RESUMO
Iridium, the most widely used anode catalyst in proton exchange membrane water electrolysis (PEMWE), must be used minimally due to its high price and limited supply. However, reducing iridium loading poses challenges due to abnormally large anode polarization. Herein, we present an anode catalyst layer (CL) based on a one-dimensional iridium nanofiber that enables a high current density operation of 3 A cm-2 at 1.86 V, even at an ultralow loading (0.07 mgIr cm-2). The performance is maintained even with a Pt coating-free porous transport layer (PTL) because our nanofiber CL circumvents the interfacial electron transport problem caused by the native oxide on the Ti PTL. We attribute this to the low work function and the low-ionomer-exposed surface of the nanofiber CL, which prevent the formation of Schottky contact at the native oxide interface. These results highlight the significance of optimizing the electronic properties of the CL/PTL interface for low-iridium-loading PEMWE.
RESUMO
Solid-state polymer-based electrolytes (SSPEs) exhibit great possibilities in realizing high-energy-density solid-state lithium metal batteries (SSLMBs). However, current SSPEs suffer from low ionic conductivity and unsatisfactory interfacial compatibility with metallic Li because of the high crystallinity of polymers and sluggish Li+ movement in SSPEs. Herein, differing from common strategies of copolymerization, a new strategy of constructing a high-entropy SSPE from multivariant polymeric ligands is proposed. As a protocol, poly(vinylidene fluoride-co-hexafluoropropylene) (PH) chains are grafted to the demoed polyethylene imine (PEI) with abundant -NH2 groups via a click-like reaction (HE-PEIgPHE). Compared to a PH-based SSPE, our HE-PEIgPHE shows a higher modulus (6.75 vs 5.18 MPa), a higher ionic conductivity (2.14 × 10-4 vs 1.03 × 10-4 S cm-1), and a higher Li+ transference number (0.55 vs 0.42). A Li|HE-PEIgPHE|Li cell exhibits a long lifetime (1500 h), and a Li|HE-PEIgPHE|LiFePO4 cell delivers an initial capacity of 160 mAh g-1 and a capacity retention of 98.7%, demonstrating the potential of our HE-PEIgPHE for the SSLMBs.
RESUMO
The interface between the catalyst and the ionomer in the catalyst layer of polymer electrolyte membrane fuel cells (PEMFCs) has been a subject of keen interest, but its effect on durability has not been fully understood due to the complexity of the catalyst layer structure. Herein, we utilize a Pt nanoparticle (NP) array electrode fabricated using a block copolymer template as the platform for a focused investigation of the interfacial change between the Nafion thin film and the Pt NP under a constant potential. A set of analyses for the electrodes treated with various potentials reveals that the Nafion thin film becomes densely packed at the intermediate potentials (0.4 and 0.7 V), indicating an increased ionomer-catalyst interaction due to the positive charges formed at the Pt surface at these potentials. Even for a practical PEMFC single cell, we demonstrate that the potential holding at the intermediate potentials increases ionomer adsorption to the Pt surface and the oxygen transport resistance, negatively impacting its power performance. This work provides fresh insight into the mechanism behind the performance fade in PEMFCs caused by potential-dependent ionomer rearrangement.
RESUMO
This paper investigates the performance of deep convolutional spiking neural networks (DCSNNs) trained using spike-based backpropagation techniques. Specifically, the study examined temporal spike sequence learning via backpropagation (TSSL-BP) and surrogate gradient descent via backpropagation (SGD-BP) as effective techniques for training DCSNNs on the field programmable gate array (FPGA) platform for object classification tasks. The primary objective of this experimental study was twofold: (i) to determine the most effective backpropagation technique, TSSL-BP or SGD-BP, for deeper spiking neural networks (SNNs) with convolution filters across various datasets; and (ii) to assess the feasibility of deploying DCSNNs trained using backpropagation techniques on low-power FPGA for inference, considering potential configuration adjustments and power requirements. The aforementioned objectives will assist in informing researchers and companies in this field regarding the limitations and unique perspectives of deploying DCSNNs on low-power FPGA devices. The study contributions have three main aspects: (i) the design of a low-power FPGA board featuring a deployable DCSNN chip suitable for object classification tasks; (ii) the inference of TSSL-BP and SGD-BP models with novel network architectures on the FPGA board for object classification tasks; and (iii) a comparative evaluation of the selected spike-based backpropagation techniques and the object classification performance of DCSNNs across multiple metrics using both public (MNIST, CIFAR10, KITTI) and private (INHA_ADAS, INHA_KLP) datasets.
RESUMO
The pulverization of lithium metal electrodes during cycling recently has been suppressed through various techniques, but the issue of irreversible consumption of the electrolyte remains a critical challenge, hindering the progress of energy-dense lithium metal batteries. Here, we design a single-ion-conductor-based composite layer on the lithium metal electrode, which significantly reduces the liquid electrolyte loss via adjusting the solvation environment of moving Li+ in the layer. A Li||Ni0.5Mn0.3Co0.2O2 pouch cell with a thin lithium metal (N/P of 2.15), high loading cathode (21.5 mg cm-2), and carbonate electrolyte achieves 400 cycles at the electrolyte to capacity ratio of 2.15 g Ah-1 (2.44 g Ah-1 including mass of composite layer) or 100 cycles at 1.28 g Ah-1 (1.57 g Ah-1 including mass of composite layer) under a stack pressure of 280 kPa (0.2 C charge with a constant voltage charge at 4.3 V to 0.05 C and 1.0 C discharge within a voltage window of 4.3 V to 3.0 V). The rational design of the single-ion-conductor-based composite layer demonstrated in this work provides a way forward for constructing energy-dense rechargeable lithium metal batteries with minimal electrolyte content.
Assuntos
Líquidos Corporais , Lítio , Eletrólitos , Íons , MetaisRESUMO
Tailoring the Li+ microenvironment is crucial for achieving fast ionic transfer and a mechanically reinforced solid-electrolyte interphase (SEI), which administers the stable cycling of Li-metal batteries (LMBs). Apart from traditional salt/solvent compositional tuning, this study presents the simultaneous modulation of Li+ transport and SEI chemistry using a citric acid (CA)-modified silica-based colloidal electrolyte (C-SCE). CA-tethered silica (CA-SiO2 ) can render more active sites for attracting complex anions, leading to further dissociation of Li+ from the anions, resulting in a high Li+ transference number (≈0.75). Intermolecular hydrogen bonds between solvent molecules and CA-SiO2 and their migration also act as nano-carrier for delivering additives and anions toward the Li surface, reinforcing the SEI via the co-implantation of SiO2 and fluorinated components. Notably, C-SCE demonstrated Li dendrite suppression and improved cycling stability of LMBs compared with the CA-free SiO2 colloidal electrolyte, hinting that the surface properties of the nanoparticles have a huge impact on the dendrite-inhibiting role of nano colloidal electrolytes.
RESUMO
High electrochemical polarization during a redox reaction in the electrode of aqueous zinc-bromine flow batteries largely limits its practical implementation as an effective energy storage system. This study demonstrates a rationally-designed composite electrode that exhibits a lower electrochemical polarization by providing a higher number of catalytically-active sites for faster bromine reaction, compared to a conventional graphite felt cathode. The composite electrode is composed of electrically-conductive graphite felt (GF) and highly active mesoporous tungsten oxynitride nanofibers (mWONNFs) that are prepared by electrospinning and simple heat treatments. Addition of the 1D mWONNFs to porous GF produces a web-like structure that significantly facilitates the reaction kinetics and ion diffusion. The cell performance achieves in this study demonstrated high energy efficiencies of 89% and 80% at current densities of 20 and 80 mA cm-2 , respectively. Furthermore, the cell can also be operated at a very high current density of 160 mA cm-2 , demonstrating an energy efficiency of 62%. These results demonstrate the effectiveness of the mWONNF/GF composite as the electrode material in zinc-bromine flow batteries.
RESUMO
Salt anions with a high donor number (DN) enable high sulfur utilization in lithium-sulfur (Li-S) batteries by inducing three-dimensional (3D) Li2 S growth. However, their insufficient compatibility with Li metal electrodes limits their cycling stability. Herein, a new class of salt anion, thiocyanate (SCN- ), is presented, which features a Janus character of electron donor and acceptor. Due to a strong Li+ coordination by SCN- and the direct interaction of SCN- with polysulfide anions, the LiSCN electrolyte has a remarkably high lithium polysulfide solubility. This electrolyte induces 3D Li2 S formation and ameliorates cathode passivation, even more than Br- , a typical high DN anion. Moreover, SCN- forms a Li3 N-enriched stable SEI layer at the surface of the Li metal electrode, enhancing cycling stability. A Li-S battery with the LiSCN electrolyte shows high current density operation (2.54 mA cmâ»2 ) with high discharge capacity (1133 mAh gâ»1 ) and prolonged cycle life (100 cycles). This work demonstrates that the cathode and anode performance in a Li-S battery can be simply and concurrently enhanced by the single salt anion.
RESUMO
A deep eutectic solvent (DES) is an ionic liquid-analog electrolyte, newly emerging due to its low cost, easy preparation, and tunable properties. Herein, a zinc-bromine battery (ZBB) with a Zn-halide-based DES electrolyte prepared by mixing ZnBr2 , ZnCl2 , and a bromine-capturing agent is reported. The water-free DES electrolyte allows a closed-cell configuration for the ZBB owing to the prevention of Br2 evaporation and H2 evolution. It is found that the Cl- anion changes the structure of the zinc-halide complex anions and demonstrated that it improves the ion mobility and electrode reaction kinetics. The DES electrolyte with the optimized ZnCl2 composition shows much higher rate capability and a cycle life 90 times longer than that of a ZnCl2 -free DES electrolyte. A pouch-type flexible ZBB battery based on the DES electrolyte exhibits swelling-free operation for more than 120 cycles and stable operation under a folding test, suggesting its potential in consumer applications such as wearable electronics.
RESUMO
Structural colors have advantages compared with chemical pigments or dyes, such as iridescence, tunability, and unfading. Many studies have focused on developing the ability to switch ON/OFF the structural color; however, they often suffer from a simple and single stimulus, remaining structural colors, and target selectivity. Herein, we present regionally controlled multistimuli-responsive structural color switching surfaces. The key part is the utilization of a micropatterned DNA-hydrogel assembly on a single substrate. Each hydrogel network contains a unique type of stimuli-responsive DNA motifs as an additional cross-linker to exhibit swelling/deswelling via stimuli-responsive DNA interactions. The approach enables overcoming the existing limitations and selectively programming the DNA-hydrogel to a decrypted state (ON) and an encrypted state (OFF) in response to multiple stimuli. Furthermore, the transitions are reversible, providing cyclability. We envision the potential of our method for diverse applications, such as sensors or anticounterfeiting, requiring multistimuli-responsive structural color switching surfaces.
Assuntos
DNA , Hidrogéis , Corantes , DNA/química , Hidrogéis/químicaRESUMO
Zinc ion batteries are promising candidates for large-scale energy storage systems. However, they suffer from the critical problems of insufficient cycling stability due to internal short-circuiting by zinc dendrites and zinc metal orphaning. In this work, a polymer of intrinsic microporosity (PIM-1) is reported as an ion regulating layer and an interface modulator, which promotes a uniform Zn plating and stripping process. According to spectroscopic analyses and computational calculations, PIM-1 enhances the reaction kinetics of a Zn metal electrode by altering the solvation structure of Zn2+ ions and increasing the work function of the Zn surface. As a result, the PIM-1 coating significantly improves the cyclability (1700 h at 0.5 mA cm-2 ) and Coulombic efficiency (99.6% at 3 mA cm-2 ) of the Zn/Zn2+ redox reaction. Moreover, the PIM-1 coated Zn operates for more than 200 h at 70% Zn utilization even under 10 mA cm-2 and 110 h at 95% Zn utilization of the Zn metal electrode. A Zn||V2 O5 full cell employing the PIM-1 layer exhibits seven times longer cycle life compared to the cell using bare Zn. The findings in this report demonstrate the potential of microporous materials as a key ingredient in the design of reversible Zn electrodes.
RESUMO
The optimal architecture of three-dimensional (3D) interface between a polymer electrolyte membrane (PEM) and catalyst layer (CL) is one of the most important issues to improve PEM fuel cells' (PEMFCs) performance. Here, we report the fabrication of hierarchical wrinkled PEM/CL interface over a large area. We fabricated the hierarchical wrinkles on a multiscale from nanometers to micrometers by bottom-up-based facile, scalable, and simple method. Notably, it allows one to go beyond the limit of the catalyst utilization by extremely enlarged interfacial area. The resulting hierarchical wrinkled PEM/CL displays a dramatically increased electrochemically active surface area (ECSA) and power performance by the enhancement factors of 89% and 67% compared with those of flat interface, which is one of the best enhancements compared to previous PEMFCs. We believe the scalability of hierarchical wrinkled interface can be exploited to design advanced 3D interfaces for high-performance PEMFCs even with ultralow Pt-loading.
RESUMO
The long-term cycling of anode-free Li-metal cells (i.e., cells where the negative electrode is in situ formed by electrodeposition on an electronically conductive matrix of lithium sourced from the positive electrode) using a liquid electrolyte is affected by the formation of an inhomogeneous solid electrolyte interphase (SEI) on the current collector and irregular Li deposition. To circumvent these issues, we report an atomically defective carbon current collector where multivacancy defects induce homogeneous SEI formation on the current collector and uniform Li nucleation and growth to obtain a dense Li morphology. Via simulations and experimental measurements and analyses, we demonstrate the beneficial effect of electron deficiency on the Li hosting behavior of the carbon current collector. Furthermore, we report the results of testing anode-free coin cells comprising a multivacancy defective carbon current collector, a LixNi0.8Co0.1Mn0.1-based cathode and a nonaqueous Li-containing electrolyte solution. These cells retain 90% of their initial capacity for over 50 cycles under lean electrolyte conditions.
RESUMO
Sharks, marine creatures that swim fast and have an antifouling ability, possess dermal denticle structures of micrometer-size. Because the riblet geometries on the denticles reduce the shear stress by inducing the slip of fluid parallel to the stream-wise direction, shark skin has the distinguished features of low drag and antifouling. Although much attention has been given to low-drag surfaces inspired from shark skin, it remains an important challenge to accurately mimic denticle structures in the micrometer scale and to finely control their structural features. This paper presents a novel method to create shark skin-mimetic denticle structures for low drag by exploiting a photoreconfigurable azopolymer. The light-designed denticle structure exhibits superior hydrophobicity and an antifouling effect as sharks do. This work suggests that our novel photoreconfiguration technology, mimicking shark skin and systematically manipulating various structural parameters, can be used in a reliable manner for diverse applications requiring low-drag surfaces.
Assuntos
Tubarões , Animais , Biomimética , Interações Hidrofóbicas e Hidrofílicas , PeleRESUMO
Aqueous zinc (Zn) metal batteries (ZMBs) are considered a promising candidate for grid-scale energy storage due to their freedom from fire hazards. However, a limited reversibility of Zn metal electrode caused by dendritic Zn growth has hindered the advent of high-capacity Zn metal batteries (>4 mAh cm-2 ). Herein, it is reported that fast electrokinetic Zn-ion transport extremely improves the Zn metal reversibility. It is revealed that a negatively charged porous layer (NPL) provides the electrokinetic Zn-ion transport by surface conduction, and consequently impedes the depletion of Zn-ion on electrode surface as indicated by numerical simulations and overlimiting current behavior. Due to the quick Zn-ion delivery, a dendrite-free and densely packed Zn metal deposit is accommodated inside its pores. With the introduction of the NPL, the cycling stability of Zn symmetric cell is enhanced by 21 times at 10 mA cm-2 /10 mAh cm-2 . Average Coulombic efficiency of 99.6% is achieved over 500 cycles for electrodeposition/stripping at 30 mA cm-2 /5 mAh cm-2 on NPL-Cu electrode. Furthermore, a high-capacity Zn/V2 O5 full cell with the NPL exhibits an extraordinary stability over 1000 cycles at a capacity of 4.8 mAh cm-2 .
RESUMO
Developing a safe and long-lasting lithium (Li) metal battery is crucial for high-energy applications. However, its poor cycling stability due to Li dendrite formation and excessive Li pulverization is the major hurdle for its practical applications. Here, we present a silica (SiO2) nanoparticle-dispersed colloidal electrolyte (NDCE) and its design principle for suppressing Li dendrite formation. SiO2 nanoclusters in the NDCE play roles in enhancing the Li+ transference number and increasing the Li+ diffusivity in the vicinity of the Li-plating substrate. The NDCE enables less-dendritic Li plating by manipulating the nucleation-growth mode and extending Sand's time. Moreover, SiO2 can interplay with the electrolyte components at the Li-metal surface, enriching fluorinated compounds in the solid electrolyte interface layer. The initial control of the Li plating morphology and SEI structure by the NDCE leads to a more uniform and denser Li deposition upon subsequent cycling, resulting in threefold enhancement of the cycle life. The efficacy of the NDCEs has been further demonstrated by the practical battery design, featuring a commercial-level cathode and thin Li-metal (40 µm) anode.
RESUMO
Two main bottlenecks encountered when implementing energy-efficient spike-timing-dependent plasticity (STDP) based sparse coding, are the complex computation of winner-take-all (WTA) operation and repetitive neuronal operations in the time domain processing. In this article, we present an energy-efficient STDP based sparse coding processor. The low-cost hardware is based on the algorithmic reduction techniques as following: First, the complex WTA operation is simplified based on the prediction of spike emitting neurons. Sparsity based approximation in spatial and temporal domain are also efficiently exploited to remove the redundant neurons with negligible algorithmic accuracy loss. We designed and implemented the hardware of the STDP based sparse coding using 65nm CMOS process. By exploiting input sparsity, the proposed SNN architecture can dynamically trade off algorithmic quality for computation energy (up to 74%) for Natural image (maximum 0.01 RMSE increment) and MNIST (no accuracy loss) applications. In the inference mode of operations, the SNN hardware achieves the throughput of 374 Mpixels/s and 840.2 GSOP/s with the energy-efficiency of 781.52 pJ/pixel and 0.35 pJ/SOP.
Assuntos
Neurônios/fisiologia , Sinapses/fisiologia , Desenho de Equipamento , Humanos , Dispositivos Lab-On-A-Chip , Aprendizado de Máquina , Modelos Neurológicos , Redes Neurais de ComputaçãoRESUMO
Micro/nano-patterned alumina surfaces are important in a variety fields such as chemical/biotechnology, surface science, and microelectro-mechanical systems. However, for patterning alumina surfaces, it still remains a challenge to have a lithographic tool that has large flexibility in design layouts, structural reconfigurability, and a simple fabrication process. In this work, a new alumina-patterning platform that uses a photo-reconfigurable azobenzene-alumina composite as an imprinting material is presented. Under far-field irradiation, the azobenzene-alumina anisotropically flows in the direction parallel to the light polarization. Accordingly, an arbitrarily designed azobenzene-alumina composite by imprinting can be deterministically reconfigured by light polarization and irradiation time. The photo-reconfigured azobenzene-alumina is then converted to pure alumina through calcination in an air atmosphere, which provides thin crack-free alumina patterns with a high structural fidelity. The novel combination of photo-reconfigurable azobenzene moieties and an alumina precursor for imprinting the material provides large flexibility in designing and controlling geometric parameters of the alumina pattern, which potentially offers significant value in various micro/nanotechnology fields.
RESUMO
Anisotropic small structures found throughout living nature have unique functionalities as seen by Gecko lizards. Here, we present a simple yet programmable method for fabricating anisotropic, submicrometer-sized bent pillar structures using photoreconfiguration of an azopolymer. A slant irradiation of a p-polarized light on the pillar structure of an azopolymer simply results in a bent pillar structure. By combining the field-gradient effect and directionality of photofluidization, control of the bending shape and the curvature is achieved. With the bent pillar patterned surface, anisotropic wetting and directional adhesion are demonstrated. Moreover, the bent pillar structures can be transferred to other polymers, highlighting the practical importance of this method. We believe that this pragmatic method to fabricate bent pillars can be used in a reliable manner for many applications requiring the systematic variation of a bent pillar structure.
RESUMO
In this paper, we present an energy and area efficient spike neural network (SNN) processor based on novel spike counts based methods. For the low cost SNN design, we propose hardware-friendly complexity reduction techniques for both of learning and inferencing modes of operations. First, for the unsupervised learning process, we propose a spike counts based learning method. The novel learning approach utilizes pre- and post-synaptic spike counts to reduce the bit-width of synaptic weights as well as the number of weight updates. For the energy efficient inferencing operations, we propose an accumulation based computing scheme, where the number of input spikes for each input axon is accumulated without instant membrane updates until the pre-defined number of spikes are reached. In addition, the computation skip schemes identify meaningless computations and skip them to improve energy efficiency. Based on the proposed low complexity design techniques, we design and implement the SNN processor using 65 nm CMOS process. According to the implementation results, the SNN processor achieves 87.4% of recognition accuracy in MNIST dataset using only 1-bit 230 k synaptic weights with 400 excitatory neurons. The energy consumptions are 0.26 pJ/SOP and 0.31 µJ/inference in inferencing mode, and 1.42 pJ/SOP and 2.63 µJ/learning in learning mode of operations.