A method for detecting a leakage channel of a GaN-HEMT device

Abstract

A kind of detection method of GaN-HEMT device leakage channel.It relates to the field of semiconductor technology.It includes the following steps:S100, a series of bias voltage is set, the device gate drive leakage current is measured, and it is recorded as IGSS;S200, a series of bias voltage is set, the device saturation drain-source current is measured, and it is recorded as IDSS;S300, according to the measured IGSS value and IDSS value, it is judged whether the device exists leakage channel;S310, when the obtained IGSS value exceeds the IGSS normal value of GaN-HEMT device in standard specification one order of magnitude, it is judged as IGSS defective product, it is explained that there is leakage channel between the two electrodes of GS;The present application can be operated using power device detector, avoids the high cost of using chip-level valuable analytical instrument required by traditional method such as EBIRCH, and compared with traditional detection method, it is convenient, fast, simple and time-saving.

Description

Technical Field

[0001] This invention relates to the field of semiconductor technology, and in particular to a method for detecting leakage current channels in GaN-HEMT devices. Background Technology

[0002] In the field of power electronic device technology, third-generation semiconductors, represented by GaN and SiC, are receiving increasing attention. GaN, in particular, has advantages such as a large bandgap, high critical breakdown field strength, and high electron mobility, and has strong application potential in power device markets such as fast charging, data centers, OBCs, and solar inverters.

[0003] Currently, the main application of GaN in power devices is GaNHEMT devices. Since Khan et al. fabricated the first AlGaN / GaN high electron mobility transistor (HEMT) in 1993, horizontal GaNHEMT devices have attracted widespread attention due to their superior electrical performance and lower power consumption compared to Si devices. In 2005, Nitronex launched the first commercially available depletion-mode RF GaNHEMT device grown on a Si substrate, and in 2009, EPC launched the first enhancement-mode Si-based GaNHEMT device.

[0004] Although GaNHEMT devices offer superior performance compared to traditional Si devices, yield remains the biggest challenge in the fabrication of the most mainstream PGaN-enhanced GaNHEMT devices. This yield is a key factor limiting the large-scale application of PGaN-enhanced GaNHEMT devices. Currently, the yield of PGaN-enhanced GaNHEMT devices in domestic foundries is generally between 50-70%, with the most significant defects being IGSS and IDSS failures. Therefore, analyzing defective GaNHEMT devices to pinpoint the specific location of leakage paths is crucial. Traditional methods for detecting leakage paths caused by IGSS and IDSS failures in GaNHEMT devices involve the EBIRCH method, which requires microscopic analysis of the chip, resulting in high costs and long processing times. Therefore, providing a simple and rapid method to determine the location of leakage paths caused by IGSS and IDSS failures in GaNHEMT devices is both necessary and urgent. Summary of the Invention

[0005] To address the above problems, this invention provides a simple, quick, and low-cost method for detecting leakage current channels in GaN-HEMT devices.

[0006] The technical solution of this invention is: a method for detecting leakage current channels in GaN-HEMT devices, comprising the following steps: S100 sets a series of bias voltages and measures the gate drive leakage current of the device, denoted as IGSS; S200 sets a series of bias voltages and measures the saturation drain-source current of the device, denoted as IDSS; S300 determines whether a device has a leakage path based on the measured IGSS and IDSS values; S310. If the measured IGSS value exceeds the normal IGSS value of GaN-HEMT device in the standard specification by one order of magnitude, it is judged as an IGSS defective product, indicating that there is a leakage current path between its GS electrodes. S320. If the measured IDSS value exceeds the normal IDSS value of GaN-HEMT device in the standard specification by one order of magnitude, it is judged as an IDSS defective product, indicating that there is a leakage current channel between the two electrodes of its GD. S400 is used to detect the leakage current between the GS and SG electrodes; S410, the two electrodes of SD are not short-circuited; S420, set a series of bias voltages, the same as in step S100, between the two electrodes of GS and the two electrodes of SG; S430, record the leakage currents IGS and ISG obtained by measuring the two electrodes of GS and SG under a series of bias voltages respectively; S500 is used to detect the leakage current between the GD and DG electrodes. S510, the two electrodes of SD are not short-circuited; S520, a series of bias voltages, the same as those in step S200, are set between the two electrodes of GD and DG. S530, record the leakage currents IGD and IDG obtained by measuring the two electrodes of GD and DG under a series of bias voltages respectively; S600 is used to detect leakage current between the DS and SD electrodes; When detecting the leakage current between the two electrodes of S610 and DS, the two electrodes of SG should not be short-circuited; when detecting the leakage current between the two electrodes of SD and DG, the two electrodes of DG should not be short-circuited. S620, a series of bias voltages, the same as those in step S200, are set between the two electrodes DS and SD; The leakage currents IDS and ISD obtained by measuring the two electrodes DS and SD under a series of bias voltages of S630 are recorded respectively. S700, by comparing and measuring 8 sets of data, the leakage path of GaN-HEMT device was determined; S710, when the device IGSS is faulty, it indicates that there is a leakage path between the GS electrodes, such as... Figure 1 As shown, the forms in which the leakage current path exists between the two electrodes of GS include: S711, leakage channel 7 of SiN layer between GS above the device; S712, leakage channel 6 between the PGaN processing layer and the AlGaN layer below the GS in the middle of the device; S713, leakage channel 5 between the ALGaN layer and GaN layer below the device GS; S720, when the device IDSS is faulty, it indicates that there is a leakage path between the two electrodes of GD or between the D electrode and the substrate, as shown in the figure. The leakage path between the D electrode and the substrate is leakage path 1. The leakage path between the two electrodes of GD can be categorized as follows: S721, leakage channel 4 of SiN layer between the two electrodes of GD above the device; S722, leakage channel 3 between the PGaN treatment layer and the AlGaN layer below the two electrodes of GD in the middle of the device; S723, leakage channel 2 between the ALGaN layer and GaN layer between the two electrodes of GD below the device; S800, determine the location of the leakage current path between GS based on the values ​​of IGS and ISG in step S430; S810, if the difference between IGS, ISG, and IGSS under the same absolute bias voltage is within 100%, and the IV curves of IGS, ISG, and IGSS always show a linear relationship under the detection conditions, the difference in the linear correlation coefficient r between different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged, it indicates that there is no [symbol missing] between the two electrodes of GS and between the two electrodes of SG. Figure 2 The Schottky contact or PN rectifier contact shown indicates that there is a leakage current path 7 between the two electrodes GS at this time. S820, if the applied voltage falls before the voltage of PGaN and the gate metal reaches the Schottky barrier voltage of the structure, the difference between the IGS value and the ISG value under the same absolute bias voltage is within 100%, and the difference between the IGS value and the IGSS value is within 100%. After the applied voltage falls on the Schottky barrier of PGaN and the gate metal to reach 1-3V, the IGS value under the same absolute bias voltage is 1 / 100-1 / 10000 of the ISG value. The ISG IV curve is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IGS IV curve remains unchanged, indicating that from the S electrode to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the S electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel 6 between the GS electrodes at this time. S830, if the difference between the IGS and ISG values ​​is within 100% under the same absolute bias voltage, and the difference between the IGS and IGSS values ​​is within 100%, the IV curves of IGS and ISG remain unchanged under the detection conditions. For the IGS IV curve, when the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V, the linear correlation coefficient r rises rapidly. For the ISG IV curve, when the voltage falling on the Schottky barrier of PGaN and the gate metal reaches 1-3V, the linear correlation coefficient r rises rapidly. The IV curves of IGS and ISG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there are a series of voltage barriers between the two GS electrodes and the two SG electrodes. This is consistent with the situation that there is a leakage channel 5 between the two GS electrodes. Therefore, it can be determined that there is a leakage channel 5 between the two GS electrodes. S900, when IDSS is faulty, first determine whether there is a leakage path between the two electrodes of GD based on whether the difference between IDG value and IDSS value is within 100%. If the difference between the IDG value and the normal value of IDSS in the specification is within 100%, it indicates that there is a leakage current path between the two electrodes of GD. If the IGD value is more than one order of magnitude smaller than the normal IDSS value in the specification, it means that the leakage path causing the IDSS failure is not between the two electrodes of GD. After confirming the existence of a leakage current path between the two electrodes of GD, the location of the leakage current path between the two electrodes is then determined based on the values ​​of IGD and IDG, including the following cases: S910, under the same absolute bias voltage, the difference between IGD and IDG values ​​is within 100%, and the IV curves of IGD and IDG always show a linear relationship under the detection conditions. The difference in the linear correlation coefficient r between different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged. This indicates that there is no direct correlation between the two electrodes of GD and DG. Figure 2 The Schottky contact or PN rectifier contact shown indicates that there is a leakage current channel 4 between the two electrodes of GD at this time. S920, if the applied voltage falls on the voltage between PGaN and the gate metal before reaching the Schottky barrier voltage of the structure, the difference between the IGD value and the IDG value under the same absolute bias voltage is within 100%. After the applied voltage falls on the Schottky barrier of PGaN and the gate metal and reaches 1-3V, the IGD value under the same absolute bias voltage is 1 / 100-1 / 10000 of the IDG value. The IV curve of IDG is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IV curve of IGD remains unchanged, indicating that from the D to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the D electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel 3 between the two electrodes of GD at this time. S930, if the difference between IGD and IDG values ​​is within 100% under the same absolute bias voltage, the IV curves of IGD and IDG remain unchanged under the detection conditions. For the IV curve of IGD, the linear correlation coefficient r rises rapidly after the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V. For the IV curve of IDG, the linear correlation coefficient r rises rapidly after the voltage falling on the Schottky barrier of PGaN and gate metal reaches 1-3V. The IV curves of IGD and IDG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there is a series of voltage barriers in both GD and DG, which is consistent with the phenomenon of leakage channel 2 in GD. Therefore, it is determined that there is leakage channel 2 between the two electrodes of GD.

[0007] If IDSS is faulty and it is determined that there is no leakage path between the two electrodes of GD, the presence of leakage path 1 between the D electrode and the substrate is determined based on the values ​​of ISD and IDS in step S630. If the difference between the values ​​of ISD, IDS, and IDSS under the same absolute bias voltage in step S630 is within 100%, it can be determined that the IDSS abnormality is due to the presence of leakage path 1 between the D electrode and the substrate. Specifically, in step S100, the gate drive leakage current IGSS of the device is measured, including: S110, DS electrodes short-circuited; A series of bias voltages are applied between S120 and GS; S130 reads the IGSS of the device under different bias voltages.

[0008] Specifically, in step S200, measuring the device's saturated drain-source current IDSS includes: S210, GS electrodes short-circuited; A series of bias voltages are applied between S220 and DS; S230 reads the IDSS of the device under different bias voltages.

[0009] Specifically, the bias voltage in step S120 is between 0 and 5V.

[0010] Specifically, the bias voltage in step S220 is between 0 and 500V.

[0011] The present invention proposes a method for detecting leakage current channels in GaN-HEMT devices, which has the following advantages: it can be operated using a power device detector, avoiding the high cost of using expensive chip-level analytical instruments required by traditional methods such as EBIRCH, and is more convenient, faster, simpler and time-saving compared to traditional detection methods. Attached Figure Description

[0012] Figure 1 This is a distribution diagram of possible leakage current paths in defective GaN-HEMT devices of the present invention. Figure 2 This is a diagram showing the contact characteristics between layers of the GaN-HEMT device of the present invention. In the figure, 1 is the leakage channel between the drain electrode and the substrate; 2 is the leakage channel between the AlGaN and GaN layers between the drain electrodes below the device; 3 is the leakage channel between the PGaN treatment layer between the drain electrodes in the middle of the device and the AlGaN layer below it; 4 is the leakage channel between the SiN layer between the drain electrodes above the device; 5 is the leakage channel between the AlGaN and GaN layers between the bottom electrodes and the gate electrodes below the device; 6 is the leakage channel between the PGaN treatment layer between the bottom electrodes and the AlGaN layer below it; and 7 is the leakage channel between the SiN layer between the bottom electrodes and the gate electrodes above the device. Detailed Implementation

[0013] Embodiments of the present invention are described in detail below, with example values ​​for each embodiment given in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0014] Five PGaN-enhanced GaN-HEMT devices prepared by our research group were used as test subjects. Specific test data are shown in Tables 1-5. This test followed a method for detecting leakage current channels in GaN-HEMT devices disclosed in this patent, including the following steps: In step S100, the measurement of the gate drive leakage current IGSS of the measuring device includes: S110, DS electrodes short-circuited; A series of bias voltages are applied between S120 and GS; S130, read the IGSS of the device under different bias voltages. The IGSS values ​​of the device under different bias voltages are shown in Table 1: Table 1: Further, in step S200, the measurement of the saturated drain-source current IDSS of the measuring device includes: S210, GS electrodes short-circuited; A series of bias voltages are applied between S220 and DS; S230, read the IDSS of the device under different bias voltages. The IDSS values ​​of the device under different bias voltages are shown in Table 2: Table 2: Furthermore, in step S300, determining whether a leakage path exists in the GaN-HEMT device using IGSS and IDSS values ​​includes: S310, the IGSS data of the samples in Table 1 are compared with the normal IGSS values ​​of GaN-HEMT devices in the standard specification. It can be seen that the IGSS of samples 1, 3 and 4 are normal, while the IGSS of samples 2 and 5 are five orders of magnitude higher than the normal value, which are IGSS defective products. This indicates that there is a leakage channel between the two electrodes of GS. S320, Table 2 shows that the IDSS data of the samples are compared with the normal IDSS values ​​of GaN-HEMT devices in the standard specification. It can be seen that the IDSS of samples 2 and 4 are normal, while the IDSS of samples 1, 3 and 5 are five orders of magnitude larger than the normal value, which means that they are defective products. This indicates that there is a leakage channel between the two electrodes of GD or between the D electrode and the substrate.

[0015] Furthermore, in step S400, the detection of leakage current between the GS and SG electrodes includes: S410 and SD should not be shorted; S420, the series of bias voltages set between GS and SG are the same as in step S100; The leakage current between electrodes GS and SG in S430 is denoted as IGS and ISG. The values ​​of IGS and ISG under different bias voltage conditions are shown in Table 3. Table 3: Furthermore, in step S500, the detection of leakage current between the GD and DG electrodes includes: S510, SD should not be shorted; S520, the series of bias voltages set between GD and DG are the same as in step S200; The leakage current between electrodes GD and DG in S530 is denoted as IGD and IDG. The values ​​of IGD and IDG under different bias voltage conditions are shown in Table 4. Table 4: Furthermore, in step S600, the detection of leakage current between the DS and SD electrodes includes: When detecting the leakage current between the two electrodes of S610 and DS, SG should not be short-circuited; when detecting the leakage current between the two electrodes of SD, DG should not be short-circuited. S620, the series of bias voltages set between DS and SD are the same as in step S200; The leakage current between electrodes DS and SD of S630 is denoted as IDS and ISD. The values ​​of IDS and ISD under different bias voltage conditions are shown in Table 5. Further, in step S700, the comparison measurement of 8 sets of data to determine the leakage current path of the GaN-HEMT device includes: S710. First, based on the results in step S300, it is determined that there is a leakage channel between the GS electrodes of samples 2 and 5 in this test, and there is a leakage channel between the GD electrodes of samples 1, 3, and 5 or between the D electrode and the substrate. Among them, there is a leakage channel between the GS electrodes and the GD electrodes of sample 5 or between the D electrode and the substrate.

[0016] The leakage current path distribution diagram of the S720 GaN-HEMT device is shown below. Figure 1 As shown, a defective IGSS indicates a leakage current path between the two electrodes of the gas and metal electrodes (GS). The forms of this leakage current path between GS can be categorized as follows: 1. Leakage channel 7 in the SiN layer between GS and G on the device; 2. Leakage channel 6 between the PGaN processing layer and the underlying AlGaN layer in the middle of the device; 3. Leakage path between the ALGaN layer and GaN layer below the device.

[0017] IDSS defects indicate the presence of a leakage path between the two electrodes of the GD or between the D electrode and the substrate. The forms of leakage paths between GDs can be categorized as follows: 1. Leakage channel 4 in the SiN layer between GD and the device; 2. Leakage channel between the PGaN processing layer and the underlying AlGaN layer in the middle of the device; 3. Leakage channel 2 between the ALGaN layer and GaN layer between the GD electrode and the substrate below the device; the leakage channel between the D electrode and the substrate is in the form of leakage channel 1.

[0018] The contact characteristics between layers of the S730 GaN-HEMT device are as follows: Figure 2As shown, there is a Schottky contact between the gate metal and PGaN, and a PN rectifier contact is formed between AlGaN and GaN. For the different leakage channels mentioned above, the values ​​of the leakage current detected are different when the absolute values ​​of the bias voltages are applied between the GS and GD electrodes. Therefore, this principle can be used to determine the form of leakage channel in the device.

[0019] S740, based on S710, it is known that there is a leakage path between the GS of samples 2 and 5. Then, based on the IGS and ISG values ​​of samples 2 and 5 in Appendix 3, the location of the leakage path between GS is determined. At this time, there are three cases: 1. If the difference between IGS, ISG, and IGSS under the same absolute bias voltage is within 100%, and the IV curves of IGS, ISG, and IGSS always show a linear relationship under the detection conditions, the difference in the linear correlation coefficient r between different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged, it indicates that there is no [correlation] between the two electrodes of GS and between the two electrodes of SG. Figure 2 The Schottky contact or PN rectifier contact shown indicates that there is a leakage current path 7 between the two electrodes GS at this time. 2. If the applied voltage falls before the voltage between PGaN and the gate metal reaches the Schottky barrier voltage of the structure, the difference between the IGS value and the ISG value is within 100% under the same absolute bias voltage, and the difference between the IGS value and the IGSS value is within 100%. After the applied voltage falls on the Schottky barrier of PGaN and the gate metal to reach 1-3V, the IGS value is 1 / 100-1 / 10000 of the ISG value under the same absolute bias voltage. The ISG IV curve is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IGS IV curve remains unchanged, indicating that from the S electrode to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the S electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel between the GS electrodes at this time. 3. If the difference between the IGS and ISG values ​​is within 100% under the same absolute bias voltage, and the difference between the IGS and IGSS values ​​is within 100%, the IV curves of IGS and ISG remain unchanged under the detection conditions. For the IGS IV curve, the linear correlation coefficient r rises rapidly after the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V. For the ISG IV curve, the linear correlation coefficient r rises rapidly after the voltage falling on the Schottky barrier of PGaN and the gate metal reaches 1-3V. The IV curves of IGS and ISG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there are a series of voltage barriers between the two GS electrodes and the two SG electrodes. This is consistent with the situation that there is a leakage channel 5 between the two GS electrodes. Therefore, it can be determined that there is a leakage channel 5 between the two GS electrodes. Comparing the data in Appendix Table 3, we can see that Sample 2 conforms to Case 1 of this analysis. It can be concluded that the IGSS defect of Sample 2 is due to the leakage channel 7 of the SiN layer between the gate and gate (GS). This is caused by residual metal bridging between the GS during the gate metal stripping process in the chip manufacturing process. Sample 5 conforms to Case 2 of this analysis. The Schottky contact barrier voltage of this device is between 1.5V and 2V. It can be concluded that the defect of Sample 5 is due to the leakage channel 6 between the PGaN processing layer and the underlying AlGaN layer between the GS. This is due to GaN wafer warping. Incomplete PGaN passivation during the PGaN passivation process of some samples led to the appearance of leakage channel 6.

[0020] S750, if IDSS is faulty, first determine if there is a leakage path between GDs. Firstly, check if the difference between the IDG and IDSS values ​​in step S500 is within 100% to determine if a leakage path exists between GDs. If the difference between the IDG value and the normal IDSS value in the specification is within 100%, it indicates a leakage path exists between GDs. If the IGD value is more than one order of magnitude smaller than the normal IDSS value in the specification, it indicates that the leakage path causing the IDSS fault is not between GDs. After determining that a leakage path exists between GDs, then determine the location of the leakage path between GDs based on the IGD and IDG values ​​in Appendix 4. As above, there are three possibilities: 1. Under the same absolute bias voltage, the difference between IGD and IDG values ​​is within 100%, and the IV curves of IGD and IDG always show a linear relationship under the detection conditions. The difference in the linear correlation coefficient r between different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged. This indicates that there is no direct correlation between the two electrodes of GD and DG. Figure 2 The Schottky contact or PN rectifier contact shown indicates that there is a leakage current channel 4 between the two electrodes of GD at this time. 2. If the applied voltage falls on the voltage between PGaN and the gate metal before reaching the Schottky barrier voltage of the structure, the difference between the IGD value and the IDG value under the same absolute bias voltage is within 100%. After the applied voltage falls on the Schottky barrier of PGaN and the gate metal and reaches 1-3V, the IGD value under the same absolute bias voltage is 1 / 100-1 / 10000 of the IDG value. The IV curve of IDG is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IV curve of IGD remains unchanged, indicating that from the D to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the D electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel between the two electrodes of GD at this time. 3. If the difference between IGD and IDG values ​​is within 100% under the same absolute bias voltage, the IV curves of IGD and IDG remain unchanged under the detection conditions. For the IV curve of IGD, the linear correlation coefficient r rises rapidly after the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V. For the IV curve of IDG, the linear correlation coefficient r rises rapidly after the voltage falling on the Schottky barrier of PGaN and gate metal reaches 1-3V. The IV curves of IGD and IDG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there is a series of voltage barriers in both GD and DG, which is consistent with the phenomenon of leakage channel 2 in GD. Therefore, it is determined that there is leakage channel 2 between the two electrodes of GD.

[0021] Comparing the data in Appendix Table 4, samples 1, 3, and 5 all meet condition 2 in this section. The Schottky contact barrier voltage of this device is between 1.5V and 2V. It can be concluded that the defects of samples 1, 3, and 5 are due to leakage channels 3 between the PGaN processing layer and the underlying AlGaN layer in the GD process. This is because the GaN wafer is warped, and the PGaN passivation process is incomplete in some samples, resulting in the appearance of leakage channels 3.

[0022] S760, if IDSS is too high and there is no leakage path between GD, determine whether there is a leakage path 1 between D and substrate based on the ISD and IDS values ​​in step S600. If the difference between the ISD, IDS, and IDSS values ​​in step S600 is within 100%, it can be determined that the IDSS abnormality is due to the existence of a leakage path 1 between D and substrate. S750 shows that the IDSS defects in samples 1, 3, and 5 are due to the presence of leakage path 3. Therefore, no leakage path 3 from D to substrate was found in the tested samples. In reality, the probability of a leakage path between D and substrate is extremely low. Leakage path 1 only occurs when some areas of the epitaxial wafer do not meet the standards, resulting in epitaxial wafer breakdown. However, epitaxial wafers undergo rigorous testing before leaving the factory, so such cases are extremely rare. Step S760 is for the detection of the case where IDSS is abnormal in S750 but no leakage path between GD is found. In Table 5, the IDS and ISD of sample 5 are two orders of magnitude larger than those of samples 1 and 3. This is because there is a leakage channel between the GD and GS of sample 5 and the underlying AlGaN layer. These two leakage channels are connected on the same horizontal plane.

[0023] According to the method for detecting leakage channels in GaN-HEMT devices disclosed in this patent, five prepared samples were tested, and the conclusions were drawn: the IGSS defect of sample 2 was due to a leakage channel 7 between the SiN layer and the GS layer; the IGSS defect of sample 5 was due to a leakage channel 6 between the PGaN treatment layer and the underlying AlGaN layer and the GS layer; the IDSS defects of samples 1, 3, and 5 were due to a leakage channel 3 between the PGaN treatment layer and the underlying AlGaN layer and the GD layer; sample 4 was a good product.

[0024] Further define the bias voltage range and step size in step S120; In step S120, the bias voltage range is generally between 0-5V. In special cases, it can be further relaxed to the standard gate breakdown voltage of the device. The voltage range step can be set according to the actual situation, but it is best not to exceed 1V.

[0025] Further define the bias voltage range and step size in step S220; In step S220, the bias voltage range is generally between 0-500V. In special cases, it can be further relaxed to the device's nominal maximum operating blocking voltage. The voltage range step can be set according to the actual situation, but it is best not to exceed 1000V.

[0026] Regarding the information disclosed in this case, the following points need to be clarified: (1) The embodiments disclosed in this case only involve the leakage channels of the five samples involved in the embodiments disclosed in this case. Other possible forms of leakage channels are not mentioned in this article due to factors such as the number of samples tested and the probability of different leakage channels appearing, but this does not affect the implementation of the detection method. (2) Where there is no conflict, the embodiments and features disclosed in this case can be combined with each other to obtain new embodiments; The above are merely specific embodiments disclosed in this case, but the scope of protection of this disclosure is not limited thereto. The scope of protection disclosed in this case shall be determined by the scope of protection of the claims.

Claims

1. A method for detecting leakage current channels in GaN-HEMT devices, characterized in that, Includes the following steps: S100 sets a series of bias voltages and measures the gate drive leakage current of the device, denoted as IGSS; S200 sets a series of bias voltages and measures the saturation drain-source current of the device, denoted as IDSS; S300 determines whether a device has a leakage path based on the measured IGSS and IDSS values; S310. If the measured IGSS value exceeds the normal IGSS value of GaN-HEMT device in the standard specification by one order of magnitude, it is judged as an IGSS defective product, indicating that there is a leakage current path between its GS electrodes. S320. If the measured IDSS value exceeds the IDSS value of GaN-HEMT device in the standard specification by one order of magnitude, it is judged as an IDSS defective product, indicating that there is a leakage current channel between the two electrodes of its GD. S400 is used to detect the leakage current between the GS and SG electrodes; S410, the two electrodes of SD are not short-circuited; S420, set a series of bias voltages, the same as in step S100, between the two electrodes of GS and the two electrodes of SG; S430, record the leakage currents IGS and ISG obtained by measuring the two electrodes of GS and SG under a series of bias voltages respectively; S500 is used to detect the leakage current between the GD and DG electrodes. S510, the two electrodes of SD are not short-circuited; S520, a series of bias voltages, the same as those in step S200, are set between the two electrodes of GD and DG. S530, record the leakage currents IGD and IDG obtained by measuring the two electrodes of GD and DG under a series of bias voltages respectively; S600 is used to detect leakage current between the DS and SD electrodes; When detecting the leakage current between the two electrodes of S610 and DS, the two electrodes of SG should not be short-circuited; when detecting the leakage current between the two electrodes of SD, the two electrodes of DG should not be short-circuited. S620, a series of bias voltages, the same as those in step S200, are set between the two electrodes DS and SD; The leakage currents IDS and ISD obtained by measuring the two electrodes DS and SD under a series of bias voltages of S630 are recorded respectively. S700, by comparing and measuring 8 sets of data, the leakage path of GaN-HEMT device was determined; S710, when the device IGSS is faulty, it indicates that there is a leakage path between the GS electrodes. The forms in which the leakage path exists between the GS electrodes include: S711, leakage channel 7 of SiN layer between GS above the device; S712, leakage channel 6 between the PGaN processing layer and the AlGaN layer below the GS in the middle of the device; S713, leakage channel 5 between the ALGaN layer and GaN layer below the device GS; S720, when the device IDSS is faulty, it indicates that there is a leakage path between the two electrodes of GD or between the D electrode and the substrate. The leakage path between the D electrode and the substrate is leakage path 1. The leakage path between the two electrodes of GD can be categorized as follows: S721, leakage channel 4 of SiN layer between the two electrodes of GD above the device; S722, leakage channel 3 between the PGaN treatment layer and the AlGaN layer below the two electrodes of GD in the middle of the device; S723, leakage channel 2 between the ALGaN layer and GaN layer between the two electrodes of GD below the device; S800, determine the location of the leakage current path between GS based on the values ​​of IGS and ISG in step S430; S810, if the difference between IGS, ISG and IGSS under the same absolute bias voltage is within 100%, and the IV curves of IGS, ISG and IGSS under the detection conditions are always linearly related, the difference of the linear correlation coefficient r of different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged, it indicates that there is no Schottky contact or PN rectifier contact between the two electrodes of GS and the two electrodes of SG. Therefore, it can be determined that there is a leakage current channel 7 between the two electrodes of GS at this time. S820, if the applied voltage falls on the voltage between PGaN and the gate metal before reaching the Schottky barrier voltage of PGaN and the gate metal, the difference between the IGS value and the ISG value under the same absolute bias voltage is within 100%, and the difference between the IGS value and the IGSS value is within 100%. After the applied voltage falls on the Schottky barrier voltage of PGaN and the gate metal to reach 1-3V, the IGS value under the same absolute bias voltage is 1 / 100-1 / 10000 of the ISG value. The ISG IV curve is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IGS IV curve remains unchanged, indicating that from the S electrode to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the S electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel 6 between the GS electrodes at this time. S830, if the difference between the IGS and ISG values ​​is within 100% under the same absolute bias voltage, and the difference between the IGS and IGSS values ​​is within 100%, the IV curves of IGS and ISG remain unchanged under the detection conditions. For the IGS IV curve, when the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V, the linear correlation coefficient r rises rapidly. For the ISG IV curve, when the voltage falling on the Schottky barrier of PGaN and the gate metal reaches 1-3V, the linear correlation coefficient r rises rapidly. The IV curves of IGS and ISG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there are a series of voltage barriers between the two GS electrodes and the two SG electrodes. This is consistent with the situation that there is a leakage channel 5 between the two GS electrodes. Therefore, it can be determined that there is a leakage channel 5 between the two GS electrodes. S900, when IDSS is faulty, first determine whether there is a leakage path between the two electrodes of GD based on whether the difference between IDG value and IDSS value is within 100%. If the difference between the IDG value and the normal value of IDSS in the specification is within 100%, it indicates that there is a leakage current path between the two electrodes of GD. If the IGD value is more than one order of magnitude smaller than the normal IDSS value in the specification, it means that the leakage path causing the IDSS failure is not between the two electrodes of GD. After confirming the existence of a leakage current path between the two electrodes of GD, the location of the leakage current path between the two electrodes is then determined based on the values ​​of IGD and IDG, including the following cases: S910, under the same absolute bias voltage, the difference between IGD and IDG values ​​is within 100%, and the IV curves of IGD and IDG under the detection conditions always show a linear relationship. The difference in the linear correlation coefficient r of different curves is within 50%, and the linear correlation coefficient r of the same curve remains unchanged. This indicates that there is no Schottky contact or PN rectifier contact between the two electrodes of GD and DG. Therefore, it can be determined that there is a leakage current channel 4 between the two electrodes of GD at this time. S920, if the applied voltage falls on the voltage between PGaN and the gate metal before reaching the Schottky barrier voltage of PGaN and the gate metal, the difference between the IGD value and the IDG value under the same absolute bias voltage is within 100%. After the applied voltage falls on the Schottky barrier voltage of PGaN and the gate metal and reaches 1-3V, the IGD value under the same absolute bias voltage is 1 / 100-1 / 10000 of the IDG value. The IV curve of IDG is divided into two stages, and the linear correlation coefficient r of the two stages is increasing. The linear correlation coefficient r of the IV curve of IGD remains unchanged, indicating that from the D to the G electrode is the positive bias voltage direction of the Schottky contact, and from the G electrode to the D electrode is the negative bias voltage direction of the Schottky contact. Therefore, it can be determined that there is a leakage channel 3 between the two electrodes of GD at this time. S930, if the difference between IGD and IDG values ​​is within 100% under the same absolute bias voltage, the IV curves of IGD and IDG remain unchanged under the detection conditions. For the IV curve of IGD, the linear correlation coefficient r rises rapidly after the voltage falling on the PN junction barrier of PGaN and AlGaN reaches 1-3V. For the IV curve of IDG, the linear correlation coefficient r rises rapidly after the voltage falling on the Schottky barrier of PGaN and gate metal reaches 1-3V. The IV curves of IGD and IDG are divided into two stages, and the linear correlation coefficient r of the two stages is increasing. This indicates that there is a series of voltage barriers in both GD and DG, which is consistent with the phenomenon of leakage channel 2 in GD. Therefore, it is determined that there is leakage channel 2 between the two electrodes of GD. If IDSS is faulty and it is determined that there is no leakage path between the two electrodes of GD, the value of ISD and IDS in step S630 is used to determine whether there is a leakage path 1 between the D electrode and the substrate. If the difference between the values ​​of ISD, IDS and IDSS under the same absolute bias voltage in step S630 is within 100%, it can be determined that the IDSS abnormality is due to the existence of a leakage path 1 between the D electrode and the substrate.

2. The method for detecting leakage current channels in a GaN-HEMT device according to claim 1, characterized in that, In step S100, the gate drive leakage current of the device is measured, including: S110, DS electrodes short-circuited; A series of bias voltages are applied between S120 and GS; S130 reads the IGSS of the device under different bias voltages.

3. The method for detecting leakage current channels in a GaN-HEMT device according to claim 1, characterized in that, In step S200, the saturation drain-source current IDSS of the device is measured, including: S210, GS electrodes short-circuited; A series of bias voltages are applied between S220 and DS; S230 reads the IDSS of the device under different bias voltages.

4. The method for detecting leakage current channels in a GaN-HEMT device according to claim 2, characterized in that, The bias voltage in step S120 is between 0 and 5V.

5. The method for detecting leakage current channels in a GaN-HEMT device according to claim 3, characterized in that, The bias voltage in step S220 is between 0 and 500V.

Images