Ionic liquid DEME-TFSI memristor with biological synapse simulation function and preparation method thereof

Abstract

The invention provides an ionic liquid DEME-TFSI memristor with a biological synapse simulation function. The ionic liquid DEME-TFSI memristor comprises a bottom electrode, an ionic liquid layer and a top electrode layer, wherein the bottom electrode is arranged at the bottommost part; the ionic liquid layer is arranged above the bottom electrode; and the top electrode layer is arranged above the ionic liquid layer. The DEME-TFSI memristor with the biological synapse simulation function belongs to a typical nonvolatile memristor, has the advantages of low cost, simple preparation conditions, easy design of the structure of the whole memristor, simple process, stable performance and the like, successfully simulates the characteristics of the biological synapse function, such as LTP / LTD, PPF, STDP and the like, and has a wide application prospect. It is shown that the DEME-TFSI memristor has application potential as a high-performance nonvolatile memory and in future computer systems and neuromorphic computation.

Description

Technical Field

[0001] The invention belongs to the field of memory, and in particular relates to an ionic liquid DEME-TFSI memristor with a biological synapse simulation function and a preparation method thereof. Background Art

[0002] Biological synapses are the basic part of signal transmission between neurons. Each biological synapse consists of presynaptic / postsynaptic membranes and synaptic cleft. The realization of synaptic characteristics depends on the transmission of neurotransmitters in the synapses connecting two neurons. In neural networks, neurons are the basic elements connected to each other through synapses, which have regulatory functions and information processing capabilities. In order to realize the functions of adaptive neural networks, neuromorphic hardware composed of electronic devices has attracted widespread attention. These electronic devices have adjustable and storable conductive states, and can spontaneously generate and regulate neural signals. Among them, memristors, as an emerging and efficient device, have become an important research object for realizing the functions of neural networks.

[0003] The dispersion of threshold voltage and high and low resistance values ​​of memristors is the main problem currently encountered, which may lead to erroneous read and write operations or even device failure. In order to improve performance stability, researchers have proposed a variety of methods, including adding interface layers, doping, adding nanocrystals, improving preparation processes, and improving operation methods. Currently, the most reported memristor materials are binary oxides such as NiO, SiOX, NbOX, and sulfur compound semiconductors. There are very few memristors based on ions as intermediate layers. As a new type of polar solvent, ionic liquids have almost no vapor pressure, are non-flammable, non-volatile, have good chemical and thermal stability, can be recycled and are environmentally friendly. Therefore, they are called "green" chemical solvents and can replace traditional volatile and toxic solvents. In addition, due to the unique properties of ionic liquids, it is widely used in memristors. Early research focused on using ionic liquids as electrolyte media to achieve the resistance switching function of memristors. As the research deepened, researchers began to explore the potential of ionic liquids in constructing artificial optoelectronic synapses. A memristor doped with an organic polymer and ionic liquid has been developed to realize low-voltage artificial photoelectric synapses. This memristor obtains a rich conductance state through light field regulation at millivolt voltages, and exhibits short-term and long-term synaptic plasticity functions, including EPSC, PPF, PPD, LTP, LTD, etc., as well as learning-forgetting-relearning characteristics. [1] DEME-TFSI ionic liquid has some unique properties. It can provide high ionic conductivity, which is very important for fast ion migration in electronic devices such as memristors. DEME-TFSI ionic liquid has a wide electrochemical window, which allows it to remain stable over a wide potential range, helping to improve the reliability and performance of electronic devices. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art and to propose an ionic liquid DEME-TFSI memristor with biological synapse simulation function and a preparation method thereof.

[0005] The technical solution of the present invention is as follows:

[0006] The first aspect of the present invention provides an ionic liquid DEME-TFSI memristor having a function of simulating a biological synapse, characterized in that it comprises a bottom electrode, an ionic liquid layer, and a top electrode layer;

[0007] Wherein, the bottom electrode is arranged at the bottom;

[0008] The ionic liquid layer is disposed above the bottom electrode;

[0009] The top electrode is disposed above the ionic liquid layer.

[0010] Preferably, the bottom electrode is one of ITO conductive glass, fluorine-doped tin dioxide (FTO) conductive substrate or aluminum-doped zinc oxide (AZO) conductive substrate;

[0011] More preferably, the bottom electrode is a fluorine-doped tin oxide (FTO) conductive substrate.

[0012] The electrode layer is one of active electrodes silver, copper, aluminum, zinc, magnesium and iron;

[0013] More preferably, the electrode layer is active electrode aluminum;

[0014] Preferably, the thickness of the electrode is 150nm to 350nm, more preferably 150nm to 300nm, and further preferably 150nm to 200nm;

[0015] The thickness of the ionic liquid layer is 50 nm to 300 nm, preferably 100 nm to 250 nm, more preferably 150 nm to 200 nm;

[0016] The ionic liquid layer is located above the bottom electrode;

[0017] The second aspect of the present invention provides a method for preparing the ionic liquid DEME-TFSI memristor described in the first aspect, the method comprising the following steps:

[0018] Step 1: Use clean water + detergent, deionized water, acetone, and alcohol to clean the conductive substrate with an ultrasonic instrument in sequence, put it in a drying oven to dry it after cleaning, and then ozone treat the conductive substrate before spin coating to obtain a bottom electrode;

[0019] Step 2: spin coating DEME-TFSI on the bottom electrode and heat treating to form an ionic liquid layer;

[0020] Step 3: growing a top electrode on the ionic liquid layer by vacuum thermal evaporation to obtain the ionic liquid memristor.

[0021] According to the method of the second aspect of the present invention, the step (1) further comprises the following steps:

[0022] Use water + detergent, deionized water, acetone, alcohol and other solvents to ultrasonically clean the conductive base of the fluorine-doped tin dioxide (FTO) conductive substrate in sequence, repeat each solvent once, dry it in a drying oven after the ultrasonic treatment, and perform ozone operation on the fluorine-doped tin dioxide (FTO) before spin coating;

[0023] Preferably, the ultrasonic cleaning time is 25 min-100 min, more preferably 25 min-60 min, further preferably 25 min-40 min;

[0024] Preferably, the drying temperature is 55°C-100°C, more preferably 55°C-75°C, further preferably 60°C.

[0025] Preferably, the drying time is 4h-10h, more preferably 4h-8h, further preferably 4h-6h;

[0026] Preferably, the ozone time is 25 min-60 min, more preferably 25 min-45 min, further preferably 25 min-35 min.

[0027] According to the method of the second aspect of the present invention, the step (2) further comprises the following steps:

[0028] The ionic liquid is drop-coated on the fluorine-doped tin dioxide (FTO) after ozone treatment, and after the drop-coating is completed, the ionic liquid layer is obtained by annealing with a rapid annealing apparatus;

[0029] Preferably, the drop coating volume is 150 μL-300 μL, more preferably 150 μL-250 μL, and further preferably 150 μL-200 μL;

[0030] Preferably, the dripping speed is 2500rpm-4000rpm, more preferably 2500rpm-3500rpm. Most preferably, it is 3000rpm, and the dripping time is 30min;

[0031] Preferably, the baking temperature is 150°C to 400°C, more preferably 150°C to 300°C, and further preferably 180°C to 250°C;

[0032] Preferably, the baking time is 5 min to 20 min, more preferably 5 min to 15 min, and most preferably 10 min.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] The present invention utilizes the uniqueness of ionic liquids, and through different operation modes, tests the IV cycle characteristics, long-term potentiation / depression (LTP / LTD), short-term potentiation (PPF), pulse timing-dependent plasticity (STDP) of the device, etc., which show that the device has good repetitive resistance switching conversion behavior and potential application in the nervous system. The handwritten digit recognition analysis based on artificial neural network was done, and after 200 trainings, the recognition accuracy can reach 89%. The results show that Al / DEME-TFSI / FTO devices have application potential as high-performance non-volatile memory in future computer systems and neuromorphic computing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings generally illustrate various embodiments by way of example and not limitation, and together with the description and claims, serve to illustrate the embodiments of the invention. Where appropriate, the same reference numerals are used throughout the drawings to refer to the same or similar parts. Such embodiments are illustrative and are not intended to be exhaustive or exclusive embodiments of the present apparatus or method.

[0036] Figure 1 : Device structure diagram;

[0037] Figure 2 (a): Single-cycle IV characteristics of the device;

[0038] Figure 2 (b): High and low configuration distribution of the device under 100 cycles of IV cycling characteristics read at 0.2 V voltage;

[0039] Figure 3 :Conductance modulation illustration;

[0040] Figure 4 : Double pulse allotropy PPF;

[0041] Figure 5 : Long-term potentiation LTP and long-term depression LTD;

[0042] Figure 6 : spike timing dependent plasticity STDP;

[0043] Figure 7 : Artificial neural network models and accuracy. DETAILED DESCRIPTION

[0044] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0045] Example 1

[0046] This example is used to illustrate the ionic liquid DEME-TFSI memristor with biological synapse simulation function and the preparation method of the present invention.

[0047] Figure 1 The schematic diagram of the structure of the ionic liquid DEME-TFSI memristor in Example 1 of the present invention is shown. The ionic liquid layer is the middle layer 2, and the ionic liquid layer is grown on the bottom electrode 1. The ionic liquid layer is a transparent material containing water molecules. The hydrogen ions decomposed under the negative gate voltage and its own cation DEME+ will accumulate at the junction of the top electrode layer 3 and the ionic liquid layer 2, enhancing the Schottky barrier at the junction, so that the device undergoes configuration switching. 4 is two probes of the Keithley B1500A semiconductor parameter analyzer, one probe is pressed on the top electrode layer 3, and the other probe is pressed on the bottom electrode 1. By applying voltages of different parameters to the probes on the top electrode layer 3 to change the group resistance state transformation during the period, the function of the device simulating biological synapses is realized.

[0048] In this embodiment, a fluorine-doped tin dioxide (FTO) conductive substrate is used as the bottom electrode and aluminum is used as the electrode layer. The preparation method of the ionic liquid DEME-TFSI memristor with biological synapse simulation function provided by the present invention is carried out.

[0049] Detailed description:

[0050] The specific steps of making the ionic liquid DEME-TFSI memristor of the present invention are as follows:

[0051] 1. Cleaning of fluorine-doped tin oxide (FTO) conductive substrate

[0052] Step 1: Place the FTO device to be cleaned on the cleaning rack and into a beaker, and prepare the rubber ring and plastic film.

[0053] Step 2: Add a small amount of detergent and clean water that has covered the FTO into the beaker with the FTO, and place it in an ultrasonic instrument for ultrasonic cleaning for 30 minutes.

[0054] Step 3: Pour all the liquid in the beaker after cleaning in step 2 into the waste liquid bucket, then add deionized water to cover the FTO and put it into an ultrasonic instrument for ultrasonic cleaning for 30 minutes.

[0055] Step 4: Pour all the liquid in the beaker after cleaning in step 3 into the waste liquid bucket, then add acetone to submerge the FTO and put it into an ultrasonic instrument for ultrasonic cleaning for 30 minutes.

[0056] Step 5: Pour all the liquid in the beaker after cleaning in step 4 into the waste liquid bucket, then add alcohol to cover the FTO and put it into an ultrasonic instrument for ultrasonic cleaning for 30 minutes, then put it in a 60°C drying oven for 4 hours.

[0057] 2. Preparation of ionic liquid DEME-TFSI liquid layer

[0058] Step 1: Test the conductivity of FTO with a multimeter and select the conductive side.

[0059] Step 2: Apply conductive adhesive to the edge of the FTO with conductive surface and treat with UV ozone for 30 minutes.

[0060] Step 3: Place the ozone-treated FTO on the drop coating instrument.

[0061] Step 4: Use a pipette to draw 200uL of ionic liquid DEME-TFSI and drop it on the FTO in the previous step.

[0062] Step 5: Set the coating speed to 3000 rpm.

[0063] Step 6: Set the dripping time to 30 seconds.

[0064] Step 7: Place the FTO after drop coating into a rapid annealing furnace.

[0065] Step 8: Evacuate the growth chamber to 1×10-6 Torr.

[0066] Step 9: Raise the temperature of the growth chamber to 200°C.

[0067] Step 10: The annealing time is 10 minutes to form an ionic liquid layer.

[0068] 3. Growth of electrode layer

[0069] Step 1: Remove the conductive glue under the ionic liquid layer.

[0070] Step 2: Place a clean mask on top of the ionic liquid layer.

[0071] Step 3: Use new conductive adhesive to stick on top of the mask in the previous step.

[0072] Step 4: Place the wafer from the previous step into the evaporation apparatus to evaporate aluminum for the electrode layer.

[0073] Step 5: Remove the conductive glue and mask from the device that has been evaporated in the previous step.

[0074] The ionic liquid DEME-TFSI memristor is fabricated.

[0075] Example 2

[0076] This example is used to illustrate the IV cycle characteristics of the device and the stability of the device.

[0077] Figure 2 The single-cycle IV characteristic curve of the device prepared in Example 1 and the high and low resistance distribution diagram of 100 cycles read at a voltage of 0.2V. During the test, the FTO was grounded and a voltage was applied to the aluminum electrode with a step size of 0.025V. Figure 2 In (a), "1" refers to the curve of the 0-2V scanning process, "2" refers to the curve of the -2V-0V scanning process, "3" refers to the curve of the 0V-2.5V scanning process, with a current limit of 1mA, and "4" refers to the curve of the 2.5V-0V scanning process. In curves 1 to 4, "1" and "2" are the process of the memristor switching from a low resistance state to a high resistance state, that is, the process of erasing information. "3" and "4" are the process of the memristor switching from a high resistance state to a low resistance state, that is, the process of writing information. Figure 2 (a) shows that the Al / DEME-TFSI / FTO structured memristor prepared in Example 1 has typical non-volatile resistance switching characteristics. Figure 2 (b) is a durability test of the device prepared in Example 1. The test results show that the device has good stability under the 100-cycle IV cycle characteristic curve, and the switching ratio ON / OFF is about 1.5 orders of magnitude.

[0078] Example 3

[0079] This example is used to illustrate that the device conductance is continuously adjustable, confirming its potential to simulate biological synapses.

[0080] Figure 3 The figure is a diagram of the continuous regulation of the conductance of the device prepared in Example 1. The device was subjected to continuous negative bias sweep, the stop voltage amplitude increased from -1V to -2.4V, and then continuous positive voltage sweep was performed, and the Icc value increased from 0.5mA to 10mA. It can be seen from the figure that the conductance of the device is continuously adjustable, showing good conductance modulation characteristics, indicating that the device prepared in Example 1 has the potential to simulate biological synapses.

[0081] Example 4

[0082] This embodiment is used to illustrate that the device has the ability to simulate biological synaptic behavior.

[0083] Different pulse voltages were applied to both ends of the Al / DEME-TFSI / FTO device prepared in Example 1, and the functions of biological synapses such as PPF, LTP / LTD and STDP were successfully simulated.

[0084] Figure 4This is a PPF simulation of the device prepared in Example 1. Double pulse heterogeneity describes the enhancement of the excitatory synaptic current (EPSC) of the first pulse by the second pulse when two consecutive pulses are successfully applied to the presynaptic membrane (top electrode). The value of EPSC is closely related to the time interval between the two pulses applied to the presynaptic membrane. The shorter the time interval between the two pulses, the more obvious this enhancement effect is. Figure 4 Exhibits PPF properties similar to those in biology.

[0085] Figure 5 This is an LTP / LTD simulation of the device prepared in Example 1. By applying continuous positive pulses and continuous negative pulses to the device, the device exhibits good LTP / LTD characteristics, which are similar to the long-term memory of the human brain.

[0086] Figure 6 This is a STDP simulation of the device prepared in Example 1. By adopting the time division multiplexing (TDM) method to design the spike-shaped pulse, the STDP characteristics are successfully exhibited during the period, indicating that the device prepared in Example 1 can simulate the STDP learning rule.

[0087] Example 5

[0088] This embodiment uses the results of LTP / LTD to construct a three-layer artificial neural network and combines the algorithm to perform handwritten digit recognition.

[0089] This example uses the LTP / LTD results of the device prepared in Example 1 to construct an artificial neural network, and uses the algorithm to simulate, and the results are as follows: Figure 7 As shown in the figure, it can be seen that as the number of training times increases, the blurriness of the picture decreases and the recognition accuracy gradually increases. After 200 training times, the recognition accuracy reaches 89%. This result proves that Al / DEME-TFSI / FTO devices have potential application prospects in the field of neuromorphic computing.

[0090] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An ionic liquid DEME-TFSI memristor with biological synapse simulation function, characterized in that: comprising a bottom electrode, an ionic liquid layer, and a top electrode layer; Wherein, the bottom electrode is arranged at the bottom; The ionic liquid layer is disposed above the bottom electrode; The top electrode is disposed above the ionic liquid layer.

2. The ionic liquid DEME-TFSI memristor according to claim 1, characterized in that: The bottom electrode is one of ITO conductive glass, fluorine-doped tin dioxide conductive substrate or aluminum-doped zinc oxide conductive substrate.

3. The ionic liquid DEME-TFSI memristor according to claim 1, characterized in that: The thickness of the ionic liquid layer is 50 nm to 300 nm.

4. The ionic liquid DEME-TFSI memristor according to claim 1, characterized in that: The material of the ionic liquid layer is DEME-TFSI.

5. The ionic liquid DEME-TFSI memristor according to claim 1, characterized in that: The thickness of the top electrode layer is 150nm-350nm.

6. The ionic liquid DEME-TFSI memristor according to claim 1, characterized in that: The top electrode material is one of silver, copper, aluminum, zinc, magnesium and iron.

7. A method for preparing an ionic liquid DEME-TFSI memristor having a biological synapse-simulating function, characterized in that: The method comprises the following steps: Step 1: Use clean water + detergent, deionized water, acetone, and alcohol to clean the conductive substrate with an ultrasonic instrument in sequence, put it in a drying oven to dry it after cleaning, and then ozone treat the conductive substrate before spin coating to obtain a bottom electrode; Step 2: spin coating DEME-TFSI on the bottom electrode and heat treating to form an ionic liquid layer; Step 3: growing a top electrode on the ionic liquid layer by vacuum thermal evaporation to obtain the ionic liquid memristor.

8. The method for preparing the ionic liquid DEME-TFSI memristor having a biological synapse simulating function according to claim 7, characterized in that: Each step of cleaning the conductive substrate in step 1 takes 30 minutes; The duration of drying the conductive substrate in step 1 is 4 to 8 hours; The duration of ozone treatment of the conductive substrate in step 1 is 30 minutes.

9. The method for preparing the ionic liquid DEME-TFSI memristor having a biological synapse simulating function according to claim 7, characterized in that: In step 2, DEME-TFSI is spin-coated in a volume of 150 μL to 300 μL; In the step 2, DEME-TFSI is spin-coated at a speed of 2500 rpm to 4000 rpm; In step 2, DEME-TFSI is spin-coated for 30 min; The temperature of the heat treatment in step 2 is 150°C to 400°C; The heat treatment time in step 2 is 5 min to 20 min.

10. The method for preparing the ionic liquid DEME-TFSI memristor having a biological synapse simulating function according to claim 7, characterized in that: In the step 3, a top electrode is grown, and the thickness of the top electrode is 150nm to 350nm.

Images