of 14.85%.

CLPVG: Circular limited penetrable visibility graph as a new network model for time series.

A visibility graph transforms time series into graphs, facilitating signal processing by advanced graph data mining algorithms. In this paper, based on the classic limited penetrable visibility graph method, we propose a novel mapping method named circular limited penetrable visibility graph, which replaces the linear visibility line in limited penetrable visibility graph with nonlinear visibility arc for pursuing more flexible and reasonable mapping of time series. Tests on degree distribution and some common network features of the generated graphs from typical time series demonstrate that our circular limited penetrable visibility graph can effectively capture the important features of time series and show higher robust classification performance than the traditional limited penetrable visibility graph in the presence of noise. The experiments on real-world time-series datasets of radio and electroencephalogram signals also suggest that the structural features provided by a circular limited penetrable visibility graph, rather than a limited penetrable visibility graph, are more useful for time-series classification, leading to higher accuracy. This classification performance can be further enhanced through structural feature expansion by adopting subgraph networks. All of these results demonstrate the effectiveness of our circular limited penetrable visibility graph model.

Nonswellable and Tough Supramolecular Hydrogel Based on Strong Micelle Cross-Linkings.

Because of the difference in osmotic pressure, most tough hydrogels swell under physiological conditions, which seriously weakens their mechanical properties, limiting their applications in biomedicine. Herein, a novel strategy based on strong and high-density micelle cross-linkings is proposed to prepare nonswellable and tough hydrogel. To realize a strong micelle cross-linker, the synergetic effect of hydrophobic and quadruple hydrogen-bonding interactions is employed by introducing an alkyl chain-protected ureido pyrimidinone moiety into a segmented copolymer backbone. The length of the alkyl is the key factor in determining the strength of the hydrophobic interaction, which was carefully tailored to gain micelles with high strength and suitable solubility. A supramolecular hydrogel was formed in situ by simply linking micelle cross-linkers with poly(ethylene glycol) chains. The strong and high-density micelle cross-linkings restrain multiple effective chains outside the micelle from stretching during swelling, and the deformability of micelle cross-linkings disperses the local stress to maintain the network with high cross-linking density upon loading. Therefore, the hydrogel exhibited an outstanding nonswelling behavior under physiological conditions and excellent mechanical properties with a compressive strength of 4 MPa. The rapid in situ gelation also facilitated injection and cell encapsulation. Meanwhile, it also showed good tissue adhesion, cytocompatibility, and suitable degradability. This novel and facile strategy can offer new insights into the exploitation of cross-linkings to prepare nonswellable hydrogels for biomedical applications.

High-Performance Additive-/Post-Treatment-Free Nonfullerene Polymer Solar Cells via Tuning Molecular Weight of Conjugated Polymers.

In recent years, rapid advances are achieved in polymer solar cells (PSCs) using nonfullerene small molecular acceptors. However, no research disclosing the influence of molecular weight (Mn ) of conjugated polymer on the nonfullerene device performance is reported. In this work, a series of polymers with different Mn s are synthesized to systematically investigate the connection between Mn and performance of nonfullerene devices for the first time. It is found that the device performance improves substantially as the Mn increases from 12 to 38 kDa and a power conversion efficiency (PCE) as high as 10.5% is realized. It has to be noted this PCE is achieved without using any additives and post-treatments, which is among the top efficiencies of additive- and post-treatment-free PSCs. Most importantly, the variation trend of the optimal active layer thickness and morphology is significantly different from the device with fullerene as acceptor. The findings clarify the effect of Mn on the performance of nonfullerene PSCs, which would benefit further efficiency improvement of nonfullerene PSCs.

Hierarchical Morphology of Polymer Blend Films Induced by Convection-Driven Solvent Evaporation.

Homogeneous thin films of polymer blends with a desired morphology are necessary because of their applications in the fields such as optoelectronics, sensors, biomedicine, and so on. The frequently employed approach for the thin film preparation, spin coating is only able to achieve a homogeneous film for a small area because of the overwhelming spin-driven solvent evaporation with increased size. Here, a convection-guided morphology formation for polystyrene:poly(methyl methacrylate) blend films is reported. In situ observation shows that the morphology changed from homogeneous deposition with a scale less than 10 µm to a self-organized cellular pattern with a scale of more than 100 µm after the fluid flow is involved. Selective dissolution of the hierarchical films reveals that the cellular morphology is attributed to the flow-field-guided deposition of sequentially generated precipitates. The coupling of phase separation and fluid convection results in the hierarchical morphology that includes Voronoi cellular division as the primary structure and the detailed heterogeneous inner-cell features as the secondary structure. Isolated modulation of either micro- or mesoscale in the hierarchical morphology could be carried out via adjusting phase interaction or the convection disturbance correspondingly, providing a flexible and straightforward strategy to construct designed hierarchical structures for polymer thin films toward desired function or property.

[Research progress on artificial meniscus implants].

Meniscus injury has been one of the most common knee injuries in current society. The research on artificial meniscus implants as substitutes in meniscus reconstruction therapy has become global focus in order to solve clinical problems such as irreparable meniscus injury and symptoms after full or partial meniscectomy. At present, researches on artificial meniscus implants mainly focus on biodegradable meniscus scaffolds and non-biodegradable meniscus substitutes. Although the commercialized meniscal implants, such as CMI ®, Actifit ® and NUsurface ®, have been applied in the clinical, none of them can perfectively restore or permanently replace the natural meniscus tissue, effectively solve the symptoms after meniscectomy, and prevent cartilage degenerative diseases. The research progress, application, advantages and disadvantages of different kinds of artificial meniscus implants are reviewed in this manuscript, and the prospect is provided.

Regiocontrolled Electrosynthesis of [60]Fullerene Bisadducts: Photovoltaic Performance and Crystal Structures of C60 o-Quinodimethane Bisadducts.

C60 o-quinodimethane bisadducts [C60(QM)2] are promising electron acceptors for bulk heterojuction (BHJ) organic solar cells (OSCs). However, previous production of C60(QM)2 often resulted in excessive regioisomers, which were difficult to purify and might consequently obscure the structure-performance study of the organofullerene acceptors. Herein, the electrosynthesis of C60(QM)2 is reported. The reaction exhibits a strong regiocontrol with generation of fewer regioisomers. Pure regioisomers of cis-2, trans-3, and e C60(QM)2 are obtained, and the structures are solved with single-crystal X-ray diffraction. Interestingly, the cis-2 and trans-3 regioisomers exhibit better photovoltaic performance than the e regioisomer in the OSCs based on poly(3-hexylthiophene) (P3HT), which can be correlated with their structural difference.

Design of Double Conductive Layer and Grid-Assistant Face-to-Face Structure for Wide Linear Range, High Sensitivity Flexible Pressure Sensors.

Recently, flexible pressure sensors have drawn great attention because of their potential application in human-machine interfaces, healthcare monitoring, electronic skin, etc. Although many sensors with good performance have been reported, researchers mostly focused on surface morphology regulation, and the effect of the resistance characteristics on the performance of the sensor was still rarely systematically investigated. In this paper, a strategy for modulating electron transport is proposed to adjust the linear range and sensitivity of the sensor. In the modulating process, we constructed a double conductive layer (DCL) and grid-assistant face-to-face structure and obtained the sensor with a wide linear range of 0-700 kPa and a high sensitivity of 57.5 kPa-1, which is one of the best results for piezoresistive sensors. In contrast, the sensor with a single conductive layer (SCL) and simple face-to-face structure exhibited a moderate linear range (7 kPa) and sensitivity (2.8 kPa-1). Benefiting from the great performance, the modulated sensor allows for clear pulse wave detection and good recognition of gait signals, which indicates the great application potential in human daily life.

Ultrawide linear range, high sensitivity, and large-area pressure sensor arrays enabled by pneumatic spraying broccoli-like microstructures.

Large-area pressure sensor arrays with a wide linear response range and high sensitivity are beneficial to map the inhomogeneous interface pressure, which is significant in practical applications. Here, we demonstrate a pneumatic spraying method to prepare large-area microstructure films (PSMF) for high performance pressure sensor arrays. The sprayed surface morphology is designable by controlling the spraying parameters. It is worth noting that the constructed "broccoli" like morphology with a swollen top and shrunken bottom inspired a new mechanism to enlarge the linear response range by decreasing the series resistance with pressure increasing. At the same time, the pneumatic sprayed "broccoli" has a rough surface due to droplet stacking, which reduces the initial current effectively. Hence, the sensor achieves a 10 000 kPa ultrawide linear response range with a high sensitivity (98.71 kPa-1), and low detection (5 Pa). The prepared sensor has a small static response error (4.4%) and 5000 cycle full-range dynamic response durability. Finally, the constructed sensor arrays can distinguish the pressure distribution in different ranges clearly, which indicates a great potential in health care, motion detection, and the tire industry.

A polyurethane-based hydrophilic elastomer with multi-biological functions for small-diameter vascular grafts.

Thrombosis and intimal hyperplasia (IH) are two major problems faced by the small-diameter vascular grafts. Mimicking the native endothelium and physiological elasticity of blood vessels is considered an ideal strategy. Polyurethane (PU) is suitable for vascular grafts in mechanics because of its molecular designability and elasticity; however, it generally lacks the endothelium-like biofunctions and hydrophilicity. To solve this contradiction, a hydrophilic PU elastomer is developed by crosslinking the hydrophobic hard-segment chains containing diselenide with diaminopyrimidine-capped polyethylene glycol (PEG). In this network, the hydrophobic aggregation occurs underwater due to the uninterrupted hard-segment chains, leading to a significant self-enhancement in mechanics, which can be tailored to the elasticity similar to natural vessels by adjusting the crosslinking density. A series of in vitro studies confirm that the hydrophilicity of PEG and biological activities of aminopyrimidine and diselenide give the PU multi-biological functions similar to the native endothelium, including stable catalytic release of nitric oxide (NO) in the physiological level; anti-adhesion and anti-activation of platelets; inhibition of migration, adhesion, and proliferation of smooth muscle cells (SMCs); and antibacterial effect. In vivo studies further prove the good histocompatibility with both significant reduction in immune response and calcium deposition. STATEMENT OF SIGNIFICANCE: Constructing small-diameter vascular grafts similar to the natural vessels is considered an ideal method to solve the restenosis caused by thrombosis and intimal hyperplasia (IH). Because of the long-term stability, bulk modification is more suitable for implanted materials, however, how to achieve the biofunctions, hydrophilicity, and elasticity simultaneously is still a big challenge. In this work, a kind of polyurethane-based elastomer has been designed and prepared by crosslinking the functional long hard-segment chains with PEG soft segments. The underwater elasticity based on hydration-induced stiffening and the multi-biological functions similar to the native endothelium are compatible with natural vessels. Both in vitro and in vivo experiments demonstrate the potential of this PU as small-diameter vascular grafts.

PathMLP: Smooth path towards high-order homophily.

Real-world graphs exhibit increasing heterophily, where nodes no longer tend to be connected to nodes with the same label, challenging the homophily assumption of classical graph neural networks (GNNs) and impeding their performance. Intriguingly, from the observation of heterophilous data, we notice that certain high-order information exhibits higher homophily, which motivates us to involve high-order information in node representation learning. However, common practices in GNNs to acquire high-order information mainly through increasing model depth and altering message-passing mechanisms, which, albeit effective to a certain extent, suffer from three shortcomings: (1) over-smoothing due to excessive model depth and propagation times; (2) high-order information is not fully utilized; (3) low computational efficiency. In this regard, we design a similarity-based path sampling strategy to capture smooth paths containing high-order homophily. Then we propose a lightweight model based on multi-layer perceptrons (MLP), named PathMLP, which can encode messages carried by paths via simple transformation and concatenation operations, and effectively learn node representations in heterophilous graphs through adaptive path aggregation. Extensive experiments demonstrate that our method outperforms baselines on 16 out of 20 datasets, underlining its effectiveness and superiority in alleviating the heterophily problem. In addition, our method is immune to over-smoothing and has high computational efficiency. The source code will be available in https://github.com/Graph4Sec-Team/PathMLP.

Improving the Mechanical Properties and Tribological Behavior of Sulfobetaine Polyurethane Based on Hydrophobic Chains to Be Applied as Artificial Meniscus.

In the context of meniscus reconstruction in knee joints, current bulk biomaterials fail to meet the clinical demands for both excellent mechanical strength and low coefficient of friction. In this research, zwitterionic polyurethanes (PUs) with varying sulfobetaine (SB) groups were synthesized as the potential materials for artificial meniscus to investigate the relationship between the structures of SB groups and the performances of PUs. Under the saturation condition of 3 mg/mL hyaluronic acid aqueous solution, PU with long-alkyl chains and SB groups (PU-hSB4) exhibited a good tensile modulus (111.5 MPa) because the hydrophobic interaction of carbon chains was able to maintain the ordered aggregations of hard segment domains. Interestingly, hydrophobic chains in the molecular chain could also improve the tribological performance of PU-hSB4 instead of resulting from the surface roughness of samples, the components of lubricants, and the counterface of samples. A thicker and relatively stable hydration layer of noncrystal water was formed on the surface of PU-hSB4, which exhibited superior resistance to external forces compared to other PUs. Even if the hydration layer was damaged, PU-hSB4 was able to resist the compression of cartilage due to its high surface modulus, thus maintaining a similar and stable coefficient of friction (0.15-0.16) to native meniscus (0.18) and excellent wear resistance. In addition, the low cytotoxicity of PU-hSB4 further demonstrated its great potential to be applied in artificial meniscus instead.

Fabrication of mechanical skeleton of small-diameter vascular grafts via rolling on water surface.

Mimicking the multilayered structure of blood vessels and constructing a porous inner surface are two effective approaches to achieve mechanical matching and rapid endothelialization to reduce occlusion in small-diameter vascular grafts. However, the fabrication processes are complex and time consuming, thus complicating the fabrication of personalized vascular grafts. A simple and versatile strategy is proposed to prepare the skeleton of vascular grafts by rolling self-adhesive polymer films. These polymer films are directly fabricated by dropping a polymer solution on a water surface. For the tubes, the length and wall thickness are controlled by the rolling number and position of each film, whereas the structure and properties are tailored by regulating the solution composition. Double-layer vascular grafts (DLVGs) with microporous inner layers and impermeable outer layers are constructed; a microporous layer is formed by introducing a hydrophilic polymer into a polyurethane (PU) solution. DLVGs exhibit a J-shaped stress-strain deformation profile and compliance comparable to that of coronary arteries, sufficient suture retention strength and burst pressure, suitable hemocompatibility, significant adhesion, and proliferation of human umbilical vein endothelial cells. Freshly prepared PU tubes exhibit good cytocompatibility. Thus, this strategy demonstrates potential for rapid construction of small-diameter vascular grafts for individual customization.

RGP: Neural Network Pruning Through Regular Graph With Edges Swapping.

Deep learning technology has found a promising application in lightweight model design, for which pruning is an effective means of achieving a large reduction in both model parameters and float points operations (FLOPs). The existing neural network pruning methods mostly start from the consideration of the importance of model parameters and design parameter evaluation metrics to perform parameter pruning iteratively. These methods were not studied from the perspective of network model topology, so they might be effective but not efficient, and they require completely different pruning for different datasets. In this article, we study the graph structure of the neural network and propose a regular graph pruning (RGP) method to perform a one-shot neural network pruning. Specifically, we first generate a regular graph and set its node-degree values to meet the preset pruning ratio. Then, we reduce the average shortest path-length (ASPL) of the graph by swapping edges to obtain the optimal edge distribution. Finally, we map the obtained graph to a neural network structure to realize pruning. Our experiments demonstrate that the ASPL of the graph is negatively correlated with the classification accuracy of the neural network and that RGP has a strong precision retention capability with high parameter reduction (more than 90%) and FLOPs reduction (more than 90%) (the code for quick use and reproduction is available at https://github.com/Holidays1999/Neural-Network-Pruning-through-its-RegularGraph-Structure).

The Importance of Expert Knowledge for Automatic Modulation Open Set Recognition.

Automatic modulation classification (AMC) is an important technology for the monitoring, management, and control of communication systems. In recent years, machine learning approaches are becoming popular to improve the effectiveness of AMC for radio signals. However, the automatic modulation open-set recognition (AMOSR) scheme that aims to identify the known modulation types and recognize the unknown modulation signals is not well studied. Therefore, in this paper, we propose a novel multi-modal marginal prototype framework for radio frequency (RF) signals (MMPRF) to improve AMOSR performance. First, MMPRF addresses the problem of simultaneous recognition of closed and open sets by partitioning the feature space in the way of one versus other and marginal restrictions. Second, we exploit the wireless signal domain knowledge to extract a series of signal-related features to enhance the AMOSR capability. In addition, we propose a GAN-based unknown sample generation strategy to allow the model to understand the unknown world. Finally, we conduct extensive experiments on several publicly available radio modulation data, and experimental results show that our proposed MMPRF outperforms the state-of-the-art AMOSR methods.

Ternary donor-insulator-acceptor systems for polymer solar cells.

A series of block copolymers with fixed length of the semiconductor-block poly(3-butylthiophene) (P3BT) and varying length of the insulator-block polystyrene (PS) are synthesized. These copolymers are blended with phenyl-C61-butyric acid methyl ester (PCBM) for the bulk heterojunction photoactive layers. With appropriate insulator-block length and donor-acceptor ratio, the power conversion efficiency increases by one order of magnitude compared with reference devices with pure P3BT/PCBM. PS blocks improve the miscibility of the active layer blends remarkably. The P3BT-b-PS crystallizes as nanorods with the P3BT core covered with the PS-block, which creates a nanoscale tunneling barrier between donor and acceptor leading to more efficient transportation of charge carriers in the semiconductors.

In situ real-time investigation of Staphylococcus aureus on hemisphere-patterned polyurethane films.

Surface patterning is a promising approach to prevent bacterial adhesion and biofilm formation without the concerns of antimicrobial resistance. To determine the parameters of a patterned surface that can affect bacterial behavior, a sphere-like coccus (Staphylococcus aureus) was investigated on a series of polyurethane films with ordered hemisphere patterns. The bacterial retention data in a growth medium indicated that the surface patterns significantly decreased bacterial adhesion and proliferation. The most notable effects were observed with the 2 µm-pattern as well as the patterned polycaprolactone and polystyrene films, and the accessible contact area of the polyurethane films, surface wettability, and spatial confinement, did not show an influence. An optical microscope with a modified incubation cell was used for in situ real-time observations of bacterial colonization, proliferation, and migration. Based on appropriate statistical analyses, it was concluded that topographical geometry played a dominant role. In combination with the retention assessment in a nongrowth medium, it was found that pattern-mediated inhibition of biofilm formation was mainly achieved by affecting bacterial proliferation rather than adhesion. This study provides new insight for designing biofilm-resistant biomimetic materials.