MO-TFT trap state capture parameter extraction method and system

Abstract

The invention discloses an MO-TFT trap state capture parameter extraction method and system, and the method comprises the steps: applying a laser excitation signal to an active layer region of an MO-TFT, and enabling the laser excitation signal to enter the active layer region of the MO-TFT for regulating and controlling the surface potential # imgabs0 # and inducing different transient light current responses; applying a bias voltage VDS between the source electrode and the drain electrode of the MO-TFT to ensure that the device works in a deep linear region; measuring the transient photocurrent response of the MO-TFT, and recording a transient photocurrent change curve; and calculating and extracting trap state capture parameters. According to the method, relaxation time and corresponding trap state capture parameters are extracted through measurement and fitting of transient current changes, and accurate characterization of the trap state characteristics of the MO-TFT is achieved. The method provided by the invention can effectively obtain trap state capture kinetic parameters, has high test precision and model applicability, and can be widely applied to reliability evaluation and optimization design of the MO-TFT device.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor device characteristic testing and optoelectronic measurement, and particularly relates to a method and system for extracting trap state capture parameters of MO-TFT. Background Art

[0002] Metal oxide thin film transistor (MO-TFT) is a thin film transistor based on metal oxide semiconductor channel material, which is widely used in fields such as high-resolution display, wearable devices, and optoelectronic detection. MO-TFT has become a key technology for next-generation electronic devices due to its high mobility, low manufacturing cost, and transparency. However, during long-term operation, MO-TFT is easily affected by trap states, resulting in threshold voltage drift, increased subthreshold swing, and decreased photo-response characteristics, thus affecting the stability and reliability of the device.

[0003] The trap states of MO-TFT mainly include bulk traps inside the channel material and interface traps at the interface. These trap states can capture or release carriers, changing the electrical characteristics of the device. Under long-term operation, trap states capture electrons or holes, leading to threshold voltage drift over time. In addition, MO-TFT is very sensitive to light illumination. Light of different wavelengths may excite oxygen vacancies or trap states, changing the optoelectronic response characteristics of the device. For light illumination excitation, the capture and release processes of trap states exhibit complex transient photocurrent changes, affecting the dynamic stability of the device. Therefore, how to accurately extract the trap state parameters of MO-TFT and establish a quantitative characterization method is a key issue for optimizing device performance and improving model accuracy.

[0004] Currently, the methods for extracting trap state capture parameters of MO-TFT mainly include static electrical testing, transient current measurement, and photo-excitation characterization, etc. Among them, static electrical testing usually uses current-voltage or capacitance-voltage measurement to estimate the trap state density from the threshold voltage drift. However, this method is difficult to distinguish trap states of different energy levels and cannot capture the dynamic characteristics of trap states. Transient current measurement triggers MO-TFT with a pulsed electrical signal and measures the transient current decay process to extract trap state capture kinetic parameters. However, this method is easily affected by test system noise, and traditional electrical signal pulses are difficult to accurately control the carrier distribution in the channel. In addition, the photo-excitation characterization method usually uses a light source with a fixed wavelength to measure the photocurrent response of the device and analyzes the capture-release process of photo-generated carriers. However, a fixed wavelength light source is difficult to accurately characterize different trap state distributions, and existing experiments usually do not consider the influence of different photon energies on the trap state filling characteristics. Therefore, existing methods still have great limitations in transient dynamic analysis, accurate extraction of trap state distribution, and energy level resolution measurement. Summary of the Invention

[0005] To overcome the above-mentioned disadvantages and deficiencies of the prior art, the purpose of the present invention is to provide a method and system for extracting trap state capture parameters of MO-TFT.

[0006] The purpose of the present invention is achieved by the following technical solutions:

[0007] A method for extracting trap state capture parameters of MO-TFT includes:

[0008] Applying a laser excitation signal to the active layer region of the MO-TFT to regulate the surface potential and inducing different transient photocurrent responses;

[0009] Applying a bias voltage V between the source and drain of the MO-TFT DS , ensuring that the device operates in the deep linear region;

[0010] Measuring the transient photocurrent response of the MO-TFT and recording the transient photocurrent change curve;

[0011] Calculating and extracting the trap state capture parameters.

[0012] Furthermore, the calculating and extracting the trap state capture parameters includes:

[0013] According to the transient photocurrent change curve, using the trap state capture kinetics model to fit and extract the trap state parameters, the trap state parameters include: trap state relaxation time τ, electron capture coefficient r n , electron capture cross-section σ n .

[0014] Furthermore, the trap state capture kinetics model is described by the Shockley-Read-Hall recombination theory, and its formula is:

[0015]

[0016] where Q f (t) is the transient free carrier charge, Q t (t) is the trap state captured charge, Q T is the total trap state charge density, and Q1 is the charge capture ratio under thermal equilibrium.

[0017] Furthermore, the calculating and extracting the trap state capture parameters includes the following steps:

[0018] Using a non-linear fitting algorithm, inversely obtaining the free charge Q ds (t) from the transient photocurrent change curve I f (t);

[0019] Separating the trap charge Q t (t) through the charge conservation equation;

[0020] Fitting Q using the relaxation equation of trapped charges t (t) curve to extract the average capture coefficient r Q ;

[0021] The electron capture coefficient r is calculated n , and combined with the hot electron velocity v n The electron capture cross-section is calculated

[0022] Furthermore, the relaxation equation of the trapped charges is as follows:

[0023]

[0024] In the formula, is the thermal equilibrium capture ratio, r n is the electron capture coefficient, τ is the relaxation time related to free charges, θ is the distribution correction factor. The core of this equation is to reveal the non-linear coupling relationship between the trapped charge Q t (t) and the free charge Q f (t) - the instantaneous concentration of free charges not only drives the trap filling process, but also modulates the relaxation time scale through the denominator term Q f (t)+Q1

[0025] Furthermore, the electron capture cross-section σ n = r n / v n , where v n is the hot electron velocity

[0026] Furthermore, the wavelength of the laser excitation signal is 300 - 1000 nm

[0027] Furthermore, a bias voltage V DS is applied between the source and drain of the MO-TFT, and the bias voltage range is 0.1 - 5 V

[0028] Furthermore, measure the transient photocurrent response of the MO-TFT and record the transient photocurrent change curve, specifically:

[0029] Use a DHPCA-100 low-noise current amplifier to amplify and convert the transient photocurrent

[0030] Use a high-bandwidth oscilloscope to record the transient current I DS (t) change curve

[0031] A system for extracting trap state capture parameters of MO-TFT, the system includes:

[0032] Laser pulse generation module: Apply a laser excitation signal to the active layer region of the MO-TFT to regulate the surface potential and induce different transient photocurrent responses;

[0033] Bias application module: Apply a bias voltage V between the source and drain of the MO-TFT DS to ensure that the device operates in the deep linear region;

[0034] Measurement module: Used to measure the transient photocurrent response of the MO-TFT and record the transient photocurrent change curve;

[0035] Data processing module: Adopt the trap state capture dynamics model to fit the measured transient photocurrent curve and extract the trap state capture parameters.

[0036] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0037] 1. Coextraction mechanism of dynamic characteristics and static parameters: The existing transient current method only focuses on a single parameter of the relaxation time. However, the present invention realizes the dynamic simultaneous solution of free charges, trap charges, and surface potential for the first time by establishing the coupling relationship between the charge conservation equation and the SRH recombination dynamics model, and can synchronously extract the trap state relaxation time, electron capture coefficient, and capture cross-section, breaking through the limitation of the single parameter of the traditional method.

[0038] 2. Design of anti-interference and highly sensitive measurement system: Aiming at the problem that the picoampere-level current signal in traditional transient tests is vulnerable to noise interference, the system adopts a three-stage noise reduction design.

[0039] 3. Non-destructive testing and process compatibility: Compared with the electrical signal transient testing method, the photoexcitation process only regulates the carrier distribution through photon energy, avoiding device damage caused by high field strength, and meeting the requirements of on-line monitoring and reliability evaluation. In addition, the adjustable spot size technology is compatible with the testing of micron-level channel devices, providing support for the process optimization of high-resolution display driving TFTs.

[0040] 4. Full-parameter automatic extraction and industrial application potential: The system integrates a laser control module, a bias voltage loading module, a signal acquisition module, and a data processing module to realize the full-process automation of "excitation - measurement - analysis". Brief Description of the Drawings

[0041] Figure 1 is a flowchart of a method for extracting trap state capture parameters of an MO-TFT according to the present invention;

[0042] Figure 2 is a schematic diagram of the system according to the present invention. Detailed Embodiments

[0043] The present invention will be further described in detail below with reference to the embodiments, but the embodiments of the present invention are not limited thereto.

[0044] Embodiment

[0045] As shown in Figure 1 Figure Figure 1 , a method for extracting trap state capture parameters of a MO-TFT (Metal Oxide Thin Film Transistor) based on a physical model of photoexcited transient response includes the following specific steps:

[0046] S1 Data acquisition and steady-state parameter extraction: Apply a laser excitation signal to the active layer region of the MO-TFT to regulate the surface potential and induce different transient photocurrent responses, specifically inducing the redistribution of band bending and the trap state capture / release process, and use a transient current test system to measure the transient response characteristics of the device.

[0047] The transient response process of the MO-TFT is essentially a comprehensive manifestation of the dynamic adjustment of band bending and the relaxation behavior of trap state charges. By measuring the transient current at different incident wavelengths, the current peak, relaxation process, and steady-state current can be obtained. The steady-state parameters are extracted through I-V and C-V measurements, and the trap state density (DOS) is extracted through low-frequency CV (LFCV).

[0048] Specifically:

[0049] Use an MW-GX series fiber-coupled semiconductor laser to generate laser light and incident it on the active layer region of the MO-TFT;

[0050] Control the fiber laser through a TTL modulation signal (using a WF1948 multi-functional waveform generator) to ensure the precise triggering of the laser pulse;

[0051] The incident wavelength of the laser can be selected from multiple different values in the range of 300 - 1000 nm to regulate the surface potential and induce different transient light responses;

[0052] Ensure the controllability of the laser irradiation area through an optical system (lens focusing), and the spot size can be adjusted between 0.5 - 10 μm to reduce measurement errors;

[0053] Adopt the design of an optical dark cavity to shield environmental light interference and ensure the reliability of the measurement.

[0054] S2 Apply a bias voltage V DS between the source and drain of the MO-TFT to ensure that the device operates in the deep linear region;

[0055] Apply a bias voltage V DS between the source and drain of the MO-TFT, and the bias voltage range is 0.1 - 5 V to ensure that the device operates in the deep linear region;

[0056] A high-stability DC power supply (voltage fluctuation <0.1%) is adopted to ensure the accuracy and stability of the bias voltage.

[0057] S3 measures the transient photocurrent response of the MO-TFT and records the transient photocurrent change curve.

[0058] A DHPCA-100 low-noise current amplifier (bandwidth ≥10 MHz, input noise <1 fA / √Hz) is adopted to amplify and convert the transient photocurrent;

[0059] A high-bandwidth oscilloscope (Wavesurfer 3034, sampling rate ≥4 GS / s, resolution ≥12 bit) is adopted to record the transient current I DS (t) change curve;

[0060] Ensure the synchronization of the TTL signal with the current measurement system to reduce the influence of time error on the extraction of trap state capture parameters.

[0061] S4 calculates and extracts the trap state capture parameters, specifically:

[0062] The transient response process of the MO-TFT is essentially a comprehensive manifestation of the dynamic adjustment of band bending and the relaxation behavior of trap state charges. When the trap state of the device channel is excited by a laser, the gate voltage V g (t) and the surface potential interact according to the voltage distribution equation:

[0063]

[0064] Among them, Φ ma is the metal-semiconductor work function difference, Q g (t) is the gate charge density, and C ox is the gate oxide capacitance. In the transient response stage, the change of the surface potential and the change of the gate charge satisfy the differential relationship:

[0065]

[0066] This relationship indicates that the instantaneous adjustment of the surface potential is directly controlled by the distribution of the gate charge. According to the law of conservation of charge, the change of the gate charge is jointly determined by the free charge Q f (t) and the trap charge Q t (t):

[0067] dQ g (t) = dQ f (t) + dQ t (t)

[0068] The expression of the free charge Q f (t) is obtained by integrating the free carrier concentration n f(x,t) is obtained:

[0069]

[0070] Where t s is the thickness of the active layer, n0 is the flat-band electron concentration, and V t = kT / q is the thermal voltage. To further simplify the calculation, a linear approximation of the exponential term is performed using Taylor expansion:

[0071]

[0072] Combined with the integral mean value theorem, the change in the potential distribution is approximated as a linear function of the surface potential change , that is:

[0073]

[0074] m is the correction factor, and its value is taken as 0.5. Finally, the differential form of the free charge change is obtained:

[0075]

[0076] This equation reveals the sensitivity of the free charge to the surface potential change and lays the foundation for the subsequent trap state dynamics modeling.

[0077] Based on the SRH theory, the transient behavior of the trap charge is determined by the dynamic balance of the electron capture and emission processes. For the continuously distributed trap states, the total trap charge density Q T can be expressed as:

[0078]

[0079] where g a (E,x) is the acceptor-type trap state density function. The instantaneous value of the trap charge Q t (t) is then:

[0080]

[0081] f[E,E F (t)] is the Fermi-Dirac distribution function, and E F (t) is the transient quasi-Fermi level. By introducing the average capture coefficient r Q , the microscopic dynamics of the carrier concentration is transformed into a macroscopic equation of the charge density. According to the SRH recombination rate equation:

[0082]

[0083] where r nis the electron capture coefficient, n1 is the thermal equilibrium concentration. Integrate the entire space and introduce the charge density Q f (t) and Q t (t), and finally the relaxation equation of the trapped charge is derived:

[0084]

[0085] In the formula, is the thermal equilibrium capture ratio, τ is the relaxation time associated with the free charge, and θ is the distribution correction factor. The core of this equation is to reveal the trapped charge Q t (t) and free charge Q f The nonlinear coupling relationship of (t) is that the instantaneous concentration of free charges not only drives the trap filling process, but also f (t)+Q1 modulation relaxation time scale.

[0086] Transient current I of MO-TFT ds (t) is directly related to the free charge through the following equation:

[0087]

[0088] Where μ is the field effect mobility, W / L is the channel width-to-length ratio, V ds is the source-drain bias. This formula shows that the decay characteristics of transient current are essentially the free charge Q f (t) Macroscopic representation of the gradual capture by the trapped states.

[0089] By combining the above equations, a complete link for extracting the trap state capture parameters from the transient current decay curve can be established:

[0090] 1. Using nonlinear fitting algorithm, I ds (t) Inversion to get Q f (t), in order to utilize the transient photocurrent decay data corresponding to different laser wavelengths to fit the model, the nonlinear fitting algorithm uses the Levenberg-Marquardt algorithm (LM algorithm) to perform nonlinear least squares optimization.

[0091] 2. Separate Q by charge conservation equation t (t);

[0092] 3. Fitting Q using the relaxation equation of trapped charge t (t) Curve extraction r Q ;

[0093] 4. Calculate the electron capture coefficient r n , combined with the hot electron velocity v n The electron capture cross section σ is calculated n =rn / v n 。

[0094] As Figure 2 shown, this embodiment also provides a system for extracting the trap state capture parameters of MO-TFT. Through a laser excitation-electrical response collaborative detection architecture, combined with high-precision signal acquisition and kinetic model inversion technology, quantitative analysis of the trap state capture parameters of the device is achieved. The system consists of a laser pulse generation module, a bias voltage application module, a measurement module, and a data processing module. The specific functions of each module are as follows:

[0095] Including:

[0096] The laser pulse generation module uses a fiber-coupled semiconductor laser (MW-GX series) to provide multi-wavelength lasers in the range of 300 - 1000 nm, and realizes precise control of the photon energy through an adjustable optical path system. The TTL trigger signal output by the waveform generator (WF1948) synchronously modulates the laser, and combines the lens focusing and fiber transmission unit to precisely control the spot size within the range of 0.5 - 10 μm, ensuring that the laser only acts on the active region of the MO-TFT channel and avoiding interference from edge leakage current to the test results.

[0097] The bias voltage application module provides a source-drain bias voltage of 0.1 - 5 V for the device through a high-precision DC power supply, so that the channel is in a deep linear operating state. The photo-generated carriers generated by laser excitation dynamically regulate the surface potential, changing the filling and release processes of trap state charges, thereby establishing a correlation mechanism between the transient photocurrent and the trap capture parameters.

[0098] The working mode is:

[0099] Source-drain bias range: 0.1 - 5 V; Output ripple: <1 mV pp ; Establishment time: <100 μs.

[0100] The measurement module: Based on a low-noise current amplifier (DHPCA-100, bandwidth ≥10 MHz) and a high-precision oscilloscope (Wavesurfer 3034, sampling rate ≥4 GS / s) to form a core detection link. The input end of the current amplifier adopts a coaxial shield structure and a ground loop optimization design to suppress the background noise to the following; The oscilloscope synchronously acquires the laser trigger signal and the transient current waveform through an opto-isolation interface, and realizes a complete record of the nanosecond-level dynamic response at a 12-bit resolution, ensuring the linearity and repeatability of the transient photocurrent signal.

[0101] This module adopts a three-stage shielding structure, and the three-stage shielding structure includes a coaxial cable, a Faraday cage, and opto-isolation.

[0102] Data processing module: A multi-parameter model including relaxation time, electron capture coefficient, and trap state-related parameters is established for the trap state capture kinetics characteristics. The transient current decay curve is inversely calculated by an optimized Levenberg-Marquardt nonlinear fitting algorithm, and the distribution weights of trap states at different energy levels are constrained by multi-wavelength laser excitation data. An adaptive iteration step size adjustment strategy is introduced into the algorithm, shortening the convergence time of traditional global fitting by 40%, and at the same time improving the credibility of parameter extraction through residual analysis and confidence interval evaluation.

[0103] The data processing module of the present invention realizes multi-wavelength data fusion processing, confidence interval analysis, residual evaluation, and parameter visualization output.

[0104] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited by the described embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A method for extracting the trap state capture parameters of a MO-TFT, characterized in that, Including: Apply a laser excitation signal to the active layer region of the MO-TFT to regulate the surface potential and induce different transient photocurrent responses; Apply a bias voltage V between the source and drain of the MO-TFT DS to ensure that the device operates in the deep linear region; Measuring the transient photocurrent response of the MO-TFT and recording the transient photocurrent change curve; Calculating and extracting the trap state capture parameters.

2. The method for extracting the trap state capture parameters of the MO-TFT according to claim 1, characterized in that The calculating and extracting of the trap state capture parameters includes: Based on the transient photocurrent change curve, using the trap state capture kinetic model, fit and extract the trap state parameters, where the trap state parameters include: trap state relaxation time τ, electron capture coefficient r n , electron capture cross-section σ n .

3. The method for extracting the trap state capture parameters of the MO-TFT according to claim 2, wherein, The trap state capture dynamics model is described by the Shockley-Read-Hall recombination theory, and its formula is: Among them, Q f (t) is the transient free carrier charge, Q t (t) is the trapped charge in the trap state, Q T is the total trap state charge density, and Q1 is the charge capture ratio in the thermal equilibrium state.

4. The method for extracting MO-TFT trap state capture parameters according to claim 2, characterized in that, The calculating and extracting of the trap state capture parameters includes the following steps: Using a non - linear fitting algorithm, the free charge Q ds (t) is inversely obtained from the transient photocurrent change curve I f (t); Separate the trapped charge Q by the charge conservation equation t (t); Fitting Q t by using the relaxation equation of trapped charges and extracting the average capture coefficient r Q ; The electron capture coefficient r is calculated n , and then combined with the hot electron velocity v n to calculate the electron capture cross section.

5. The method for extracting the MO-TFT trap state capture parameters according to claim 4, wherein The relaxation equation of the trapped charge: In the formula, is the thermal equilibrium capture ratio, r n is the electron capture coefficient, τ is the relaxation time related to free charges, θ is the distribution correction factor. The core of this equation is to reveal the non-linear coupling relationship between the trapped charge Q t (t) and the free charge Q f (t) - the instantaneous concentration of free charges not only drives the trap filling process, but also modulates the relaxation time scale through the denominator term Q f (t) + Q1.

6. The method for extracting the trap state capture parameters of the MO-TFT according to claim 4, wherein, The electron capture cross section σ n = r n / v n , where v n is the hot electron velocity.

7. The method for extracting the trap state capture parameters of the MO-TFT according to any one of claims 1-6, characterized in that The wavelength of the laser excitation signal is 300 - 1000 nm.

8. The method for extracting the trap state capture parameters of the MO-TFT according to any one of claims 1-6, characterized in that, A bias voltage V is applied between the source and drain of the MO-TFT DS , and the bias voltage range is 0.1 to 5 V.

9. The method for extracting the trap state capture parameters of the MO-TFT according to any one of claims 1-6, characterized in that, Measuring the transient photocurrent response of the MO-TFT and recording the transient photocurrent change curve, specifically: Using a DHPCA-100 low-noise current amplifier to amplify and convert the transient photocurrent; Use a high-bandwidth oscilloscope to record the transient current I DS (t) variation curve.

10. A MO-TFT trap state capture parameter extraction system, characterized in that, Including: Laser pulse generation module: Apply a laser excitation signal to the active layer region of the MO-TFT to regulate the surface potential and induce different transient photocurrent responses; Bias voltage application module: Apply a bias voltage V between the source and drain of the MO-TFT DS , ensuring that the device operates in the deep linear region; Measurement module: used to measure the transient photocurrent response of the MO-TFT and record the transient photocurrent change curve; Data processing module: using the trap state capture dynamics model to fit the measured transient photocurrent curve and extract the trap state capture parameters.

Images