Edge Termination Design and Simulation for Bulk GaN Rectifiers

 

K. H. Baik (1), Y. Irokawa (2), F. Ren (3), S. J. Pearton (1), S. S. Park (4) and S.K.Lee (4)

(1)   Department of Materials Science and Engineering, University of Florida, Gainesville FL 32611, USA

(2)   Toyota Central Research and Development Laboratories, Inc., Nagakute, Aichi, 480-1192, Japan

(3)   Department of Chemical Engineering, University of Florida, Gainesville FL 32611, USA

(4)   Samsung Advanced Institute of Technology, Suwon 440-600, South Korea

 

ABSTRACT

            GaN bulk rectifiers show excellent on-state resistances (in the milli-ohm.cm-2 range),forward turn-on voltages of ~1.8V and reverse-recovery times of <50 ns. A key requirement is to develop effective edge termination techniques in order to prevent premature surface-induced breakdown. We have performed a simulation study of the effects of varying the dielectric passivation material (SiO2, SiNX, AlN, Sc2O3 or MgO), the thickness of this material, the extent of metal overlap onto the dielectric and the ramp oxide angle on the resulting reverse breakdown voltage (VB) of bulk rectifiers. We find that SiO2 produces the highest VB of the materials investigated, that there is an optimum metal overlap distance for a given oxide thickness and small oxide ramp angles produce the highest VB.

 

 

INTRODUCTION

            There is strong interest in the development of ultra high power inverter modules based on GaN and other wide bandgap semiconductors.(1-5) These would have application in pulsed power for avionics and electric ships, in solid-state drivers for heavy electric motors and in advanced power management and control electronics. Schottky rectifiers are a key element of inverter modules because of their high switching speeds and low switching losses. While excellent reverse blocking voltages (VB) have been achieved in lateral GaN rectifiers (VB up to ~9.7 kV), these devices have limited utility because of their low on-state current.(6,7) Recently a number of reports have appeared of vertical geometry GaN rectifiers fabricated on free-standing substrates.(8-10) These have shown excellent forward current characteristics, with total currents > 1.7A for 7 mm diameter devices and low forward turn-on voltages (VF ~1.8V). The reverse breakdown voltages in these structures are still limited by avalanche breakdown at defects and/or surfaces. The rapid progress in improving both defect density and purity of these free-standing substrates makes them the most promising approach for achieving both high VB and on-state currents,(11,12) in comparison to methods such as Metal Organic Chemical Vapor Deposition of lightly-doped stand-off layers.(5, 13-16)

            A simple model for avalanche breakdown in GaN resulting from impact ionization produces the relation (3)

VB ~ 1.98x1015 N-0.75

where N is the doping concentration in the GaN. Currently, all GaN rectifiers show performance limited by the presence of defects and by breakdown initiated in the depletion region near the electrode corners. In SiC rectifiers, a wide variety of edge termination methods have been employed to smooth out the electric field distribution around the rectifying contact periphery, including mesas,(17) high resistivity layers created by ion implantation,(18,19) field plates(20,21) and guard rings.(22) The situation is far less developed for GaN, with just a few reports of combined guard rings/field-plate termination (8,9,23)

            In this paper we report on design and simulation of bulk GaN rectifiers with field plate edge termination. The effect of different dielectric materials used for the field plate was examined, as well as factors such as the extent of metal electrode overlap, dielectric thickness and ramp oxide angle.The simulations show that the VB could be increased by up to a factor of two over the value in unterminated rectifiers.

 

Experimental

            The structure at the top of Figure 1 was used as our standard for simulation and is based on the available free-standing GaN bulk substrates, which currently have a doping density of ~1016 cm-3. While their thickness is ~200 µm, the depetion depth is limited by the background doping and we used a thickness of 30 µm in the simulations. The parameters we investigated in this study were the dielectric material, its thickness and the extent of metal overlap onto the field plate. The simulations were carried out using the MEDICITM code. The back n-ohmic contact resistance was assumed to be 10-6Wcm2, which is consistent with our past experimental data, and an interface state density of 5x1011 eV-1cm-2 was assumed for the dielectric/GaN interface. Once again, this is based on our past experimental results.(24) After designing the particular basic structure, a mesh of nodes is created to allow the solutions to the transport equations to be obtained. A typical mesh grid is shown at the bottom of Figure 1. The program includes Shockley-Reed-Hall and Auger recombination, an incomplete ionization model and an average of the available high-field saturation and avalanche models. We assumed a conduction band density of states of 2.6x1018 cm-3 for the n-type GaN, a surface recombination velocity of 103 cm.sec-1 and Shockley-Reed-Hall lifetime of 1 ns.

 

RESULT AND DISCUSSION

            Figure 2 shows data from an unterminated rectifier. The maximum electric field occurs directly under the corner of the Schottky contact and emphasizes that avalanche breakdown is more likely to initiate at that location. The breakdown voltage was 980 V, obtained from the simulated reverse current-voltage (I-V) characteristics shown at the bottom of the figure.

            Figure 3 shows the calculated breakdown voltages obtained for a 0.7 µm thick SiO2 field plate on top of the rectifier, as a function of the extent of the overlap of the Schottky contact onto the SiO2. Note that the VB values increase rapidly for metal overlaps up to ~10 µm, with a maximum increase of ~63% in breakdown voltage relative to the unterminated device. Beyond an overlap of 10 µm, there is no further improvement in breakdown voltage from this given thickness of SiO2 field plate. We believe this is due to the fact that the lateral spread of the depletion layer becomes comparable to the depth of this layer, so that extending the field plate into undepleted regions does not affect the breakdown behavior. (21)

            The effect of SiO2 thickness at a given metal overlap distance of 10 µm is shown in Figure 4. There is an almost linear increase in VB with increasing oxide thickness up to 0.7 µm. At thickness >1µm, the simulations showed that the electric field inside the oxide began to increase. One must therefore choose a thickness such that the field strength inside the oxide does not exceed its breakdown strength. The fact that very thick oxide layers do not lead to an improvement in VB is an advantage from a practical viewpoint because such layers would require very long deposition times and introduce problems such as stress.

            There was also no particular advantage to use of SiNx as the dielectric, compared to SiO2, in either a bilayer or by itself. Figure 5 shows the effect on VB of addition of SiNN on top of a 0.7 µm thick SiO2 field plate (top) and a comparison of the effects of SiO2 and SiNN field plate thickness (bottom). In all cases the metal overlap distance was set at 10 µm.

            Other dielectrics that have demonstrated reasonably low interface state densities on GaN include AlN, MgO and Sc2O3.(25-27) Figure 6 shows a comparison of the VB values obtained for 0.7 µm thick dielectric films of different materials, for a fixed metal overlap distance of 10 µm. SiO2 produces the highest breakdown voltage rectifiers for these conditions because of its large bandgap and low dielectric constant. However, in real devices it should be considered that reliability is of utmost importance and it is not necessarily the case that SiO2 would be the best choice with this consideration in mind. For example, Sc2O3 appears to produce the most effective passivation of surface states on GaN/AlGaN heterostructure field effect transistors.(28) Obviously, much more work needs to be done to establish experimentally the relative tradeoff between VB and long-term device stability.

            An additional parameter that can be simulated is the ramp angle at the edge of the oxide field plate where it meets the Schottky contact. Figure 7 shows the effect of this angle on VB of a GaN rectifier with a 1 µm thick SiO2 field plate and 10 µm mesa overlap. Note that the VB at ramp angle < 4° is roughly double that of the unpassivated device. This is consistent with past reports for 6H-SiC Schottky devices.(29) The ramp angle is relatively straightforward to control by using a sloped photoresist mask when etching the openings in the SiO2 dielectric layer.

 

SUMMARY AND CONCLUSION

The main findings of this simulation study can be summarized as follows:

(i)                  the use of an optimized SiO2 field plate edge termination can increase the reverse breakdown volatge of bulk GaN rectifiers by up to a factor of two compared to unterminated devices.

(ii)                the dielectric material, thickness and ramp angle all influence the resulting VB of the rectifier by determining where the maximum field strength occurs in the device structure. The key aspect in designing the field plate edge termination is to shift the region of the high field region away from the periphery of the rectifying contact.

 

ACKNOWLEDGEMENTS

The work at UF is partially supported by NSF CTS-991173 and EPRI (Ben Damsky).

REFERENCES

1.      G. T. Heydt and B. J. Skromme, Mat. Res. Soc.. Symp. Proc. Vol. 483 3 (1998)

 

  1. E. R. Brown, Solid. State Electrochem. 43 1918 (1998)

 

  1. M. Trivedi and K. Shenai, J. Appl. Phys. 85 6889 (1999)

 

  1. S. J. Pearton, F. Ren, A. P. Zhang and K. P. Lee, Mar. Sci. Eng. R30 55 (2000)

 

  1. Z. Z. Bandic, D. M. Bridger, E. C. Piquette, T. C. McGill, R. P. Vaudo, V. M. Phanse and J. M. Redwing, Appl. Phys. Lett. 74 1266 (1999)

 

  1. A. P. Zhang, J. W. Johnson, F. Ren, J. Han, A. J. Polyakov, N. B. Smirnov, A. V. Govorkov, J. M. Redwing, K. P. Lee and S. J. Pearton, Appl. Phys. Lett. 78 823 (2001)

 

  1. A. P. Zhang, G. Dang, F. Ren, J. Han, A. Y. Polyakov, N. B. Smirnov, A. V. Govorkov, J. M. Redwing, X. A. Cao and S. J. Pearton, Appl. Phys. Lett. 76 1767 (2000)

 

  1. J. W. Johnson, J. R. LaRoche, F. Ren, B. P. Gila, M. E. Overberg, C. R. Abernathy, J. I. Chyi, C. C. Chuo, T. E. Nee, C. M. Lee, K. P. Lee, S. S. Park, J. I. Park and S. J. Pearton, Solid-State Electrochem. 45 405 (2001)

 

  1. J. W. Johnson, A. P. Zhang, W. B. Luo, F. Ren, S. J. Pearton, S. S. Park, Y. J. Park and J.-I. Chyi, IEEE Trans Electron. Dev. 49 32 (2002)

 

  1. J. W. Johnson, B. Luo, F. Ren, D. Palmer, S. J. Pearton, S. S. Park, Y. J. Park, Solid-State Electron. 46 911 (2002)

 

  1. S. S. Park, I. W. Park and S. H. Choh, Jap. J. Appl. Phys. 39 L1141 (2000)

 

  1. H. Morkoc, Mar. Sci. Eng. R33 135 (2001)

 

  1. A. P. Zhang, G. T. Dang, F. Ren, H. Cho, K. P. Lee, S. J. Pearton, J.-I. Chti, T. E. Nee and C. C. Chuo, Solid-State Electron. 44 619 (1999)

 

  1. F. Ren, A. P. Zhang, G. Dang, X. A. Cao, H. Cho, S. J. Pearton, J.I. Chyi, C. M. Lee and C. C. Chuo, Solid-State Electron 44 619 (1999)

 

  1. T. G. Zhu, D. J. Lambert, B. S. Shelton, M. M. Wong, V. Chowdhurg and R. D. Dupis, Appl. Phys. Lett. 77 2918 (2000)

 

  1. B. S. Shelton, T. G. Zhu, D. J. Lambert and R. D. Dupis, IEEE Trans. Electron. Dev. 48 1498 (2001)
  2. P. G. Neudeck, D. J. Larkin, A. Rowell and L. G. Matus, Appl. Phys. Lett. 64 1386 (1994)

 

  1. D. T. Morisette, J. A. Cooper, Jr., M. R. Melloch, G. M. Dolng, P. M. Shenoy, M. Zakari and J. Gladish, IEEE Trans. Electron. Dev. 48 349 (2001)

 

  1. A.Itoh, T. Kimoto and H. Matsunami, IEEE Electron. Dev. Lett. 17 139 (1996)

 

  1. V. Saxena, J. N. Su and A. J. Steckl, IEEE Trans. Electron. Dev. 46 456 (1999)

 

  1. M. C. Tarplee, V. P. Madangarli, Q. Zhang and T. S. Sudarshan, IEEE Trans. Electron. Dev. 48 2659 (2001)

 

  1. N. V. Dyakonova, P. A. Ivanov, V. A. Koglov, M. E. Levinshtein, J. W. Palmour, S. L. Pumyantsev and R. Singh, IEEE Trans. Electron. Dev. 46 2188 (1999)

 

  1. A. P. Zhang, G. Dang, X. A. Cao, H. Cho, F. Ren, J. Han, J.-I. Chyi, C. M. Lee, T. E. Nee, C. C. Chuo, G. C. Chi, S. N. G. Chu, R. G. Wilson and S. J. Pearton, MRS Internet J. Nitride Semicon. Res. 551 W11.67 (2000)

 

  1. R. Mehandru, B. P. Gila, J. Kim, J. W. Johnson, K. P. Lee, B. Luo, A. H. Onstine, C. R. Abernathy, S. J. Pearton and F. Ren, Electrochem. Solid-State Lett. 5 G51 (2002)

 

  1. M.Hong,H.M.Ng,J.Kwo,A.R.Kortran,J.N.Baillargeon,S.N.G.Chu,J.P.Mannaerts,A.Y.Cho,F.Ren,C.R.Abernathy and S.J.Pearton,presented at 197th Electrochem.Soc.meeting,May 2000,Toronto,Canada.

 

  1. B.P.Gila,J.W.Johnson,R.Mehandru,B.Luo,A.H.Onstine,K.K.Allums,V.Krishnamoorthy,S.Bates,C.R.Abernathy,F.Ren and S.J.Pearton,Phys.Stat.SolidiA 188 239(2001)

 

  1. T.S.Lay,M.Hong,J.Kwo,J.P.Mannaerts,W.H.Hung and D.J.Huang,Solid-State Electron.45 1679(2001).

 

  1. B.Luo,J.W.Johnson,J.Kim,R.Mehandru,F.Ren,B.Gila,A.H.Onstine,C.R.Abernathy,S.J.Pearton,A.G.Baca,R.D.Briggs,R.J.Shul,C.Monier and J.Han,Appl.Phys.Lett.80 1661(2002)

 

  1. G. Brezeane, J. Fernandez, J. Millan, J. Rebello, M. Badila, G. Dilimot, Mat. Sci. Forum 264-268 941 (1998)

 

 

 

 

 

 

Figure captions

Figure 1. Schematic of simulated bulk GaN rectifier structure (top) and mesh grid employed for simulations (bottom).

 

Figure 2. Potential contours at breakdown point for unterminated rectifier (top) and reverse I-V characteristic (bottom).

 

Figure 3. Effect of metal overlap distance in VB for rectifiers with 0.7 µm thick SiO2 field plate.

 

Figure 4. Effect of SiO2 thickness on VB for rectifiers with 10 µm metal overlap.

 

Figure 5. Effect on VB of rectifiers with a SiO2/SiNx bilayer field plate (top) and comparison of the effets of SiO2 or SiNX thickness for single-material field plate (bottom). The metal overlap was 10 µm in all cases.

 

Figure 6. Dependence of VB on the material used for field termination. The thickness was 0.7 µm and the metal overlap was 10 µm in all cases.

 

Figure 7. Influence of oxide ramp angle on VB for a 1 µm thick SiO2 field plate.

 

 

 

 

 

 

 

 

 

 

 

 


 

 

 

 


Figure 1

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5.

 

 

 

 

 

 

 

 

Oxide

Nitride

AlN

MgO

Sc2O3

e

3.9

7.5

8.5

9.8

14

Eg (eV)

9

4.7

6.2

8

6.3

 

 

 

 

 

 

 

Figure 6.

 

 

 

 

 

 

 

 

 

Figure 7.