Ga2O3 SBD trap test method based on forward current transient response

Abstract

The invention discloses a Ga2O3 SBD trap test method based on forward current transient response, and belongs to the field of semiconductor device reliability. Comprising the following steps: giving a filling voltage interval and a filling time interval of a Ga2O3 SBD trap filling stage; giving a bias condition and a test time interval of a trap test stage; giving a normalization processing method and range of the forward current transient response; a characterization range of a Ga2O3 SBD device level trap time constant is given. The transient response curve of the forward current of the Ga2O3SBD under the constant test voltage is collected, and the transient current variation caused by the trap is evaluated through a normalization processing method. And repeatedly testing under the condition of multiple groups of filling voltages, establishing a time constant spectrum of the trap by using a structure function method, and extracting a time constant parameter of the trap in the Ga2O3SBD to realize in-situ and online Ga2O3SBD trap test. The method is suitable for evaluating the trap effect of Ga2O3 SBD devices from different manufacturers, and can realize spectral valuation characterization of Ga2O3 SBD traps and in-situ and nondestructive testing of trap parameters.

Description

Technical Field

[0001] The present invention relates to electrical property testing and trap parameter characterization of a potassium oxide Schottky barrier diode (Ga2O3 SBD), and belongs to the field of reliability testing of novel semiconductor devices. Background Art

[0002] Gallium oxide (Ga2O3) Schottky barrier diodes (SBDs) have become ideal candidate materials for the next generation of high-voltage, high-efficiency power electronic devices due to their ultra-wide bandgap (~4.9eV), high critical breakdown electric field (~8MV / cm) and excellent high-temperature resistance. In the fields of electric vehicle fast charging systems, smart grids, industrial motor drives, and aerospace power management, Ga2O3SBDs are expected to achieve higher power density and lower energy loss than traditional SiC and GaN devices. Studies have shown that Ga2O3SBDs can still maintain a low leakage current (<1μA / mm2) under a reverse bias of 1200V. 2 ), making it extremely competitive in high-voltage applications. However, the trap effect in Ga2O3 materials and devices can significantly reduce their electrical performance and even cause device failure, severely limiting their practical applications. The traps in Ga2O3SBD mainly come from crystal defects, impurity doping and Schottky interface states. Studies have shown that high-density traps may cause a significant increase in the forward on-resistance of the SBD, while causing the reverse leakage current to increase by 1-2 orders of magnitude. In addition, in dynamic switching tests, the charge storage effect caused by traps will significantly prolong the reverse recovery time, reducing the efficiency of Ga2O3SBD in high-frequency switching applications, thereby affecting the long-term reliability of the device.

[0003] At present, the trap testing methods for Ga2O3SBD mainly include deep level transient spectroscopy (DLTS), capacitance-voltage (CV) testing and thermally stimulated current (TSC) technology. Among them, DLTS can detect deep traps with energy levels of 0.1 to 1.0 eV, but the detection sensitivity for shallow energy level traps less than 0.1 eV is insufficient; the CV testing method can reflect the interface state density of the device, but it is difficult to distinguish between bulk traps and interface states. In addition, TSC can analyze the thermal activation energy of the trap, but the test time is long and the resolution is limited. Therefore, there is an urgent need to develop more accurate and efficient trap testing technology to optimize the reliability and performance of Ga2O3SBD.

[0004] This technology addresses the trapping problem in Ga2O3SBD devices and proposes a device-level trap testing method based on the forward current transient response. By evaluating the change in forward current caused by the release of Ga2O3SBD traps under different bias voltages, an RC network and deconvolution technique are used to establish the Ga2O3SBD's trap peak spectrum. From this, the time constant corresponding to each trap type is obtained, thereby enabling in-situ testing of Ga2O3SBD traps. This characterization method is simple and convenient to operate, providing an effective technical means for analyzing the trap changes of Ga2O3SBDs under practical application conditions. Summary of the Invention

[0005] CNTFET interface traps, gate dielectric traps, etc. may cause reliability issues such as threshold voltage drift and increased hysteresis window. Under various stresses such as electrical, thermal, and radiation, the reliability issues caused by the CNTFET trap effect are becoming increasingly prominent. The generation and increase of defects and traps may lead to device performance degradation and failure. Interface traps in Ga2O3 materials and devices will cause their electrical performance to degrade, manifesting as increased forward on-resistance, increased reverse leakage current, and deterioration of dynamic switching characteristics. Under long-term electrical stress, high-temperature operation or radiation environment, the accumulation of traps may further aggravate device reliability issues and even cause premature failure. At present, deep level transient spectroscopy (DLTS) and capacitance-voltage (CV) test methods for Ga2O3SBD traps have insufficient detection sensitivity for shallow energy level traps, and it is difficult to accurately capture the charging and discharging dynamics of the traps. To address these issues, this paper proposes a Ga2O3 SBD trap testing method based on the forward current transient response. By precisely analyzing the device's forward current relaxation characteristics under different bias voltages, a spectral representation of the Ga2O3 SBD traps is constructed, accurately distinguishing trap peaks and obtaining time constant information. This method effectively reveals the trap filling and emission mechanisms of Ga2O3 SBDs under electrical stress, providing an accurate test basis for optimizing device reliability and improving device performance.

[0006] The technical solution adopted by the present invention is a Ga2O3SBD trap testing method based on forward current transient response. The specific implementation process is as follows:

[0007] S1. Place the sample to be tested on the constant temperature platform in the vacuum chamber of the probe station, and set the temperature of the constant temperature platform to T through the temperature control system. In step S1, the range of T can be set to 20℃≤T≤30℃ to evaluate the forward current transient response of Ga2O3 SBD at room temperature. Adjust the vacuum pump to control the pressure P in the vacuum chamber, where P can be set to 10 -6 mbar≤P≤10 -5mbar, and remain unchanged during the test to ensure constant humidity in the sample's environment. The probe station is adjusted, and two probes are used to contact the anode and cathode of the Ga2O3 SBD, respectively. The cathode of the sample under test is grounded, thus completing the electrical connection between the sample under test and the electrical parameter tester.

[0008] S2. Without applying power, let the Ga2O3 SBD sample stand for t a minutes, where t a The setting range is 15min≤t a ≤30min. At this time, the Ga2O3 SBD sample under test is in the initial state without electrical stress, and the sample temperature is consistent with the temperature T of the set constant temperature platform, completing the preparation work before the test.

[0009] S3. Adjust the semiconductor parameter tester to apply bias voltage to the device, which mainly includes two stages: trap filling and testing. In the trap filling stage, set the filling voltage V F and filling time t f During the filling period, a voltage of V is applied to the Ga2O3 SBD sample. F The filling voltage of the device makes the carriers inside the device captured by the trap energy level. F The range is -15V≤V F ≤5V, filling time t f The range is 10ms≤t f ≤100s. After completing t f After the time is filled, it enters the trap test phase and sets the test voltage V M At this time, V M The range is V on ≤V M , where V on is the turn-on voltage of the Ga2O3 SBD sample under test. During the trap release process, the test voltage V M Should be greater than or equal to the turn-on voltage V on To ensure that the sample under test is in the on state, the trap release status can be reflected by the transient change of the forward conduction current. M Under bias conditions, the transient response of the Ga2O3SBD forward current is monitored in logX logarithmic form, and the test time is set to t M The logX logarithmic acquisition mode of the forward current transient response data can be set to log5, log10, log20, log50 or log100, and the test time t m The range is 1s≤t m ≤1000s, the data acquisition accuracy of the forward current transient curve can reach 4ms.

[0010] S4. The current data corresponding to the end time is recorded as I(end). At the same time, the obtained transient current curve is normalized based on I(end). That is, the measured transient current I(t) corresponding to each time point is divided by I(end) to obtain the normalized transient current I NOR (t)=I(t) / I(end). Among them, the range of I(t) is 1mA≤I(t)≤100mA, and the range of I(end) is I(t)≤I(end)≤1A. After normalization, the normalized data corresponding to the end time I NOR (end) is 1, and the normalized data corresponding to the start time is I NOR (0) is I(0) / I(end), from which the transient current change △I NOR =(1-I(0) / I(end))×100%. In which, the normalized data I NOR The range of (t) is 0≤I NOR (t)≤1, transient current change △I after normalization NOR The range is 0≤△I NOR ≤ 1. Based on the acquired transient current response curve, the Bayesian deconvolution technique is used to establish the time constant spectrum and identify the trap peak.

[0011] S5. Based on the normalized transient current response curve I NOR , the peak spectrum can be established using the structure function method, where the time represented by the horizontal axis is expressed in log10 logarithmic form, and the relative amplitude of the trap represented by the vertical axis is expressed in linear coordinate form. From this, the time constant of the Ga2O3 SBD trap can be read. The trap causes the transient current curve to change in accordance with the e exponential function, and the number of e exponents N contained in the curve is the number of trap types. The horizontal axis coordinate corresponding to the trap peak is the time constant of the corresponding trap, and the trap time constant corresponding to the i-th trap peak DPi is recorded as t i . Where 1≤i≤N, 1s≤t i ≤1000s.

[0012] S6. After the transient current curve test is completed, the tested sample is left to stand for t without power supply. b minutes, where t b The setting range is 15min≤t b ≤30min. After the static state is completed, the filling voltage V F Change M times and repeat steps S1 to S5 to obtain (M+1) sets of time constant spectra under filling voltage, where M is a positive integer. F Under different filling voltages V FThe time constant of the lower trap remains basically unchanged, thereby identifying the traps in the Ga2O3 SBD and obtaining the time constant information of the traps.

[0013] The present invention proposes a Ga2O3 SBD trap testing method based on forward current transient response. Targeting device-level trap testing of Ga2O3 SBDs, the method can obtain the forward current transient response curve of the device under test under a constant bias voltage, and utilize the structure function method with peak spectrum technology to obtain the Ga2O3 SBD trap parameters. The present invention has the following beneficial effects: the Ga2O3 SBD trap testing method of the present invention is simple to operate and has the characteristics of being in-situ and non-destructive. The test voltage and test time conditions involved in the present invention are both adjustable, and are highly applicable to trap testing of Ga2O3 SBD devices from different manufacturers and sources, providing a fast and convenient testing method for trap testing of Ga2O3 devices and reliability assessment of Ga2O3 devices in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 : This is a diagram of the test device involved in the present invention, in which: 1-Keysight B1505 semiconductor parameter tester, 2-probe station, 3-tested sample, 4-constant temperature platform, 5-vacuum pump, 6-temperature control system.

[0015] Figure 2 : The test timing diagram and transient current response schematic diagram involved in the present invention.

[0016] Figure 3 : The trap time constant spectrum involved in the present invention. DETAILED DESCRIPTION

[0017] The present invention will be described in more detail below with reference to the accompanying drawings and specific embodiments.

[0018] Taking Ga2O3 SBD device as an example, the device electrical testing device involved in the present invention is as follows: Figure 1 As shown, it includes a semiconductor parameter tester, a probe station, a Ga2O3 SBD sample to be tested, a constant temperature platform, a vacuum pump and a temperature control system, wherein the Ga2O3 SBD to be tested is β-Ga2O3 SBD. The test process of the method involved in the present invention is as follows Figure 2 As shown, the following steps are included:

[0019] Step 1: Place the sample on a constant temperature platform, set the temperature of the constant temperature platform to 25°C through the temperature control system, and adjust the vacuum pump to a pressure of 10 -6mbar to ensure that the humidity of the sample environment remains unchanged during the test. The probe station is adjusted, and two probes are used to contact the anode and cathode of the Ga2O3 SBD respectively. The cathode of the sample under test is grounded, thus completing the electrical connection between the sample under test and the electrical parameter tester.

[0020] Step 2: Without power, let the sample stand for 15 minutes. At this time, the sample is in its initial state without electrical stress. At the same time, the temperature of the sample is consistent with the set constant temperature platform temperature, completing the preparation work before the test.

[0021] Step 3: Adjust the semiconductor parameter tester to apply bias voltage to the device. The measurement sequence is as follows: Figure 2 As shown in (a), it mainly includes two stages: trap filling and testing. In the trap filling stage, the filling voltage V F Set to -4V, fill time t f Set to 15s; after 15s, the test phase begins and the start voltage V on is 0.75V, the test voltage V M Set to 1V, monitor the transient response of the forward current under this condition, and test time t M 300s.

[0022] Step 4: Obtain transient current data in log10 logarithmic form, with a start time of 4ms and an end time of 300s. The current data corresponding to the end time is recorded as I(end). At the same time, the obtained transient current curve is normalized based on I(end). That is, the measured transient current I(t) corresponding to each time point is divided by I(end) to obtain the normalized transient current I NOR (t)=I(t) / I(end). After normalization, the normalized data corresponding to the end time I NOR (end) is 1, and the normalized data corresponding to the start time is I NOR (0) is I(0) / I(end), from which the transient current change △I NOR =(1-I(0) / I(end))×100%. by Figure 2 Taking the curve shown in (b) as an example, the normalized data corresponding to the starting time of the curve I NOR (0) is 0.97128, so the transient current change of this curve is 2.872%.

[0023] Step 5: Based on the normalized transient current response curve I NORThe structure function method can be used to establish a peak spectrum and read the time constant of the Ga2O3 SBD trap from it. The trap-induced transient current curve changes follow an e-exponential function, and the number of e-exponents included in the curve is the same as the number of traps. Figure 3 The result of time constant spectrum processing is shown in the form of log10, where the horizontal axis represents the time, and the vertical axis represents the relative amplitude of the trap, which is expressed in the form of linear coordinates. The peak spectrum obtained by the structure function method has three trap peaks, indicating that the curve contains three e-exponential functions and identifies three types of traps. In addition, the horizontal axis coordinate corresponding to the trap peak is the time constant of the corresponding trap. The three trap peaks are named DP1, DP2 and DP3 in the order of time constant from small to large. F Under the bias condition of -4V, the time constants of the three traps are 40ms, 1.8s and 90s respectively.

[0024] Step 6: After the transient current curve test, the sample will be left standing for 15 minutes without power. F Change to -6V, -8V and -10V in sequence, and repeat steps 1 to 5 to obtain 4 sets of time constant spectra under filling voltage. The results are as follows Figure 3 As shown. At different filling voltages V F Under the same conditions, the time constant spectrum shows three trap peaks, and the peak positions are basically unchanged, that is, the trap time constant is basically unchanged. Therefore, the traps of Ga2O3 SBD are identified and the time constant information of the traps is obtained.

Claims

1. A Ga2O3 SBD trap testing method based on forward current transient response, characterized in that: The measuring method comprises the following steps: Step 1: Place the sample under test on the constant temperature platform in the vacuum chamber of the probe station, and set the temperature of the constant temperature platform to T through the temperature control system; adjust the vacuum pump to control the pressure P in the vacuum chamber and keep it unchanged during the test; adjust the probe station, use two probes to contact the anode and cathode of the Ga2O3SBD respectively, where the cathode of the sample under test is grounded, thereby completing the electrical connection between the sample under test and the electrical parameter tester; Step 2: Without applying power, let the Ga2O3 SBD sample stand for t a minutes; at this time, the Ga2O3 SBD sample under test is in the initial state without electrical stress, and the sample temperature is consistent with the temperature T of the set constant temperature platform, completing the preparation work before the test; Step 3: Adjust the semiconductor parameter tester to apply bias voltage to the device, which mainly includes two stages: trap filling and testing. In the trap filling stage, set the filling voltage V F and filling time t f During the filling period, a voltage of V is applied to the Ga2O3 SBD sample. F The filling voltage of the device makes the carriers inside the device captured by the trap energy level; after completing t f After the time is filled, it enters the trap test phase and sets the test voltage V M ; During the trap release process, the test voltage V M Should be greater than or equal to the turn-on voltage V on To ensure that the sample under test is in the on state, the trap release situation is reflected by the transient change of the forward conduction current; M Under bias conditions, the transient response of the Ga2O3 SBD forward current is monitored in logX logarithmic form, and the test time is set to t M ; Step 4: Record the current data corresponding to the end time as I(end). At the same time, normalize the obtained transient current curve based on I(end). That is, divide the measured transient current I(t) corresponding to each time point by I(end) to obtain the normalized transient current I NOR (t) = I(t) / I(end); after normalization, the normalized data corresponding to the end time I NOR (end) is 1, and the normalized data corresponding to the start time is I NOR (0) is I(0) / I(end), from which the transient current change △I NOR =(1-I(0) / I(end))×100%; Based on the acquired transient current response curve, the Bayesian deconvolution technique is used to establish the time constant spectrum and identify the trap peak; Step 5: Based on the normalized transient current response curve I NOR , the peak spectrum is established using the structure function method, where the horizontal axis represents the time in log10 logarithmic form, and the vertical axis represents the relative amplitude of the trap in linear coordinate form; the time constant of the Ga2O3 SBD trap can be read from it; the transient current curve caused by the trap satisfies the e exponential function change, and the number of e exponents N contained in the curve is the number of trap types; the horizontal axis coordinate corresponding to the trap peak is the time constant of the corresponding trap, and the trap time constant corresponding to the i-th trap peak DPi is recorded as t i ; Step 6: After the transient current curve test is completed, the sample is left to stand for t without power supply. b minutes; after the static state is completed, the voltage V F Change M times and repeat steps 1 to 5 to obtain (M+1) sets of time constant spectra under different filling voltages V F Under the same conditions, the time constant spectrum shows N trap peaks, and the peak positions are basically unchanged, that is, the trap time constant is basically unchanged, thereby identifying the traps of Ga2O3 SBD and obtaining the trap time constant information.

2. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 1, the sample to be tested is a Ga2O3 Schottky barrier diode. The initial test temperature T range is 20℃≤T≤30℃ to evaluate the forward current transient response of Ga2O3 SBD at room temperature. The pressure P in the vacuum chamber ranges from 10 -6 mbar≤P≤10 -5 mbar, and remain unchanged during the test to ensure that the humidity of the environment where the sample is located is constant.

3. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 2, the Ga2O3 SBD is left to stand for a period of time t before testing. a The setting range is 15min≤t a ≤30min, and t a During the static time, the Ga2O3 SBD under test is in an unpowered state.

4. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 3, the Ga2O3 SBD trap filling voltage V F The range is -15V≤V F ≤5V, filling time t f The range is 10ms≤t f ≤100s; Trap test voltage V M The range is V on ≤V M , where V on is the turn-on voltage of the Ga2O3 SBD sample tested.

5. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 3, the curve collected during the Ga2O3 SBD trap test is the filling voltage V F Switch to the test voltage V M After that, the forward current of Ga2O3 SBD changes with time.

6. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 3, the forward current transient response data is collected in logX logarithmic mode, where the logarithmic collection can be set to log5, log10, log20, log50 or log100, and the test time t m The range is 1s≤t m ≤1000s, the data acquisition accuracy of the forward current transient curve is up to 4ms.

7. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 4, the test range of transient current I(t) is 1mA≤I(t)≤100mA, and the range of current I(end) corresponding to the end time is I(t)≤I(end)≤1A; the normalized data is defined as I NOR (t)=I(t) / I(end), where I NOR (0) is in the range of 0≤I NOR (0)≤1, transient current change △I after normalization NOR The range is 0≤△I NOR ≤1.

8. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 5, the trap causes the transient current curve to change in accordance with an e-exponential function, and the number N of e-exponents included in the curve is the number of trap types, where N is a positive integer.

9. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 5, the trap characterization method of Ga2O3 SBD is based on the forward current transient curve. The peak spectrum characterization of Ga2O3 SBD trap form is constructed using the structure function method. The horizontal axis coordinate corresponding to the trap peak is the time constant of the corresponding trap. The trap time constant corresponding to the i-th trap peak DPi is recorded as t i , where 1≤i≤N, 1s≤t i ≤1000s.

10. The Ga2O3SBD trap testing method based on forward current transient response according to claim 1, characterized in that: In step 6, the static time t after the transient current curve test is completed b The range is 15min≤t b ≤30min; at different filling voltages, the transient curve test times are (M+1) times, where M is a positive integer.

Images