Poly(Chalcogeno-Viologen-Triphenylamine) Material Polymer Memristor for Biological Synapse Simulation and Neuromorphic Computing and Its Preparation

Abstract

The present invention belongs to the field of organic / polymeric information storage devices, and specifically provides a poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and neuromorphic computing and its preparation. The device structure is a classic sandwich structure, with the bottom electrode being ITO, the active layer being a PCVTPA thin film containing redox structural units of viologen and triphenylamine bridged by chalcogen elements (S, Se, Te) (including PSVTPA, PSeVTPA, PTeVTPA), and the top electrode being Al. By applying a continuously scanned cyclic voltage to the device, its current-voltage curve exhibits memristive performance based on history dependence, which can be used to simulate synaptic potential / inhibition, human learning and memory functions, the transition from short-term synaptic plasticity to long-term plasticity, and can also perform decimal arithmetic functions.

Description

Technical Field

[0001] The present invention belongs to the field of memristors, and particularly relates to the preparation of a biomimetic memristor based on chalcogeno-viologen polymers, and the prepared memristive device has the functions of biological synapse simulation and neuromorphic computing at the same time. Background Art

[0002] Human memory mainly benefits from the natural evolution of neural networks, which exhibits several prominent characteristics, such as large-scale parallel processing, in-memory computing architecture, event-driven operation, etc. Since the discovery of the first true memristor by HP Labs in 2008, efforts have been made to develop new memristive functional materials and devices to construct artificial neural networks for neuromorphic computing and simulate physiological functions. The memristor-based brain-inspired intelligent computing system can not only significantly improve the computing power of modern computer systems through large-scale parallelization and extremely low power consumption, but also overcome the von Neumann bottleneck (i.e., the limited throughput between the memory and the central processing unit) when dealing with data-intensive tasks. A large number of inorganic materials have been used to construct memristor devices and still play a leading role in manufacturing memristors with excellent processing efficiency and huge storage capacity. Similar to these inorganic materials, some polymer functional materials, including metal-containing polymers, polymer-based multi-component redox systems, and pure polymers, have also been found to exhibit excellent memristor performance in recent years. Through these polymer memristors with mechanical flexibility and light weight, non-linear transmission characteristics similar to biological synapses can be observed. More importantly, their memristive properties can be easily adjusted through innovative molecular design and synthesis strategies.

[0003] Integrating programmable multi-level storage, synaptic biomimicry, and neuromorphic computing into the same electronic device is quite ideal. For this purpose, we introduce chalcogen (S, Se, Te) bridged viologen units into the polymer, which can greatly improve the conductivity, electrochemistry, and electrochromic properties of the material. Taking PTeVTPA as an example, the Al / PTeVTPA / ITO device exhibits excellent multi-level storage and history-dependent memristive switching performance. This device can not only be used to simulate synaptic potential / inhibition, human learning and memory functions, and the transition from short-term synaptic plasticity to long-term plasticity, but also perform decimal arithmetic operations. This work is expected to provide a new idea for constructing high-performance synaptic biomimicry and neuromorphic computing systems in the near future. Summary of the Invention

[0004] Therefore, based on the above advantages of memristors, the purpose of the present invention is to provide a method for biological synapse simulation and neuromorphic computing of a polymer memristor that integrates data storage and / or information processing functions.

[0005] The second object of the present invention is to provide a poly(chalcogeno-viologen-triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing and its applications.

[0006] The third object of the present invention proposes a method for constructing a poly(chalcogeno-viologen-triphenylamine) material polymer memristor for biological synapse simulation and neuromorphic computing.

[0007] Another object of the present invention is also to provide a method for synthesizing the active layer polymer of such a memristor.

[0008] The technical solution of the present invention:

[0009] A poly(chalcogeno-viologen-triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing, the active layer of which is synthesized from a chalcogen element-bridged viologen and a triphenylamine polymer. The structural composition of the polymer memristor from bottom to top is:

[0010] (1) Indium tin oxide electrode ITO;

[0011] (2) Active layer: polymer PCVTPA thin film, and the polymer PCVTPA thin film is PSVTPA, PSeVTPA or PTeVTPA;

[0012] (3) Aluminum electrode.

[0013] Furthermore, the structure of the polymer PCVTPA thin film is shown as follows:

[0014]

[0015] The poly(chalcogeno-viologen-triphenylamine) material polymer memristor of the present invention has both biological synapse simulation and / or neuromorphic computing functions.

[0016] The poly(chalcogeno-viologen-triphenylamine) material polymer memristor of the present invention has a biological synapse simulation function. The device has different responses to different forms of input pulses, and can be used to simulate the inhibition of synaptic potential, the learning and memory functions of humans, and the transition function from short-term synaptic plasticity to long-term plasticity.

[0017] The poly(chalcogeno-viologen-triphenylamine) material polymer memristor of the present invention has a neuromorphic computing function, including four arithmetic functions; there is a linear symmetric relationship between the current shown by the device and the number of scans, and the four arithmetic functions are realized by using this characteristic.

[0018] The poly(thiogroup viologen - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing according to the present invention, and the implementation methods for biological synapse simulation (Examples 1 and 2).

[0019] The poly(thiogroup viologen - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing according to the present invention, and the implementation method for neuromorphic computing (Example 3).

[0020] The present invention also provides a preparation method of the polymer PCVTPA in the above - mentioned poly(thiogroup viologen - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing:

[0021] (1) Synthesis of thiophene[2,3 - c:5,4 - c']bipyridine, selenophene[2,3 - c:5,4 - c']bipyridine or tellurophene[2,3 - c:5,4 - c']bipyridine;

[0022]

[0023] (2) Synthesis of N,N - bis(4 - (bromomethyl)phenyl)aniline M2;

[0024]

[0025] (3) Synthesis of the active material layer: PCVTPA;

[0026]

[0027] The mixture of M1 and M2 in N,N - dimethylformamide is heated at 60 °C - 70 °C for 5 - 7 days (preferably heated at 60 °C for 5 days); the precipitate is separated by vacuum filtration, washed with CH2Cl2, the product is collected, and dried under high vacuum at 50 - 65 °C (preferably 65 °C) to obtain a powdery solid. Subsequently, it is dispersed in acetonitrile, an excess of methyl trifluoroacetate is added, and then stirred overnight; after the reaction is completed, the above reaction mixture is added to ice methanol, the precipitate is separated by vacuum filtration, washed with methanol, and dried under high vacuum at 50 - 65 °C (preferably 65 °C) to obtain the polymer material PCVTPA.

[0028] The present invention also provides a method for constructing the above - mentioned poly(thiogroup viologen - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing. The specific steps include:

[0029] 1) The ITO substrate is ultrasonically cleaned with ethanol, acetone, and isopropanol for 15 min in sequence, and dried with nitrogen;

[0030] 2) Dissolve the PCVTPA film in an acetonitrile solution and spin-coat it on the ITO bottom electrode, then dry it. The thickness is about 200 - 300 nm (preferably 200 nm).

[0031] 3) Deposit the top electrode Al on the surface of the active layer, and the deposition thickness is about 100 - 200 nm (preferably 100 nm).

[0032] The present invention also provides an application of the poly(chalcogeno viologen - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing in biological synapse simulation and neuromorphic computing.

[0033] The specific operation method for realizing the biological synapse simulation and / or neuromorphic computing function of the present invention is as follows

[0034] For the Al / PTeVTPA / ITO device, applying the voltage of 0V → 1V → 0V repeatedly seven times can generate 7 different current states, and there is an increasing trend based on history dependence between different current states. The specific operation method for realizing the data processing function of the present invention is as follows. Under continuous negative voltage (0V → -0.5V → 0V) and positive voltage (0V → 0.5V → 0V) scans of the Al / PTeVTPA / ITO device, the current of the device can be continuously regulated. During the forward scan voltage process of 0V → 0.5V → 0V, the absolute value of the current of the device will slowly increase with the increase of the scan times; during the negative scan voltage process of 0V → -0.5V → 0V, the absolute value of the current of the device will slowly decrease with the increase of the scan times, and there is a good linear relationship between the magnitude of the absolute value of the current and the scan times. At the same time, the other two devices (Al / PSeVTPA / ITO, Al / PSVTPA / ITO) can also show the same performance. Based on this, we realized the functions of biological synapse simulation and neuromorphic computing of the device by using pulse testing.

[0035] The present invention provides a poly(chalcogen violet - triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing, which realizes the technology of integrating data storage and information processing into one; the fabricated device has very good linearity and symmetry in its memristive performance, so it can be applied to bionic simulation and neuromorphic computing. This represents a certain breakthrough compared to the single applications of other devices; in the prior art (including the related technologies previously published by this research group), the performance that the devices can achieve mainly involves data storage functions, and the materials related to memristive performance also require additional processing during application due to their poor linearity and symmetry, increasing the technical difficulty. Based on this, the present invention realizes storage and processing within one device and can achieve the same applications without additional processing; the present invention solves the problems of insufficient linearity and poor symmetry in the memristive performance of the devices in the prior art, which leads to the need for additional data processing during their application to achieve corresponding applications, while our invention solves this problem, has good linearity and symmetry, and can be directly applied.

[0036] By applying a continuously scanned cyclic voltage to the device, its current - voltage curve exhibits history - dependent memristive performance, which can be used to simulate synaptic potential / inhibition, human learning and memory functions, the transition from short - term synaptic plasticity to long - term plasticity, and can also perform decimal arithmetic functions. By designing and selecting polymer structural units, the electrical properties of the polymer are regulated to realize high - performance electronic devices for biological synapse simulation and neuromorphic computing, providing a powerful means for devices that can simultaneously meet storage and computing requirements.

[0037] The present invention has the following advantages:

[0038] 1. The memristor mentioned in the present invention simultaneously has the functions of biological synapse simulation and / or neuromorphic computing.

[0039] 2. The storage device with PCVTPA as the active layer has stable performance.

[0040] 3. The device with PTeVTPA as the active layer has excellent linearity in its memristive performance and has the potential to break through the current bottleneck of chips. It has broad application prospects in the future field of artificial intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Schematic diagram of the history - dependent memristive performance of the device with the Al / PCVTPA / ITO structure: a) Al / PSVTPA / ITO; b) Al / PSeVTPA / ITO; c) Al / PTeVTPA / ITO.

[0042] Figure 2 Bio - mimetic performance of the Al / PTeVTPA / ITO device: Simulation of frequency - dependent synaptic facilitation. Figures a) current and b) current change (ΔI = I n - I1) versus 10 voltage - pulse stimulations at different frequencies; Figure c) current versus 5 - pulse stimulations with different pulse durations; d) simulation of short - term potentiation and paired - pulse facilitation.

[0043] Figure 3 Further simulation of biological synapses by the Al / PTeVTPA / ITO device: Simulation of human brain memory and forgetting. a) The response of the device to pulses corresponding to synaptic inhibition and enhancement. b) Memory retention performance of experiments (symbols) and fitting (solid lines) after different numbers of pulse stimulations. d - h) Demonstration of the "learning - forgetting - relearning" process.

[0044] Figure 4 Four - arithmetic operations of the Al / PTeVTPA / ITO device: a) Linearity of device performance; b) Realization of 10 - 10 = 0; c) Addition and commutative law; d) Subtraction and commutative law; (e) Multiplication and commutative law; (f) Division operation of 6 / 4 = 1.5. Specific implementation manners

[0045] The following describes the embodiments of the present invention in detail with reference to the accompanying drawings, so that the memristor performance of the device prepared by the present invention is more intuitive and easy to understand.

[0046] Preparation Example 1: A synthesis method of a polymer active layer of a polymer memristor that simultaneously has biological synapse simulation and / or neuromorphic computing functions, including:

[0047]

[0048] 1) Synthesis of thiophene - benzo[2,3 - c:5,4 - c']dipyridine (M1a)

[0049] 3,3'-Dibromo-4,4'-bipyridine (628 mg, 2.0 mmol) was dissolved in dry THF (60 mL), and a solution of n-butyllithium (2.5 M in hexane, 1.68 mL, 4.2 mmol) was added dropwise via syringe at -94 °C. After stirring for 1 h at the same temperature, S2Cl2 (284 mg, 2.1 mmol) was added dropwise to the above solution. Then the reaction mixture was allowed to warm to room temperature and stirred overnight. After evaporating the solvent under reduced pressure, the residue was dissolved in H2O (30 mL), NH4OH (25%, 30 mL), and NH4Cl (saturated, 30 mL). The mixture was extracted several times with CHCl3 (50 mL each time), and the combined organic phases were dried over anhydrous Na2SO4 and concentrated to give a brown oil, which was purified by column chromatography (SiO2, petroleum ether / ethyl acetate, 2:1) to afford 150 mg (40%) of a pale yellow solid. 1H-NMR (400 MHz, CDCl3): δ / ppm = 9.30 (s, 2H), 8.76 (d, 2H), 8.14 (d, 2H).

[0050] 2) Synthesis of selenopheno[2,3-c:5,4-c']bipyridine (M1b)

[0051] 3,3'-Dibromo-4,4'-bipyridine (628 mg, 2.0 mmol) was dissolved in dry THF (60 mL), and a solution of SeCl2 (315 mg, 2.1 mmol) was added dropwise via syringe at -94 °C. After stirring for 1 h at the same temperature, the above solution was added dropwise to the solution. Then the reaction mixture was allowed to warm to room temperature and stirred overnight. After evaporating the solvent under reduced pressure, the residue was dissolved in H2O (30 mL), NH4OH (25%, 30 mL), and NH4Cl (saturated, 30 mL). The mixture was extracted several times with CHCl3 (50 mL each time), and the combined organic phases were dried over anhydrous Na2SO4 and concentrated to give a brown oil, which was purified by column chromatography (SiO2, petroleum ether / ethyl acetate, 2:1) to afford 89 mg (20%) of a bright yellow solid. 1 1H-NMR (400 MHz, CDCl3): δ / ppm = 9.30 (s, 2H), 8.76 (d, 2H), 8.14 (d, 2H).

[0052] 3) Synthesis of telluropheno[2,3-c:5,4-c']bipyridine (M1c)

[0053] 3,3'-Dibromo-4,4'-bipyridine (628 mg, 2.0 mmol) was dissolved in dry THF (60 mL), and a solution of n-butyllithium (2.5 M in hexane, 1.68 mL, 4.2 mmol) was added dropwise via syringe at -94 °C. After stirring for 1 h at the same temperature, SeCl2 (315 mg, 2.1 mmol) was added dropwise to the above solution. Then the reaction mixture was allowed to warm to room temperature and stirred overnight. After evaporating the solvent under reduced pressure, the residue was dissolved in H2O (30 mL), NH4OH (25%, 30 mL), and NH4Cl (saturated, 30 mL). The mixture was extracted several times with CHCl3 (50 mL each time), and the combined organic phases were dried over anhydrous Na2SO4 and concentrated to give a brown oil, which was purified by column chromatography (SiO2, petroleum ether / ethyl acetate, 2:1) to afford 89 mg (20%) of a bright yellow solid. 1 1H-NMR (400 MHz, CDCl3): δ / ppm = 9.30 (s, 2H), 8.76 (d, 2H), 8.14 (d, 2H).

[0054]

[0055] 4) Synthesis of N,N-bis(4-(bromomethyl)phenyl)aniline (M2)

[0056] N,N-Bis(4-(formyl)phenyl)aniline (1.36 g, 4.5 mmol) and sodium borohydride (0.378 g, 10.0 mmol) were stirred under dark at room temperature overnight in a mixed solvent of 120 mL of dry dichloromethane and ethanol (1:1). Then 120 mL of water was added to the above reaction mixture. The resulting mixture was extracted with CHCl3 at least three times (120 mL each time), and the combined organic phases were dried over anhydrous Na2SO4. After evaporating the solvent under reduced pressure, the residual white powder was directly suspended in 50 mL of dry ether, and then a solution of phosphorus tribromide (514 μL, 5.4 mmol) in 10 mL of dry ether was added dropwise to the above suspension at 0 °C within 5 min. The reaction mixture was allowed to warm to room temperature and then stirred overnight in the dark. A mixture of ice water (60 mL) and saturated NaHCO3 solution (30 mL) was added to the above reaction system. The reaction mixture was extracted with CH2Cl2 at least three times (60 mL each time), and the combined organic phases were dried over anhydrous Na2SO4. After evaporating the solvent under reduced pressure, a viscous light green solid (1.88 g, 96%) was obtained. 1 1H NMR (400 MHz, CD2Cl2): δ / ppm = 7.31 - 7.26 (m, 6H), 7.11 - 7.09 (m, 3H), 7.03 - 7.01 (d, 4H), 4.52 (s, 4H). EI: calcd for C20 H 17 Br2N: 430.97; found: m / z = 430.97 (M + , 100%).

[0057]

[0058] 5) Active material layer: Synthesis of PCVTPA

[0059] A mixture of M1 (1.0 mmol) and M2 (1.0 mmol) in N,N-dimethylformamide (30 mL) was heated at 60 °C for 5 days. The precipitate was separated by vacuum filtration and washed with CH2Cl2 at least three times (30 mL each time). The product was collected and dried under high vacuum at 65 °C to obtain a powdery solid, which was then dispersed in acetonitrile (30 mL), and an excess of methyl trifluoroacetate was added, followed by stirring overnight. After the reaction was completed, the above reaction mixture was slowly added to 500 mL of ice-cold methanol. The precipitate was separated by vacuum filtration and washed with methanol at least three times (50 mL each time). The product was collected and dried under high vacuum at 65 °C to obtain the polymer material PCVTPA.

[0060] Preparation Example 2:

[0061] A method for constructing a polymeric memristor of poly(chalcogeno-viologen-triphenylamine) material for biosynapse simulation and / or neuromorphic computing, the specific steps including:

[0062] 1) The ITO substrate was ultrasonically cleaned with ethanol, acetone, and isopropanol for 15 min in sequence and dried with nitrogen;

[0063] 2) The PCVTPA thin film prepared in Preparation Example 1 was dissolved in an acetonitrile solution and spin-coated on the ITO bottom electrode, followed by drying to a thickness of about 200 nm;

[0064] 3) The top electrode Al was deposited on the surface of the active layer to a deposition thickness of about 100 nm.

[0065] Example 1:

[0066] As Figure 2 , in order to explore the effects of pulse stimulation frequency and duration on current, we successfully mimicked spike frequency-dependent plasticity (SRDP) of biosynapses ( Figure 2 of a, Figure 2 of b) by increasing the frequency of voltage pulses applied to the Al / PTeVTPA / ITO device. During the experiment, the stimulation frequency varied from 1, 2, 5, 10, 20, while the number of pulse stimulations was fixed at 10. From Figure 2As can be seen from a, the more frequently the device acting as a biological synapse is stimulated, the higher the device current observed in this study. At a frequency of 1 or 2 Hz, the change in the device current with 10 pulse stimulations is very small. As the frequency is further increased, the observed current change begins to become more obvious or significant. At a frequency of 20 Hz and 10 stimulations, compared with 1 stimulation, the change in the device current reaches 51.99 A. These results indicate that the memristor can be used as a high-pass filter. Figure 2 c shows that when we applied a frequency of 1 Hz and 5 stimulations, by changing the pulse duration from 5 ms, 10 ms, 20 ms to 60 ms, the device current remained almost stable, indicating that the device has good stability. Then we set a double-pulse stimulation, by changing the time interval between the two pulses from 0 - 2000 ms, and recorded the change value of the current, from Figure 2 As can be seen from d, by fitting with the formula I(t) = I0 + A exp(-t / τ), two characteristic time scales are shown, τ1 is 71 ms and τ2 is 280 ms. These two values are very close to those shown by biological synapses.

[0067] Example 2:

[0068] As Figure 3 , synaptic potentiation and depression, which are considered the neurobiological basis of the brain's memory function, can be achieved through action potential spikes. When the device is applied with continuous negative voltage pulses or positive stimuli respectively, the weight of the synapse can be effectively inhibited or potentiated ( Figure 4 a). During biological stimulation, the processes of synaptic enhancement and weakening are competitive. Figure 3 b shows the retention curves of synaptic weights under different numbers of the same voltage pulse stimulations. From this, it can be seen that the synaptic weight undergoes rapid decay at the beginning and then gradually tends to increase. This result means that the number of the same voltage pulse stimulations with the same amplitude, period and duration can greatly affect the memory loss or retention performance of the device. Basically, human memory mainly comes from short-term plasticity (STP) and long-term plasticity (LTP). Long-term memory can be maintained for a long time (such as several days, months, or even years), while short-term memory can only be maintained for a very short time (such as a few seconds or minutes). As Figure 3 shown in c, as the number of pulse stimulations increases, the relaxation time constant changes from 4.51 s @ 10 pulses to 4.98 s @ 20 pulses to 6.81 s @ 30 pulses to 7.40 s @ 40 pulses to 9.35 s @ 50 pulses and then to 10.36 s @ 60 pulses. This finding makes it possible to transition from short-term memory to long-term memory in our device. Moreover, the process of "learning - forgetting - re - learning" in human daily life has also been successfully explored in this study (Figure 3 d- Figure 3 g). Apply a stimulus of 50 consecutive voltage pulses to the device. The current of the device gradually increases as the number of pulses increases ( Figure 3 d). This is similar to the "learning" process observed in the human learning process. Once the power is cut off, the current of the device gradually drops to an intermediate state within 200 seconds ( Figure 3 e). This process can be regarded as a "forgetting" process. Then we apply a stimulus of 35 consecutive pulses to the device again, and the current returns to the end level of the first learning stage ( Figure 3 f). Another "re-forgetting" process occurs ( Figure 3 f). Compared with the device current (about 0.36 mA) detected at the end of the first "forgetting" process, the device current observed at the end of the "re-forgetting" process reaches 0.41 mA ( Figure 3 g), which is much higher than the former. This indicates that the speed of "re-forgetting" is greatly slowed down compared with the first "forgetting" speed. Then, with only 15 voltage pulses, the current of the device returns to the same high level as the end level of the first learning stage ( Figure 3 h). These results show that the voltage pulses applied to the device will become fewer and fewer over time.

[0069] Example 3:

[0070] As Figure 4 , from the above discussion, we can see that our device shows a continuous resistive switching effect at any device current level. It can be used to perform arithmetic addition, subtraction, multiplication, and division. When a stimulus of consecutive voltage pulses is applied to the device, as Figure 4 shown in a of Figure 4 a-c, the relationship between the number of pulses and the device current read at 0.5 V is completely linear whether in the positive sweep or negative sweep. Figure 4b). After applying a stimulus of 10 consecutive positive pulses (0.5 V and 10 ms) to the device, the observed device current reached 0.43 mA. Thus, when a current of ~0.43 mA is read, the number 10 will be calculated in future computations. By setting the device current of 0.2 mA as the initial state of the device, the decimal digits from 0 - 10 can be indexed proportionally. When a stimulus of 10 consecutive negative pulses is then applied, the device current returns to the initial state. By such operations, one can easily achieve the subtraction operation of 10 - 10 = 0 and calibrate the device as shown for precise decimal operations. In this device, a positive voltage that causes an increase in the device current is applied for addition operations, while a negative voltage applied to the device corresponds to the subtraction function. By monitoring the current change in the initial state of the device, the number of input pulse stimuli can be calculated. Applying 6 consecutive positive pulses (0.5 V and 10 ms), and then another series of 4 positive pulses with the same duration and amplitude as the former, results in a device current of approximately 0.43 mA, which fully confirms that 6 + 4 = 10( Figure 4 c). When we reverse the order of the two sets of input pulse signals, under the same experimental conditions, the device current also reaches ~0.43 mA. These results verify the conversion rule such as 6 + 4 = 4 + 6 = 10. After preloading 10 positive pulses (0.5 V and 10 ms), the device current reaches ~0.43 mA. Applying 4 consecutive negative pulses and then a subsequent series of 6 consecutive negative pulses causes the device current to drop back to the initial state again. This confirms that 10 - 6 - 4 = 0. Reversing the loading order of the input signals also causes the device current to drop to the value of the initial state (10 - 4 - 6 = 0), which confirms that 10 - 6 - 4 = 10 - 4 - 6 = 0. These findings demonstrate the commutative subtraction operation. Similarly, the multiplication operation based on cumulative addition also follows the commutative law. The result obtained by applying two sets of 5 positive pulses is the same as that obtained by applying five sets of 2 positive pulses, ensuring that 2×5 = 5×2 = 10. Decimal division is based on the combination of subtraction and addition operations. For example, we can perform the division operation of 64 with our device( Figure 4 d). The device is first reset to the initial state, then 6 consecutive positive voltage pulses are applied, followed by another series of 4 negative pulses for the operation of 6 - 4. As a result, a remainder of 2 and an integer quotient of 1 are achieved. Considering that the remainder 2 is less than the divisor 4, we add two series of 9 positive voltage pulses to the device and replace 2 with "2 + 2×9" (2×10). Subsequently, subtracting -4 five times (20 - 4 - 4 - 4 - 4) causes the device current to return to the initial state again. Then, the calculation of the decimal division terminates. These results indicate that the decimal point is at the tenth place and the quotient 6 / 4 = 1.5 (1 + 0.5).

[0071] The test results of Al / PSVTPA / ITO and Al / PSeVTPA / ITO are basically consistent with those of Examples 1-3. The memristive properties of the three devices are similar, and they are all similar symmetric curves. Details will not be elaborated here.

Claims

1. A poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing, characterized in that: Its active layer is polymerized from a viologen bridged by chalcogen elements and a triphenylamine polymer. The structural composition of the polymer memristor from bottom to top is as follows: (1) Indium tin oxide electrode ITO; (2) Active layer: Polymer PCVTPA thin film, and the polymer PCVTPA thin film is PSVTPA, PSeVTPA or PTeVTPA; (3) Aluminum electrode.

2. The poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing according to claim 1, wherein, The structure of the polymer PCVTPA thin film is shown as follows: n is the repeating unit.

3. The poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing according to claim 1, wherein It has both biological synapse simulation and neuromorphic computing functions.

4. The poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing according to claim 1, wherein It has the function of biological synapse simulation. The device has different responses to different forms of input pulses and can be used to simulate the inhibition of synaptic potential, the learning and memory functions of humans, and / or the transition function from short-term synaptic plasticity to long-term plasticity.

5. The poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing according to claim 1, wherein It has the function of neuromorphic computing, including four arithmetic operations; there is a linear symmetric relationship between the current shown by the device and the number of scans, and the four arithmetic operations function is realized by using this characteristic.

6. The poly(chalcogeno-viologen-triphenylamine) material-based polymeric memristor for biological synapse simulation and / or neuromorphic computing according to any one of claims 1-5, characterized in that, The polymer PCVTPA is prepared by the following steps: (1) Synthesis of thiophene[2,3-c:5,4-c']bipyridine, selenophene[2,3-c:5,4-c']bipyridine or tellurophene[2,3-c:5,4-c']bipyridine; (2) Synthesis of N,N-bis(4-(bromomethyl)phenyl)aniline M2; (3) Synthesis of the active material layer: PCVTPA; The mixture of M1 and M2 in N,N-dimethylformamide is heated at 60 °C - 70 °C for 5 - 7 days; the precipitate is separated by vacuum filtration, washed with CH2Cl2, the product is collected, and dried under high vacuum at 50 - 65 °C to obtain a powdery solid, which is then dispersed in acetonitrile, and an excess of methyl trifluoroacetate is added, and then stirred overnight; after the reaction is completed, the above reaction mixture is added to ice methanol, the precipitate is separated by vacuum filtration, washed with methanol, and dried under high vacuum at 50 - 65 °C to obtain the polymer material PCVTPA.

7. The preparation method of the poly(chalcogeno-viologen-triphenylamine) material-based polymer memristor for biological synapse simulation and / or neuromorphic computing according to any one of claims 1-6, characterized in that, The specific steps include: 1) The ITO substrate is ultrasonically cleaned successively with ethanol, acetone, and isopropanol, and dried with nitrogen; 2) PCVTPA is dissolved in an acetonitrile solution, and then spin-coated on the ITO substrate and dried to obtain a PCVTPA thin film with a thickness of 200 - 300 nm; 3) The top electrode Al is deposited on the surface of the active layer, and the deposition thickness is 100 - 200 nm.

8. Application of the poly(chalcogen viologen-triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing according to any one of claims 1 - 6 or the memristor prepared by the preparation method of the poly(chalcogen viologen-triphenylamine) material polymer memristor for biological synapse simulation and / or neuromorphic computing according to claim 7 in biological synapse simulation and / or neuromorphic computing.

Images