Column Row Convolutional Neural Network: Reducing Parameters for Efficient Image Processing.

Recent advancements in deep learning have achieved significant progress by increasing the number of parameters in a given model. However, this comes at the cost of computing resources, prompting researchers to explore model compression techniques that reduce the number of parameters while maintaining or even improving performance. Convolutional neural networks (CNN) have been recognized as more efficient and effective than fully connected (FC) networks. We propose a column row convolutional neural network (CRCNN) in this letter that applies 1D convolution to image data, significantly reducing the number of learning parameters and operational steps. The CRCNN uses column and row local receptive fields to perform data abstraction, concatenating each direction's feature before connecting it to an FC layer. Experimental results demonstrate that the CRCNN maintains comparable accuracy while reducing the number of parameters and compared to prior work. Moreover, the CRCNN is employed for one-class anomaly detection, demonstrating its feasibility for various applications.

Tailoring Three-Dimensional Cross-Linked Networks Based on Water-Soluble Polymeric Materials for Stable Silicon Anode.

For stable battery operation of silicon (Si)-based anodes, utilizing cross-linked three-dimensional (3D) network binders has emerged as an effective strategy to mitigate significant volume fluctuations of Si particles. In the design of cross-linked network binders, careful selection of appropriate cross-linking agents is crucial to maintaining a balance between the robustness and functionality of the network. Herein, we strategically design and optimize a 3D cross-linked network binder through a comprehensive analysis of cross-linking agents. The proposed network is composed of poly(vinyl alcohol) grafted poly(acrylic acid) (PVA-g-PAA, PVgA) and aromatic diamines. PVgA is chosen as the polymer backbone owing to its high flexibility and facile synthesis using an ecofriendly water solvent. Subsequently, an aromatic diamine is employed as a cross-linker to construct a robust amide network that features a resonance-stabilized high modulus and enhanced adhesion. Comparative investigations of three cross-linkers, 2,2'-bis(trifluoromethyl)benzidine, 3,3'-oxidianiline, and 4,4'-oxybis[3-(trifluoromethyl)aniline] (TFODA), highlight the roles of the trifluoromethyl group (-CF3) and the ether linkage. Consequently, PVgA cross-linked with TFODA (PVgA-TFODA), featuring both -CF3 and -O-, establishes a well-balanced 3D network characterized by heightened elasticity and improved binding forces. The optimized Si and SiOx/graphite composite electrodes with the PVgA-TFODA binder demonstrate impressive structural stability and stable cycling. This study offers a novel perspective on designing cross-linked network binders, showcasing the benefits of a multidimensional approach considering chemical and physical interactions.

Defect-Engineered Semiconducting van der Waals Thin Film at Metal-Semiconductor Interface of Field-Effect Transistors.

The significance of metal-semiconductor interfaces and their impact on electronic device performance have gained increasing attention, with a particular focus on investigating the contact metal. However, another avenue of exploration involves substituting the contact metal at the metal-semiconductor interface of field-effect transistors with semiconducting layers to introduce additional functionalities to the devices. Here, a scalable approach for fabricating metal-oxide-semiconductor (channel)-semiconductor (interfacial layer) field-effect transistors is proposed by utilizing solution-processed semiconductors, specifically semiconducting single-walled carbon nanotubes and molybdenum disulfide, as the channel and interfacial semiconducting layers, respectively. The work function of the interfacial MoS2 is modulated by controlling the sulfur vacancy concentration through chemical treatment, which results in distinctive energy band alignments within a single device configuration. The resulting band alignments lead to multiple functionalities, including multivalued transistor characteristics and multibit nonvolatile memory (NVM) behavior. Moreover, leveraging the stable NVM properties, we demonstrate artificial synaptic devices with 88.9% accuracy of MNIST image recognition.

P- and N-type InAs nanocrystals with innately controlled semiconductor polarity.

InAs semiconductor nanocrystals (NCs) exhibit intriguing electrical/optoelectronic properties suitable for next-generation electronic devices. Although there is a need for both n- and p-type semiconductors in such devices, InAs NCs typically exhibit only n-type characteristics. Here, we report InAs NCs with controlled semiconductor polarity. Both p- and n-type InAs NCs can be achieved from the same indium chloride and aminoarsine precursors but by using two different reducing agents, diethylzinc for p-type and diisobutylaluminum hydride for n-type NCs, respectively. This is the first instance of semiconductor polarity control achieved at the synthesis level for InAs NCs and the entire semiconductor nanocrystal systems. Comparable field-effective mobilities for holes (3.3 × 10-3 cm2/V·s) and electrons (3.9 × 10-3 cm2/V·s) are achieved from the respective NC films. The mobility values allow the successful fabrication of complementary logic circuits, including NOT, NOR, and NAND comprising photopatterned p- and n-channels based on InAs NCs.

Enhancing Efficiency and Stability of Tin Halide Perovskite Light-Emitting Diodes via Engineered Alkali/Multivalent Metal Salts.

Sn-based perovskite light-emitting diodes (PeLEDs) have emerged as promising alternatives to Pb-based PeLEDs with their rapid increase in performance owing to the various research studies on inhibiting Sn oxidation. However, the absence of defect passivation strategies for Sn-based perovskite LEDs necessitates further research in this field. We performed systematic studies to investigate the design rules for defect passivation agents for Sn-based perovskites by incorporating alkali/multivalent metal salts with various cations and anions. From the computational and experimental analyses, sodium trifluoromethanesulfonate (NaTFMS) was found to be the most effective passivation agent for PEA2SnI4 films among the explored candidate agents owing to favorable reaction energetics to passivate iodide Frenkel defects. Consequently, the incorporation of NaTFMS facilitates the formation of uniform films with relatively large crystals and reduced Sn4+. The NaTFMS-containing PEA2SnI4 PeLEDs demonstrate an improved luminance of 138.9 cd/m2 and external quantum efficiency (EQE) of 0.39% with an improved half-lifetime of more than threefold. This work provides important insight into the design of defect passivation agents for Sn-based perovskites.

Approaching Theoretical Limits in the Performance of Printed P-Type CuI Transistors via Room Temperature Vacancy Engineering.

Development of a novel high performing inorganic p-type thin film transistor could pave the way for new transparent electronic devices. This complements the widely commercialized n-type counterparts, indium-gallium-zinc-oxide (IGZO). Of the few potential candidates, copper monoiodide (CuI) stands out. It boasts visible light transparency and high intrinsic hole mobility (>40 cm2 V-1 s-1 ), and is suitable for various low-temperature processes. However, the performance of reported CuI transistors is still below expected mobility, mainly due to the uncontrolled excess charge- and defect-scattering from thermodynamically favored formation of copper and iodine vacancies. Here, a solution-processed CuI transistor with a significantly improved mobility is reported. This enhancement is achieved through a room-temperature vacancy-engineering processing strategy on high-k dielectrics, sodium-embedded alumina. A thorough set of chemical, structural, optical, and electrical analyses elucidates the processing-dependent vacancy-modulation and its corresponding transport mechanism in CuI. This encompasses defect- and phonon-scattering, as well as the delocalization of charges in crystalline domains. As a result, the optimized CuI thin film transistors exhibit exceptionally high hole mobility of 21.6 ± 4.5 cm2 V-1 s-1 . Further, the successful operation of IGZO-CuI complementary logic gates confirms the applicability of the device.

Ion trap and release dynamics enables nonintrusive tactile augmentation in monolithic sensory neuron.

An iontronic-based artificial tactile nerve is a promising technology for emulating the tactile recognition and learning of human skin with low power consumption. However, its weak tactile memory and complex integration structure remain challenging. We present an ion trap and release dynamics (iTRD)-driven, neuro-inspired monolithic artificial tactile neuron (NeuroMAT) that can achieve tactile perception and memory consolidation in a single device. Through the tactile-driven release of ions initially trapped within iTRD-iongel, NeuroMAT only generates nonintrusive synaptic memory signals when mechanical stress is applied under voltage stimulation. The induced tactile memory is augmented by auxiliary voltage pulses independent of tactile sensing signals. We integrate NeuroMAT with an anthropomorphic robotic hand system to imitate memory-based human motion; the robust tactile memory of NeuroMAT enables the hand to consistently perform reliable gripping motion.

Physically defined long-term and short-term synapses for the development of reconfigurable analog-type operators capable of performing health care tasks.

Extracting valuable information from the overflowing data is a critical yet challenging task. Dealing with high volumes of biometric data, which are often unstructured, nonstatic, and ambiguous, requires extensive computer resources and data specialists. Emerging neuromorphic computing technologies that mimic the data processing properties of biological neural networks offer a promising solution for handling overflowing data. Here, the development of an electrolyte-gated organic transistor featuring a selective transition from short-term to long-term plasticity of the biological synapse is presented. The memory behaviors of the synaptic device were precisely modulated by restricting ion penetration through an organic channel via photochemical reactions of the cross-linking molecules. Furthermore, the applicability of the memory-controlled synaptic device was verified by constructing a reconfigurable synaptic logic gate for implementing a medical algorithm without further weight-update process. Last, the presented neuromorphic device demonstrated feasibility to handle biometric information with various update periods and perform health care tasks.

Chiroptical Synaptic Heterojunction Phototransistors Based on Self-Assembled Nanohelix of π-Conjugated Molecules for Direct Noise-Reduced Detection of Circularly Polarized Light.

High-performance chiroptical synaptic phototransistors are successfully demonstrated using heterojunctions composed of a self-assembled nanohelix of a π-conjugated molecule and a metal oxide semiconductor. To impart strong chiroptical activity to the device, a diketopyrrolopyrrole-based π-conjugated molecule decorated with chiral glutamic acid is newly synthesized; this molecule is capable of supramolecular self-assembly through noncovalent intermolecular interactions. In particular, nanohelix formed by intertwinded fibers with strong and stable chiroptical activity in a solid-film state are obtained through hydrogen-bonding-driven, gelation-assisted self-assembly. Phototransistors based on interfacial charge transfer at the heterojunction from the chiroptical nanohelix to the metal oxide semiconductor show excellent chiroptical detection with a high photocurrent dissymmetry factor of 1.97 and a high photoresponsivity of 218 A W-1 . The chiroptical phototransistor demonstrates photonic synapse-like, time-dependent photocurrent generation, along with persistent photoconductivity, which is attributed to the interfacial charge trapping. Through the advantage of synaptic functionality, a trained convolutional neural network successfully recognizes noise-reduced circularly polarized images of handwritten alphabetic characters with better than 89.7% accuracy.

Repairable Macroscopic Monodomain Arrays from Block Copolymers Enabled by Photoplastic and Photodielectric Effects.

Upon exposure to UV light (120 mW/cm2, λ = 365 nm), a trans-cis isomerization occurs in a cylinder-forming, azobenzene-containing block copolymer of polydimethylsiloxane-b-poly((4(phenyldiazenyl)phenoxy)hexyl acrylate) (PDMS-b-PPHA) that enables the generation of monodomains of healable, long-range ordered arrays of nanoscopic domains over macroscopic distances. The trans-cis isomerization gives rise to a significant increase in the dielectric constant (from 6.52 to 19.8 at 100 Hz, photodielectric behavior) and a dramatic decrease in the Tg (from 54 to 1 °C, photoplastic behavior) of the PPHA block. By combining these characteristics with an in-plane electric field, macroscopic monodomains of near-perfectly aligned cylindrical microdomains are achieved at low temperatures, and a damage repair is clearly uncovered, where the 300 nm wide scratches can be completely healed at 40 °C, leaving a smooth, uniformly thick film where the continuity and orientation of the aligned microdomains are restored. Subsequent exposure to visible light causes a cis-trans isomerization, increasing the matrix Tg to 54 °C, producing highly oriented and aligned PDMS cylindrical microdomains in a PPHA matrix.

Graphene barristors for de novo optoelectronics.

Graphene-based vertical Schottky-barrier transistors (SBTs), renowned as graphene barristors, have emerged as a feasible candidate to fundamentally expand the horizon of conventional transistor technology. The remote tunability of graphene's electronic properties could endorse multi-stimuli responsive functionalities for a broad range of electronic and optoelectronic applications of transistors, with the capability of incorporating nanochannel architecture with dramatically reduced footprints from the vertical integrations. In this Feature Article, we provide a comprehensive overview of the progress made in the field of SBTs over the last 10 years, starting from the operating principles, materials evolution, and processing developments. Depending on the types of stimuli such as electrical, optical, and mechanical stresses, various fields of applications from conventional digital logic circuits to sensory technologies are highlighted. Finally, more advanced applications toward beyond-Moore electronics are discussed, featuring recent advancements in neuromorphic devices based on SBTs.

Humanlike spontaneous motion coordination of robotic fingers through spatial multi-input spike signal multiplexing.

With advances in robotic technology, the complexity of control of robot has been increasing owing to fundamental signal bottlenecks and limited expressible logic state of the von Neumann architecture. Here, we demonstrate coordinated movement by a fully parallel-processable synaptic array with reduced control complexity. The synaptic array was fabricated by connecting eight ion-gel-based synaptic transistors to an ion gel dielectric. Parallel signal processing and multi-actuation control could be achieved by modulating the ionic movement. Through the integration of the synaptic array and a robotic hand, coordinated movement of the fingers was achieved with reduced control complexity by exploiting the advantages of parallel multiplexing and analog logic. The proposed synaptic control system provides considerable scope for the advancement of robotic control systems.

Multiplexed Complementary Signal Transmission for a Self-Regulating Artificial Nervous System.

Neuromorphic engineering has emerged as a promising research field that can enable efficient and sophisticated signal transmission by mimicking the biological nervous system. This paper presents an artificial nervous system capable of facile self-regulation via multiplexed complementary signals. Based on the tunable nature of the Schottky barrier of a complementary signal integration circuit, a pair of complementary signals is successfully integrated to realize efficient signal transmission. As a proof of concept, a feedback-based blood glucose level control system is constructed by incorporating a glucose/insulin sensor, a complementary signal integration circuit, an artificial synapse, and an artificial neuron circuit. Certain amounts of glucose and insulin in the initial state are detected by each sensor and reflected as positive and negative amplitudes of the multiplexed presynaptic pulses, respectively. Subsequently, the pulses are converted to postsynaptic current, which triggered the injection of glucose or insulin in a way that confined the glucose level to a desirable range. The proposed artificial nervous system demonstrates the notable potential of practical advances in complementary control engineering.

Monolithic Tandem Vertical Electrochemical Transistors for Printed Multi-Valued Logic.

Organic electrochemical transistors (OECTs) have recently emerged as a feasible candidate to realize the next generation of printable electronics. Especially, their chemical versatility and the unique redox-based operating principle have provided new possibilities in high-functioning logic circuitry beyond the traditional binary Boolean logic. Here, a simple strategy to electrochemically realize monolithic multi-valued logic transistors is presented, which is one of the most promising branches of transistor technology in the forthcoming era of hyper Moore's law. A vertically stacked heterogeneous dual-channel architecture is introduced with a patterned reference electrode, which enables a facile manifestation of stable and equiprobable ternary logic states with a reduced transistor footprint. The dual-ion-penetration mechanism coupled with ultrashort vertical channel even allows a very-high accessing frequency to multiple logic states reaching over 10 MHz. Furthermore, printed arrays of ternary logic gates with full voltage swing within 1 V are demonstrated.

A biomimetic ocular prosthesis system: emulating autonomic pupil and corneal reflections.

The human light modulation response allows humans to perceive objects clearly by receiving the appropriate amount of light from the environment. This paper proposes a biomimetic ocular prosthesis system that mimics the human light modulation response capable of pupil and corneal reflections. First, photoinduced synaptic properties of the quantum dot embedded photonic synapse and its biosimilar signal transmission is confirmed. Subsequently, the pupillary light reflex is emulated by incorporating the quantum dot embedded photonic synapse, electrochromic device, and CMOS components. Moreover, a solenoid-based eyelid is connected to the pupillary light reflex system to emulate the corneal reflex. The proposed ocular prosthesis system represents a platform for biomimetic prosthesis that can accommodate an appropriate amount of stimulus by self-regulating the intensity of external stimuli.

Neurorobotic approaches to emulate human motor control with the integration of artificial synapse.

The advancement of electronic devices has enabled researchers to successfully emulate human synapses, thereby promoting the development of the research field of artificial synapse integrated soft robots. This paper proposes an artificial reciprocal inhibition system that can successfully emulate the human motor control mechanism through the integration of artificial synapses. The proposed system is composed of artificial synapses, load transistors, voltage/current amplifiers, and a soft actuator to demonstrate the muscle movement. The speed, range, and direction of the soft actuator movement can be precisely controlled via the preset input voltages with different amplitudes, numbers, and signs (positive or negative). The artificial reciprocal inhibition system can impart lifelike motion to soft robots and is a promising tool to enable the successful integration of soft robots or prostheses in a living body.

Alkyl Chain Engineering of Low Bandgap Non-Fullerene Acceptors for High-Performance Organic Solar Cells: Branched vs. Linear Alkyl Side Chains.

In this work, we report the synthesis and photovoltaic properties of IEBICO-4F, IEHICO-4F, IOICO-4F, and IDICO-4F non-fullerene acceptors (NFAs) bearing different types of alkyl chains (2-ehtylhexyl (EH), 2-ethylbutyl (EB), n-octyl (O), and n-decyl (D), respectively). These NFAs are based on the central indacenodithiophene (IDT) donor core and the same terminal group of 2-(5,6-difluoro-3-oxo-2,3-dihydro-1H-inden-1-ylidene)malononitrile (IC-2F), albeit with different side chains appended to the thiophene bridge unit. Although the side chains induced negligible differences between the NFAs in terms of optical band gaps and molecular energy levels, they did lead to changes in their melting points and crystallinity. The NFAs with branched alkyl chains exhibited weaker intermolecular interactions and crystallinity than those with linear alkyl chains. Organic solar cells (OSCs) were fabricated by blending these NFAs with the p-type polymer PTB7-Th. The NFAs with appended branched alkyl chains (IEHICO-4F and IEBICO-4F) possessed superior photovoltaic properties than those with appended linear alkyl chains (IOICO-4F and IDICO-4F). This result can be ascribed mainly to the thin-film morphology. Furthermore, the NFA-based blend films with appended branched alkyl chains exhibited the optimal degree of aggregation and miscibility, whereas the NFA-based blend films with appended linear alkyl chains exhibited higher levels of self-aggregation and lower miscibility between the NFA molecule and the PTB7-Th polymer. We demonstrate that changing the alkyl chain on the π-bridging unit in fused-ring-based NFAs is an effective strategy for improving their photovoltaic performance in bulk heterojunction-type OSCs.

Non-von Neumann multi-input spike signal processing enabled by an artificial synaptic multiplexer.

Multiplexing is essential for technologies that require processing of a large amount of information in real time. Here, we present an artificial synaptic multiplexing unit capable of realizing parallel multi-input control system. Ion gel was used as a dielectric layer of the artificial synaptic multiplexing unit because of its ionic property, allowing multigating for parallel input. A closed-loop control system that enables multi-input-based feedback for actuator bending control was realized by incorporating an ion gel-based artificial synaptic multiplexing unit, an actuator, and a bending angle sensor. The proposed multi-input control system could simultaneously process input and feedback signals, offering a breakthrough in industries in which the processing of vast amounts of streaming data is essential.

Comb-type polymer-hybridized MXene nanosheets dispersible in arbitrary polar, nonpolar, and ionic solvents.

Solution-based processing of two-dimensional (2D) nanomaterials is highly desirable, especially for the low-temperature large-area fabrication of flexible multifunctional devices. MXenes, an emerging family of 2D materials composed of transition metal carbides, carbonitrides, or nitrides, provide excellent electrical and electrochemical properties through aqueous processing. Here, we further expand the horizon of MXene processing by introducing a polymeric superdispersant for MXene nanosheets. Segmented anchor-spacer structures of a comb-type polymer, polycarboxylate ether (PCE), provide polymer grafting-like steric spacings over the van der Waals range of MXene surfaces, thereby reducing the colloidal interactions by the order of 103, regardless of solvent. An unprecedented broad dispersibility window for Ti3C2Tx MXene, covering polar, nonpolar, and even ionic solvents, was achieved. Furthermore, close PCE entanglements in MXene@PCE composite films resulted in highly robust properties upon prolonged mechanical and humidity stresses.

A general fruit acid chelation route for eco-friendly and ambient 3D printing of metals.

Recent advances in metal additive manufacturing (AM) have provided new opportunities for prompt designs of prototypes and facile personalization of products befitting the fourth industrial revolution. In this regard, its feasibility of becoming a green technology, which is not an inherent aspect of AM, is gaining more interests. A particular interest in adapting and understanding of eco-friendly ingredients can set its important groundworks. Here, we demonstrate a water-based solid-phase binding agent suitable for binder jetting 3D printing of metals. Sodium salts of common fruit acid chelators form stable metal-chelate bridges between metal particles, enabling elaborate 3D printing of metals with improved strengths. Even further reductions in the porosity between the metal particles are possible through post-treatments. A compatibility of this chelation chemistry with variety of metals is also demonstrated. The proposed mechanism for metal 3D printing can open up new avenues for consumer-level personalized 3D printing of metals.