J. Semicond. > 2015, Volume 36 > Issue 7 > 074005

SEMICONDUCTOR DEVICES

Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress

Lei Shi, Shiwei Feng, Kun Liu and Yamin Zhang

+ Author Affiliations
DOI: 10.1088/1674-4926/36/7/074005

PDF

Abstract: The phenomenon of self-changing on the device parameters and characteristics after a step voltage stress was applied to the gate is studied in AlGaN/GaN high electron mobility transistors. The device was measured every 5 min after the stress was removed. The large-signal parasitic source (drain) resistance, transfer characteristics, threshold voltage, drain-source current, gate-source (drain) reverse current-voltage characteristics changed spontaneously after the removal of the stress. The time constant of the self-changing was about 25-27 min. The gate-source (drain) capacitance-voltage characteristics were constant during this process. Electrons were trapped by the surface states and traps in the AlGaN barrier layer when the device was under stress. The traps in the AlGaN barrier layer then released electrons in less than 10 s. The surface states released electrons continuously during the entire measurement stage, leading to the self-changing of mearsurement result.

Key words: AlGaN/GaN HEMTssurface statesself-changingstress

AlGaN/GaN high electron mobility transistors (HEMTs) have shown extremely excellent performance in high frequency and high power applications because of their superior properties, such as their high polarization-induced 2-D electron gas (2DEG) density, high electron saturation velocity, and high critical breakdown electric field[1, 2, 3, 4]. There are many researchers studying the degradation mechanism and reliability problems, especially the reliability of the gate. Degradation caused by a reverse bias between the gate and drain (source) is an important research topic. There have been several reports on the reliability of the electrodes under electric stress. Joh et al. proposed a degradation mechanism that involved the formation of defects through the inverse piezoelectric effect, which was caused by a high vertical electric field at the drain edge of the gate[5, 6]. Dammann et al. proposed that the degradation of the drain and gate currents were likely caused by hot-electron induced trap generation under high drain voltages[7]. It was found that the surface states between the gate and drain could trap electrons that tunneled from the gate under a high drain-gate voltage in the OFF-state[8, 9, 10].

Most research has focused on the behaviour of devices under stress. More study is needed on the intermediate/long-term behaviour of devices after the stress is removed. In this work, self-changing of the device parameters and characteristics have been observed over a relatively long period of time after the removal of a reverse bias voltage stress. The effects of the surface states and traps in the AlGaN barrier layer are analyzed during this process, and the time constant of the change is obtained by exponential fit. The effects caused by a reverse bias on the device reliability are studied from a different perspective.

The samples were grown by metal-organic chemical vapor deposition (MOCVD) on SiC substrates, which had an epitaxial structure consisting of a GaN buffer layer followed by an AlGaN barrier layer, as shown in Figure 1. Electron beam-evaporation was used to deposit the Ti/Al/Ni/Au source/drain ohmic contacts and Ni/Au gate electrodes. The samples had gate lengths of 1.1 μm and widths of 150 μm, respectively. The spacings of the gate-drain and gate-source were 2.4 μm and 1.5 μm, respectively. The surfaces of the devices were passivated with a 100 nm thick layer of Si3N4.

Figure  1.  Cross section of device structure.

A negative bias was applied to the gate (VG), from 8 to 70 V in 1 V steps, while the source and drain were grounded (VS=VD = 0 V). Each voltage lasted for 100 s, then the voltage stress was removed for 200 s before the next voltage was applied. The 70 V bias stress was removed at time = 0 s, then the device parameters and characteristics were measured every 5 min from time = 0 to 120 min.

An Agilent B1500A semiconductor parameter analyzer was used to perform the stress and measurements. The parameters and characteristics measured included the large-signal parasitic source resistance (RS), the large-signal parasitic drain resistance (RD), the transfer characteristics between the drain-source current IDS and the gate-source voltage VGS (the drain-source voltage VDS = 5 V, VGS = -2 to -0.5 V, step = 0.005 V), the drain-source current (IDmax) (VGS = 0 V, VDS = 5 V), the reverse current-voltage (I-V) characteristics of the gate-source (VGS = 5 to 0 V, step = 0.05 V), the reverse I-V characteristics of the gate-drain (the drain-source voltage VGD = 5 to 0 V, step = 0.05 V), the capacitance-voltage (C-V) characteristics of the gate-source (VGS = 5 to 0 V, step = 0.05 V), and the C-V characteristics of the gate-drain (VGD = 5 to 0 V, step = 0.05 V). The temperature was maintained at 20 . To reduce electromagnetic interference, the measurements were taken in electromagnetic shielding equipment.

Figure 2 shows the changing trends for RS and RD, which were normalized by the values measured before the test. RS and RD decreased after the stress was removed. This decrease was fast in the early stage, slowing in the later stage. To obtain a quantitative analysis of the changing process, RS and RD were fitted to exponentials (Figure 2), showing that the time constants (t1) for RS were similar to those for RD.

Figure  2.  The changes of the normalized RS and RD with time (time = 0-120 min, step = 5 min). The data were fitted with exponential function: y=y0+A1exp((xx0)/t1), where y represents RS and RD, and x represents time. The time constants (t1) for curves C and D were 25.2 min and 25.6 min, respectively.

To analyze this phenomenon, possible effects from the stress need to be taken into consideration. There are three main mechanisms in GaN HEMTs: the hot-electron effect, inverse piezoelectric effect, and gate electron injection[5, 6, 7, 8, 9, 10]. In consideration of the three reasons: (1) the pinch-off voltage of the sample is about 2 V, its magnitude is far less than the stress voltage, which is in the range of 8 to 70 V. Thus, 2DEG in the channel was depleted with the stress applied on the gate; (2) VDS = 0 when the device was stressed; (3) the temperature was maintained at 20 , the channel hot-electron effect was not significant in this experiment. The latter two mechanisms impacted on the device during the stress stage. The degradation under stress and the self-changing of the device parameters and characteristics after stress may have resulted from one of the two mechanisms, a combination of both, or other sources such as inherent traps in the AlGaN barrier layer. Whatever the causes are, the electrons were trapped by the surface states at the AlGaN barrier layer and/or by traps in the AlGaN barrier layer. Hot-electrons trapped in the GaN layer were ignored because the hot-electron effect was negligible.

Because the temperature was maintained at 20 and there was no drain-source current (IDS) during the stress stage, the ohmic contact of the source did not degrade under stress[11]. The changes in RS, caused by changes in the resistance of the 2DEG channel under the gap between the gate and source, meant that the 2DEG density decreased in this area. Note that a gate current injection method was used to test RS and RD, as shown in the inset, a small current IS (ID) was injected from gate to source (drain) and VDS (VSD) was measured simultaneously, then the values of RS and RD can be found from formulae: RS=VDS/IS, RD=VSD/ID[12, 13]. This method led to detrapping, which contributed to the release of electrons that were trapped in the AlGaN barrier layer. Except the measurements of RS and RD, the other measurements were performed in the form of pulses, and the bias in these measurements was less than -5 V, which was far less than the stress voltage, so the influences of these measurements themselves were not significant on the results. Joh and Del Alamo showed that most traps in the AlGaN barrier layer released electrons in the range of 101 to 101 s[14]. However, in Figure 1, the time constant (t1) in which RS quickly decreased was much longer than 101 s, and changes in the RS lasted longer than 120 min. Changes in the 2DEG density during the process could not be explained by the traps in the AlGaN barrier layer that released electrons. Now the surface states that caused this phenomenon will be considered. Electrons trapped by the surface states between the gate and the source depleted 2DEG in the channel under this area during the stress stage[8]. The surface states gradually released electrons after the voltage stress had been removed. Most of the electrons were released in t1. Correspondingly, the 2DEG density was recovered. Changes in the RD were caused by the surface states between the gate and drain, which also released electrons, this is the same as the analysis of RS.

Figure 3 shows the transfer characteristics, IDS-VGS curves moved to the left continuously after the stress. The change of threshold voltage (VGS(th)) (VGS at IDS = 1 mA/mm) was obtained from the transfer characteristics. VGS(th) and IDmax showed similar trends to RS and RD, except that the magnitude of VGS(th) and IDmax increased, while RS and RD decreased. Furthermore, their time constants, t1, were similar, showing that they were induced by the same mechanism. The changes of VGS(th) and IDmax lasted much longer than 10 s. The same as the analysis of RS and RD, it is considered that it was not the traps in the AlGaN barrier layer, but the surface states that caused their self-changing phenomena.

Figure  3.  Transfer characteristics at time = 0, 10, 20, 30, 60, 120 min. (VDS = 5 V, VGS = 2 to 0.5 V, step = 0.005 V). The inset shows VGS(th) (VGS at IDS = 1 mA/mm) and normalized IDmax (VDS = 5 V, VGS = 0 V) (time = 0-120 min, step is 5 min). The curves were fitted with the exponential equation: y=y0+A1exp((xx0)/t1), where y represents VGS(th), IDmax and x represents time. The time constants, t1 for curves G and H were 27.7 min and 27.2 min, respectively.

A band diagram shows the effect of electrons trapped by surface states in Figure 4. Here, the conducting band discussed is around the position of the 2DEG channel. The conducting band bent upward because the electrons trapped in the surface states during stress stage induced an electric field, shown in Figure 4(a). The surface states gradually released electrons after the voltage stress was removed, so the effect of the electric field decreased, the conducting band bent downward, shown in Figure 4(b). When time = 0 min, the 2DEG density reached the threshold value at VGS=VGS(th)(0min). When time = 120~min, 2DEG density exceeded the threshold value at VGS=VGS(th)(0min) because of the lower position of conducting band. To make the 2DEG density reach the threshold value by raising the conducting band up, the magnitude of VGS needs to increase to VGS(th)(120min). From 0 to 120 min, the conducting band bent downword continuously with the decreasing of the electric field intensity. Thus, the magnitude of VGS(th) increased correspondingly. The change of IDmax, which was caused by the change of 2DEG density, could be explained by the same mechanism above.

Figure  4.  Band diagram of the AlGaN/GaN heterostructure showing the effect of electrons trapped by surface states. (a) Time = 0 min. (b) Time = 120 min.

Figure 5 shows the self-changing phenomenon on the reverse I-V characteristics of the gate-source. The magnitude of the reverse gate-source current (IGS) increased with time. The reverse I-V characteristics of the gate-drain were similar to that of the gate-source, they were not plotted for simplicity. The time constants of reverse gate-drain current (IGD) were about 27 min. Reverse IGS and IGD showed similar trends to RS, RD, VGS(th), IDmax. Furthermore, their time constants, t1, were all about 25-27 min, showing that they were induced by the same mechanism again. The electric field caused by electrons trapped in the surface states had opposite direction relative to reverse IGS (IGD) in the measurement. So the flow of reverse IGS and IGD were impeded by the electric field. As the analysis above, the electric field decreased during the 120 min, thus, the reverse IGS and IGD increased correspondingly.

Figure  5.  Reverse I-V characteristics between the gate and source at time = 0, 10, 20, 30, 60, 120 min. (VGS = 5 to 0 V, step = 0.05~V). The inset shows the IGS-time curves at VGS = 5 V and 4.5~V (time = 0-120 min, step is 5 min). The curves were fitted with the exponential equation: y=y0+A1exp((xx0)/t1), where y represents IGS and x represents time. The time constants, t1 for curves K and L were 25.7 min and 26.5 min, respectively.

C-V meausurement could show traps variation in the AlGaN barrier layer more directly relative to I-V measurement. Despite the fact that the distance from the drain to the gate was longer than that from the source to the gate, the HEMT structure was essentially symmetric. The capacitance of the gate-source (drain) (Cgs (Cgd)) is depicted by the cross sectional layout of the sample structure in Figure 6. Cgs1 (Cgd1) is the inter-electrode capacitance between the electrodes. Cgs2 (Cgd2) is the capacitance from the gate to the source (drain) through the passivation layer and the surface of the AlGaN barrier layer. Cgs3 (Cgd3) is the capacitance from the gate to the source (drain) through the main body of the AlGaN barrier layer and the 2DEG channel. Cgs and Cgd both contain the three components above. The main parameters are Cgs3 and Cgd[15].

Figure  6.  Schematic diagram of the gate-source capacitance and the gate-drain capacitance in the sample.

In Figure 7, with different bias voltages, the C-V characteristic curves included three regions, which are indicated by the dotted lines. They are region I (the cutoff region), region II (the moderate 2DEG region) and region III (the strong 2DEG region)[15, 16]. In regions II and III, the influence of the capacitance between the gate and drain was introduced to Cgs because the 2DEG channel connected drain and source. Thus, region I was focused on, which represented the characteristics of the gate-source. In region I, Cgs3 was the capacitance from the gate to the source through the main body of the AlGaN barrier layer without the 2DEG channel. The inset in Figure 7 shows the properties of region I. The slopes of lines O and P had a magnitude of 105 and showed that Cgs was constant at different times in region I. This meant that for Cgs3, the main part of Cgs was constant, correspondingly. If the amount of traps in the AlGaN barrier layer under the gap between the gate and the source changed during the measurements, Cgs3 would change correspondingly. It was considered that the traps in this area had released electrons in the interval between the stress and the C-V measurement, which lasted for no longer than 10 s, then Cgs3 was constant in the measurement stage. C-V characteristics of the gate-drain were not plotted for simplicity. Like Cgs3, Cgd3 was constant too, which could be analysed in the same way. The C-V measurements showed that the surface states were the only possible reason for the self-changing phenomenon in these measurements.

Figure  7.  C-V characteristics between the gate and source at time = 0, 10, 20, 30, 60, 120 min. (VGS = 5 V, step = 0.05 V). The inset shows the Cgs-time curves at VGS = 5 V and 4.5 V (time = 0-120~min, step = 5 min). The curves were linear and were fitted with the equation: y=a+bx, where y represents Cgs and x represents time. The slopes (b) of lines O and P were 1.22 × 105 and 1.40 × 105, respectively.

For AlGaN/GaN HEMTs, a reverse step voltage stress between the gate and source (drain) degraded the sample. The phenomenon of self-changing happened on the device parameters and characteristics after the stress. Electrons were trapped by the surface states and traps in the AlGaN barrier layer during the stress stage. After removal of the stress, the traps in the AlGaN barrier released electrons over a short period of time no longer than 10 s, assisted by the RS and RD test. The surface states released electrons over a period of more than 120 min, with most being released in the first 25 to 27~min. The behaviours of the RS, RD, transfer characteristics, threshold voltage, drain-source current, reverse I-V characteristics and C-V characteristics between the gate and source (drain) after stress were consistent, and were not caused by the traps in the AlGaN barrier layer, but by the surface states. The effects of most degradation mechanisms were immediate or short-term, while the surface states had a long-term effect on the AlGaN/GaN HEMTs. More than 7200 s was required to find out the degradation caused by the surface states. The finding could be used to ascertain the effect of voltage stress on HEMTs exactly.



[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Fig. 1.  Cross section of device structure.

Fig. 2.  The changes of the normalized RS and RD with time (time = 0-120 min, step = 5 min). The data were fitted with exponential function: y=y0+A1exp((xx0)/t1), where y represents RS and RD, and x represents time. The time constants (t1) for curves C and D were 25.2 min and 25.6 min, respectively.

Fig. 3.  Transfer characteristics at time = 0, 10, 20, 30, 60, 120 min. (VDS = 5 V, VGS = 2 to 0.5 V, step = 0.005 V). The inset shows VGS(th) (VGS at IDS = 1 mA/mm) and normalized IDmax (VDS = 5 V, VGS = 0 V) (time = 0-120 min, step is 5 min). The curves were fitted with the exponential equation: y=y0+A1exp((xx0)/t1), where y represents VGS(th), IDmax and x represents time. The time constants, t1 for curves G and H were 27.7 min and 27.2 min, respectively.

Fig. 4.  Band diagram of the AlGaN/GaN heterostructure showing the effect of electrons trapped by surface states. (a) Time = 0 min. (b) Time = 120 min.

Fig. 5.  Reverse I-V characteristics between the gate and source at time = 0, 10, 20, 30, 60, 120 min. (VGS = 5 to 0 V, step = 0.05~V). The inset shows the IGS-time curves at VGS = 5 V and 4.5~V (time = 0-120 min, step is 5 min). The curves were fitted with the exponential equation: y=y0+A1exp((xx0)/t1), where y represents IGS and x represents time. The time constants, t1 for curves K and L were 25.7 min and 26.5 min, respectively.

Fig. 6.  Schematic diagram of the gate-source capacitance and the gate-drain capacitance in the sample.

Fig. 7.  C-V characteristics between the gate and source at time = 0, 10, 20, 30, 60, 120 min. (VGS = 5 V, step = 0.05 V). The inset shows the Cgs-time curves at VGS = 5 V and 4.5 V (time = 0-120~min, step = 5 min). The curves were linear and were fitted with the equation: y=a+bx, where y represents Cgs and x represents time. The slopes (b) of lines O and P were 1.22 × 105 and 1.40 × 105, respectively.

[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
1

Effect of the post-gate annealing on the gate reliability of AlGaN/GaN HEMTs

Changxi Chen, Quan Wang, Wei Li, Qian Wang, Chun Feng, et al.

Journal of Semiconductors, 2021, 42(9): 092802. doi: 10.1088/1674-4926/42/9/092802

2

Influence of field plate on surface-state-related lag characteristics of AlGaN/GaN HEMT

Yong Lei, Hai Lu

Journal of Semiconductors, 2015, 36(7): 074007. doi: 10.1088/1674-4926/36/7/074007

3

A gate current 1/f noise model for GaN/AlGaN HEMTs

Yu'an Liu, Yiqi Zhuang

Journal of Semiconductors, 2014, 35(12): 124005. doi: 10.1088/1674-4926/35/12/124005

4

Hysteresis analysis of graphene transistor under repeated test and gate voltage stress

Jie Yang, Kunpeng Jia, Yajuan Su, Yang Chen, Chao Zhao, et al.

Journal of Semiconductors, 2014, 35(9): 094003. doi: 10.1088/1674-4926/35/9/094003

5

Fluorine-plasma surface treatment for gate forward leakage current reduction in AlGaN/GaN HEMTs

Wanjun Chen, Jing Zhang, Bo Zhang, Kevin Jing Chen

Journal of Semiconductors, 2013, 34(2): 024003. doi: 10.1088/1674-4926/34/2/024003

6

A Ku band internally matched high power GaN HEMT amplifier with over 30% of PAE

Ge Qin, Chen Xiaojuan, Luo Weijun, Yuan Tingting, Pu Yan, et al.

Journal of Semiconductors, 2012, 33(1): 014003. doi: 10.1088/1674-4926/33/1/014003

7

Effect of varying layouts on the gate temperature for multi-finger AlGaN/GaN HEMTs

Wang Jianhui, Wang Xinhua, Pang Lei, Chen Xiaojuan, Jin Zhi, et al.

Journal of Semiconductors, 2012, 33(9): 094004. doi: 10.1088/1674-4926/33/9/094004

8

Extrinsic and intrinsic causes of the electrical degradation of AlGaN/GaN high electron mobility transistors

Fang Yulong, Dun Shaobo, Liu Bo, Yin Jiayun, Cai Shujun, et al.

Journal of Semiconductors, 2012, 33(5): 054005. doi: 10.1088/1674-4926/33/5/054005

9

A Ku-band high power density AlGaN/GaN HEMT monolithic power amplifier

Ge Qin, Chen Xiaojuan, Luo Weijun, Yuan Tingting, Pang Lei, et al.

Journal of Semiconductors, 2011, 32(8): 085001. doi: 10.1088/1674-4926/32/8/085001

10

High performance AlGaN/GaN HEMTs with 2.4 μm source–drain spacing

Wang Dongfang, Wei Ke, Yuan Tingting, Liu Xinyu

Journal of Semiconductors, 2010, 31(3): 034001. doi: 10.1088/1674-4926/31/3/034001

11

Noise performance in AlGaN/GaN HEMTs under high drain bias

Pang Lei, Pu Yan, Liu Xinyu, Wang Liang, Liu Jian, et al.

Journal of Semiconductors, 2009, 30(8): 084004. doi: 10.1088/1674-4926/30/8/084004

12

Impact of UV/ozone surface treatment on AlGaN/GaN HEMTs

Yuan Tingting, Liu Xinyu, Zheng Yingkui, Li Chengzhan, Wei Ke, et al.

Journal of Semiconductors, 2009, 30(12): 124001. doi: 10.1088/1674-4926/30/12/124001

13

Annealing before gate metal deposition related noise performance in AlGaN/GaN HEMTs

Pang Lei, Pu Yan, Liu Xinyu, Wang Liang, Li Chengzhan, et al.

Journal of Semiconductors, 2009, 30(5): 054001. doi: 10.1088/1674-4926/30/5/054001

14

60Co γ-rays irradiation effect in DC performance of AlGaN/GaN high electron mobility transistors

Gu Wenping, Chen Chi, Duan Huantao, Hao Yue, Zhang Jincheng, et al.

Journal of Semiconductors, 2009, 30(4): 044002. doi: 10.1088/1674-4926/30/4/044002

15

Post-Gate Process Annealing Effects of Recessed AlGaN/GaN HEMTs

Liu Guoguo, Huang Jun, Wei Ke, Liu Xinyu, He Zhijing, et al.

Journal of Semiconductors, 2008, 29(12): 2326-2330.

16

DC Characteristics of AlGaN/GaN HEMTs with a Field Plate Gate

Wei Ke, Liu Xinyu, He Zhijing, Wu Dexin

Journal of Semiconductors, 2008, 29(3): 554-558.

17

An 8GHz Internally Matched AlGaN/GaN HEMT Power Amplifier with RC Stability Network

Zeng Xuan, Chen Xiaojuan, Liu Guoguo, Yuan Tingting, Chen Zhongzi, et al.

Journal of Semiconductors, 2008, 29(8): 1445-1448.

18

Performance Improvement of AlGaN/GaN HEMTs by Surface Treatment Prior to Si3N4 Passivation

Li Chengzhan, Liu Dan, Zheng Yingkui, Liu Xinyu, Liu Jian, et al.

Journal of Semiconductors, 2008, 29(2): 329-333.

19

Development of High Performance AlGaN/GaN HEMTswith Low Ohmic Contact

Liu Jian, Li Chenzhan, Wei Ke, He Zhijing, Liu Guoguo, et al.

Chinese Journal of Semiconductors , 2006, 27(S1): 262-265.

20

Degradation Under High-Field Stress and Effects of UV Irradiation on AlGaN/GaN HEMTs

Wang Chong, Zhang Jincheng, Hao Yue, Yang Yan

Chinese Journal of Semiconductors , 2006, 27(8): 1436-1440.

1. Ai, L., Tian, F., Xu, A. et al. Effect of InxAl1-xAs graded buffer materials on pseudomorphic InP HEMT. Frontiers in Materials, 2022. doi:10.3389/fmats.2022.980077
2. Ajayan, J., Nirmal, D. 22 nm In0:75Ga0:25As channel-based HEMTs on InP/GaAs substrates for future THz applications. Journal of Semiconductors, 2017, 38(4): 044001. doi:10.1088/1674-4926/38/4/044001
3. Chong, W., Wei, X., He, Y. et al. Effects of the fluorine plasma treatment on low-density drain AlGaN/GaN HEMT. Journal of Semiconductors, 2016, 37(6): 064009. doi:10.1088/1674-4926/37/6/064009
  • Search

    Advanced Search >>

    GET CITATION

    Lei Shi, Shiwei Feng, Kun Liu, Yamin Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. Journal of Semiconductors, 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005
    L Shi, S W Feng, K Liu, Y M Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. J. Semicond., 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005.
    shu

    Export: BibTex EndNote

    Article Metrics

    Article views: 2290 Times PDF downloads: 21 Times Cited by: 3 Times

    History

    Received: 05 September 2014 Revised: Online: Published: 01 July 2015

    Catalog

      Email This Article

      User name:
      Email:*请输入正确邮箱
      Code:*验证码错误
      Lei Shi, Shiwei Feng, Kun Liu, Yamin Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. Journal of Semiconductors, 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005 ****L Shi, S W Feng, K Liu, Y M Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. J. Semicond., 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005.
      Citation:
      Lei Shi, Shiwei Feng, Kun Liu, Yamin Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. Journal of Semiconductors, 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005 ****
      L Shi, S W Feng, K Liu, Y M Zhang. Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress[J]. J. Semicond., 2015, 36(7): 074005. doi: 10.1088/1674-4926/36/7/074005.

      Mechanism of the self-changing parameters and characteristics in AlGaN/GaN high-electron mobility transistors after a step voltage stress

      DOI: 10.1088/1674-4926/36/7/074005
      Funds:

      Project supported by the National Natural Science Foundation of China (Nos. 61376077, 61201046, 61204081), and the Beijing Natural Science Foundation (Nos. 4132022, 4122005).

      • Received Date: 2014-09-05
      • Accepted Date: 2015-02-13
      • Published Date: 2015-01-25

      Catalog

        /

        DownLoad:  Full-Size Img  PowerPoint
        Return
        Return