The ability of high‐order tuning of the synaptic plasticity in an artificial synapse can offer significant improvement in the processing time, low‐power recognition, and learning capability in a neuro‐inspired computing system. Inspired by light‐assisted dopamine‐facilitated synaptic activity, which achieves rapid learning and adaptation by lowering the threshold of the synaptic plasticity, a two‐terminal organolead halide perovskite (OHP)‐based photonic synapse is fabricated and designed in which the synaptic plasticity is modified by both electrical pulses and light illumination. Owing to the accelerated migration of the iodine vacancy inherently existing in the coated OHP film under light illumination, the OHP synaptic device exhibits light‐tunable synaptic functionalities with very low programming inputs (≈0.1 V). It is also demonstrated that the threshold of the long‐term potentiation decreases and synaptic weight further modulates when light illuminates the device, which is phenomenologically analogous to the dopamine‐assisted synaptic process. Notably, under light exposure, the OHP synaptic device achieves rapid pattern recognition with ≈81.8% accuracy after only 2000 learning phases (60 000 learning phases = one epoch) with a low‐power consumption (4.82 nW/the initial update for potentiation), which is ≈2.6 × 103 times lower than when the synaptic weights are updated by only high electrical pulses.
Modern artificial neural network technology using a deterministic computing framework is faced with a critical challenge in dealing with massive data that are largely unstructured and ambiguous. This challenge demands the advances of an elementary physical device for tackling these uncertainties. Here, we designed and fabricated a SiOx nanorod memristive device by employing the glancing angle deposition (GLAD) technique, suggesting a controllable stochastic artificial neuron that can mimic the fundamental integrate‐and‐fire signaling and stochastic dynamics of a biological neuron. The nanorod structure provides the random distribution of multiple nanopores all across the active area, capable of forming a multitude of Si filaments at many SiOx nanorod edges after the electromigration process, leading to a stochastic switching event with very high dynamic range (≈5.15 × 1010) and low energy (≈4.06 pJ). Different probabilistic activation (ProbAct) functions in a sigmoid form are implemented, showing its controllability with low variation by manufacturing and electrical programming schemes. Furthermore, as an application prospect, based on the suggested memristive neuron, we demonstrated the self‐resting neural operation with the local circuit configuration and revealed probabilistic Bayesian inferences for genetic regulatory networks with low normalized mean squared errors (≈2.41 × 10‐2) and its robustness to the ProbAct variation.
Multiple building components with
a variety of switching capabilities
are required to implement memory-driven parallel computing architecture
and multifunctional device applications. In this regard, there is
a high demand for implementing multiple switching modes based on only
one active material in a simple device form. In this study, we demonstrated
the ability of a solution-processable two-terminal Ag (or Al)/NiO
x
/ITO memristor that exhibits triple-switching
characteristics depending on the different voltage regimes. Notably,
the device exhibits an analog bipolar switching behavior under a low
programming voltage (≤1.0 V), enabling essential synaptic functions,
as well as a high recognition accuracy of 87.42% in a single-layer
neural network. After the electroforming process to form the oxygen
vacancy-based conductive filament in the NiO
x
layer, the device concurrently exhibited digital bipolar switching
and unipolar threshold switching behaviors at different voltage regimes
without compliance current. These three switching characteristics
are related to the transition of dominant switching mechanisms depending
on the operating scheme, which is investigated using various material
and chemical characterization techniques during switching, including
cross-sectional scanning transmission electron microscopy, atomic
force microscopy, ultraviolet photoelectron spectroscopy, and X-ray
photoelectron spectroscopy.
Microbial fuel cells (MFCs) can convert chemical energy to electricity using microbes as catalysts and a variety of organic wastewaters as substrates. However, electron loss occurs when fermentable substrates are used because fermentation bacteria and methanogens are involved in electron flow from the substrates to electricity. In this study, MFCs using glucose (G-MFC), propionate (P-MFC), butyrate (B-MFC), acetate (A-MFC), and a mix (M-MFC, glucose:propionate:butyrate:acetate = 1:1:1:1) were operated in batch mode. The metabolites and microbial communities were analyzed. The current was the largest electron sink in M-, G-, B-, and A-MFCs; the initial chemical oxygen demands (COD(ini)) involved in current production were 60.1% for M-MFC, 52.7% for G-MFC, 56.1% for B-MFC, and 68.3% for A-MFC. Most of the glucose was converted to propionate (40.6% of COD(ini)) and acetate (21.4% of COD(ini)) through lactate (80.3% of COD(ini)) and butyrate (6.1% of COD(ini)). However, an unknown source (62.0% of COD(ini)) and the current (34.5% of COD(ini)) were the largest and second-largest electron sinks in P-MFC. Methane gas was only detected at levels of more than 10% in G- and M-MFCs, meaning that electrochemically active bacteria (EAB) could out-compete acetoclastic methanogens. The microbial communities were different for fermentable and non-fermentable substrate-fed MFCs. Probably, bacteria related to Lactococcus spp. found in G-MFCs with fermentable substrates would be involved in both fermentation and electricity generation. Acinetobacter-like species, and Rhodobacter-like species detected in all the MFCs would be involved in oxidation of organic compounds and electricity generation.
In a microbial fuel cell (MFC), exoelectrogens, which transfer electrons to the electrode, have been regarded as a key factor for electricity generation. In this study, U-tube MFC and plating methods were used to isolate exoelectrogens from the anode of an MFC. Disparate microorganisms were identified depending on isolation methods, despite the use of an identical source. Denaturing gel gradient electrophoresis (DGGE) analysis showed that certain microorganisms became dominant in the U-tube MFC. The predominant bacterium was similar to Ochrobactrum sp., belonging to the Alphaproteobacteria, which was shown to be able to function as an exoelectrogen in a previous study. Three isolates, one affiliated with Bacillus sp. and two with Paenibacillus sp., were identified using the plating method, which belonged to the Gram-positive bacteria, the Firmicutes. The U-tube MFCs were inoculated with the three isolates using the plating method, operated in the batch mode and the current was monitored. All of the U-tube MFCs inoculated with each isolate after isolation from plates produced lower current (peak current density: 3.6–16.3 mA/m2) than those in U-tube MFCs with mixed culture (48.3–62.6 mA/m2). Although the isolates produced low currents, various bacterial groups were found to be involved in current production.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.