Neuroelectronics and Memristive Nanomaterials Research Laboratory (NEUROMENA)

Contract number

075-15-2022-1123

Time span of the project

2022-2024

Laboratory's website

since August 2023 Kazantsev Viktor Borisovich

2022 - 2023 Park‬ Bae Ho

As of 01.12.2023

Number of staff members

scientific publications

Objects of intellectual property

Name of the project:

Hybrid neuroelectronics for robotic complexes and artificial intelligence systems based on biocompatible memristive nanomaterials

Research directions: Materials technology

Goals of project:

Developing electronic components for hybrid neuromorphic systems based on biocompatible memristive nanomaterials, technologies of the synthesis of films of metal oxides and new gradient oxide nanocomposites, oxide nanostructures and 2D materials with improved operational parameters, designing structures and technologies for manufacturing memristive crossbars and algorithms for programming neural networks based on them, modeling, manufacturing and of prototypes of hybrid neuromorphic systems performing the main cognitive functions of robotic complexes, artificial intelligence systems and next-generation computer architectures based on neural processors as well as creating a world-class laboratory of neural electronics and memristive nanomaterials at the Southern Federal University under the supervision of the leading scientist from South Korea.

Project objective:

Modeling the processes of the synthesis of biocompatible memristive nanomaterials based on oxide films of metals, oxide films of gradient nanocomposites, oxide nanostructures and 2D materials produced by electrochemical, pulsed laser and magnetron deposition as well as by chemical precipitation from the gas phase;
Modeling resistive switching in biocompatible memristive nanomaterials based on oxide films of metals, oxide films of gradient nanocomposites, oxide nanostructures and 2D materials;
Researching the laws of synthesis of biocompatible memristive nanomaterials based on oxide films of metals, oxide films of gradient nanocomposites, oxide nanostructures and 2D materials;
Developing methodologies and programs for controlling the measurement of the main parameters of resistive switching in memristive structures;
Developing basic mathematical models of the neural-like dynamics of biocompatible memristive structures;
Researching the structure, phase and element composition of biocompatible memristive nanomaterials based on oxide films of metals, oxide films of gradient nanocomposites, oxide nanostructures and 2D materials;
Researching the laws of impact of the dimensional effects, the structural and electrophysical parameters on resistive switching in biocompatible memristive nanomaterials;
Studying the reproducibility and temporal stability of resistive switching in biocompatible memristive nanomaterials;
Researching multilevel resistive switching in biocompatible memrisitve nanomaterials;
Researching the processes of the generation and transmission of signals in models of neurons and synapses built on the basis of biocompatible memristive structures;
Researching of the impact of the materials of the electrodes of crossbar structures, the composition and temperature of the environment on resistive switching in prototypes of hybrid neuromorphic structures based on biocompatible memristive nanomaterials;
Developing and optimizing the designs of memristive crossbar structures for hybrid neuromorphic systems of robotic complexes and RеRAM neural processors;
Designing technological processes of the formation of memristive crossbar structures for hybrid neuromorphic systems of robotic complexes and RеRAM neural processors;
Manufacturing prototypes of hybrid neuromorphic structures of robotic complexes and RеRAM neural processors based on biocompatible memristive nanomaterials;
Modeling processes of signal transmission and processing in neural networks, relying on biologically feasible memristive devices;
Programming prototypes of hybrid neuromorphic structures of robotic complexes and RеRAM neural processors based on biocompatible memristive nanomaterials;
Researching the operational parameters of prototypes of hybrid neuromorphic strucrues of robotic complexes and RеRAM neural processors based on biocompatible memristive nanomaterials.

Scientific results:

A mathematical model of the formation of titanium oxide nanostructures by local anodic oxidation was developed. Based on the developed mathematical model, a theoretical study of the influence of the formation modes of oxide nanostructures on their geometric parameters, as well as the distribution of the chemical composition of the oxide and oxygen vacancies within the nanostructure, was conducted.

A model for the formation of biocompatible memristive nanomaterials based on metal oxide films obtained by pulsed laser deposition was developed, allowing for the calculation of the film thickness distribution profile on the substrate surface, taking into account the constructive and technological parameters of the equipment used.

The modeling of resistive switching in biocompatible memristive nanomaterials based on metal oxide films, gradient nanocomposite oxide films, oxide nanostructures, and 2D materials was performed. A mathematical model of the formation and destruction of nanoscale conduction channels (filaments) in the titanium oxide nanostructure was developed, considering the processes of generation and migration of oxygen vacancies within the nanostructure under the influence of an external electric field.

Experimental studies on the synthesis patterns of biocompatible memristive nanomaterials based on metal oxide films and oxide nanostructures by pulsed laser deposition were conducted. The influence of the number of pulses, thickness, oxygen pressure, annealing temperature, and laser pulse frequency on the parameters of metal oxide films was established.

A mathematical model of neuro-like dynamics of biocompatible memristive structures and a neural generator model, representing a modified FitzHugh-Nagumo neuron with complex threshold excitation and memristor nonlinearity, were developed. The proposed neural generator model with memristive nonlinearity reproduces the complex threshold response dynamics, which is advantageous for designing spike-based neuromorphic systems.

Studies on the structure, phase, and elemental composition of biocompatible memristive nanomaterials based on metal oxide films obtained by pulsed laser deposition using X-ray diffraction (XRD), energy-dispersive X-ray spectroscopy (EDX), and X-ray photoelectron spectroscopy (XPS) methods were conducted, establishing the influence patterns of pulsed laser deposition modes on the structural properties of BaTiO₃, LiNbO₃, ZnO, and VO₂.

The resistive switching in transparent memristor structures based on ZnO and TiO₂ films obtained by magnetron sputtering at room temperature in an oxygen-free environment on glass substrates for use in neuromorphic systems, including dynamic machine vision systems, transparent, and flexible electronics, was investigated. Studies were also conducted on the formation of memristor structures by magnetron sputtering with varying ZnO film thicknesses and different electrode materials on silicon substrates. Based on the obtained experimental results, a technological route for forming transparent memristor structures by magnetron sputtering (MF and RF modes) at room temperature in an oxygen-free environment was developed and optimized.

Regularities of the influence of dimensional effects on resistive switching in titanium oxide nanostructures obtained by LAO method have been established. The analysis of the data showed that this method allows for the controlled formation of memristor structures with switching voltages in a wide range of values. The result demonstrates the absence of parasitic connections during resistive switching in non-forming titanium ONC and allows for significant simplification of the memristor array manufacturing technology by forming a common oxide layer for all ReRAM elements, rather than for each element individually.

A study of the temporal stability of the surface charge of TiO₂ films was conducted. Based on the results, time dependence of voltage and charge region size were constructed. It was shown that the voltage decreased from 320±21 mV to 22±5 mV over 90 minutes, and the structure size increased from 3.43±0.12 μm to 4.12±0.14 μm.

Experimental studies of multilevel resistive switching in nanocrystalline films of zinc oxide, titanium oxide, and hafnium oxide have been conducted. Analysis of the results showed that TiO₂, HfO₂, and ZnO films exhibit multilevel resistive switching. Increasing the voltage sweep and current compliance leads to an increase in the number of resistive states.

The influence of top electrode materials on the resistive switching of memristor structures based on nanocrystalline zinc oxide films has been investigated. It was shown that memristor structures with a top electrode made of carbon (C) exhibit the highest RHRS/RLRS ratio, while structures with a top electrode made of platinum (Pt) demonstrate the lowest switching voltage values. Thus, the choice of top electrode material should be determined based on the initial requirements when manufacturing neuromorphic structures: if the priority is maximizing the RHRS/RLRS ratio, then the top electrodes should be made of carbon (C); if the priority is energy efficiency, then they should be made of platinum (Pt).

The influence of temperature on the main parameters of memristor structures based on nanocrystalline zinc oxide film has been investigated. It was shown that memristor structures based on nanocrystalline zinc oxide film are capable of functioning in the temperature range from 25 °C to 120 °C, which is important for the operation of future elements of neuromorphic structures in adverse environmental conditions.

The influence of top electrode materials on the resistive switching of titanium oxide-based memristor structures has been studied. Analysis of the results showed that the nanostructures formed on the surface of the Ti film exhibit a stable resistive switching effect without the need for additional forming operations. The material of the top contact significantly affected the current-voltage characteristics of the Ti/TiO2/top electrode structure. It was shown that using Pt as the top contact resulted in a symmetrical current-voltage characteristic, while using TiN and C as top contacts resulted in asymmetrical current-voltage characteristics. Using Pt as the top contact reduces the power consumption of the Ti/TiO2/top electrode structure compared to TiN, and also reduces power consumption compared to C.

An adaptive model of a memristive device Au/Zr/SiO₂/TiN/Ti/SiO₂/Si demonstrating stochastic resistive switching characteristics under the influence of a neuromorphic signal, which ensures adaptive coupling and synchronization of neuromorphic FitzHugh-Nagumo memristive oscillators, has been experimentally investigated.

Education and personnel occupational retraining:

During the laboratory's operation, its employees have successfully completed 20 courses of additional professional education and advanced training courses.

A new main professional educational program of higher education for students in the field of study 11.04.04 "Electronics and Nanoelectronics," specializing in "Neuromorphic and Neurohybrid Artificial Intelligence Systems," has been developed and is being implemented at the Southern Federal University. Qualification: Master's degree.

The laboratory staff have successfully habilitated 6 bachelor's theses, 6 master's theses, and 2 candidate dissertations in the field of scientific research.

Cooperation:

Chandigarh University
Lovely Professional University
N. I. Lobachevsky State University of Nizhny Novgorod
Molecular Electronics Research Institute
Immanuel Kant Baltic Federal University
Vladimir State University
Skolkovo Institute of Science and Technology

Hide Show full

Avilov V. I., Tominov R. V., Vakulov Z. E., Zhavoronkov L. G., Smirnov V. A.

Titanium oxide artificial synaptic device: Nanostructure modeling and synthesis, memristive cross-bar fabrication, and resistive switching investigation // Nano Research. 2023. №16. Vol. 7. PP. 10222-10233.

Tominov R. V., Vakulov Z. E., Avilov V. I., Shikhovtsov I.A., Varganov V. I., Kazantsev V. B., Gupta L. R., Prakash C., Smirnov V.A.

Approaches for Memristive Structures Using Scratching Probe Nanolithography: Towards Neuromorphic Applications // Nanomaterials. 2023. № 13. Vol. 10. PP. 1583.

Kipelkin I., Gerasimova S., Guseinov D., Pavlov D., Vorontsov V., Mikhaylov A., Kazantsev V.

Mathematical and experimental model of neuronal oscillator based on memristor-based nonlinearity // Mathematics. 2023. № 11. Vol. 5. PP. 1268.

Koryazhkina M. N., Filatov D. O., Tikhov S. V., Belov A. I., Serov D. A., Kryukov R. N., Zubkov S. Y., Vorontsov V. A., Pavlov D. A., Gryaznov E. G., Orlova E. S.

Electrical characteristics of CMOS-compatible SiO x-based resistive-switching devices // Nanomaterials. 2023. № 13. Vol. 14. PP. 2082.

Stasenko S. V., Mikhaylov A. N., Kazantsev V. B.

Control of Network Bursting in a Model Spiking Network Supplied with Memristor—Implemented Plasticity // Mathematics. 2023. № 11. Vol. 18. PP. 3888.

Prakash C., Gupta L. R., Mehta A., Vasudev H., Tominov R., Korman E., Fedotov A., Smirnov V., Kesari K. K.

Computing of neuromorphic materials: an emerging approach for bioengineering solutions // Materials Advances. 2023. № 4. Vol. 23. PP. 5882-5919.

Program for calculating deposition modes of nanocrystalline films deposited by pulsed laser deposition method: software for computers, Russian Federation No. 2022668785, filed on 07.10.2022; published on 12.10.2022.

Khakhulin D. A., Vakulov Z. E., Smirnov V. A.

Program for calculating the parameters of resistive switching in memristor structures based on metal oxides: software for computers, Russian Federation No. 2022669961, filed on 13.10.2022; published on 26.10.2022.

Ugriumov I. S., Tominov R. V., Vakulov Z. E., Parshina N. V., Shikhovtsov I. A., Avilov V. I., Smirnov V. A.

Program for automating parametric analysis of the reproducibility of resistive switching in memristive nanomaterials and structures: software for computers, Russian Federation No. 2023682931, filed on 18.10.2023; published on 01.11.2023.

Khakhulin D. A., Dzyuba D. A., Smirnov V. A.

Hosting organization	Field of studies	City	Invited researcher	Time span of the project
Laboratory for Ultra Wide-band Gap Semiconductors National University of Science and Technology MISIS - (NUST MISIS)	Material Technology	Moscow	Кузнецов Андрей Юрьевич Sweden	2022-2024
Laboratory for Ion-selective membranes M.V.Lomonosov Moscow State University - (MSU)	Material Technology	Moscow	Ameduri Bruno Michel France	2022-2024
Laboratory for Advanced Steels for Agricultural Equipment Russian State Agrarian University - Moscow Timiryazev Agricultural Academy - (RSAU – MTAA named after K.A. Timiryazev)	Material Technology	Moscow	Kaibyshev Rustam Oskarovich Russia	2021-2023