Experimental evaluation of impact ionization coefficients in GaN

Kunihiro, Kazuaki; Kasahara, K.; Takahashi, Yuji; Ohno, Yasuo

doi:10.1109/55.806100

Cited by 151 publications

(70 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, avalanche breakdown is observed in vertical GaN PN diodes in recent experimental studies, 4-7,9 exhibiting a positive temperature coefficient of the breakdown voltage and hence proper modeling is vital for vertical power devices. The GaN impact ionization coefficient, which gives an estimate of the avalanche voltage has been studied both experimentally 36,37 and by Monte Carlo simulations. 24,25,27,[80][81][82] The avalanche multiplication is calculated by using the impact ionization coefficients according to equation (13).…”

Section: E Impact Ionization Parametersmentioning

confidence: 99%

“…Although experimentally determined values of electron mobility [32][33][34][35] are reported there are few available experimental data for hole mobility. The Chynoweth law 36 is accepted as an accurate representation of the avalanche effect in GaN but impact ionization coefficients reported from Monte Carlo simulations 24,25,27 as well as experimental results 36,37 differ from one work to another. In case of high doping concentrations incomplete ionization of donor (silicon) 38,39 and acceptor (magnesium) [40][41][42][43] needs to be taken into account and their corresponding activation energies also show a wide spread.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Modeling and simulation of bulk gallium nitride power semiconductor devices

et al. 2016

View full text Add to dashboard Cite

Bulk gallium nitride (GaN) power semiconductor devices are gaining significant interest in recent years, creating the need for technology computer aided design (TCAD) simulation to accurately model and optimize these devices. This paper comprehensively reviews and compares different GaN physical models and model parameters in the literature, and discusses the appropriate selection of these models and parameters for TCAD simulation. 2-D drift-diffusion semi-classical simulation is carried out for 2.6 kV and 3.7 kV bulk GaN vertical PN diodes. The simulated forward current-voltage and reverse breakdown characteristics are in good agreement with the measurement data even over a wide temperature range.

show abstract

Section: E Impact Ionization Parametersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Modeling and simulation of bulk gallium nitride power semiconductor devices

et al. 2016

View full text Add to dashboard Cite

show abstract

“…54 An impact-ionized hole current 55,56 or gate-edge electroluminescence (EL) (Refs. 11 and 43) should be observed for the impact ionization induced at the gate edge.…”

Section: E Trapping Mechanism Of Non-localized Hot Electronsmentioning

confidence: 99%

Non-localized trapping effects in AlGaN/GaN heterojunction field-effect transistors subjected to on-state bias stress

Hu¹,

Hashizume²

2012

Journal of Applied Physics

View full text Add to dashboard Cite

Evidence of space charge limited flow in the gate current of AlGaN/GaN high electron mobility transistors Appl. Phys. Lett. 100, 223504 (2012) Off-state breakdown and dispersion optimization in AlGaN/GaN heterojunction field-effect transistors utilizing carbon doped buffer Appl. Phys. Lett. 100, 223502 (2012) Charge transport and trap characterization in individual GaSb nanowires J. Appl. Phys. 111, 104515 (2012) The asymmetrical degradation behavior on drain bias stress under illumination for InGaZnO thin film transistors Appl. Phys. Lett. 100, 222901 (2012) Mechanism of random telegraph noise in junction leakage current of metal-oxide-semiconductor field-effect transistor J. Appl. Phys. 111, 104513 (2012) Additional information on J. Appl. Phys. For AlGaN/GaN heterojunction field-effect transistors, on-state-bias-stress (on-stress)-induced trapping effects were observed across the entire drain access region, not only at the gate edge. However, during the application of on-stress, the highest electric field was only localized at the drain side of the gate edge. Using the location of the highest electric field as a reference, the trapping effects at the gate edge and at the more distant access region were referred to as localized and non-localized trapping effect, respectively. Using two-dimensional-electron-gas sensing-bar (2DEG-sensing-bar) and dual-gate structures, the non-localized trapping effects were investigated and the trap density was measured to be $1.3 Â 10 12 cm À2 . The effect of passivation was also discussed. It was found that both surface leakage currents and hot electrons are responsible for the non-localized trapping effects with hot electrons having the dominant effect. Since hot electrons are generated from the 2DEG channel, it is highly likely that the involved traps are mainly in the GaN buffer layer. Using monochromatic irradiation (1.24-2.81 eV), the trap levels responsible for the non-localized trapping effects were found to be located at 0.6-1.6 eV from the valence band of GaN. Both trap-assisted impact ionization and direct channel electron injection are proposed as the possible mechanisms of the hot-electron-related non-localized trapping effect. Finally, using the 2DEG-sensing-bar structure, we directly confirmed that blocking gate injected electrons is an important mechanism of Al 2 O 3 passivation. V C 2012 American Institute of Physics.

show abstract

“…Impact ionization [2] as well as charge trapping [8] has been discussed in literature as possible origin for hole currents in GaN-based devices. Therefore, the possibility of hole emission from traps needs to be excluded as mechanism for the here observed interband EL to prove the presence of impact ionization.…”

Section: Resultsmentioning

confidence: 99%

“…[6]. Initially, the commonly used method of gate current analysis for detecting impact ionization [2] was applied to attempt to probe hole currents. The hole current generated by, e.g., impact ionization is expected to be in the range of nA [1], which requires a very low gate leakage current level in order to distinguish between hole current by impact ionization and other gate leakage current contributions.…”

mentioning

confidence: 99%

Evidence for impact ionisation in AlGaN/GaN HEMTs with InGaN back-barrier

et al. 2011

View full text Add to dashboard Cite

Electroluminescence (EL) spectroscopy in combination with drift-diffusion simulations was used to prove the presence of impact ionization in AlGaN/GaN HEMTs illustrated on InGaN back-barrier devices. Regardless of the level of gate leakage current, which is dominated by contributions such as surface leakage current and others, EL enabled to reveal hole generation due to impact ionization. Hole currents as low as 10pA were detectable by the optical technique used.Introduction: AlGaN/GaN HEMTs represent a promising technology for high power RF applications. Besides charge carrier trapping, impact ionization due to high electric fields in the device channel [1-3] can limit device reliability and consequently the use of GaNbased HEMTs. The presence of impact ionization has been debated for GaN HEMTs with controversial evidence due to high leakage in many devices [3], illustrating the limitations of the commonly applied method of gate current analysis for GaN HEMTs.Therefore, a new methodology is needed to probe impact ionization in AlGaN/GaN HEMTs independently of the gate leakage current level. A bell-shaped dependence of the gate current versus gate-source voltage has been used in AlGaAs/GaAs HEMTs [4] to prove impact ionization, supported by studies of interband electroluminescence (EL).So far neither has been unambiguously observed in GaN-based HEMTs. This letter Ref. [6]. Initially, the commonly used method of gate current analysis for detecting impact ionization [2] was applied to attempt to probe hole currents. The hole current generated by, e.g., impact ionization is expected to be in the range of nA [1], which requires a very low gate leakage current level in order to distinguish between hole current by impact ionization and other gate leakage current contributions. The inset of Fig. 1 shows, however, that the leakage current level of the devices used is too high in order to draw conclusions on the presence of a hole current. EL spectroscopy was used to probe the hole current independently from the level of gate leakage current. EL spectra emitted from the source-drain gap were recorded using a Renishaw RM system, while the device was operated at a drain-source and gate-source bias of 20 V and 0 V, respectively. Results:The EL spectrum shown in Fig. 1 reveals a tail in the red-infrared spectral range typically observed in GaN-based HEMTs, which is related to hot carrier relaxation in the active device region [7]. Furthermore, a peak of the EL signal is apparent at the bandgap energy of InGaN, evidencing the recombination of electrons and holes in the InGaN layer. This clearly proves the presence of holes in the InGaN layer of the device under operation, in addition to electrons from normal device operation. The InGaN device layer 3 acts as charge carrier collecting layer in the device (inset of Fig. 2), and as optical hole probe due to its high optical quantum efficiency. We note that EL around the GaN bandgap energy tends to be optically less efficient, possibly explaining why this has not been observed to...

show abstract

Experimental evaluation of impact ionization coefficients in GaN

Cited by 151 publications

References 14 publications

Modeling and simulation of bulk gallium nitride power semiconductor devices

Modeling and simulation of bulk gallium nitride power semiconductor devices

Non-localized trapping effects in AlGaN/GaN heterojunction field-effect transistors subjected to on-state bias stress

Evidence for impact ionisation in AlGaN/GaN HEMTs with InGaN back-barrier

Contact Info

Product

Resources

About