Three-Dimensional Crossbar Arrays of Self-rectifying Si/SiO2/Si Memristors

Abstract

Memristors are promising building blocks for the next-generation memory and neuromorphic computing systems. Most memristors use materials that are incompatible with the silicon dominant complementary metal-oxide-semiconductor technology, and require external selectors in order for large memristor arrays to function properly. Here we demonstrate a fully foundry-compatible, all-silicon-based and self-rectifying memristor that negates the need for external selectors in large arrays. With a p-Si/SiO 2 /n-Si structure, our memristor exhibits repeatable unipolar resistance switching behaviour (10 5 rectifying ratio, 10 4 ON/OFF) and excellent retention at 300°C. We further build three-dimensinal crossbar arrays (up to five layers of 100 nm memristors) using fluid-supported silicon membranes, and experimentally confirm the successful suppression of both intra-and inter-layer sneak path currents through the built-in diodes. The current work opens up opportunities for low-cost mass production of three-dimensional memristor arrays on large silicon and flexible substrates without increasing circuit complexity.

Retention Secured Nonlinear and Self‐Rectifying Analog Charge Trap Memristor for Energy‐Efficient Neuromorphic Hardware

“…In this regard, charge trap memristors (CTM) are promising memristive systems due to their inherent forming‐free and self‐rectifying characteristics. [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 ] Also, they show analog conductance change characteristics at low operation current range, enabling high‐density MCA applications. [ 12 , 13 ] The CTM mechanism is identical to the charge‐trap flash (CTF) memory, in which reversible charge trapping and de‐trapping to the charge trap layer alter the threshold voltage of the transistor.…”

2D-Material-Based Volatile and Nonvolatile Memristive Devices for Neuromorphic Computing

“…A common method is to choose multiple-device combinations, such as one transistor–one memristor (1T1M), one selector–one memristor (1S1M), and one diode–one memristor (1D1M), in neuromorphic hardware systems . Moreover, 2D material-based synaptic devices with self-rectifying properties should be utilized to avoid the “sneak path” problem in neuromorphic hardware systems. , Their applications in RC and ANN for neuromorphic computing depend on their retention time. Exact control of the decay time of these memristive devices on the basis of a specific requirement is necessary.…”

To the Issue of the Memristor’s HRS and LRS States Degradation and Data Retention Time

“…Also in this work, an increase in endurance was observed with an increase in the LRS resistance (below the operating current). A memristor whose operating principle is based on the diffusion of silver ions [92] has shown endurance >10 8 switching cycles. According to the authors, the result obtained is associated with minimizing the switching voltage, since the endurance readings are reduced on devices with a large limiting current.…”

“…Thus, a bit of information in the memristor memory cell is stored in the form of structural changes in the local region of the dielectric enclosed between the two conducting electrodes. Memristors with only two levels of electrical resistance (one-bit), integrated into the cross-bar architecture [2][3][4][5][6], and in 3D configura-tions [7,8], represent the foundation for future ultralarge integrated circuits of fast nonvolatile memory ReRAM with a long retention time. At the same time, multilevel (multibit) memristors with a set of intermediate discrete levels of electrical resistance of the cells (from 4 to 20) [9][10][11][12][13][14][15][16][17] offer the possibility of creating systems with a parallel computing mechanism and synaptic plasticity function, which is necessary for constructing recurrent neural networks and next generation artificial intelligence architectures [18][19][20][21].…”