Method for predicting mechanical fatigue life of flexible thin film transistor based on low-frequency noise

Abstract

The invention relates to the technical field of semiconductors, and discloses a low-frequency noise-based flexible thin film transistor mechanical fatigue life prediction method, which comprises the following steps of: obtaining a to-be-predicted flexible thin film transistor; calculating a transfer characteristic curve and a low-frequency noise power spectrum of the to-be-predicted flexible thin film transistor; acquiring first data of the to-be-predicted flexible thin film transistor based on a transfer characteristic curve, calculating second data of the to-be-predicted flexible thin film transistor based on the first data and the low-frequency noise power spectrum, and inputting the mechanical stress intensity, the cycle number and the second data into a trained life prediction model, outputting a curve graph of the flat-band voltage power spectral density degradation amount and the contact noise power spectral density degradation amount changing along with the cycle; and obtaining the fatigue life of the to-be-predicted flexible thin film transistor under a specific mechanical stress intensity condition according to the curve graph. According to the invention, the accuracy and reliability of prediction can be improved, and the time and cost required for carrying out a complete fatigue life test on the device can be reduced.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a method for predicting the mechanical fatigue life of a flexible thin film transistor based on low-frequency noise. Background Art

[0002] New display is one of the key development directions of strategic emerging industries in the era of national informatization and intelligence. As an important development goal in the field of new display, high-resolution flexible display has long been the focus of many leading domestic companies. Among them, flexible thin film transistor (Flexible Thin Film Transistor, referred to as FTFT) is a special form of thin film transistor (TFT), and it is also the basic unit for realizing flexible integrated circuit switch control, so it has become a key driving component for flexible display. In practical applications, flexible TFT usually needs to work under complex mechanical stress; as the working time increases, the above mechanical stress will gradually cause microscopic defects and damage in the gate oxide layer, active layer, source and drain contact area and various interfaces of the flexible TFT, thereby causing the device threshold voltage, transconductance and mobility. Degradation of electrical parameters, and then causing the failure of the flexible display system. Therefore, in order to ensure the long-term stable operation of flexible display, it is necessary to determine the service life of flexible TFT under different mechanical stresses and evaluate its reliability. In the prior art, the following methods are usually used to determine the service life of flexible TFTs: 1. A mechanical fatigue tester is used to apply cyclic mechanical stress to the flexible device, and the degradation of the device's electrical properties before and after the fatigue test is observed to evaluate the device's mechanical fatigue life; however, the time period required for a complete mechanical fatigue test is long, which will consume a lot of testing time and expenses, resulting in low quality inspection efficiency and high quality inspection costs for flexible TFTs; 2. Since the physical characterization technology of the microstructure in the device (focused ion beam analysis, transmission electron microscopy, etc.) requires a long time and high costs, the mechanical fatigue life of the device is usually evaluated based on the degradation of the macroscopic electrical parameters of the flexible TFT, but the macroscopic electrical parameters cannot intuitively reflect the changes in the microscopic defects and potential damage inside the device, thereby restricting the accuracy and credibility of the service life calculation results; 3. The fatigue life of the flexible device is also predicted and estimated by establishing a linear relationship between mechanical stress and fatigue life (stress-life relationship curve); in actual production life, there may be a nonlinear relationship between mechanical stress and the fatigue life of the flexible TFT, which will lead to inaccurate prediction results of the stress-life relationship curve. Summary of the invention

[0003] The purpose of the present invention is to overcome the problems existing in the prior art and provide a method for predicting the mechanical fatigue life of a flexible thin-film transistor based on low-frequency noise. The present invention can improve the accuracy and reliability of the prediction and reduce the time and cost required for a complete fatigue life test of the device.

[0004] In order to achieve the above object, the present invention provides a method for predicting the mechanical fatigue life of a flexible thin film transistor based on low-frequency noise, comprising: Obtain a flexible thin film transistor to be predicted; Calculating the transfer characteristic curve and low-frequency noise power spectrum of the flexible thin film transistor to be predicted; Acquire first data of the flexible thin film transistor to be predicted based on the transfer characteristic curve, wherein the first data includes at least one of a threshold voltage, a transconductance and a field effect mobility; Calculate second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum, wherein the second data includes at least one of an initial value of a flat band voltage power spectrum density and an initial value of a contact noise power spectrum density; Inputting the mechanical stress intensity, the number of cycles and the second data into the trained life prediction model, and outputting a curve graph showing the variation of the flat band voltage power spectrum density degradation amount and the contact noise power spectrum density degradation amount with the number of cycles; The number of cycles during which the degradation amount of the flat band voltage power spectrum density or the degradation amount of the contact noise power spectrum density exceeds the degradation amount threshold is recorded, and the number of cycles is used as the fatigue life of the flexible thin film transistor to be predicted under the mechanical stress intensity condition.

[0005] Furthermore, the flexible thin film transistor to be predicted is one or more combinations of an amorphous silicon thin film transistor, a low-temperature polysilicon thin film transistor, an organic thin film transistor and an oxide thin film transistor.

[0006] Furthermore, the flicker noise is taken as a noise source to calculate the low-frequency noise power spectrum.

[0007] Further, the calculating the second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum includes: (1) Using a noise test system to measure the flicker noise data of the flexible thin film transistor to be predicted; (2) Based on the flicker noise data, separating and constructing a flicker noise model of the flexible thin film transistor to be predicted, the flicker noise model including a first flicker noise model of drain current and a second noise model of contact resistance; (3) Calculating second data of the flexible thin film transistor to be predicted based on the flicker noise model and the first data.

[0008] Furthermore, the formula for calculating the initial value of the flat band voltage power spectrum density is as follows:

[0009] in, is a flicker noise power spectrum density based on the drain current of the flexible thin film transistor to be predicted; is transconductance; is the Coulomb scattering coefficient; is the gate oxide capacitance per unit area; The gate bias voltage applied when calculating the flicker noise power spectral density of the flexible thin film transistor to be predicted; is the threshold voltage; is the initial value of the flat-band voltage power spectrum density; The formula for calculating the initial value of the contact noise power spectrum density is as follows:

[0010] in, is the drain current density, is the initial value of contact noise power spectral density.

[0011] Furthermore, the mechanical stress intensity includes tensile stress intensity and bending stress intensity, wherein the tensile stress intensity is represented by the magnitude of the pulling force, and the bending stress intensity is represented by the magnitude of the bending radius.

[0012] Furthermore, the life prediction model is based on a back propagation neural network, and the life prediction model specifically includes an input layer, a first hidden layer, a second hidden layer and an output layer connected in sequence. The number of nodes in the first hidden layer and the second hidden layer is determined by the following formula:

[0013] In the above formula, is the number of nodes in the input layer, is the number of nodes in the output layer, is the number of nodes in the hidden layer.

[0014] Furthermore, training the life prediction model specifically includes: (1) Obtain flexible thin film transistor samples; (2) Calculating the first transfer characteristic curve and the first low-frequency noise power spectrum of the flexible thin film transistor sample; (3) Calculate the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density based on the first transfer characteristic curve and the first low-frequency noise power spectrum; (4) Using a uniform orthogonal experimental method to group the flexible thin film transistor samples, and establishing mechanical stress intensity groups and cycle groups based on the grouping results; (5) Apply corresponding mechanical stress and cycle number to each group of flexible thin film transistor samples after grouping to conduct fatigue life test; (6) Measure and record the second low-frequency noise power spectrum of each group of flexible thin film transistor samples after the test, and calculate the degradation amount of the first flat band voltage power spectrum density and the first contact noise power spectrum density of each group of flexible thin film transistor samples based on the second low-frequency noise power spectrum, the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density; (7) Constructing a training set, the training set includes mechanical stress intensity, cycle number, the initial value of the first flat band voltage power spectrum density, the initial value of the first contact noise power spectrum density, the first flat band voltage power spectrum density degradation amount and the first contact noise power spectrum density degradation amount; (8) Iteratively train the life prediction model using the training set to obtain a trained life prediction model.

[0015] Furthermore, before using the training set to iteratively train the life prediction model, the training set is normalized and / or differentiated.

[0016] Furthermore, the degradation threshold is , that is, the flat-band voltage power spectrum density degradation exceeds the initial value of the flat-band voltage power spectrum density or the contact noise power spectrum density degradation exceeds the contact noise power spectrum density initial value The number of cycles is taken as the fatigue life of the flexible thin film transistor to be predicted under the mechanical stress intensity conditions.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention uses a trained life prediction model to predict the fatigue life of a flexible thin film transistor under given mechanical stress conditions, thereby effectively shortening the time and cost required for a complete fatigue life test of the flexible thin film transistor; 2. The present invention uses low-frequency noise characteristics that are more sensitive to microscopic defects and damages in the device as information parameters to predict the fatigue life of the device under mechanical stress, thereby improving the accuracy and credibility of the prediction; 3. Use a neural network model with strong nonlinear mapping capabilities and high accuracy to replace the traditional stress-life relationship model, effectively improving the accuracy and reliability of the model prediction results and reducing the computational workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a method for predicting mechanical fatigue life of a flexible thin film transistor based on low-frequency noise according to Embodiment 1 of the present invention; Figure 2 is a structural diagram of a metal oxide thin film transistor according to Example 1 of the present invention; Figure 3 is a structural schematic diagram of a noise power spectral density testing system according to Embodiment 1 of the present invention; Figure 4 is a schematic diagram of the structure of the life prediction model of Example 1 of the present invention; Figure 5 This is a training and prediction flow chart of the life prediction model of Example 1 of the present invention. DETAILED DESCRIPTION

[0019] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0020] For flexible TFTs, the industry often applies cyclic mechanical stress to them and observes the degradation of the device's electrical performance before and after fatigue testing to evaluate the device's mechanical fatigue life. However, the test requires a long period of time and will consume a lot of experimental time and costs. Secondly, the industry evaluates the mechanical fatigue life of the device based on the degradation of the macroscopic electrical parameters of the flexible TFT, but the macroscopic electrical parameters cannot directly reflect the changes in the microscopic defects and potential damage inside the device, thus restricting the accuracy and reliability of the service life calculation results. Finally, the industry predicts the fatigue life of flexible TFTs by establishing a stress-life relationship curve model, but the nonlinear relationship between mechanical stress and fatigue life will lead to inaccurate prediction results of the model. Based on this, the embodiment of the present invention is based on the low-frequency noise characteristics that are highly sensitive to the microscopic damage of the device, and uses a neural network model with strong nonlinear mapping capabilities and high accuracy to predict the number of failure cycles of the device under different mechanical stresses, which not only reduces the experimental time and cost, but also effectively improves the accuracy and reliability of the fatigue life prediction model.

[0021] Example 1 like Figure 1 As shown, a method for predicting mechanical fatigue life of a flexible thin film transistor based on low-frequency noise in a preferred embodiment of an embodiment of the present invention includes: S1: Obtain the flexible thin film transistor to be predicted; In a feasible embodiment, the flexible thin film transistor to be predicted includes one or more combinations of amorphous silicon thin film transistors, low temperature polycrystalline silicon thin film transistors, organic thin film transistors and oxide thin film transistors. In this embodiment, a metal oxide thin film transistor is selected as the flexible thin film transistor to be predicted. The structure of the metal oxide thin film transistor is shown in FIG. Figure 2, mainly including: gate, gate oxide layer, source / drain, etc.

[0022] S2: Calculate the transfer characteristic curve and low-frequency noise power spectrum of the flexible thin film transistor to be predicted; In a feasible embodiment, obtaining the transfer characteristic curve of the flexible thin film transistor to be predicted requires setting test conditions, which are as follows: Applied gate bias The range is: ; Applied source-drain voltage for: ; In this embodiment, the flicker noise ( noise) as the noise source to calculate the low-frequency noise power spectrum and obtain the The test conditions of the noise power spectrum are as follows: Applied gate bias The range is: ; Applied source-drain voltage for: ; It should be noted that TFT The noise amplitude is usually weak, so it is necessary to use a high-precision test device to test it and amplify the collected data. The test device structure of the noise power spectrum is as follows: Figure 3 As shown in the figure, it specifically includes: low noise amplifier, spectrum analyzer, analog-to-digital converter (ADC) and semiconductor parameter tester. Among them, the source test unit of the semiconductor parameter tester can provide accurate test current or test voltage for the device under test; then, the current noise is amplified by the low noise amplifier and sent to the ADC for collection; finally, the collected data is passed to the spectrum analyzer and fast Fourier transform (FFT) is performed to calculate Noise power spectrum. The noise test speed of this test structure can reach 20s / bias, and the minimum test noise current accuracy is as low as 10-27A2 / Hz, which can ensure Noise test frequency range and test accuracy; It should be noted that the above measurements are intended to predict the performance of flexible thin film transistors. Noise power spectrum, the applied gate bias With threshold voltage The difference ( - ) is much larger than the source-drain voltage , to ensure that the device works in the linear region; due to The noise is more significant in the low frequency band, so the sampling frequency is set to 10Hz.

[0023] S3: acquiring first data of the flexible thin film transistor to be predicted based on the transfer characteristic curve, the first data including at least one of a threshold voltage, a transconductance and a field effect mobility; In a feasible embodiment, the threshold voltage of the flexible thin film transistor to be predicted is obtained by obtaining the transfer characteristic curve according to step S2. , transconductance Field Effect Mobility .

[0024] S4: calculating second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum, the second data including at least one of an initial value of a flat band voltage power spectrum density and an initial value of a contact noise power spectrum density; In a feasible embodiment, calculating the second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum includes: (1) Using a noise test system to measure the flicker noise data of the flexible thin film transistor to be predicted, the test frequency range is 1 Hz ~ 10,000 Hz; (2) Based on the flicker noise data, separating and constructing a flicker noise model of the flexible thin film transistor to be predicted, the flicker noise model including a first flicker noise model of drain current and a second noise model of contact resistance; (3) Calculating second data of the flexible thin film transistor to be predicted based on the flicker noise model and the first data.

[0025] Furthermore, the formula for calculating the initial value of the flat band voltage power spectrum density is as follows:

[0026] in, is a flicker noise power spectrum density based on the drain current of the flexible thin film transistor to be predicted; is transconductance; is the Coulomb scattering coefficient; is the gate oxide capacitance per unit area; The gate bias voltage applied when calculating the flicker noise power spectral density of the flexible thin film transistor to be predicted; is the threshold voltage; is the initial value of the flat-band voltage power spectrum density; The formula for calculating the initial value of the contact noise power spectrum density is as follows:

[0027] in, is the drain current density, is the initial value of contact noise power spectral density.

[0028] S5: input the mechanical stress intensity, the number of cycles and the second data into the trained life prediction model, and output a curve diagram showing the variation of the flat band voltage power spectrum density degradation and the contact noise power spectrum density degradation with the number of cycles; In a feasible embodiment, the mechanical stress intensity includes tensile stress intensity and bending stress intensity, wherein the tensile stress intensity is represented by the magnitude of the pulling force, and the bending stress intensity is represented by the magnitude of the bending radius.

[0029] Furthermore, the life prediction model is based on a back propagation neural network. The life prediction model specifically includes an input layer, a first hidden layer, a second hidden layer and an output layer connected in sequence. The double hidden layer design can enhance the fitting accuracy and robustness of the neural network prediction model to nonlinear data. The life prediction model is shown in Figure 4 , The number of nodes in the first hidden layer and the second hidden layer is determined by the following formula:

[0030] In the above formula, is the number of nodes in the input layer, is the number of nodes in the output layer, is the number of nodes in the hidden layer. Based on the empirical formula, the number of input layer nodes in this embodiment is 4, and the number of output layer nodes is 2. Then the initial number of nodes in each hidden layer is is 3, and it will be adjusted in the subsequent neural network training process. The ReLU function is selected as the hidden layer activation function, and its expression is:

[0031] Using the identity function as the output layer activation function, its expression is:

[0032] Further, the process of training the life prediction model is shown in Figure 5 , specifically including: (1) Obtain flexible thin film transistor samples; In this embodiment, the materials, process parameters and sizes of all flexible thin film transistor samples are completely consistent; n random sub-samples are drawn from the sampling matrix according to the simple random sampling principle, where n ≥ 20; (2) Calculating the first transfer characteristic curve and the first low-frequency noise power spectrum of the flexible thin film transistor sample; It can be understood that the implementation process of this step is consistent with S2 and will not be repeated here.

[0033] (3) Calculate the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density based on the first transfer characteristic curve and the first low-frequency noise power spectrum; It can be understood that the implementation process of this step is consistent with S4 and will not be repeated here.

[0034] (4) Using a uniform orthogonal experimental method to group the flexible thin film transistor samples, and establishing mechanical stress intensity groups and cycle groups based on the grouping results; Specifically, mechanical stress intensity groups were established according to the grouping results ( F 1, F 2,……, F i) and cycle group ( N 1, N 2,……, N j).

[0035] (5) Apply corresponding mechanical stress and cycle number to each group of flexible thin film transistor samples after grouping to conduct fatigue life test; Specifically, the design of the fatigue life test is as follows: Set the type and intensity of mechanical stress, and take 4 levels for each stress intensity; The 20 samples were randomly divided into 4 groups, each containing 5 devices, corresponding to a set of stress levels; The mechanical stress is applied to the device by using a fatigue life test device to carry out a mechanical stress fatigue life test.

[0036] (6) Measure and record the second low-frequency noise power spectrum of each group of flexible thin film transistor samples after the test, and calculate the degradation amount of the first flat band voltage power spectrum density and the first contact noise power spectrum density of each group of flexible thin film transistor samples based on the second low-frequency noise power spectrum, the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density; Specifically, the first flat band voltage power spectrum density degradation is calculated The formula is as follows:

[0037] in, For flexible thin film transistor samples F Mechanical stress strength of i, N The first flat-band voltage power spectrum density after the j cycle experiment is: is the initial value of the first flat band voltage power spectrum density; Calculate the first contact noise power spectral density degradation The formula is as follows:

[0038] in, For flexible thin film transistor samples F Mechanical stress strength of i, N The degradation of the first contact noise power spectrum density after the j-cycle experiment, is the initial value of the first contact noise power spectral density.

[0039] (7) Constructing a training set, the training set includes mechanical stress intensity, cycle number, the initial value of the first flat band voltage power spectrum density, the initial value of the first contact noise power spectrum density, the first flat band voltage power spectrum density degradation amount and the first contact noise power spectrum density degradation amount; Specifically, the training set is normalized and / or differentiated to eliminate adverse effects caused by singular sample data and obtain stable training data.

[0040] (8) Iteratively train the life prediction model using the processed training set to obtain a trained life prediction model.

[0041] Specifically, the iterative training includes: initializing the weight value of each node in the hidden layer of the network; inputting the input signal in the training data set into the neural network model, and calculating the model predicted output value; comparing the model predicted output value with the expected output value in the training data set, and calculating the prediction error; iteratively updating the weight value of each node in the hidden layer of the neural network according to the prediction error; when the prediction error meets the requirements, the iteration ends and the weight value of each node of the neural network is determined.

[0042] S6: recording the number of cycles during which the degradation amount of the flat band voltage power spectrum density or the degradation amount of the contact noise power spectrum density exceeds the degradation amount threshold, and using the number of cycles as the fatigue life of the flexible thin film transistor to be predicted under specific mechanical stress intensity conditions.

[0043] In a feasible embodiment, the degradation threshold is , that is, the flat-band voltage power spectrum density degradation exceeds the initial value of the flat-band voltage power spectrum density or the contact noise power spectrum density degradation exceeds the contact noise power spectrum density initial value The number of cycles is taken as the fatigue life of the flexible thin film transistor to be predicted under the mechanical stress intensity conditions.

[0044] In this embodiment, the degradation threshold is .

[0045] Example 2 An embodiment of the present invention also provides a computer-readable storage medium, on which is stored a computer program for a method for predicting the mechanical fatigue life of a flexible thin-film transistor based on low-frequency noise. When the computer program for the method for predicting the mechanical fatigue life of a flexible thin-film transistor based on low-frequency noise is processed, the steps of the method for predicting the mechanical fatigue life of a flexible thin-film transistor based on low-frequency noise are implemented.

[0046] In summary, an embodiment of the present invention provides a method for predicting the mechanical fatigue life of a flexible thin-film transistor based on low-frequency noise, which calculates the initial values ​​of the flat-band voltage power spectrum density and the contact noise power spectrum density based on the 1 / f noise power spectrum of the flexible TFT; performs a mechanical stress fatigue life test on the flexible TFT under mechanical stresses of different types and intensities; obtains the flat-band voltage power spectrum density and the contact noise power spectrum density of the flexible TFT after a certain number of cycles, and calculates the degradation amount relative to the initial value; uses the mechanical stress intensity, the number of cycles, the initial values ​​of the flat-band voltage power spectrum density and the contact noise power spectrum density of the flexible TFT as input signals, and uses the degradation amount of the flat-band voltage power spectrum density and the contact noise power spectrum density as output signals to establish a neural network model as a fatigue life prediction model for the device; uses the fatigue life prediction model to quickly, accurately and non-destructively predict and evaluate the mechanical fatigue life of flexible TFTs of the same batch. Compared with the prior art, the present invention uses a trained life prediction model to predict the fatigue life of flexible thin-film transistors under given mechanical stress conditions, thereby effectively shortening the time and cost required for a complete fatigue life test on flexible thin-film transistors; it also uses low-frequency noise characteristics that are more sensitive to microscopic defects and damage in the device as information parameters to predict the fatigue life of the device under mechanical stress, thereby improving the accuracy and credibility of the prediction; it also uses a neural network model with strong nonlinear mapping capabilities and high accuracy to replace the traditional stress-life relationship model, thereby effectively improving the accuracy and reliability of the model prediction results and reducing the computational workload.

[0047] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and substitutions can be made without departing from the technical principles of the present invention. These improvements and substitutions should also be regarded as the scope of protection of the present invention.

Claims

1. A method for predicting the mechanical fatigue life of a flexible thin film transistor based on low-frequency noise, characterized in that: The method comprises the following steps: Obtain a flexible thin film transistor to be predicted; Calculating the transfer characteristic curve and low-frequency noise power spectrum of the flexible thin film transistor to be predicted; Acquire first data of the flexible thin film transistor to be predicted based on the transfer characteristic curve, wherein the first data includes at least one of a threshold voltage, a transconductance and a field effect mobility; Calculate second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum, wherein the second data includes at least one of an initial value of a flat band voltage power spectrum density and an initial value of a contact noise power spectrum density; Inputting the mechanical stress intensity, the number of cycles and the second data into the trained life prediction model, and outputting a curve graph showing the variation of the flat band voltage power spectrum density degradation amount and the contact noise power spectrum density degradation amount with the number of cycles; The number of cycles during which the degradation amount of the flat band voltage power spectrum density or the degradation amount of the contact noise power spectrum density exceeds the degradation amount threshold is recorded, and the number of cycles is used as the fatigue life of the flexible thin film transistor to be predicted under the mechanical stress intensity condition.

2. The method for predicting mechanical fatigue life of flexible thin film transistors based on low-frequency noise according to claim 1, characterized in that: The flexible thin film transistor to be predicted is one or more combinations of an amorphous silicon thin film transistor, a low-temperature polysilicon thin film transistor, an organic thin film transistor and an oxide thin film transistor.

3. The method for predicting mechanical fatigue life of flexible thin film transistors based on low-frequency noise according to claim 1, characterized in that: The low-frequency noise power spectrum is calculated using flicker noise as the noise source.

4. The method for predicting mechanical fatigue life of flexible thin film transistors based on low-frequency noise according to claim 3, characterized in that: The calculating the second data of the flexible thin film transistor to be predicted based on the first data and the low-frequency noise power spectrum includes: (1) Using a noise test system to measure the flicker noise data of the flexible thin film transistor to be predicted; (2) Based on the flicker noise data, separating and constructing a flicker noise model of the flexible thin film transistor to be predicted, the flicker noise model including a first flicker noise model of drain current and a second noise model of contact resistance; (3) Calculating second data of the flexible thin film transistor to be predicted based on the flicker noise model and the first data.

5. The method for predicting mechanical fatigue life of flexible thin film transistors based on low-frequency noise according to claim 4, characterized in that: The formula for calculating the initial value of the flat band voltage power spectrum density is as follows: in, is a flicker noise power spectrum density based on the drain current of the flexible thin film transistor to be predicted; is transconductance; is the Coulomb scattering coefficient; is the gate oxide capacitance per unit area; The gate bias voltage applied when calculating the flicker noise power spectral density of the flexible thin film transistor to be predicted; is the threshold voltage; is the initial value of the flat-band voltage power spectrum density; The formula for calculating the initial value of the contact noise power spectrum density is as follows: in, is the drain current density, is the initial value of contact noise power spectral density.

6. The mechanical fatigue life of a flexible thin film transistor based on low frequency noise according to claim 1 The prediction method is characterized in that The mechanical stress intensity includes tensile stress intensity and bending stress intensity, wherein the tensile stress intensity is represented by the magnitude of the pulling force, and the bending stress intensity is represented by the magnitude of the bending radius.

7. The mechanical fatigue life of a flexible thin film transistor based on low frequency noise according to claim 1 The prediction method is characterized in that The life prediction model is based on a back propagation neural network, and specifically comprises an input layer, a first hidden layer, a second hidden layer and an output layer connected in sequence. The number of nodes in the first hidden layer and the second hidden layer is determined by the following formula: In the above formula, is the number of nodes in the input layer, is the number of nodes in the output layer, is the number of nodes in the hidden layer.

8. The mechanical fatigue life of a flexible thin film transistor based on low frequency noise according to claim 1 The prediction method is characterized in that Training the lifespan prediction model specifically includes: (1) Obtain flexible thin film transistor samples; (2) Calculating the first transfer characteristic curve and the first low-frequency noise power spectrum of the flexible thin film transistor sample; (3) Calculate the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density based on the first transfer characteristic curve and the first low-frequency noise power spectrum; (4) Using a uniform orthogonal experimental method to group the flexible thin film transistor samples, and establishing mechanical stress intensity groups and cycle groups based on the grouping results; (5) Apply corresponding mechanical stress and cycle number to each group of flexible thin film transistor samples after grouping to conduct fatigue life test; (6) Measure and record the second low-frequency noise power spectrum of each group of flexible thin film transistor samples after the test, and calculate the degradation amount of the first flat band voltage power spectrum density and the first contact noise power spectrum density of each group of flexible thin film transistor samples based on the second low-frequency noise power spectrum, the initial value of the first flat band voltage power spectrum density and the initial value of the first contact noise power spectrum density; (7) Constructing a training set, the training set includes mechanical stress intensity, cycle number, the initial value of the first flat band voltage power spectrum density, the initial value of the first contact noise power spectrum density, the first flat band voltage power spectrum density degradation amount and the first contact noise power spectrum density degradation amount; (8) Iteratively train the life prediction model using the training set to obtain a trained life prediction model.

9. The mechanical fatigue life of a flexible thin film transistor based on low frequency noise according to claim 8 The prediction method is characterized in that Before using the training set to iteratively train the life prediction model, the training set is normalized and / or differentiated.

10. A method for predicting mechanical fatigue life of a flexible thin film transistor based on low frequency noise according to any one of claims 1 to 9, characterized in that: The degradation threshold is , that is, the flat-band voltage power spectrum density degradation exceeds the initial value of the flat-band voltage power spectrum density or the contact noise power spectrum density degradation exceeds the contact noise power spectrum density initial value The number of cycles is taken as the fatigue life of the flexible thin film transistor to be predicted under the mechanical stress intensity conditions.

Images