Nociceptive Memristor

Abstract

The biomimetic characteristics of the memristor as an electronic synapse and neuron have inspired the advent of new information technology in the neuromorphic computing. The application of the memristors can be extended to the artificial nerves on condition of the presence of electronic receptors which can transfer the external stimuli to the internal nerve system. In this work, nociceptor behaviors are demonstrated from the Pt/HfO /TiN memristor for the electronic receptors. The device shows four specific nociceptive behaviors; threshold, relaxation, allodynia, and hyperalgesia, according to the strength, duration, and repetition rate of the external stimuli. Such nociceptive behaviors are attributed to the electron trapping/detrapping to/from the traps in the HfO layer, where the depth of trap energy level is ≈0.7 eV. Also, the built-in potential by the work function mismatch between the Pt and TiN electrodes induces time-dependent relaxation of trapped electrons, providing the appropriate relaxation behavior. The relaxation time can take from several milliseconds to tens of seconds, which corresponds to the time span of the decay of biosignal. The material-wise evaluation of the electronic nociceptor in comparison with other material, which did not show the desired functionality, Pt/Ti/HfO /TiN, reveals the importance of careful material design and fabrication.

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges

“…For example, intrinsic plasticity involves the modification of neuronal excitability through regulation of voltage-dependent ion-channels and can be viewed as a change in the input-output function of a neuron. [199,200] Firstly, the regulation of dendritic channels amplifies EPSP.…”

A True Random Number Generator Using Threshold‐Switching‐Based Memristors in an Efficient Circuit Design

“…[10,11] The memristor's nonuniformity, however, mainly due to the complicated and stochastic physical phenomena occurring in it during switching, has been one of the biggest issues with regard to its use as a next-generation memory. As Internet of things (IoT) is showing explosive growth, and a software random number generator, also known as "pseudorandom number generator (PRNG)," can easily be attacked due to its www.advelectronicmat.de energy below the conduction band edge, which was explained by Kim et al [10] The mechanism can be understood from the presence of shallow trap levels in HfO 2 and an internal electric field, caused by the work function difference between two electrodes (Pt and TiN). [12][13][14][15][16][17] TRNG is a hardware device that produces random bits from the intrinsic stochasticity of its physical process.…”

Kernel Application of the Stacked Crossbar Array Composed of Self‐Rectifying Resistive Switching Memory for Convolutional Neural Networks

“…While the HfO 2− x /TiN interface constituted a quasi‐Ohmic contact that provided a fluent carrier (electron) migration between the abundant trap sites in the HfO 2− x film and TiN electrode upon a positive bias application, the Pt/HfO 2− x interface constituted a Schottky‐type interfacial contact, suppressing the electron injection from the Pt to HfO 2− x upon a negative bias application to the Pt TE. The details of such electron trapping or detrapping‐based ReRAM are reported in previous studies . In short, the ReRAM was set (switching from HRS to LRS) upon the sufficiently high positive bias application to the Pt, which trapped the electrons within the HfO 2− x layer, whereas the device was reset (switching from LRS to HRS), upon the application of a sufficiently negative bias, which detrapped the electrons within the HfO 2− x layer.…”