A bio-memristor for multi-level storage and its preparation method

Abstract

The present invention relates to a bio-memristor for multi-level storage and a preparation method thereof. The bio-memristor is sequentially composed of an electrode layer, a memristive functional layer, and a conductive layer. The memristive functional layer is sequentially composed of a GO layer, an SF layer (with a crystallinity of 35-60%), and a GO layer. The thickness of each GO layer is 15-25 nm, the thickness of the SF layer is 30-80 nm, and the thickness of the electrode layer is 200-300 nm. The electrode layer is an Al electrode layer or an Ag electrode layer. The preparation method comprises the following steps: spin-coating a GO dispersion on the upper surface of the conductive layer to form a GO layer, spin-coating an SF solution, then soaking the solution in an ethanol-water solution and drying it to form an SF layer, and then spin-coating a GO dispersion on the upper surface of the SF layer to form a GO layer, and then depositing the electrode layer. The product of the present invention can achieve the device's transition between dual-resistance switching and triple-resistance switching by regulating the limiting current. The preparation method of the present invention is simple and easy to operate.

Description

Technical Field

[0001] The present invention belongs to the technical field of composite materials and relates to a bio-memristor for multi-level storage and a preparation method thereof. Background Art

[0002] Neuromorphic computing is currently gaining intensive research in a range of emerging fields, including artificial intelligence and high-density storage. This biomimetic computing approach can simultaneously perform data storage and computation, overcoming several bottlenecks of the traditional von Neumann computer architecture. Memristors offer significant advantages in building neuromorphic computing devices due to their simple structure, ease of integration, and structural similarity to the synaptic system in the human brain. During the resistance switching process of a memristor, the high resistance state (HRS) and low resistance state (LRS) can serve as the "1" and "0" of binary code for information storage. Currently, most multi-level storage memristors achieve multi-resistance switching by adjusting the limiting current. Different limiting currents enable the device to achieve different LRSs, which can be used as different codes for information storage, thus endowing the device with multi-level storage capabilities. To date, materials that have achieved multi-level storage capabilities have primarily focused on organic and inorganic materials. However, these materials suffer from issues such as difficulty in degradation and poor biocompatibility, limiting their application in bioelectronics and implantable devices. Biomaterial-based memristors offer inherent advantages in neuromorphic computing due to their biocompatibility, biodegradability, and sustainability. Silk fibroin (SF), derived from natural silk, offers excellent mechanical properties, lightweight properties, and affordability, making it a promising biomemristor construction material for practical applications. However, few biomaterials can currently produce memristors with multiple resistance state transitions within a single cycle.

[0003] Reference 1 (Adv. Funct. Mater., 2012(22), 4493-4499) first demonstrated that the SF film exhibited non-volatile resistance switching behavior in a sandwich device based on ITO and aluminum, with a high-to-low resistance ratio of 10 and a data retention time of 10. 3 s, which proposed that the carrier capture / removal caused by the oxidation and reduction process of SF is the main cause of the resistance switching memory effect.

[0004] In the SF-based memristor device prepared in Reference 2 (Adv. Funct. Mater., 2015(25), 3825-3831), two types of resistance switching behaviors can be achieved by adjusting the limiting current. By applying different scan voltages to the device, the device can switch back and forth between LRS and HRS. Based on this resistance switching method, two application modes, charge information storage and resistance switching, can be realized, and a high resistance switching ratio (about 10) of random access memory is shown. 7 ) and a long retention time (>4500s). The authors of the paper then prepared an ultra-lightweight SF-based memristor and a transparent transient bio-memristor, both of which had good performance, indicating that SF-based memristors have broad application prospects.

[0005] While the aforementioned pure SF-based memristors have made some progress, their memristive performance still has significant room for improvement. For example, the aforementioned SF-based memristors can only switch once, from HRS to LRS. With the advancement of materials processing and nanotechnology, research on the functionalization of SF materials has become increasingly in-depth. Through different levels of functionalization, researchers are enabling SF to maintain its inherent advantages while also incorporating other excellent properties.

[0006] Reference 3 (Nanotechnology, 2013(24), 345202) prepared a SF composite memristor by doping gold nanoparticles into SF. The resulting memristor has bipolar behavior and a resistance switching ratio greater than 10. 6 The resistance switching mechanism is the formation and breaking of conductive filaments, but the device has poor cycling stability and can only be switched about 10 times.

[0007] Reference 4 (Small, 2017 (13), 1702390) used wool keratin (WK) and gold nanoclusters (AuNCs) to mesoscopically functionalize silk fibroin and prepare a biocompatible and partially degradable WK@AuNCs-SF biomemristor. It was used to simulate the working mechanism of neural synapses using the conductance changes of potassium and sodium ion channels to achieve information transmission. When a pulse signal is applied, the WK@AuNCs-SF biomemristor transmits information through Ag. + The migration of nanostructured nanostructured nanostructures changes its conductivity, which is similar to the working mechanism of neural synapses and can be used as an active medium for building biological synaptic devices. Compared with pure SF-based memristors, WK@AuNCs-SF memristors have better overall performance and the device's cycle stability has been improved to about 100 times, but its resistance switching ratio is only 10 2 , there is still room for improvement.

[0008] Reference 5 (Adv. Funct. Mater., 2019, 1904777) used silver nanoclusters (AgNCs) and bovine serum albumin (BSA) to modify SF, significantly improving the memristive performance of SF. The mechanism is that AgNCs@BSA acts as an electron potential well, completely changing the transmission behavior of charged particles in the SF film. The resulting silk composite memristor has a rewritable endurance of 100 times and a resistance switching ratio of 10. 3 , and showed unique synaptic characteristics and synaptic learning ability. However, in order to expand its application in fields such as information storage, the number of erasable and rewritable times and the on-off ratio still need to be further improved.

[0009] Reference 6 (Organic Electronics, 2017, 276-284) used the method of compounding cadmium selenide (CdSe) quantum dots on the surface of SF film to regulate the interface structure of SF and prepare a memristor with multi-level resistance transition characteristics. By analyzing the energy bands of CdSe quantum dots and SF, it was found that there are many hole capture centers in the interface energy gap between the two. The space charge limited current mechanism followed in the charge transfer process is the main reason for the device to show multiple resistance states. In the forward scan, the device clearly shows three different resistance state switches from low to high at 0.17V, 0.66V and 0.96V, and all three resistance states can be maintained stably for a long time. In addition, compared with other memristors, this device has low power consumption, good repeatability, low operating voltage, and a resistance switching ratio of more than 10 4 This work once again confirms that functionalization is a feasible strategy to improve the performance of SF-based memristors, but the memristive performance still needs to be improved.

[0010] Reference 7 (Advanced Electronic Materials, 2022, 2101139) uses the method of doping silver nitrate (AgNO3) in SF film to prepare a volatile memristor that can be operated at low voltage and is highly uniform. Unlike traditional non-volatile SF-based memristors, volatile SF-based memristors use tyrosine (Tyr) residues in SF to reduce silver ions in situ to uniformly dispersed silver nanoparticles (AgNPs), providing a stable current path for persistent resistance switching operations. The SF-based memristor has good device uniformity and reliable stability (>100 cycles), and a low threshold voltage of 0.17V. Most importantly, the device has multi-level storage performance and can be switched by setting 5 different I cc Five different LRSs were obtained, which gave the device better storage, artificial synapse-like and neuron computing functions, providing new candidate materials for future bio-integrated neuromorphic electronics.

[0011] Reference 8 (Nano materials, 2019 (9), 518) successfully prepared a three-layer ITO / GO@PBD / PMMA / GO@PBD / Ni memristive device by sandwiching polymethyl methacrylate (PMMA) between a novel double-layer nanocomposite film composed of 2-(4-tert-butylphenyl)-5-(4-biphenyl)-1,3,4-oxadiazole (PBD)-doped graphene oxide (GO). Compared with the single-layer ITO / GO@PBD / Ni device, the device exhibits ternary non-volatile resistive switching behavior, with three stable resistance states in the same switching cycle. That is, two sudden resistance state transitions occur when the applied voltage is from 0V to -6V. When the voltage reaches -0.9V, the device transitions from HRS to intermediate resistance state (IRS), and then transitions from IRS to LRS when the voltage increases to -2V. When the applied voltage was subsequently switched to a forward voltage, a transition from LRS to IRS was observed at 3.8 V. Then, when the voltage was increased to 6 V, the device transitioned from IRS to HRS. Therefore, by designing a device structure consisting of an insulating polymer PMMA sandwiched between double graphene oxide embedded in a PBD nanocomposite layer, the storage density of the memristor was greatly improved, showing great potential for development in fields such as logic computing and synaptic function simulation.

[0012] Although the ITO / GO@PBD / PMMA / GO@PBD / Ni device in Reference 8 can realize three resistance states in one switching cycle, it is not possible to achieve the three resistance states by adjusting I cc By adjusting the I cc Changing the resistance value of a device often requires multiple "write" and "read" cycles. This results in a device that can store less data in a single cycle, has a more limited function, and consumes more power to switch resistance states. Ideally, multiple resistance states can be switched within a single resistance switching cycle. Devices with this characteristic are crucial for achieving higher efficiency and higher density storage.

[0013] Therefore, it is of great significance to study a bio-memristor for multi-level storage and its preparation method to achieve multi-level storage within the same cycle in the memristor. At the same time, it is also possible to realize the conversion between dual-resistance state and multi-resistance state by adjusting test parameters (such as limiting current, etc.) to expand the application field. Summary of the Invention

[0014] The purpose of the present invention is to solve the above-mentioned problems and provide a bio-memristor for multi-level storage and a preparation method thereof.

[0015] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0016] A bio-memristor for multi-level storage, which is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence, and the crystallinity of the SF layer is 35-60%;

[0017] The thickness of each GO layer is 15 to 25 nm, the thickness of the SF layer is 30 to 80 nm, and the thickness of the electrode layer is 200 to 300 nm; the electrode layer is an Al electrode layer or an Ag electrode layer; the thickness of the electrode layer will affect the resistance of the entire device, the starting voltage of the device, the heat generated by the device during operation, etc.; if the electrode layer is too thick, the device resistance becomes larger, the starting voltage becomes higher, and the heat generated is large, thereby affecting the performance of the device; if the electrode layer is too thin, the electrode is easily oxidized and its conductivity is affected.

[0018] As the preferred technical solution:

[0019] In the bio-memristor for multi-level storage as described above, the thickness of the conductive layer is 0.6 to 1.0 mm.

[0020] In the bio-memristor for multi-level storage as described above, the conductive layer is an ITO film or a PET conductive film.

[0021] A bio-memristor for multi-level storage as described in any of the above items can realize the mutual conversion between the dual-resistance switching and the three-resistance switching of the device by regulating the limiting current. cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc When the current is >0.01A, the bio-memristor used for multi-level storage exhibits three-state resistance switching behavior. Existing memristors are unable to achieve mutual conversion between two-state resistance switching behavior and three-state resistance switching behavior by adjusting the limiting current value. This invention is the first to prepare a memristor with such characteristics, filling a gap in the existing technology.

[0022] When the regulating current limit value I cc When the A is >0.01, the bio-memristor used for multi-level storage exists in three resistance states: high, medium, and low within one switching cycle;

[0023] Bio-memristors with a resistance-on / off ratio greater than 10 for multi-level storage 4 , data retention time is greater than 10 3 s, and still has good stability after 100 cycles.

[0024] The present invention also provides a method for preparing a bio-memristor for multi-level storage as described in any of the above items, comprising: spin-coating a GO dispersion on the upper surface of the conductive layer to form a GO layer, spin-coating an SF solution, and then immersing and drying the SF layer in an ethanol aqueous solution with a volume concentration of 75-90%. Then, spin-coating a GO dispersion on the upper surface of the SF layer to form a GO layer, and depositing an electrode layer to obtain a bio-memristor for multi-level storage.

[0025] The main components of SF solution are water, protein, amino acids, and some metal ion impurities. Amino acid molecules contain functional groups related to oxygen ions, such as carboxyl (COOH), which can generate oxygen ions and oxygen vacancies. Furthermore, as ampholytes, amino acids also have the ability to capture and release protons. When spin-coated into a film, the SF molecules in the film are essentially random coils. Post-treatment with aqueous ethanol effectively promotes SF crystallization, resulting in a denser arrangement of SF molecules, thereby reducing leakage paths and ineffective defects and improving carrier mobility. Furthermore, the addition of aqueous ethanol promotes the regular formation of conductive filaments, improving device stability and durability to a certain extent. Furthermore, the dense arrangement of SF molecules reduces the free volume of the molecules at the microscale and the volume of the SF film at the macroscale, which helps reduce device power consumption. The ethanol concentration and immersion time affect the crystallinity of the SF film.

[0026] While there have been a few reports on SF-based multi-level storage memristors, most SF-based memristors have relatively small resistance switches, poor cyclic stability, and an inability to switch between multiple resistance states within a single switching cycle, limiting their application in multi-level data storage, integrated sensing, storage, and computing, and wearable flexible devices. Research has shown that improving the multi-level storage performance of SF-based memristors requires material modification, composite doping with other materials, or designing functional layer structures. Graphene-based materials have excellent resistance switching properties, are non-reactive with SF, and are easy to process. Multi-layered memristor functional layers can be fabricated without destroying the SF structure, effectively improving the performance of SF-based memristors.

[0027] As the preferred technical solution:

[0028] As described above, the concentration range of all GO dispersions is 0.5-1.0 mg / ml; each spin coating of the GO dispersion is carried out in three stages, with the spin coating speed increasing from the first stage to the third stage; because the GO dispersion has poor hydrophilicity and is difficult to adhere well to the conductive layer and SF layer, if a higher spin coating speed is used in the first stage, the GO dispersion will be thrown away, resulting in the loss of the GO film in the functional layer, affecting the resistive performance of the device; the higher spin coating speed is used in the third stage to dry the GO layer film, so that the device remains relatively dry after the spin coating is completed.

[0029] As described above, each time the GO dispersion is spin-coated, the spin-coating speed in the first stage is 200-1000 rpm, and the spin-coating time is 10-30 s; the spin-coating speed in the second stage is 1000-1800 rpm, and the spin-coating time is 10-30 s; the spin-coating speed in the third stage is 2000-3000 rpm, and the spin-coating time is 5-20 s; the spin-coating speed and spin-coating time of the spin coater will affect the uniformity of the functional layer, and thus affect the stability, durability and consistency of the entire device. If the memristive functional layer is too thin, the resulting device is easily broken down during operation; if the memristive functional layer is too thick, the device may require a large voltage to be converted to LRS, which consumes a lot of power and may even lose the resistive state switching function.

[0030] In the method described above, the concentration of the SF solution is 2.0-3.0 wt%; when spin-coating the SF solution, the spin-coating speed is 1200-3000 rpm, and the spin-coating time is 30-60 s. Because SF has good hydrophilicity and adhesion, it can adhere well to the surface of the GO membrane, and there is no need to worry that too high a spin-coating speed will cause all the SF solution to be thrown away. At the same time, the thickness of the SF layer must not be too thick, so a longer spin-coating time is required.

[0031] As described above, the soaking time is 60 to 90 minutes, and the drying time is 20 to 30 minutes. Within the soaking time range of the ethanol selected in the present invention, the crystallinity of the obtained SF membrane is relatively high. If the soaking time is further extended, the crystallinity of the SF membrane does not change much.

[0032] In the above-mentioned method, the electrode layer is deposited by thermal evaporation, electron beam evaporation or magnetron sputtering.

[0033] The SF solution is prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% NaCO solution, washing with deionized water, dissolving the SF solution in a lithium bromide (LiBr) solution, and then centrifuging, filtering, dialysis, and concentrating the solution to 2.0-3.0 wt% at 4°C to obtain the SF solution. The NaCO solution has a 0.5 wt% NaCO concentration, and the relationship between the NaCO concentration and the degumming rate is nonlinear. When the NaCO concentration is within the 0.2-0.5 wt% range, the SF degumming rate increases with increasing NaCO concentration. After that, the degumming rate remains essentially unchanged despite continued increases in NaCO concentration. Therefore, a 0.5 wt% NaCO solution is selected based on comprehensive considerations.

[0034] The principles of the present invention are as follows:

[0035] Compared to inorganic materials, memristor devices based on organic materials have higher power consumption and poorer stability and durability than their inorganic counterparts. Natural biomaterials, while offering excellent biocompatibility, biodegradability, and flexibility, generally have relatively large molecular weights. Their uncontrollable thermochemical reactions and disordered chain arrangement result in poor device stability and durability, and low switching efficiency. As previously mentioned, by doping biomaterials with functional molecules, nanomaterials, quantum dots, or metal ions without destroying their original properties, they can impart specialized functions and effectively control biomemristors.

[0036] The memristor prepared by the bio-memristor method for multi-level storage of the present invention has the advantages of good stability and durability, long data retention time, etc., and also has a multi-level storage function, which not only realizes non-volatile dual-state resistance switching behavior, but also realizes three-state resistance switching behavior in the same cycle, and can also be adjusted by I cc The device is switched between dual-state resistance switching and triple-state resistance switching for the following reasons:

[0037] (1) The GO / SF / GO film contains a large number of oxygen-containing functional groups. The introduced GO does not react with SF. Therefore, SF adds a large number of new oxygen-containing functional groups on the basis of ensuring the number of inherent functional groups, which will be conducive to the generation of more oxygen ions and oxygen vacancies, thereby improving the conductivity of the device;

[0038] (2) The main components of SF solution are water, protein, amino acid and some metal ion impurities. An amino acid molecule contains functional groups related to oxygen ions, such as carboxyl (COOH). In addition, amino acids, as ampholytes, also have the ability to capture and release protons. In order to confirm the composition of the conductive filaments in the device, the present invention studied the temperature dependence of the resistance in the GO / SF / GO memristor and found that the resistance under HRS and LRS conditions decreased with increasing temperature, indicating that the conductive filaments have semiconductor properties in both states, forming conductive filaments dominated by oxygen vacancies rather than metal filaments.

[0039] (3) The GO in the suspension used to prepare the memristor functional layer is a micron-sized layered structure (about 1 μm in diameter). The smaller size is conducive to the uniform dispersion of GO in the solution, making the prepared functional layer film more uniform and smooth, and improving the performance stability of the functional layer; in addition, the layered structure can also increase the actual contact area with the SF layer, which is conducive to the formation of more oxygen ions and oxygen vacancies, improving the ion transport performance of the functional layer, and thus giving the device better electrical properties, such as good stability and durability, and long data retention time;

[0040] (4)I ccis an important factor affecting the operation of the memristor. Its size change may affect the resistance state transition and even the memristive mechanism of the memristor during operation. For the GO / SF / GO memristor device, when I cc When the voltage is >0.01A, the SET process (from HRS to LRS) occurs in the negative voltage scan range. When a negative voltage is applied, the device changes from HRS to LRS. When a positive voltage is applied, the device returns from LRS to HRS. It is worth noting that in the RESET process (from LRS to HRS), the IV characteristic curve of the device shows three resistance states: high, medium, and low (such as Figure 1 , where RESET-1 is the process from LRS to IRS, RESET-2 is the process from IRS to HRS, and 1-4 is the direction of the entire memristive curve, that is, in the negative direction 1-2 is the transition from HRS to LRS, and in the positive direction 3-4 is LRS and then back to HRS), and it shows a highly stable resistance switching effect with good repeatability; when I cc When ≤0.01A, the memristor changes to dual-resistance mode (such as Figure 2 , for memristors at different I cc The memristor curve below, 1-4 is the direction of the entire memristor curve, that is, in the positive direction 1-2 is the transition from HRS to LRS, and in the negative direction 3-4 is LRS and then back to HRS), which is consistent with I cc Compared with the memristor curve when >0.01A, the scanning direction is opposite, and the SET process occurs in the positive voltage scanning interval, which may be due to the change of the memristor's memristive mechanism at this time.

[0041] Study Different I cc The device conduction mechanism below can explain the two different resistance switching behaviors of the memristor. In order to confirm and distinguish the two different conduction mechanisms, the present invention analyzes the double logarithmic graphs of the IV curves during SET and RESET respectively.

[0042] like Figure 3 As shown, when I ccWhen the current is ≤0.01A, the initial slope of the HRS during the SET process is Slope≈1.16 (Region 1). At low voltages, the number of injected electrons is small, and current transport in this region follows Ohm's law. Because traps exist in both the GO and SF materials, the injected electrons are captured by both functional materials. As the positive bias applied to the top electrode increases (Region 2), the number of injected electrons increases significantly, and most of the injected electrons are captured by traps in the functional layer material. At this point, the slope of the fitted curve is Slope≈1.76, which conforms to Child's law, where current and voltage have a square relationship. Subsequently, when the voltage is further increased, the electron-filled traps form conductive filaments connecting the upper and lower electrodes. Subsequent injected electrons can now freely move through the conductive filaments in the functional layer, causing the functional layer's resistance to drop sharply and the current to increase sharply (Region 3), causing the memristor to instantly switch to LRS. These results indicate that the resistance switching process follows a space charge limited current (SCLC) mechanism, where the capture and release of electrons at specific energy levels is responsible for the resistance change. Furthermore, Region 4 (slope ≈ 1.03) in the LRS also conforms to the same conduction mechanism.

[0043] On the other hand, when I cc When the device is >0.01A, the three-state resistance switching process occurs, and the SET process occurs in the negative voltage region. After analysis, the electron transport in the SET process still follows the SCLC mechanism. Figure 4 As shown in the left figure, in the RESET-1 process in the positive voltage region (i.e., the process from LRS to IRS), the slope of the curve of the device at LRS is Slope≈0.95 (Region 1, i.e., Region 1). Obviously, ohmic conduction is still the main conductive mechanism at this time. However, unlike the above-mentioned dual-resistance resistance switching process, after the RESET-1 process occurs, the device is in IRS, and the fitting slope of the curve is Slope≈2.11 (Region 2, i.e., Region 2). Then, after the device undergoes the RESET-2 process, the device will come to HRS, and the curve fitting slope is Slope≈1.57 (Region 3, i.e., Region 3). Obviously, the slopes of Regions 2 and 3 are different from the slopes of the curves in Region 1, and they do not conform to the conduction characteristics of the SCLC mechanism.

[0044] After analysis, we believe that the resistance switching mechanism during the RESET process is a thermal electron emission mechanism that belongs to the same charge trapping / detrapping mechanism, namely PF emission mechanism. We investigated the relationship between current and voltage in regions 2 and 3 respectively and found that during the two RESET processes, the relationship between current (I) and voltage (V) obeys the PF emission mechanism, namely: When V RESET (the voltage value at the beginning of the RESET process) is applied to the device. cc The large amount of electrons injected into the functional layer is enough to create a sufficiently high internal field. At this time, heat is generated in the device with the help of the electric field. Under the thermal disturbance, the charges trapped in the trap are excited to enter the conduction band of the oxide and conduct, resulting in PF emission. In addition, the IV data of the device are fitted (such as Figure 4 The middle and right figures show the raw data and fitted curves, illustrating the carrier transport model for these two sequential processes. In Region 2 and Region 3, the curves are well fitted to straight lines with slopes of ≈1.15 and ≈0.94, respectively, indicating that the device's RESET process can be attributed to the PF emission mechanism. These results indicate that, unlike the SCLC conduction mechanism of the SET process in the dual-resistance switching mode, PF emission is the dominant conduction type in this state, further clarifying the distinct resistance switching mechanisms of the two resistance switching modes.

[0045] Beneficial effects:

[0046] (1) The present invention provides a method for preparing a flexible bio-memristor for multi-level storage, which is simple, easy to operate, and low in cost, and can effectively prepare a high-performance bio-memristor with a large switching ratio, good stability, and long data retention time;

[0047] (2) The bio-memristor for multi-level storage of the present invention can switch between multiple resistance states within one scanning voltage cycle, significantly improving the storage density of the memristor device and realizing the multi-level storage function of data;

[0048] (3) The present invention provides a bio-memristor for multi-level storage, which endows SF-based memristors with multi-level storage and multi-bit logic operation capabilities. This multi-level storage performance outperforms most SF-based memristors and is comparable to many inorganic memristors, demonstrating the great potential of this device in multi-level storage.

[0049] (4) The bio-memristor for multi-level storage of the present invention is conducive to enriching the functions of the device, simplifying the structure of the neuromorphic computing and storage system, improving the data storage efficiency, and promoting the development of memristive devices towards lightweight, high efficiency, and multifunctionality. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 For I cc Resistance switching behavior of GO / SF / GO-based memristors during one voltage sweep cycle at >0.01A;

[0051] Figure 2 For I cc ≤0.01A at different I cc Below is the IV characteristic curve of GO / SF / GO based memristor;

[0052] Figure 3 For I cc Analysis of the conduction mechanism of GO / SF / GO-based memristors at ≤0.01A;

[0053] Figure 4 For I cc Analysis of the conduction mechanism of GO / SF / GO-based memristors at >0.01A: The left figure is a double-logarithmic fitting diagram of the IV characteristic curve of the device under HRS, IRS, and LRS during the SET process; the middle and right figures are fitting diagrams based on the PF emission mechanism when the device is in regions 2 and 3, respectively. DETAILED DESCRIPTION

[0054] Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiments are only used to illustrate the present invention and are not used in limiting the scope of the present invention.In addition, should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall equally within the scope limited by the appended claims of the application.

[0055] The testing methods for each performance in the following embodiments are as follows:

[0056] The memristor's memristive characteristics were measured using a Keithley 4200A-SCS semiconductor tester in DC mode at room temperature (25°C) and atmospheric pressure. The test results were output as IV curves. Under a constant read voltage (0.2V), the time the memristor's resistance remained stable in its high and low resistance states was measured, representing the data retention duration. The ratio of the high and low resistance states was the resistance switching ratio. When a continuous sweep voltage was applied, the good overlap of the IV curves within each cycle indicated good cycling stability.

[0057] Example 1

[0058] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0059] (1) Prepare raw materials:

[0060] Conductive layer: ITO film, manufactured by South China Xiangcheng Technology Co., Ltd., with a square resistance of ≤6 ohms, a film thickness of 185nm, and a transmittance of ≥84%;

[0061] An ethanol aqueous solution with a volume concentration of 85%;

[0062] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0063] A 2.8 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, and then dissolving the SF in a lithium bromide (LiBr) solution. The solution was then centrifuged, filtered, dialyzed, and concentrated (i.e., the SF solution was concentrated at 4°C) to obtain the SF solution.

[0064] (2) Preparation of bio-memristors for multi-level storage:

[0065] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0066] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 200 rpm and a spin coating time of 30 s; the second stage had a spin coating speed of 1000 rpm and a spin coating time of 30 s; the third stage had a spin coating speed of 2000 rpm and a spin coating time of 20 s;

[0067] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 3000 rpm for 30 seconds, then immersed in ethanol aqueous solution for 60 minutes and dried for 30 minutes to form an SF layer;

[0068] The thickness of the prepared SF layer was 30 nm and the crystallinity was 40%;

[0069] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0070] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 200 rpm and a spin coating time of 30 s; the second stage had a spin coating speed of 1000 rpm and a spin coating time of 30 s; the third stage had a spin coating speed of 2000 rpm and a spin coating time of 20 s;

[0071] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the prepared electrode layer is an Al electrode layer with a thickness of 200 nm.

[0072] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-resistance switching and triple-resistance switching by regulating the limiting current. When the regulating limiting current value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc >0.01A, the bio-memristor used for multi-level storage has a three-state resistance switching behavior; when the limiting current value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage exists in three resistance states: high, medium, and low within one switching cycle; the resistance switching ratio is 1.2×10 4 , the data retention time is 1.8×10 3 s, and still has good stability after 50 cycles.

[0073] Example 2

[0074] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0075] (1) Prepare raw materials:

[0076] Conductive layer: PET conductive film, manufactured by South China Xiangcheng Technology Co., Ltd., model (ITO)-PET, square resistance of 5 ohms, thickness of 0.05mm, transmittance ≥ 80%;

[0077] An ethanol aqueous solution with a volume concentration of 80%;

[0078] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0079] A 2.5 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, dissolving the SF in a lithium bromide (LiBr) solution, and then centrifuging, filtering, dialysis, and concentrating the SF solution (i.e., concentrating the SF solution at 4°C).

[0080] (2) Preparation of bio-memristors for multi-level storage:

[0081] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0082] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 200 rpm and a spin coating time of 30 s; the second stage had a spin coating speed of 1000 rpm and a spin coating time of 30 s; the third stage had a spin coating speed of 2000 rpm and a spin coating time of 20 s;

[0083] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 3000 rpm for 30 seconds, then immersed in ethanol aqueous solution for 60 minutes and dried for 30 minutes to form an SF layer;

[0084] The thickness of the prepared SF layer was 35 nm and the crystallinity was 40%;

[0085] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0086] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 200 rpm and a spin coating time of 30 s; the second stage had a spin coating speed of 1000 rpm and a spin coating time of 30 s; the third stage had a spin coating speed of 2000 rpm and a spin coating time of 20 s;

[0087] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the prepared electrode layer is an Ag electrode layer with a thickness of 200 nm.

[0088] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-resistance switching and triple-resistance switching by regulating the limiting current. When the regulating limiting current value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc >0.01A, the bio-memristor used for multi-level storage has a three-state resistance switching behavior; when the limiting current value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage exists in three resistance states: high, medium, and low within one switching cycle; the resistance switching ratio is 0.8×10 4 , the data retention time is 1.3×10 3s, and still has good stability after 45 cycles.

[0089] Example 3

[0090] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0091] (1) Prepare raw materials:

[0092] Conductive layer: ITO film, manufactured by South China Xiangcheng Technology Co., Ltd., with a square resistance of ≤6 ohms, a film thickness of 185nm, and a transmittance of ≥84%;

[0093] An ethanol aqueous solution with a volume concentration of 85%;

[0094] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0095] A 2.8 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, and then dissolving the SF in a lithium bromide (LiBr) solution. The solution was then centrifuged, filtered, dialyzed, and concentrated (i.e., the SF solution was concentrated at 4°C) to obtain the SF solution.

[0096] (2) Preparation of bio-memristors for multi-level storage:

[0097] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0098] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 600 rpm and a spin coating time of 20 s; the second stage had a spin coating speed of 1400 rpm and a spin coating time of 25 s; the third stage had a spin coating speed of 2600 rpm and a spin coating time of 15 s;

[0099] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 2000 rpm for 50 seconds, then immersed in ethanol aqueous solution for 60 minutes and dried for 30 minutes to form an SF layer;

[0100] The thickness of the prepared SF layer was 54 nm and the crystallinity was 45%;

[0101] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0102] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 600 rpm and a spin coating time of 20 s; the second stage had a spin coating speed of 1400 rpm and a spin coating time of 25 s; the third stage had a spin coating speed of 2600 rpm and a spin coating time of 15 s;

[0103] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the electrode layer is an Al electrode layer with a thickness of 240 nm.

[0104] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-resistance switching and triple-resistance switching by regulating the limiting current. When the regulating limiting current value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc >0.01A, the bio-memristor used for multi-level storage has a three-state resistance switching behavior; when the limiting current value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage exists in three resistance states: high, medium, and low within one switching cycle; the resistance switching ratio is 2.0×10 4 , the data retention time is 6.0×10 3 s, and still has good stability after 100 cycles.

[0105] Example 4

[0106] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0107] (1) Prepare raw materials:

[0108] Conductive layer: PET conductive film, manufactured by South China Xiangcheng Technology Co., Ltd., model (ITO)-PET, square resistance of 5 ohms, thickness of 0.05mm, transmittance ≥ 80%;

[0109] An ethanol aqueous solution with a volume concentration of 80%;

[0110] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0111] A 2.5 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, dissolving the SF in a lithium bromide (LiBr) solution, and then centrifuging, filtering, dialysis, and concentrating the SF solution (i.e., concentrating the SF solution at 4°C).

[0112] (2) Preparation of bio-memristors for multi-level storage:

[0113] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0114] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 600 rpm and a spin coating time of 20 s; the second stage had a spin coating speed of 1400 rpm and a spin coating time of 25 s; the third stage had a spin coating speed of 2600 rpm and a spin coating time of 15 s;

[0115] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 2000 rpm for 50 seconds, then immersed in ethanol aqueous solution for 60 minutes and dried for 30 minutes to form an SF layer;

[0116] The thickness of the prepared SF layer was 52 nm and the crystallinity was 45%;

[0117] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0118] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 600 rpm and a spin coating time of 20 s; the second stage had a spin coating speed of 1400 rpm and a spin coating time of 25 s; the third stage had a spin coating speed of 2600 rpm and a spin coating time of 15 s;

[0119] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the prepared electrode layer is an Ag electrode layer with a thickness of 240 nm.

[0120] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-resistance switching and triple-resistance switching by regulating the limiting current. When the regulating limiting current value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc>0.01A, the bio-memristor used for multi-level storage has a three-state resistance switching behavior; when the limiting current value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage has three resistance states: high, medium, and low within one switching cycle; the resistance switching ratio is 1.6×10 4 , the data retention time is 5.4×10 3 s, and still has good stability after 95 cycles.

[0121] Example 5

[0122] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0123] (1) Prepare raw materials:

[0124] Conductive layer: ITO film, manufactured by South China Xiangcheng Technology Co., Ltd., with a square resistance of ≤6 ohms, a film thickness of 185nm, and a transmittance of ≥84%;

[0125] An ethanol aqueous solution with a volume concentration of 85%;

[0126] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0127] A 2.8 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, and then dissolving the SF in a lithium bromide (LiBr) solution. The solution was then centrifuged, filtered, dialyzed, and concentrated (i.e., the SF solution was concentrated at 4°C) to obtain the SF solution.

[0128] (2) Preparation of bio-memristors for multi-level storage:

[0129] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0130] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 1000 rpm and a spin coating time of 15 s; the second stage had a spin coating speed of 1800 rpm and a spin coating time of 20 s; the third stage had a spin coating speed of 3000 rpm and a spin coating time of 10 s;

[0131] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 1200 rpm for 60 seconds, then immersed in ethanol aqueous solution for 90 minutes and dried for 30 minutes to form an SF layer;

[0132] The thickness of the prepared SF layer was 74 nm and the crystallinity was 60%;

[0133] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0134] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 1000 rpm and a spin coating time of 15 s; the second stage had a spin coating speed of 1800 rpm and a spin coating time of 20 s; the third stage had a spin coating speed of 3000 rpm and a spin coating time of 10 s;

[0135] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the electrode layer is an Al electrode layer with a thickness of 280 nm.

[0136] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-resistance switching and triple-resistance switching by regulating the limiting current. When the regulating limiting current value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc >0.01A, the bio-memristor used for multi-level storage has a three-state resistance switching behavior; when the limiting current value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage has three resistance states: high, medium, and low within one switching cycle; the resistance switching ratio is 2.5×10 4 , the data retention time is 4.0×10 3 s, and still has good stability after 85 cycles.

[0137] Example 6

[0138] A method for preparing a bio-memristor for multi-level storage comprises the following steps:

[0139] (1) Prepare raw materials:

[0140] Conductive layer: PET conductive film, manufactured by South China Xiangcheng Technology Co., Ltd., model (ITO)-PET, square resistance of 5 ohms, thickness of 0.05mm, transmittance ≥ 80%;

[0141] An ethanol aqueous solution with a volume concentration of 80%;

[0142] GO dispersion: Manufacturer: Hangzhou Gaoxin Technology Co., Ltd., GX-GO-1, monolayer rate over 99%, oxygen content 30%-40%, thickness ~1nm, stable dispersion solvents include water, DMF, ethylene glycol, etc.

[0143] A 2.5 wt% SF solution was prepared by boiling peeled silkworm cocoons twice in a 0.5 wt% Na2CO3 aqueous solution, washing with deionized water, dissolving the SF in a lithium bromide (LiBr) solution, and then centrifuging, filtering, dialysis, and concentrating the SF solution (i.e., concentrating the SF solution at 4°C).

[0144] (2) Preparation of bio-memristors for multi-level storage:

[0145] (2.1) Spin coating a GO dispersion on the upper surface of the conductive layer to form a GO layer;

[0146] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 1000 rpm and a spin coating time of 15 s; the second stage had a spin coating speed of 1800 rpm and a spin coating time of 20 s; the third stage had a spin coating speed of 3000 rpm and a spin coating time of 10 s;

[0147] (2.2) The upper surface of the product obtained in step (2.1) was spin-coated with SF solution at a speed of 1200 rpm for 60 seconds, then immersed in ethanol aqueous solution for 90 minutes and dried for 30 minutes to form an SF layer;

[0148] The thickness of the prepared SF layer was 80 nm and the crystallinity was 60%;

[0149] (2.3) Spin-coating a GO dispersion on the upper surface of the SF layer in step (2.2) to form a GO layer;

[0150] The spin coating of GO dispersion was carried out in three stages: the first stage had a spin coating speed of 1000 rpm and a spin coating time of 15 s; the second stage had a spin coating speed of 1800 rpm and a spin coating time of 20 s; the third stage had a spin coating speed of 3000 rpm and a spin coating time of 10 s;

[0151] (2.4) On the upper surface of the GO layer obtained in step (2.3), an electrode layer is deposited by thermal evaporation to obtain a bio-memristor for multi-level storage; wherein the prepared electrode layer is an Ag electrode layer with a thickness of 280 nm.

[0152] The prepared bio-memristor for multi-level storage is composed of an electrode layer, a memristive functional layer and a conductive layer in sequence, wherein the memristive functional layer is composed of a GO layer, a SF layer and a GO layer in sequence; the bio-memristor for multi-level storage can realize the mutual conversion between the device's dual-state resistance switching and triple-state resistance switching by regulating the limiting current. When the regulating limiting current value Icc is ≤ 0.01A, the bio-memristor for multi-level storage has a dual-state resistance switching behavior; when the regulating limiting current value Icc is greater than 0.01A, the bio-memristor for multi-level storage has a triple-state resistance switching behavior; when the regulating limiting current value Icc is greater than 0.01A, the bio-memristor for multi-level storage has three resistance states: high, medium and low within one switching cycle; the resistance switching ratio is 1.8×104, the data retention time is 3.6×103s, and it still has good stability after 80 cycle cycles.

Claims

1. A bio-memristor for multi-level storage, characterized in that: The electrode layer, the memristive functional layer and the conductive layer are sequentially composited, wherein the memristive functional layer is sequentially composited by the GO layer, the SF layer and the GO layer, and the crystallinity of the SF layer is 35-60%. The thickness of each GO layer is 15-25 nm, the thickness of the SF layer is 30-80 nm, and the thickness of the electrode layer is 200-300 nm; The electrode layer is an Al electrode layer or an Ag electrode layer; When the regulating current limit value I cc When the current value is less than 0.01A, the bio-memristor used for multi-level storage has a dual-state resistance switching behavior; when the limiting current value I cc At >0.01A, the bio-memristor for multi-level storage exhibits three-state resistance switching behavior; When the regulating current limit value I cc When the resistance is >0.01A, the bio-memristor used for multi-level storage exists in three resistance states: high, medium, and low within one switching cycle.

2. The bio-memristor for multi-level storage according to claim 1, characterized in that: The thickness of the conductive layer is 0.6 to 1.0 mm.

3. The bio-memristor for multi-level storage according to claim 1, characterized in that: The conductive layer is an ITO film or a PET conductive film.

4. A bio-memristor for multi-level storage according to any one of claims 1 to 3, characterized in that: Bio-memristors with a resistance-on / off ratio greater than 10 for multi-level storage 4 , data retention time is greater than 10 3 s.

5. A method for preparing a bio-memristor for multi-level storage according to any one of claims 1 to 4, characterized in that: The upper surface of the conductive layer is sequentially spin-coated with a GO dispersion to form a GO layer, and then the SF solution is spin-coated and immersed in an ethanol aqueous solution with a volume concentration of 75-90% and dried to form an SF layer. Then, the upper surface of the SF layer is sequentially spin-coated with a GO dispersion to form a GO layer, and an electrode layer is deposited to obtain a bio-memristor for multi-level storage.

6. The method according to claim 5, characterized in that The concentration range of all GO dispersions was 0.5-1.0 mg / ml; each spin coating of GO dispersion was carried out in three stages, with the spin coating speed increasing from the first stage to the third stage.

7. The method according to claim 6, characterized in that Each time the GO dispersion is spin-coated, the first stage spin-coating speed is 200-1000 rpm, and the spin-coating time is 10-30 s; the second stage spin-coating speed is 1000-1800 rpm, and the spin-coating time is 10-30 s; the third stage spin-coating speed is 2000-3000 rpm, and the spin-coating time is 5-20 s.

8. The method according to claim 5, characterized in that The concentration of the SF solution is 2.0-3.0 wt %; when the SF solution is spin-coated, the spin-coating speed is 1200-3000 rpm, and the spin-coating time is 30-60 s.

9. The method according to claim 5, characterized in that The soaking time is 60 to 90 minutes, and the drying time is 20 to 30 minutes.

10. The method according to claim 5, characterized in that The electrode layer is deposited by thermal evaporation, electron beam evaporation or magnetron sputtering.

Images