From Ferroelectric Material Optimization to Neuromorphic Devices

Abstract

Due to the voltage driven switching at low voltages combined with nonvolatility of the achieved polarization state, ferroelectric materials have a unique potential for low power nonvolatile electronic devices. The competitivity of such devices is hindered by compatibility issues of well‐known ferroelectrics with established semiconductor technology. The discovery of ferroelectricity in hafnium oxide changed this situation. The natural application of nonvolatile devices is as a memory cell. Nonvolatile memory devices also built the basis for other applications like in‐memory or neuromorphic computing. Three different basic ferroelectric devices can be constructed: ferroelectric capacitors, ferroelectric field effect transistors and ferroelectric tunneling junctions. In this article first the material science of the ferroelectricity in hafnium oxide will be summarized with a special focus on tailoring the switching characteristics towards different applications.The current status of nonvolatile ferroelectric memories then lays the ground for looking into applications like in‐memory computing. Finally, a special focus will be given to showcase how the basic building blocks of spiking neural networks, the neuron and the synapse, can be realized and how they can be combined to realize neuromorphic computing systems. A summary, comparison with other technologies like resistive switching devices and an outlook completes the paper.

Near-90° Switch in the Polar Axis of Dion–Jacobson Perovskites by Halide Substitution

“…Ferroelectric materials have broad applications in memory storage and switching devices, actuators, and nonlinear optics. − Two-dimensional hybrid organic–inorganic perovskites are regarded as a promising candidate for next-generation ferroelectrics owing to their outstanding ambient stability and high structural versatility. − The flexibility in tuning either the organic or inorganic components in hybrid perovskites allows a rational way to engineer ferroelectricity in 2D hybrid perovskites. , Generally, the order–disorder transition of the organic cations is coupled to the displacive-type transition of the metal halide octahedral framework via hydrogen bonds. , The concerted complex displacements of ions result in the appearance of spontaneous polarization along a polar axis . Therefore, tuning the strength of the electrostatic effect between the organic cations and inorganic halides affects the extent of the polar off-centering displacements.…”

Self‐Curable Synaptic Ferroelectric FET Arrays for Neuromorphic Convolutional Neural Network

“…[16] To circumvent the constraints of existing neuromorphic systems based on two-terminal memristors, three-terminal synaptic transistors have been developed. [17][18][19][20] These individually programmable transistors eliminate crosstalk and sneak path current between adjacent devices, enabling the weight update process to be conducted in a selective and parallel manner. In comparison to other types of three-terminal synaptic transistors such as charge trap-based synapses, [17,18] hafniumoxide-based ferroelectric field-effect transistors (FeFETs) offer the benefits of low program bias, quick switching speed, high scalability, and complementary-metal-oxide-semiconductor compatibility.…”

“…In comparison to other types of three-terminal synaptic transistors such as charge trap-based synapses, [17,18] hafniumoxide-based ferroelectric field-effect transistors (FeFETs) offer the benefits of low program bias, quick switching speed, high scalability, and complementary-metal-oxide-semiconductor compatibility. [19,20] However, two significant challenges must be resolved to successfully integrate FeFET into neuromorphic computing as a synaptic device. First, as the synaptic weight is updated iteratively during in situ training, the synaptic device should handle multiple program operations.…”

An electric-field-driven ferroelectric nanodomain structure and its multilevel data storage application

“…[9][10][11][12][13][14][15] Ferroelectric-based solid-state synapses show promise for achieving a highly efficient biomimetic neural network. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] The ferroelectric nanodomain structure is adjusted to control spike-timing-dependent plasticity or multilevel data storage in ferroelectric field effect transistors, memristors, and tunnel junctions. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] As future biomimetic neural synaptic networks will consist of billions of ferroelectric material-based synapses, a clear understanding of the electric-field-driven ferroelectric nanodomain structure is necessary for their application.…”

“…[6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] The ferroelectric nanodomain structure is adjusted to control spike-timing-dependent plasticity or multilevel data storage in ferroelectric field effect transistors, memristors, and tunnel junctions. [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] As future biomimetic neural synaptic networks will consist of billions of ferroelectric material-based synapses, a clear understanding of the electric-field-driven ferroelectric nanodomain structure is necessary for their application. Additionally, it is also meaningful to explore the typical function of multilevel data storage in synapses.…”