Лаборатория

Лаборатория стохастических
мультистабильных систем

Laboratory of Stochastic
Multistable Systems

Scientific results 2019

1. In 2019, the following scientific results were obtained:


1. Computer simulation of stochastic microscopic phenomena responsible for the rearrangement of the atomic structure of oxide materials under the impact of an electric field and current (Joule heating) has been carried out. An original approach has been developed for modeling the patterns and parameters of resistive switching based on the kinetic Monte-Carlo (KMC) method of modeling the stochastic migration of oxygen vacancies / ions in metal-oxide memristive structures. An algorithm and a program for the KMC calculation of electroforming and switching processes have been developed, which correctly describe the experimentally observed I – V characteristics, their variance from cycle to cycle and the response of the memristor to impulse influences.

1.1. The efficiency of the computer model is demonstrated for experimentally realized Au / oxide / TiN structures, in which polycrystalline films of yttrium-stabilized zirconium dioxide (ZrO2(Y)) are used as the oxide material. It is shown that the proposed algorithm provides dynamic accounting of the varying distribution of the electric potential and temperature with the growth of a filament of arbitrary shape, as well as  the effect of oxygen exchange at the interface.

1.2. The efficiency of the computer model has been demonstrated on experimentally realized amorphous SiOx films with a columnar structure obtained by magnetron sputtering. It is shown that a feature of the proposed algorithm is its flexibility, as the approach which it is based on can be applied to various oxide memristive structures with the introduction (if necessary) of corrections, taking into account the features of these structures. Moreover, this approach does not require time-consuming calculations and is not critical to the computing power of a computer.


2. Experimental verification of the developed predictive macro models is carried out.

2.1. Experimental verification of a stochastic model of a memristive device based on a multilayer structure Au/ Ta/ ZrO2(Y)/ Ta2O5/ TiN/ Ti is carried out. It is shown that the proposed stochastic model is able to adequately reproduce the fundamental properties of resistive switching, such as the hysteresis I – V characteristic and its dependence on the sweep frequency of the control voltage.

2.2. Measurements of the evolution of the probability distributions of the resistance of a memristor made on the basis of ZrO2(Y)/Ta2O5 и SiO2, as well as a virtual memristor in the contact of the AFM probe to the surface of the ZrO2(Y) films have been carried out. No fundamental differences were observed in the evolution of distributions under the influence of external noise in both structures. This indicates that the observed regularities in the behavior of memristors characterize the fundamental properties of the memristor of a stochastic system, regardless of the material of the functional dielectric or the size of the contacts.

It was found that the distributions under the influence of an external noise signal change with time and come to a stationary state. The effect of transitional bimodality is found.

2.3. Theoretical estimates have been made and the properties of the phenomenon of stochastic resonance and resonance activation in memristive structures based on ZrO2(Y)(10 nm) / Ta2O5(10 nm) and in the corresponding ZrO2(Y) films have been experimentally confirmed. It is shown that the mechanism of occurrence of stochastic resonance is threshold and is due to the nonlinear dependence of the resistance of the dielectric layer on the concentration of defects. It was found that in the phenomenon of resonance activation, the amplitude of resistive switching has a maximum at a frequency corresponding to the characteristic frequency of hopping of O2– ions to neighboring oxygen vacancies in ZrO2(Y).


3. An experimental study of degradation processes in memristive structures based on Si3N4 и ZrO2(Y) thin films has been carried out.

An experimental study of degradation processes in multilayer memristive structures based on ZrO2(Y) (10 nm) / Ta2O5 (10 nm) as part of microdevices was carried out by supplying sequences of reading and programming voltage pulses. It was found that the maximum number of switchings depends on the parameters of a particular memristor and their variance. Most of the structures demonstrate resistive switching with the number of switching cycles 1n105 – 11106 without changing the parameters of the switching voltage pulses.


4. A new memristor macromodel has been developed, taking into account influence of external and internal noises and it has been compared with the micromodels of physicochemical phenomena responsible for resistive switching.

Based on the analysis of the micromodel of diffusion jumps, an estimate of the current jump caused by a single act of oxygen ion diffusion inside the filament was obtained: i0  7710–13 A. Analysis of the spectral characteristics of current fluctuations showed good agreement between the initial experimental data and the simulation results.

Various noise effects in the form of white and colored Gaussian noise have been analyzed. The absence of a steady probability distribution of the memristor resistance in the framework of the considered models was found.


5. Algorithms have been developed and new neural network architectures based on memristive devices have been tested.

6. Published 12 articles in scientific journals indexed in the Web of Science database (including 7 articles in journals from Q1).

7. Laboratory staff has taken part in 12 international conferences, scientific seminars, symposia on the subject of laboratory work (at the expense of subsidies).

8. In October 2019, on the basis of the Center for Scientific Culture in Erice (Italy), an international scientific conference "New Trends in Nonequilibrium Stochastic Multistable Systems and Memristors" (NES2019) was organized and held.

9. 3 applications for the grant of a patent of the Russian Federation and 1 application for the grant of an international patent (PCT) were filed and registered. RF patent for invention No. 2706207 dated 11/14/2019 was received (application filled in 2018).

10. Purchase of building materials for the current repair of the laboratory premises (room 518, building 4) was carried out. Furniture and office equipment were purchased to equip the laboratory.

11. 11 members of the research team were trained.

12. The topology and technological route for the production of a chip with an array of memristive devices have been developed.

13. Laboratory staff took part in 5 international conferences, scientific seminars, symposia on the subject of laboratory work (at the expense of additional funds).

14. 2 events on the topic "Stochastic multistable systems" (at the expense of additional funds) have been organized and hold:
  • Regular scientific and educational seminar on the basis of the laboratory of stochastic multistable systems of the UNN.
  • Subsection "Stochastic multistable systems" in the framework of the XXIII scientific conference on radiophysics dedicated to the 100th anniversary of the birth of N.А. Zheleztsov.

2. Publication in scientific journals indexed in the "web of science" database

№ 

Title of the article

Surname

Name

Title
of the journal

Year, month
(volume, issue)

Impact factor of publication

1

Measurement of the activation energies of oxygen ion diffusion in yttria stabilized zirconia by flicker noise spectroscopy

Yakimov
Filatov
Gorshkov
Antonov
Antonov
Kluev
Carollo
Spagnolo

Arkady
Dmitry
Oleg
Dmitry
Ivan
Alexey
Angelo
Bernardo

Applied Physics Letters

2019, 06
(114)

3.521

2

Noise-induced resistive switching in a memristor based on ZrO2(Y)/Ta2O5 stack

Filatov
Vrzheshch
Tabakov
Novikov
Belov
Antonov
Koryazhkina
Mikhaylov
Gorshkov
Dubkov
Spagnolo

Dmitry
Daria
Oleg
Alexey
Alexey
Ivan
Maria
Alexey
Oleg
Alexander
Bernardo

Journal of Statistical Mechanics: Theory and Experiment

2019, December
(2019, №12)

2.371

3

Memory effect and generation-recombination noise of magnetic monopoles in spin ice

Kluev
Yakimov
Spagnolo

Alexey
Arkady
Bernardo

Journal of Statistical Mechanics: Theory and Experiment

2019, September
(2019, No.9)

2.371

4

On quantumness in multi-parameter quantum estimation

Carollo
Spagnolo
Dubkov

Angelo
Bernardo
Alexander

Journal of Statistical Mechanics: Theory and Experiment

2019, September
(2019, No.9)

2.371

5

Haldane model at finite temperature

Spagnolo
Dubkov
Carollo

Bernardo
Alexander
Angelo

Journal of Statistical Mechanics: Theory and Experiment

2019, September
(2019, No.9)

2.371

6

Yttria-stabilized zirconia cross-point memristive devices for neuromorphic computing

Demin
Belov
Korolev
Griaznov
Pavlov
Gorshkov
Mikhaylov

Viacheslav
Alexey
Dmitriy
Eugeny
Dmitry
Oleg
Alexey

Microelectronic Engineering

2019, 07
(215)

1.654

7

Adaptive Properties of Spiking Neuromorphic Networks with Synapses Based on Memristive Elements

Demin

Viacheslav

Technical Physics Letters

2019, 05
(45, 4)

0.773

8

Mechanisms of Current Transport and Resistive Switching in Capacitors with Yttria-Stabilized Hafnia Layers

Tikhov
Gorshkov
Belov
Antonov
Koryazhkina
Mikhaylov

Stanislav
Oleg
Alexey
Ivan
Maria
Alexey

Technical Physics

2019, 06
(64, 6)

0.637