Sulphide passivation of GaN based Schottky diodes

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Sulphide passivation of GaN based Schottky diodes

Ashish Kumar*, Trilok Singh, Mukesh Kumar, R. SinghDepartment of Physics, Indian Institute of Technology Delhi, New Delhi 110016, India

a r t i c l e i n f o

Article history:Received 8 August 2013Received in revised form2 December 2013Accepted 17 December 2013Available online 18 January 2014

Keywords:GaNPassivationPLXPSSchottky barrier diode

a b s t r a c t

Wet chemical passivation of n-GaN surface was carried out by dipping GaN samples in ammoniumsulphide diluted in aqueous and alcoholic solvent base solutions. Photoluminescence (PL) investigationsindicated that sulphide solution effectively led to the reduction of GaN surface states. Increased bandedge PL peak showed that S2� ions are more active in alcohol based solvents. X-ray photoelectronspectroscopy revealed reduction in surface oxides by introduction of sulphide species. Ni/n-GaN Schottkybarrier diodes were fabricated on passivated surfaces. Remarkable improvement in the Schottky barrierheight (0.98 eV for passivated diodes as compared to 0.75 eV for untreated diodes) has been observed.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Gallium nitride (GaN) is a wide bandgap (3.4 eV, direct) semi-conductor which affords it special properties for applications inoptoelectronic, high power and high frequency devices and high-electron mobility transistors. For all these device applications, theelectrical and surface properties of metal/GaN interfaces areimportant to understand in order to control the complex devices[1e5]. There have been many reports about the origin of surfacestates in GaN based materials, for example threading dislocations(TDs), nitrogen vacancies (VN), and oxygen impurities [6e9]. Thesedefects states introduce excess reverse leakage current in GaNSchottky diodes [9,10]. VN and oxygen impurities as shallow surfacedonors have also been reported to cause gate leakage current [8,9].Effect on device performance have been investigated for varioustechniques used to engineer electrically active surface states e

surface passivation, dielectric layer insertion, surface treatmentswith chemicals or plasma, and post-gate-annealing [11e18]. In or-der to avoid the formation of a native-oxide layer and to improvedevice characteristics, passivation of the surface states is needed toprotect the surface from chemical and/or electronic degradationdue to oxidation and to reduce the density of electronically activesurface states. Surface passivation results in the termination of thedangling bonds at the GaN surface. This leads to a reduction in thesurface recombination velocity (SRV), which is described as the rate

of loss of charge carriers at the surface of a material. The reductionin SRV indicates a decrease in the surface defect density. Surfacepassivation to some extent, also reduces the Fermi level pinning.The passivation of interface states remains an important problemfor GaN based semiconductor devices. Different surface passivationschemes have been reported for the reduction of surface statedensity in GaN. NH4OH, (NH4)2Sx, Chlorination and Si3N4 have beenused for passivation by various groups [3e5,19e20]. But theconclusive reports related to X-ray photoelectron spectroscopy(XPS) investigation of GaN surfaces passivated by sulphide solutionare still not studied in detail.

2. Experimental

GaN epitaxial layers (0001) used in this study (grown on c-planesapphire substrate by metal organic chemical vapor deposition(MOCVD) technique) were commercially procured. The GaNepitaxial layers were 3.6 mm thick and un-intentionally doped(ND

þ w 3 � 1016 cm�3 by Hall measurements). For passivationstudies, samples (1�1 cm2) were cut from GaN/sapphire wafers. Infirst step, all samples were cleaned ultrasonically in trichloro-ethylene (TCE), acetone and isopropyl alcohol, respectively, for10 min each, and then rinsed in de-ionized water. In second step,GaN epitaxial samples were treated in HCl:H2O (1:1) solution for60 s followed by rinsing in DI water and dried using N2 blow.Ammonium sulphide [(NH4)2Sx] passivation was done thereafter inthird step at room temperature. For the purpose (NH4)2Sx solutionin two different solvents, water and t-butanol (t-C4H9OH) were

* Corresponding author. Tel.: þ91 9718686667.E-mail addresses: [email protected], [email protected] (A. Kumar).

Contents lists available at ScienceDirect

Current Applied Physics

journal homepage: www.elsevier .com/locate/cap

1567-1739/$ e see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.cap.2013.12.021

Current Applied Physics 14 (2014) 491e495

Author's personal copy

prepared and named as AMS (aq.) and AMS (alc.), respectively.Samples were kept in two solutions for 100 s time period and thendried using N2 blow. For nomenclature, samples after 2nd step arecalled as “Untreated” and after 3rd step as “treated/passivated withAMS (aq. or alc.), respectively. Details of samples and solutionsprepared are given in Table 1. The GaN epitaxial layers wereexamined using the atomic force microscope (AFM) in the contactmode to measure any change in the topography or texture of thesurface after passivation. We have used a Multi Mode ScanningProbe Microscope (Veeco) for all morphological investigations. XPSmeasurements were taken with Mg Ka (1253.6 eV) source. Theroom-temperature photoluminescence (PL) spectra was taken us-ing a diode laser (266 nmwavelength) laser as an excitation source.The device fabrication was done on untreated and passivated n-GaN surface, the details of which can be found elsewhere [21]. Thecurrentevoltage (IeV) characteristics of the devices weremeasuredusing Keithley Semiconductor analyzer (model SPS 4200).

3. Results and discussion

Fig. 1 shows AFM images for (a) HCl etched epitaxial GaN (Nosulphide treatment) and (b) GaN sample passivated in AMS (alc.)solution. The measured root mean square roughness (Rrms) valuesare approx. 2.4 nm (after 1min etching in HCl sol.) and 1.6 nm (afterAMS (alc.) passivation for 100 s). The acid treated sample has higherRMS roughness as compared to wafer cut sample (not shown here).After passivation, roughness found to be decreased as compared toetched sample. This may be possibly due to formation of sulphidelayer over GaN surface which effectively cover whole surfacereducing the overall roughness.

The room-temperature photoluminescence (PL) spectra of theuntreated n-GaN and AMS treated n-GaN, showed that the band-edge emission (3.4 eV) intensity of all the AMS treated n-GaN

samples are higher than that of the untreated n-GaN sample.Particularly, a 3.6-fold increase on the band-edge emission in-tensity is observed for sample treated with AMS (alc.) solution for100 s as shown in (Fig. 2). This AMS (alc.) surface treatment led tothe removal of the native oxide (i.e. GaOx) on the n-GaN surface,reduction of surface states, and formation of sulfurated layer nearthe n-GaN surface. Termination of the dangling bonds at the GaNsurface leads to a reduction in the surface recombination velocity(SRV), which is defined as the rate of loss of charge carriers at thesurface of a material. An estimate of the SRV (cm/s), using themodified dead layer model [22e24], is given by

SRV ¼ Dp

Lp

�1þ aLp � b

b� 1

�(1)

where Dp is the hole diffusion coefficient, Lp is the hole diffusionlength, a is the GaN absorption coefficient, and b is the band-edgeemission intensity ratio between the AMS (alc.) treated n-GaNsample and untreated n-GaN sample. A SRV of 2.1 � 104 cm/s isobtained by using parameters as Dp ¼ 0.8 cm2/s, Lp ¼ 8 � 105 cm,a ¼ 1 �105/cm [22,25], and b ¼ 3.6 is in our case (see Table 1). Thevalue is more of an upper limit since an infinite f is assumedwith ascleaned n-GaN surface. The reduction in SRV indicates a decrease inthe surface defect density. The decrease of the SRV with AMS (alc.)treatment may be due to the accumulation of majority carriers and

Table 1Sample treatment details and Surface recombination velocity (SRV) calculations.

Sample name Chemical treatment, Intensityratio (b)

SRV(�104 cm/s)

Un-treated Etched in HCl:H2O (1:1), for 60 s. ___ ___No passivation treatment

AMS (aq.) (NH4)2Sx (80%) þ H2O (20%),for 60 s.

1.6 12.3

AMS (aq.) (NH4)2Sx (80%) þ H2O (20%),for 100 s.

2.6 4.0

AMS (alc.) (NH4)2Sx (20%) þ t-C4H9OH(80%), for 100 s.

3.6 2.1

Fig. 1. AFM images for a) Untreated and b) 100 s. AMS (alc.) passivated GaN.

Fig. 2. Room-temperature PL spectra obtained from AMS treated n-GaN surfaces.

A. Kumar et al. / Current Applied Physics 14 (2014) 491e495492

Author's personal copy

the repulsion of minority carriers near the n-GaN surface [22e24].In addition, the decrease of the SRV with AMS (alc.) treatment maybe also due to the removal of the native oxide and sulfurpassivation.

Further, XPS technique has been used to investigate the surfacepassivation effects at the GaN surface. Fig. 3 shows the deconvo-luted (Gaussian function fitting after subtracting the backgroundwith Shirley function) core level (CL) spectra for Ga (3d) for (a)Untreated GaN, (b) AMS (aq.) treated for 100 s, and (c) AMS (alc.)treated for 100 s sample of GaN. The deconvoluted Ga (3d) CLspectra for untreated GaN shows dominant peak at BE of 20.2 eVand other smaller peak at BE of 20.7 eV. The BE value of thedominant peak is chemical shifted from the reported BE value ofthe elemental gallium [26]. This BE shift is caused by subtle changeof the inner electron binding energy due to different chemical en-vironments that is related to atomic valences. Therefore, the BE20.2 eV verifies the bonding between gallium and nitrogen atoms[27]. The smaller peak at BE of 20.7 eVwas attributed to GaeO bond[26]. As there is a strong tendency for GaN surfaces to be stabilizedby Ga atoms in the surface layers due to the small lattice constantand high anioneanion bond strength for GaN compared to those ofconventional IIIeV semiconductors [17,18]. Further, particularly inthe present case of c-plane ((0001) orientation) for wurtzite GaN,which is a polar GaN, there is strong tendency of Ga-rich surfacestoichiometry and it can be considered as Ga-face like surface onthe basis of relatively higher Ga atom surface densities [28e31]. Thepresence of metallic Ga at the GaN surface makes the surface

susceptible for the oxidation when expose to air environment.Fig. 3(b) shows the deconvoluted Ga (3d) core level spectra for theGaN treated with AMS (aq.). In this case, the unfitted peak featurewere formed toward slightly higher BE than the previous untreatedGaN case while the carbon peak features were unshifted in eachcase. Moreover, in this case, FWHM of the unfitted peak featurewasappeared slightly higher. Therefore, the Ga 3d core level spectradeconvolute the extra peak at BE of 21.1 eVwhich correspond to thebond formation between gallium and sulfur atoms. Moreover theGaO peak contribution was reduced. Fig. 3(c) shows the deconvo-luted Ga (3d) core level spectra for the treated GaN with AMS (alc.).Deconvolution shows further reduction of the GaO contribution inGa 3d core level (CL) spectra. This indicates that GaN treated withAMS (alc.) efficiently passivated the surface than the GaN treatedwith AMS (aq.) and supported the PL observations. Since we knowthat dielectric constant of alcoholic solutions is lesser compared toaqueous solution andmagnitude of columbic forces between ions isinversely proportional to dielectric constant of the medium. TheS2� ions will be more active in alcoholic solvent. These S2� ionseffectively better passivate GaN surface states (Ga-dangling bonds)and oxides in t-butanol solvent as compared to aqueous solvent[32]. Reduction of surface states is also supported by the results ofPL investigations where band-edge peak intensity increases afterpassivation.

We have fabricated Ni/GaN Schottky diodes on these epitaxiallayers. Two sets of sample were prepared simultaneously, one seton untreated GaN epilayer and second set on AMS (alc.) passivated(100 s) surface by regular shadow masking method in electron-beam deposition system. Electrical characterizations of these di-odes were carried out and IeV characteristics are plotted below inFig. 4. The currentevoltage characteristics of Schottky barrier di-odes (SBD) are given by thermionic emission theory [33,34]. Forbias voltage V � 3kT/q, the conventional diode equation is

I ¼ IO exp�qVnkT

�(2)

Here, A** is the effective Richardson constant, fap the apparentor measured barrier height, n is ideality parameter, A is diode areaand other symbols have their usual meanings. The experimental IeV data are plotted as log I vs V and SBH and n is calculated fromintercept and slope of the linear fit to the linear part of forwardcharacteristics as given by equations (3) and (4).

Fig. 3. XPS spectra for Ga 3d core level for (a) Untreated GaN, (b) AMS (aq.) treated for100 s, and (c) AMS (alc.) treated for 100 s sample of GaN.

Fig. 4. IeV characteristics of Ni/GaN Schottky diodes fabricated on Untreated and AMS(alc.) passivated surface.

A. Kumar et al. / Current Applied Physics 14 (2014) 491e495 493

Author's personal copy

IO ¼ AA**T2 exp��qfap

kT

�(3)

n ¼ q

kT�

dVdln I

� (4)

The measured fap, n and reverse leakage current (IR at �1 V) arelisted in Table 2. The experimental values of SBH (fap) and n varyfrom 0.76 eV to 1.4 (300 K) for untreated Ni/GaN diode to 0.98 eVand 1.3, for AMS (alc.) passivated SBD respectively. According to theSchottkyeBardeen model, Schottky barrier height approaches toBardeen limit [35], fB ¼ Eg�fCNL (Fermi pinned interface) if surfaceor interface state density is large. Schottky barrier height tends tothe Ideal SBH, fS ¼ fM�c, as the density of interface states reduces.Here symbols have usual meanings. Taking fM (Ni) ¼ 5.1 eV, Eg(GaN)¼ 3.4 eV, c (GaN)¼ 4.1 eV and fCNL ¼ 2.65 eV [21,36e38], thecalculated fB and fS are 0.75 eV and 1.0 eV, respectively. Theexperimental values of SBH (fap) for pristine Ni/GaN diode (0.76 eV)are close to Bardeen limit (0.75 eV) which improves to 0.98 eV forpassivated SBD. According to Bardeen model, Schottky barrierheight tends to the Schottky limit (Ideal SBH) as the density ofinterface states reduces, and to the Bardeen limit (Fermi pinning) ifsurface states density increases [35]. Improvement in themeasuredSBH value and decrease in reverse leakage current of passivated Ni/n-GaN SBD at 300 K is a clear indicator of improved metal-semiconductor interface and reduced density of surface states.Thus AMS (alc.) solution effectively passivates GaN surface states. Areduction in surface states would also lead to thermionic emissionthrough Schottky barrier as compared to other current transport(tunneling, generation–recombination etc.) [35]. This improvesideality factor and it can also be seen clearly in calculated idealityfactor values (1.4 in pristine SBD to 1.3 in passivated SBD).

In conclusion, these results showed that the chemical passiv-ation effectively improves interface quality in our diodes. We foundthat AMS surface treatment led to the removal of the native oxideexisted on the n-GaN and the occupation of nitrogen related va-cancies by the sulfur which protected the surface from furtheroxidation. Therefore, implementation of the passivation recipe indevice fabrication processes potentially leads to the improved de-vice performance.

Acknowledgments

Mr. Ashish Kumar acknowledges research fellowship fromUniversity Grant Commission, India for carrying out research work.We are thankful to Dr. S. R. Burman from IUC, Indore for his help inXPS measurement. We also acknowledge the important contribu-tion from Dr. Seema Vinayak from SSPL, Delhi for her help inexperiments.

References

[1] J.E. Northrup, J. Neugebauer, Theory of GaN(10(1)over-bar-0) and (11(2)over-bar-0) surfaces, Phys. Rev. B 53 (1996) 10477e10480.

[2] P. Waltereit, O. Brandt, A. Trampert, H.T. Grahn, J. Menniger, M. Ramsteiner,M. Reiche, K.H. Ploog, Nitride semiconductors free of electrostatic fields forefficient white light-emitting diodes, Nature 406 (2000) 865e868.

[3] H. Kim, L. Schuette, J. Lee, W. Lu, J.C. Mabon, Passivation of surface andinterface states in AIGaN/GaN HEMT structures by annealing, J. ElectronMater. 36 (2007) 1149e1155.

[4] T. Hashizume, S. Ootomo, S. Oyama, M. Konishi, H. Hasegawa, Chemistry andelectrical properties of surfaces of GaN and GaN/AlGaN heterostructures,J. Vac. Sci. Technol. B 19 (2001) 1675e1681.

[5] C.T. Lee, Y.J. Lin, D.S. Liu, Schottky barrier height and surface state density ofNi/Au contacts to (NH4)(2)S-x-treated n-type GaN, Appl. Phys. Lett. 79 (2001)2573e2575.

[6] P.J. Hansen, Y.E. Strausser, A.N. Erickson, E.J. Tarsa, P. Kozodoy, E.G. Brazel,J.P. Ibbetson, U. Mishra, V. Narayanamurti, S.P. DenBaars, J.S. Speck, Scanningcapacitance microscopy imaging of threading dislocations in GaN films grownon (0001) sapphire by metalorganic chemical vapor deposition, Appl. Phys.Lett. 72 (1998) 2247e2249.

[7] D.M. Schaadt, E.J. Miller, E.T. Yu, J.M. Redwing, Lateral variations in thresholdvoltage of an AlxGa1-xN/GaN heterostructure field-effect transistor measuredby scanning capacitance spectroscopy, Appl. Phys. Lett. 78 (2001) 88e90.

[8] T. Hashizume, J. Kotani, A. Basile, M. Kaneko, Surface control process of AlGaNfor suppression of gate leakage currents in AlGaN/GaN heterostructure fieldeffect transistors, Jpn. J. Appl. Phys. 2 (45) (2006) L111eL113.

[9] T. Nanjo, T. Oishi, M. Suita, Y. Abe, Y. Tokuda, Effects of a thin Al layer insertionbetween AlGaN and Schottky gate on the AlGaN/GaN high electron mobilitytransistor characteristics, Appl. Phys. Lett. 88 (2006).

[10] A. Kumar, A. Hahnel, D. Kanjilal, R. Singh, Electrical and microstructural an-alyses of 200 MeV Ag14þ ion irradiated Ni/GaN Schottky barrier diode, Appl.Phys. Lett. 101 (2012).

[11] A. Kumar, R. Kaur, S. Vinayak, R. Singh, Barrier height enhancement of Ni/GaNSchottky diode using novel Ru based passivation scheme, Appl. Phys. Lett.(2013) in press.

[12] H. Kim, J. Lee, D.M. Liu, W. Lu, Gate current leakage and breakdown mecha-nism in unpassivated AlGaN/GaN high electron mobility transistors by post-gate annealing, Appl. Phys. Lett. 86 (2005).

[13] S.T. Bradley, S.H. Goss, J. Hwang, W.J. Schaff, L.J. Brillson, Pre-metallizationprocessing effects on Schottky contacts to AlGaN/GaN heterostructures,J. Appl. Phys. 97 (2005).

[14] N. Miura, T. Oishi, T. Nanjo, M. Suita, Y. Abe, T. Ozeki, H. Ishikawa, T. Egawa,Effects of interfacial thin metal layer for high-performance Pt-Au-based Schottky contacts to AlGaN-GaN, IEEE Trans. Electron Devices 51 (2004)297e303.

[15] J. Lee, D.M. Liu, H. Kim, W. Lu, Postprocessing annealing effects on directcurrent and microwave performance of AlGaN/GaN high electron mobilitytransistors, Appl. Phys. Lett. 85 (2004) 2631e2633.

[16] S. Arulkumaran, T. Egawa, H. Ishikawa, M. Umeno, T. Jimbo, Effects ofannealing on Ti, Pd, and Ni/n-A1(0). (11)Ga(0). (89)N Schottky diodes, IEEETrans. Electron Devices 48 (2001) 573e580.

[17] Y.K. Su, P.C. Chang, C.H. Chen, S.J. Chang, C.L. Yu, C.T. Lee, H.Y. Lee, J. Gong,P.C. Chen, C.H. Wang, Nitride-based MSM UV photodetectors withphoto-chemical annealing Schottky contacts, Solid State Electron 49 (2005)459e463.

[18] H.H. Lee, R.J. Racicot, S.H. Lee, Surface passivation of GaAs, Appl. Phys. Lett. 54(1989) 724e726.

[19] P.S. Chen, T.H. Lee, L.W. Lai, C.T. Lee, Schottky mechanism for Ni/Au contactwith chlorine-treated n-type GaN layer, J. Appl. Phys. 101 (2007).

[20] C.L. Hinkle, M. Milojevic, B. Brennan, A.M. Sonnet, F.S. Aguirre-Tostado,G.J. Hughes, E.M. Vogel, R.M. Wallace, Detection of Ga suboxides and theirimpact on III-V passivation and Fermi-level pinning, Appl. Phys. Lett. 94(2009).

[21] A. Kumar, S. Vinayak, R. Singh, Micro-structural and temperature dependentelectrical characterization of Ni/GaN Schottky barrier diodes, Curr. Appl. Phys.13 (2013) 1137e1142.

[22] Y.J. Lin, W.X. Lin, C.T. Lee, F.T. Chien, Changes in optical and electrical prop-erties and surface recombination velocity of n-type GaN due to (NH4)(2)S-xtreatment, Solid State Commun. 137 (2006) 257e259.

[23] L.W. Tu, W.C. Kuo, K.H. Lee, P.H. Tsao, C.M. Lai, A.K. Chu, J.K. Sheu, High-dielectric-constant Ta2O5/n-GaN metal-oxide-semiconductor structure, Appl.Phys. Lett. 77 (2000) 3788e3790.

[24] J.F. Kauffman, G.L. Richmond, Dependence of luminescence decays from Gaaselectrolyte contacts on excitation power and applied bias e examination ofthe modified dead layer model, J. Appl. Phys. 73 (1993) 1912e1917.

[25] J.I. Pankove, GaN: from fundamentals to applications, Mat. Sci. Eng. B Solid 61-2 (1999) 305e309.

[26] D.S. Li, M. Sumiya, S. Fuke, D.R. Yang, D.L. Que, Y. Suzuki, Y. Fukuda, Selectiveetching of GaN polar surface in potassium hydroxide solution studied by X-ray photoelectron spectroscopy, J. Appl. Phys. 90 (2001) 4219e4223.

[27] S. Anantathanasarn, S. Ootomo, T. Hashizume, H. Hasegawa, Surfacepassivation of GaAs by ultra-thin cubic GaN layer, Appl. Surf. Sci. 159 (2000)456e461.

[28] A.R. Smith, R.M. Feenstra, D.W. Greve, J. Neugebauer, J.E. Northrup, Re-constructions of the GaN(000(1)over-bar) surface, Phys. Rev. Lett. 79 (1997)3934e3937.

[29] T.K. Zywietz, J. Neugebauer, M. Scheffler, The adsorption of oxygen at GaNsurfaces, Appl. Phys. Lett. 74 (1999) 1695e1697.

Table 2Calculated Schottky diode parameters of passivated diodes.

Diodename

Treatment,measurement

Idealityfactor

Barrierheight (eV)

Rev. Leak.Curr. (IR)at VR ¼ �1 V

Untreated No 1.4 0.76 6E-6AMS (alc.) (NH4)2Sx (alc.)

for 100 s1.3 0.98 6E-7

A. Kumar et al. / Current Applied Physics 14 (2014) 491e495494

Author's personal copy

[30] S.C. Cruz, S. Keller, T.E. Mates, U.K. Mishra, S.P. DenBaars, Crystallographicorientation dependence of dopant and impurity incorporation in GaN filmsgrown by metalorganic chemical vapor deposition, J. Cryst. Growth 311(2009) 3817e3823.

[31] J. Neugebauer, T. Zywietz, M. Scheffler, J.E. Northrup, C.G. Van de Walle, Cleanand As-covered zinc-blende GaN (001) surfaces: novel surface structures andsurfactant behavior, Phys. Rev. Lett. 80 (1998) 3097e3100.

[32] C. Huh, S.W. Kim, H.S. Kim, I.H. Lee, S.J. Park, Effective sulfur passivation of ann-type GaN surface by an alcohol-based sulfide solution, J. Appl. Phys. 87(2000) 4591e4593.

[33] E.H. Rhoderick, R.H. Williams, Metal-semiconductor Contacts, 2 ed., OxfordUniversity Press, New York, 1988.

[34] S.M. Sze, K.K. Ng, Physics of Semiconductor Devices, Wiley-Interscience,Hoboken, N.J, 2007.

[35] J. Bardeen, Phys. Rev. 71 (1947) 717.[36] A.Y. Polyakov, N.B. Smirnov, A.V. Govorkov, A.V. Markov, S.J. Pearton,

N.G. Kolin, D.I. Merkurisov, V.M. Boiko, C.R. Lee, I.H. Lee, Fast neutron irradi-ation effects in n-GaN, J. Vac. Sci. Technol. B 25 (2007) 436e442.

[37] V.N. Brudnyi, A.V. Kosobutsky, N.G. Kolin, The charge neutrality level and thefermi level pinning in A(3)N (BN, AlN, GaN, InN) nitrides, Russ. Phys. Jþ 51(2008) 1270e1278.

[38] A. Kumar, S. Arafin, M. Amann, R. Singh, Temperature dependence of electricalcharacteristics of Pt/GaN Schottky diode fabricated by UHV e-beam evapo-ration, Nanoscale Res. Lett. 8 (2013) 481.

A. Kumar et al. / Current Applied Physics 14 (2014) 491e495 495