Ferroelectric memristor with vertical self-assembly nano-composite structure as well as preparation method and application of ferroelectric memristor

Abstract

The invention provides a ferroelectric memristor with a vertical self-assembly nano-composite structure and a preparation method and application of the ferroelectric memristor. The structure of the ferroelectric memristor sequentially comprises a Si substrate, a strontium titanate buffer layer, a lanthanum strontium manganese oxide bottom electrode layer, a composite structure ferroelectric active layer formed by barium strontium titanate and neodymium oxide, and a top electrode layer from bottom to top, barium strontium titanate and neodymium oxide in the ferroelectric active layer of the composite structure are alternately arranged to form a vertical self-assembly columnar nano composite structure. Neodymium oxide is introduced into the ferroelectric active layer barium strontium titanate to form a vertical self-assembly nano-composite structure, extra interfaces, defects and stress are introduced while the dielectric constant is reduced, domain walls can be effectively fixed, out-of-plane lattice constants can be increased, oxygen vacancy movement can be controlled, the stability of switching voltage and the performance of the whole memristor are remarkably improved, and the memristor can be widely applied to the field of memristor devices. And the conductivity state can be stably adjusted. The multi-level storage of the ferroelectric thin film device is facilitated, and the ferroelectric thin film device can be applied to cranial nerve-like calculation and high-efficiency storage.

Description

Technical Field

[0001] The invention relates to the technical field of memristors, in particular to a vertical self-assembled nano-composite structure ferroelectric memristor, a preparation method and application thereof. Background Art

[0002] With the rapid development of information technology and artificial intelligence, the traditional von Neumann architecture leads to significant computing energy consumption and low data processing efficiency when processing dense data due to the separation of storage and computing. Brain-like neural computing systems are considered to be an important direction to replace traditional computing architectures in the future. Inspired by the fact that the human brain processes data through large-scale parallel operations of neurons and synapses, this new type of brain-like neural computing system has emerged. These systems directly receive external stimulation signals and build neural networks by continuously updating synaptic weights, aiming to simulate the learning and memory characteristics of the human brain and consume less energy when completing high-intensity tasks. Ferroelectric memristors have attracted much attention because they can accurately control the resistance state through polarization reversal and do not require electroforming activation behavior. However, existing ferroelectric memristor devices inevitably produce oxygen defects (oxygen vacancies) during the preparation process. These oxygen vacancies will lead to the random formation of conductive filaments, causing problems such as high leakage current, low durability and unstable switching voltage. In addition, most ferroelectric materials currently need to be grown on single crystal substrates, which is not conducive to silicon-based integration and limits their application.

[0003] To overcome these problems, many emerging studies have focused on creating unique interface effects by constructing heterogeneous structures to control or enhance the mobility of carriers. Notably, vertically self-assembled spontaneous phase sequence nanocomposites (VSNs) constructed by exploiting the structural incompatibility of different oxides have shown promise in forming channels to control the movement of oxygen vacancies, expanding the tunable range of ferroelectric domains, enhancing remnant polarization, and stabilizing the modulation of devices. Summary of the invention

[0004] The purpose of the present invention is to provide a vertical self-assembled nanocomposite structure ferroelectric memristor, a preparation method and application thereof, so as to solve the problem of poor device performance of the existing ferroelectric memristor caused by oxygen defects.

[0005] The present invention is achieved in that:

[0006] A vertical self-assembled nanocomposite ferroelectric memristor comprises, from bottom to top, a Si substrate, a strontium titanate buffer layer, a lanthanum strontium manganese oxide bottom electrode layer, a composite structure ferroelectric active layer formed by barium strontium titanate and neodymium oxide, and a top electrode layer; in the composite structure ferroelectric active layer, barium strontium titanate and neodymium oxide are arranged alternately to form a vertical self-assembled columnar nanocomposite structure.

[0007] Preferably, the molar ratio of barium strontium titanate to neodymium oxide in the composite structure ferroelectric active layer is 1:1, and the barium strontium titanate is specifically Ba 0.6 Sr 0.4 TiO3.

[0008] Preferably, the material of the lanthanum strontium manganese oxide bottom electrode layer is La 0.67 Sr 0.33 MnO3.

[0009] Preferably, the Si substrate is a P-type Si substrate.

[0010] The method for preparing the vertical self-assembled nanocomposite structure ferroelectric memristor comprises the following steps:

[0011] (1) Pretreatment of Si substrate;

[0012] (2) using pulsed laser deposition technology to sequentially deposit a strontium titanate buffer layer, a lanthanum strontium manganese oxide bottom electrode layer, and a composite structure ferroelectric active layer on a Si substrate; the target material used for depositing the composite structure ferroelectric active layer is a composite target material formed by sintering barium strontium titanate and neodymium oxide;

[0013] (3) A top electrode layer is deposited on the ferroelectric active layer of the composite structure using magnetron sputtering deposition technology.

[0014] Preferably, the barium strontium titanate is Ba 0.6 Sr 0.4 TiO3, and the neodymium oxide is Nd2O3.

[0015] Preferably, the specific preparation process of the composite target is: Ba 0.6 Sr 0.4 TiO3 and Nd2O3 were mixed in a 1:1 molar ratio and prepared by pure oxygen sintering at 1200°C.

[0016] Preferably, the temperature when depositing the composite structure ferroelectric active layer is 600°C to 800°C.

[0017] Preferably, the oxygen pressure is 1Pa-13Pa when depositing the composite structure ferroelectric active layer.

[0018] The vertical self-assembled nanocomposite structure ferroelectric memristor provided by the present invention can be applied to brain-like neural computing and high-performance storage, for example, it can be applied in a multi-factor security encryption system.

[0019] The present invention introduces Nd2O3 into the ferroelectric active layer barium strontium titanate to form a vertical self-assembled nanocomposite structure, which introduces additional interfaces, defects and stresses while reducing the dielectric constant, can effectively fix the domain wall to increase the out-of-plane lattice constant, and control the movement of oxygen vacancies, significantly improve the stability of the switching voltage and the overall memristor performance, so that it can stably adjust the conductivity state, thereby achieving stable multi-level regulation and neuromorphic computing. The main application of this 1-3 dimensional ferroelectric film related to storage characteristics has two directions: one is to provide more stress for the ferroelectric film and enhance its out-of-plane lattice constant. This enhanced out-of-plane lattice constant often brings higher polarization intensity; the other application direction is based on the structural characteristics corresponding to its own composite material. Under suitable deposition conditions, it can grow a nano-scale columnar structure. This unique 1-3 dimensional structure plays a vital role in controlling the movement of oxygen vacancies. Therefore, the 1-3 dimensional ferroelectric film designed by the present invention is more suitable for multi-level storage and AI smart chips.

[0020] The present invention relates to the application fields of ferroelectric memristor devices and neural networks in electronic devices and material sciences, and in particular, in terms of structure, the vertical self-assembled nanocomposite architecture (VSNs) silicon-based integrated device of epitaxial growth controls the movement of oxygen defects through interface channels, effectively improves the residual polarization intensity and switching voltage stability of the device, and achieves low power consumption and high durability; the introduction of Nd2O3 improves the lattice matching, forms a vertical self-assembled nanocomposite structure, and further enhances the ferroelectric performance and stability of the device. These are conducive to the multi-level storage of ferroelectric thin film devices, and can be applied in brain-like neural computing and high-performance storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the structure of the ferroelectric memristor in the present invention.

[0022] Figure 2 This is a TEM image of the BSTN ferroelectric active layer prepared in Example 1 of the present invention.

[0023] Figure 3 It is a TEM image of the thickness of each layer of the sample prepared in Example 1 of the present invention.

[0024] Figure 4 It is a comparison diagram of the hysteresis loops of the sample in Example 1 of the present invention and the sample in Comparative Example 1.

[0025] Figure 5 It is the XRD diagram of the samples in Example 1, Comparative Example 2 and Comparative Example 3 of the present invention.

[0026] Figure 6 It is the XRD diagram of the samples in Example 1, Comparative Example 4, Comparative Example 5 and Comparative Example 6 of the present invention.

[0027] Figure 7 It is a schematic diagram of a 16×16 cross-bar memristor array based on BSTN material proposed in the present invention.

[0028] Figure 8 It is a schematic diagram of the circuit structure of the present invention in which a memristor based on BSTN material is applied to a multi-factor security encryption system.

[0029] Fig. 9 These are the response current image results of different devices being triggered ten times individually under nine conductance states of the present invention.

[0030] Fig.10 It is a schematic diagram of a password signal sequence input by a user in a specific embodiment of the present invention.

[0031] Fig.11 It is the accuracy and loss function result after neural network operation in the specific embodiment of the present invention. DETAILED DESCRIPTION

[0032] The present invention proposes a silicon-based integrated ferroelectric memristor based on vertical self-assembled nanocomposite structures (VSNs), the structure of which includes, from bottom to top, a substrate, a buffer layer, a bottom electrode layer, a ferroelectric active layer and a top electrode layer. Figure 1 As shown, in the present invention, the substrate is a Si substrate, more specifically, the substrate is a P-type Si substrate. The buffer layer is a strontium titanate material (SrTiO3, abbreviated as STO). The bottom electrode layer is a lanthanum strontium manganese oxide material (abbreviated as LSMO), more specifically, the bottom electrode layer is La 0.67 Sr 0.33 MnO3. The top electrode layer is a Pd electrode. The ferroelectric active layer is a vertical self-assembled nanocomposite structure (VSNs) formed by barium strontium titanate (BSTO) and neodymium oxide (Nd2O3, NdO) in a 1:1 molar ratio. Barium strontium titanate is specifically Ba 0.6 Sr 0.4 TiO3, ferroelectric active layer (Ba 0.6 Sr 0.4 TiO3) 0.5 (Nd2O3) 0.5 Abbreviated as BSTN, BSTN is deposited on La in a highly directional growth manner. 0.67 Sr 0.33 On MnO3 / SrTiO3 / P-Si substrate.

[0033] The present invention effectively improves the residual polarization strength and switching voltage stability of the device by controlling the movement of oxygen vacancies, achieving low power consumption and high durability. The introduction of Nd2O3 improves lattice matching, forms a vertical self-assembled nanocomposite structure, and further enhances the ferroelectric performance and stability of the device.

[0034] Example 1

[0035] The present invention uses pulsed laser deposition technology to prepare a ferroelectric memristor based on VSNs-BSTN. The following is a detailed preparation process:

[0036] (1) Substrate pretreatment. The substrate used is a boron-doped silicon substrate (thickness 400±10μm). During pretreatment, the silicon oxide on the substrate needs to be corroded. The silicon oxide layer on the silicon surface is usually a disordered amorphous layer, which will hinder the orderly crystal growth of the film and cause the quality of the film to deteriorate. After removing the silicon oxide layer, the single crystal surface of silicon is exposed, which is conducive to the lattice matching of the film and the substrate, thereby improving the crystal quality of the film. First, soak the substrate in an acetone solution at room temperature to remove the organic matter on the substrate, and then soak it again in anhydrous ethanol for 5 minutes to remove the residual acetone solution; after taking out, soak the substrate in a 3:1 deionized water and hydrofluoric acid diluted solution for 1.5 minutes, at this time, the silicon oxide will be corroded, and finally use deionized water for ultrasonic cleaning for 10 minutes. After taking out, blow dry with a high-purity nitrogen gun.

[0037] (2) After the substrate is pretreated, the substrate is pasted on the sample stage of the pulse laser deposition system with quick-drying silver glue. The purpose of using silver glue is to make the sample contact more closely and enhance the uniformity of thermal conductivity. Then put it into the vacuum chamber, close the chamber door of the vacuum chamber, and open the mechanical pump, the side pump valve, and the quartz gauge vacuum gauge to perform preliminary vacuum treatment. After the pressure of the vacuum chamber is reduced to below 5Pa, close the side pump valve and perform higher vacuum environment treatment. First, open the fore valve in the system to ensure the normal exhaust of the molecular pump, then start the molecular pump, check the working status of the molecular pump, and whether the air-cooled radiator is working. After it is correct, open the gate valve to a small angle after the molecular pump speed is greater than 100Hz, open the ionization vacuum gauge and check whether the pressure drops normally. After it is correct, open all the gate valves so that the cavity can be processed by the molecular pump for a faster high vacuum.

[0038] After using a molecular pump to reduce the pressure of the vacuum chamber to 9E-5, turn off the ionization vacuum gauge to prevent temperature rise and oxygen damage to the ionization vacuum gauge. Turn on the heating unit of the pulsed laser deposition system, set the heating step for the sample stage, and control the heating rate between the current temperature and 500°C to 20°C / min. After reaching the temperature, reduce the rate to 10°C / min until 750°C. After the temperature stabilizes, introduce oxygen and turn on the excimer pulse laser for preheating and charging. Close the sample baffle and adjust the laser light path to ensure that the laser is accurately irradiated on the corresponding material. After the laser is charged for 470s, prepare to pre-sputter the material to be deposited to remove surface impurities. When pre-sputtering the pre-sintered target (STO, LSMO, BSTN), the laser energy is controlled at 1.3J / cm 2, frequency is 5 Hz, using KrF excimer laser (λ=248nm). After pre-sputtering 100 pulses respectively, coordinately adjust the gate valve angle and oxygen flow rate (generally the gate valve angle is 8° and the oxygen flow rate is 25 sccm), control the pressure at 1 Pa, open the sample stage baffle and prepare for formal growth.

[0039] The first material to be grown is strontium titanate (STO). This is to take advantage of the better bonding effect between strontium titanate and the silicon substrate to prevent the lattice mismatch problem caused by direct growth. STO is sputtered for a total of 4500 pulses, followed by 10 minutes of deposition. This is to ensure the quality of material growth and play a certain shaping role. After STO is deposited, the bottom electrode material lanthanum strontium manganese oxide (LSMO) is deposited. The laser parameters remain unchanged, and the oxygen atmosphere is adjusted to 26Pa (the gate valve angle is 7.2°, and the oxygen flow rate is 20sccm). After the sputtering pulses are changed to 9000 times, LSMO begins to grow and is deposited for 10 minutes. Finally, the temperature is raised to 775° (the comparison conditions are 700℃ and 625℃), and the pressure is changed to 1Pa again (the parameters are the same as when STO is deposited, and the comparison conditions are 3Pa, 5Pa, and 7Pa), and 9000 pulses of strontium titanate barium composite neodymium oxide target (Ba 0.6 Sr 0.4 TiO3 and Nd2O3 are mixed in a 1:1 molar ratio and sintered in pure oxygen at 1200℃), and the deposition is still 10 minutes to form the final columnar VSNs ferroelectric active layer. The above completes all the pulsed laser deposition growth. Finally, the oxygen atmosphere is raised to 3000Pa, the plug valve is closed, and the molecular pump is no longer allowed to evacuate the vacuum chamber, and the molecular pump is turned off to reduce the speed. Change the oxygen flow rate to 100sccm, perform oxygen cleaning, observe the reading of the quartz gauge, and when the pressure environment reaches 3000, first turn off the switch on the oxygen cylinder and change the oxygen flow rate to 0. After the actual flow rate drops, close the inflation valve and start cooling, and the speed is controlled to 5℃ / min. This is to provide sufficient oxygen atmosphere to prevent excessive oxygen defects. After the molecular pump speed drops to 0, close the front valve and the mechanical pump in turn, and wait for the cooling to end.

[0040] The structure of the ferroelectric active layer of the prepared strontium barium titanate composite neodymium oxide is as follows Figure 2 shown. Figure 2 The figure shows that barium strontium titanate (shown as B in the figure) and neodymium oxide (shown as N in the figure) grow vertically in an alternating arrangement, indicating the successful growth of a vertical interface self-assembled structure.

[0041] (3) After the pulse laser deposition system has cooled down, the sample is taken out and placed on the sample stage of the vacuum chamber of the magnetron sputtering. The mask is also fixed and covered with silver glue. The vacuum environment is prepared. After the pressure is reduced to 1.8E-4, the DC source of the magnetron sputtering is turned on, and the argon bottle is opened to allow argon (25sccm) to enter the vacuum chamber. The throttle valve of the magnetron sputtering system is opened to control the argon environment to 1.2Pa. The power of the DC source is set to 10W, the sample baffle is closed and the DC switch is turned on to observe whether the glow discharge phenomenon is normally generated in the chamber. After the glow on the metal Pd target is stable, the baffle is opened for formal material growth. The Pd material will be deposited on the top of the sample through the holes on the mask, forming a complete and usable electrical test structure. The growth time is controlled at 12min. After stabilization for 10min, the magnetron sputtering system is turned off and the mask and sample are taken out.

[0042] The thickness of each film layer deposited in this embodiment is: 37nm of strontium titanate buffer layer; 23nm of lanthanum strontium manganese oxide bottom electrode layer; 33nm of BSTN ferroelectric active layer; and 20nm of Pd top electrode layer. The transmission electron microscopy (TEM) results of the sample are shown in Figure 1. Figure 3 As shown, Figure 3 In the figure, there is still 3nm of SiO on the Si substrate. x (It should be the residue after corrosion).

[0043] The present invention forms a vertical self-assembled nanostructure through pulsed laser deposition, which enhances the control ability of oxygen vacancies at the interface. The structure realizes the restriction of oxygen vacancy movement at the vertical interface and improves the performance of the ferroelectric memristor. Moreover, it is also compatible with Si-based, solving the problem that most existing ferroelectric materials cannot achieve Si-based integration.

[0044] Comparative Example 1

[0045] Compared with Example 1, the ferroelectric active layer prepared in this comparative example is pure barium strontium titanate (BSTO for short), that is, the target material used in preparing the ferroelectric active layer is Ba 0.6 Sr 0.4 TiO3 target material, and other aspects are the same as those in Example 1.

[0046] Hysteresis loop measurements were performed on the samples in Example 1 and Comparative Example 1. The results are as follows: Figure 4 As shown. Figure 4 It can be seen that the residual polarization intensity of the device with VSNs structure prepared in Example 1 is much higher than that of the pure layered BSTO in Comparative Example 1, which is six times that of Comparative Example 1, confirming the superior performance of the vertical self-assembled nanocomposite structure of the present invention. The BSTN memristor in the present invention exhibits a wide ferroelectric domain adjustable range (0-180°), a residual polarization intensity (~21.04μC / cm 2) is significantly higher than that of barium strontium titanate (BSTO) ferroelectric material (~3.35μC / cm 2 ). In addition, the BSTN memristor 9 The performance remains stable after switching cycles, demonstrating ultra-high durability; the power consumption of the device in a single switching event is as low as 0.57pJ, which has extremely low power consumption.

[0047] Obviously, the present invention can effectively control the random movement of oxygen vacancies, confine them to the interface of vertical self-assembly, increase the adjustable range of ferroelectric domains, and utilize the additional stress provided by neodymium oxide to increase the out-of-plane lattice constant, thereby increasing the residual polarization strength and obtaining stable switching voltage.

[0048] Comparative Example 2

[0049] Compared with Example 1, the deposition temperature of the BSTN ferroelectric active layer in this comparative example is changed to 700° C., and the other conditions are the same as those in Example 1.

[0050] Comparative Example 3

[0051] Compared with Example 1, the deposition temperature of the BSTN ferroelectric active layer in this comparative example is changed to 625° C., and the other conditions are the same as those in Example 1.

[0052] XRD test was performed on the samples in Example 1, Comparative Example 2 and Comparative Example 3, and the results were as follows: Figure 5 As shown. Figure 5 It can be seen that for devices formed at different deposition temperatures, the relative intensity of the BSTO peak increases and the half-peak width decreases as the temperature increases, indicating that the crystallinity of BSTO is improved at higher temperatures. However, Nd2O3 is basically not affected by the deposition temperature.

[0053] Comparative Example 4

[0054] Compared with Example 1, the oxygen pressure of the BSTN ferroelectric active layer in this comparative example is changed to 5 Pa, and the rest is the same as Example 1.

[0055] Comparative Example 5

[0056] Compared with Example 1, the oxygen pressure of the BSTN ferroelectric active layer in this comparative example is changed to 9 Pa, and the rest is the same as Example 1.

[0057] Comparative Example 6

[0058] Compared with Example 1, the oxygen pressure of the BSTN ferroelectric active layer in this comparative example is changed to 13 Pa, and the rest is the same as Example 1.

[0059] XRD test was performed on the samples in Example 1 and Comparative Examples 4 to 6. The results are as follows: Figure 6 As shown. Figure 6 It can be seen that the half-maximum full width of the BSTO peak is the shortest when the oxygen pressure is 1Pa, and Nd2O3 is not greatly affected by changes in oxygen pressure. Therefore, the BSTN crystal quality is best when the oxygen pressure is 1Pa.

[0060] The memristor provided by the present invention has huge application potential and can be widely used in brain-like computing, AI chips, and multi-factor security encryption systems.

[0061] The use of the memristor of the present invention is described in detail below by taking a multi-factor security encryption system as an example.

[0062] First, a 16×16 cross-bar memristor array based on BSTN material was prepared using a simple mask process and no post-annealing process was required. A 16×16 cross-bar memristor array was grown on a 10×10 mm P-Si substrate, where the BSTN film was sandwiched between the bottom electrode LSMO (BEs) and the top electrode Pd (TEs) metal, as shown in the schematic diagram. Figure 7 As shown. BEs are defined as bit lines, TEs are defined as source lines, and voltage terminals and ground terminals are applied. This crossbar array is suitable for large-scale integration, which also ensures the uniformity of performance to a certain extent. The fabricated crossbar memristor can be clearly observed by scanning electron microscopy (SEM) to have clear boundaries and high quality crossbar memristor arrays at different magnifications.

[0063] The verification of the interactive application scenario of the multi-factor security encryption system is carried out by intelligent identification and verification using 3×3 devices in the BSTN memristor cross array and the built 3×3 password acquisition array and identification module. Compared with the ordinary single password system, the multi-factor security encryption system based on the BSTN memristor in the present invention is composed of three factors: different password bit sequences (random, resettable password key positions), different user passwords and specific user accounts, and has additional multiple password leakage protection. Therefore, the system has a more powerful password protection capability and improves the security performance in the application system. The BSTN memristor cross array is combined with artificial intelligence data acquisition and other technologies to enable the designed system to accurately perceive the password signals input by different users.

[0064] like Figure 8As shown, by transmitting the collected results of the hand pressing signal to the multiplexer, the signal results will act on the nine devices in the BSTN ferroelectric memristor in the form of applying a fixed read voltage pulse. The conductivity value of the device can also be regulated by pulse stimulation, and the device can be regulated to correspond to nine different resistance states, so that the output current of each device will have different response sizes. The current response is also output in the form of a pulse signal. After being transmitted to the microcontroller unit (MCU) through the analog-to-digital converter, different password signals are converted into corresponding mechanical signals, and the mechanical signals are finally processed by the microcontroller unit and transmitted to the host computer for information recognition.

[0065] In addition to the different login passwords of experimental users, the password sequence can also be changed by changing the conductance state position of the device in the memristor array. Fig. 9 Shown in the figure are the response current images of different devices triggered ten times separately under nine conductivity states. It can be seen that due to the relatively consistent pressing habits of the users, there are certain deviations, and the current response of each triggering is relatively stable and does not fluctuate too much.

[0066] Fig.10 It shows the password signal sequence input by a specific user. The password information consists of a six-digit digital signal, and the total signal acquisition time of the trigger input is 20 seconds. Then, the multi-value storage capacity and long-term synaptic plasticity of the BSTN device are used to transmit the user password data to the neuromorphic computing, and a three-layer artificial neural network is constructed with the help of Python software to identify nine different response current signals. There are nine types of response current data in the test database, and each type of data has 5,000 sets of values ​​with different pressures. The input layer of the neural network has a total of 500 neurons corresponding to each current response information, the hidden layer has 180 neurons, and the output layer has 9 neurons corresponding to the 9 response currents in the database. The weight matrix between adjacent layers is stored in the device-based cross switch, which is regarded as part of the "neural core" and is used to perform vector-matrix multiplication and outer product update operations. The recognition accuracy in the neural network is the most important feature. After the initial iteration, the recognition accuracy is 12.96%. After 500 iterations, the accuracy reaches 97.58%. The loss layer also gradually decreases with the increase of recognition accuracy, such as Fig.11 As shown (blue is recognition accuracy, purple is loss function). And in the last part of training, the relationship between training and recognition accuracy also tends to fluctuate and converge within a relatively small range (as Epoch increases from 300 to 500, recognition accuracy increases from 95.27% to 97.58%). Finally, the login and authentication process of a specific user is placed in one video, and the verification process when using the wrong account is placed in another video.

[0067] All of the above outstanding test results prove that BSTN ferroelectric memristors have great application potential in the protection of secure cryptographic systems, and user information has been effectively authenticated and identified.

Claims

1. A vertical self-assembled nanocomposite structure ferroelectric memristor, characterized in that: The structure includes, from bottom to top, a Si substrate, a strontium titanate buffer layer, a lanthanum strontium manganese oxide bottom electrode layer, a composite structure ferroelectric active layer formed by strontium barium titanate and neodymium oxide, and a top electrode layer; in the composite structure ferroelectric active layer, strontium barium titanate and neodymium oxide are arranged alternately to form a vertical self-assembled columnar nanocomposite structure.

2. The vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 1, characterized in that: The molar ratio of barium strontium titanate to neodymium oxide in the composite structure ferroelectric active layer is 1:1, and barium strontium titanate is specifically Ba 0.6 Sr 0.4 TiO3.

3. The vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 1, characterized in that: The material of the lanthanum strontium manganese oxide bottom electrode layer is La 0.67 Sr 0.33 MnO3.

4. The vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 1, characterized in that: The Si substrate is a P-type Si substrate.

5. A method for preparing the vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 1, characterized in that: The steps include: (1) Pretreatment of Si substrate; (2) using pulsed laser deposition technology to sequentially deposit a strontium titanate buffer layer, a lanthanum strontium manganese oxide bottom electrode layer, and a composite structure ferroelectric active layer on a Si substrate; the target material used for depositing the composite structure ferroelectric active layer is a composite target material formed by sintering barium strontium titanate and neodymium oxide; (3) A top electrode layer is deposited on the ferroelectric active layer of the composite structure using magnetron sputtering deposition technology.

6. The method for preparing a vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 5, characterized in that: The barium strontium titanate is Ba 0.6 Sr 0.4 TiO3, and the neodymium oxide is Nd2O3.

7. The method for preparing a vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 6, characterized in that: The specific preparation process of the composite target is: Ba 0.6 Sr 0.4 TiO3 and Nd2O3 were mixed in a 1:1 molar ratio and prepared by pure oxygen sintering at 1200°C.

8. The method for preparing a vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 5, characterized in that: The temperature when depositing the composite structure ferroelectric active layer is 600°C to 800°C.

9. The method for preparing a vertical self-assembled nanocomposite structure ferroelectric memristor according to claim 5, characterized in that: The oxygen pressure is 1Pa-13Pa when depositing the composite structure ferroelectric active layer.

10. Application of the vertical self-assembled nanocomposite ferroelectric memristor according to any one of claims 1 to 4 and the vertical self-assembled nanocomposite ferroelectric memristor prepared by the method according to any one of claims 5 to 9 in a multi-factor security encryption system.

Images