# Fileset

[Okumura_2023_Jpn._J._Appl._Phys._62_064001 (1).pdf](https://mdr.nims.go.jp/filesets/6afe2ca4-4658-4758-a0aa-4df0d6a78ce9/download)

## Creator

Hironori Okumura, Yohei Ogawara, Manabu Togawa, Masaya Miyahara, Tadaaki Isobe, Kosuke Itabashi, Jiro Nishinaga, [Masataka Imura](https://orcid.org/0000-0002-4236-9549)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Degradation of vertical GaN diodes during proton and xenon-ion irradiation](https://mdr.nims.go.jp/datasets/cd1d0636-1f8e-456d-9c03-deaa1ed1af7a)

## Fulltext

Degradation of vertical GaN diodes during proton and xenon-ion irradiationJapanese Journal of AppliedPhysics     REGULAR PAPER • OPEN ACCESSDegradation of vertical GaN diodes during protonand xenon-ion irradiationTo cite this article: Hironori Okumura et al 2023 Jpn. J. Appl. Phys. 62 064001 View the article online for updates and enhancements.You may also likeStopping power and radial dosedistribution for 42 MeV bromine ionsM N Varma, J W Baum and A V Kuehner-Scattering kernels for fast neutron therapytreatment planningGregory B Moffitt, Landon S Wootton,Björn Hårdemark et al.-Light absorption enhancement andradiation hardening for triple junction solarcell through bioinspired nanostructuresThomas Vasileiou, José M Llorens,Jerónimo Buencuerpo et al.-This content was downloaded from IP address 144.213.253.16 on 23/08/2024 at 08:56https://doi.org/10.35848/1347-4065/acddb4/article/10.1088/0031-9155/25/4/003/article/10.1088/0031-9155/25/4/003/article/10.1088/1361-6560/ab9a85/article/10.1088/1361-6560/ab9a85/article/10.1088/1748-3190/ac095b/article/10.1088/1748-3190/ac095b/article/10.1088/1748-3190/ac095bDegradation of vertical GaN diodes during proton and xenon-ion irradiationHironori Okumura1* , Yohei Ogawara1, Manabu Togawa2,3, Masaya Miyahara2,3, Tadaaki Isobe4 , Kosuke Itabashi2,3,Jiro Nishinaga5, and Masataka Imura61Faculty of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8573 Japan2High Energy Accelerator Research Organization, Tsukuba, Ibaraki 305-0801, Japan3International Center for Quantum-field Measurement Systems for Studies of the Universe and Particles, Tsukuba, Ibaraki 305-0801, Japan4The Nishina Center for Accelerator-Based Science, RIKEN, Wako, Saitama 351-0198 Japan5Research Center for Photovoltaics, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki 305-8568, Japan6Research Center for Functional Materials, National Institute for Materials Science, Tsukuba, Ibaraki 305-0044, Japan*E-mail: okumura.hironori.gm@u.tsukuba.ac.jpReceived June 2, 2022; revised April 19, 2023; accepted June 11, 2023; published online June 30, 2023We investigated the material stability of a vertical GaN Schottky barrier diode (SBD) against proton irradiations by making real-time measurements.The reverse current gradually decreased with increasing proton fluence. The current of the GaN SBD was reduced by 18% after proton irradiationswith a displacement-damage dose (Dd) of 1012 MeV g−1. We also examined signal and current degradation occurring in a vertical GaN-on-GaNp–n diode (PND) during xenon-ion irradiations. The signal gradually decreased with increasing xenon-ion fluence. Xenon-ion irradiations ofDd = 1012 MeV g−1 reduced the collected charge in the PND by 11%. This signal degradation was close to the current degradation in the GaN SBDcaused by the proton irradiations. We found that irradiations with Dd > ∼1012 MeV g−1 degraded the performance of the GaN devices.© 2023 The Japan Society of Applied Physics1. IntroductionRadiation detectors are used in high-energy physics, astrophy-sics, spectroscopy, security, and radiotherapy. Semiconductorradiation detectors have advantages of compactness, fast timingcharacteristics, and high energy resolution. In a semiconductormaterial, incident radiation produces electron–hole pairs. Theelectrons and holes diffuse to the electrodes of the detector whenan electric field is applied to them, generating a signal pulse.However, some forms of radiation produce Frenkel-type defectsby displacing atoms from their lattice sites, leaving vacancies,interstitials, and complexes.1,2) Point defects form deep ac-ceptor/donor levels, decreasing the effective donor/acceptorconcentration and carrier mobilities. Prolonged exposure ofsemiconductor detectors to radiation reduces their chargecollection efficiency (CCE), causing a loss in energy and spatialresolution. A semiconductor material with a large latticedisplacement energy is preferable for the long-term stability ofa detector in a high-radiation environment.Wide-bandgap semiconductors are suitable for harsh en-vironments because of their low intrinsic carrier density andlarge lattice displacement energy.3–7) GaN is an especiallyattractive material for radiation detectors because of its highcarrier mobility, which enhances charge carrier collection,and high atomic number, which leads to many electron–holepairs, or a large signal. A high-sensitivity GaN radiationdetector requires a vertical structure with a GaN substratebecause the signal-to-noise ratio increases with increasingdepletion width and decreasing leakage current. There arereports on charged particle detection using vertical GaNSchottky barrier diodes (SBDs) and p–n junction diodes.8–11)The degradation of semiconductor devices by chargedparticles is linearly correlated to the non-ionizing energyloss (NIEL) due to the nuclear stopping component.12) Thedisplacement damage depends on the type, fluence, andenergy of the radiation. Degradation is often determined bysubjecting the device to a 1–100MeV proton irradiationbecause protons effectively produce a more measurablechange in device characteristics in comparison with electronsor neutrons.13) In a vertical GaN PND, 2.5 MeV protons at afluence of 4 × 1013 protons cm−2 cause the breakdownvoltage to decrease by 23%,14) and reduce the hole diffusionlength by 55%.15) Still, there are few reports on the relationbetween displacement damage and device characteristics forGaN.In this study, we investigated displacement damage in GaNby irradiating a vertical SBD with protons. We also inves-tigated signal and current degradation in vertical GaN PNDsby using xenon swift-heavy ions.2. Prolonged exposure of GaN SBD to protons2.1. Experimental procedureWe used a 2 inch n-type GaN wafer with a threadingdislocation density of 5 × 106 cm−2, thickness of 300 μm,and [Si] = 2 × 1018 cm−3. A GaN SBD was directlyfabricated without epitaxial growth using the GaN wafer,which was cut into a 10 mm × 10 mm piece. A 20 nmTi/100 nm Al/10 nm Ni/50 nm Au electrode was evaporatedon the backside of the sample by using electron-beamdeposition, followed by thermal annealing at 800 °C for1 min in a nitrogen ambient to form an ohmic contact. A Ni(25 nm)/Au (25 nm) electrode with a diameter of 1 mm wasevaporated on the sample surface using electron-beamdeposition. Current–voltage (I–V ) characteristics were mea-sured at RT using a parameter analyzer (Keysight B1500A)before the proton irradiations.We built a real-time system for measuring the electricalproperties, as shown in Fig. 1(a). The GaN SBD wasconnected to a source-measure unit (ADCMT 6247G), whichwas controlled remotely via a laptop. The cathode electrodewas directly soldered on an Al plate. The anode electrode wasbonded to a flexible board using 40 μm thick Al wire. Thesample was mounted on a XY-axis robotic scanningsystem.16) The optimized scanning speed and area were∼20 mm s−1 and 32 mm × 32 mm, which helped to reducethe scan non-uniformity to less than 5%. The scanningsystem was isolated from the environment in a temperature-controlled chamber that was maintained at −15 °C by a cold064001-1 © 2023 The Author(s). Published on behalf of The Japan Society of Applied Physics by IOP Publishing LtdJapanese Journal of Applied Physics 62, 064001 (2023) REGULAR PAPERhttps://doi.org/10.35848/1347-4065/acddb4Content from this work may be used under the terms of the Creative Commons Attribution 4.0 license. Any further distribution of thiswork must maintain attribution to the author(s) and the title of the work, journal citation and DOI.https://crossmark.crossref.org/dialog/?doi=10.35848/1347-4065/acddb4&domain=pdf&date_stamp=2023-06-30https://orcid.org/0000-0002-5464-9169https://orcid.org/0000-0002-5464-9169https://orcid.org/0000-0001-5163-030Xhttps://orcid.org/0000-0001-5163-030Xmailto:okumura.hironori.gm@u.tsukuba.ac.jphttps://doi.org/10.35848/1347-4065/acddb4https://creativecommons.org/licenses/by/4.0/nitrogen gas flow to reduce the effect of thermal agitation onthe electrical characteristics.We used a high-intensity proton beam with a momentumof 70 MeV and ∼300 nA beam current provided by the 930-type azimuthal varying field cyclotron at the cyclotron andradioisotope center facility of the Tohoku University. Thescanning system enabled the proton beam, with a size of3.5 mmf, to uniformly irradiate the sample plane and indepth. The delivered proton fluence (fp) was estimated fromthe activity of Al foils after proton irradiations using gammarays from 22Na decays. fp of 9.2 × 1015 cm−2 is converted to1.4 × 1016 neq (1 MeV neutron equivalent) cm−2 because a70MeV proton induces displacement damage of 1.52 neq.17)The stopping power (dE/dx) in GaN was calculated to be40MeV cm−1 for 70MeV protons in a stopping and range ofions in matter (SRIM) simulation, indicating that the incidentprotons penetrated the GaN SBD. The equivalent totalirradiation dose for the GaN SBD is estimated to be 9.6MGy using the equation ,e dEdxpfrwhere e is elementary chargeand ρ is density (= 6.15 g cm−3 for GaN).2.2. Electrical properties of GaN SBDThe I–V characteristics of the GaN SBD before the protonirradiations are shown in Fig. 1(b). The GaN SBD showedrectifier behavior. The forward I–V characteristics wereanalyzed using a thermionic emission model with aSchottky barrier height of 1.0 eV and Richardson constantof 24 Acm−2 K−2. The specific on-resistance (Ron) andideality factor (n) were 6.3 Ωcm2 and 1.5, respectively. Ronand n were larger than in the other reports due to the largeleakage current.18,19) The leakage current of the GaN SBDcould be decreased by using a small anode electrode, a mesastructure, and a lightly donor-doped GaN epitaxial layer.The I–V characteristics of the GaN SBD just after theproton irradiations are also shown in Fig. 1(b). The GaN SBDshowed no rectifier behavior due to radiation damage. Weconsider that this degradation owes to radiation-inducedacceptor-like defects being generated in the n-typeregion.20) Particularly, a gallium-vacancy defect (VGa) hasan acceptor level at 1.0 eV above the VB maximum and actsas a compensation center for n-type GaN.21,22) Although thedisplacement energies of both gallium and nitrogen atoms aremuch less than the energy transferred to primary knock-onatoms in a high-fluence proton irradiation,13) VGa-relateddefects may have primarily formed in the n-type GaNbecause of the low formation energy. Further investigationusing deep-level transient spectroscopy are needed to clarifythe origin of the proton-induced defects.2.3. Degradation transition of GaN SBDWe made real-time measurements of the reverse current ofthe GaN SBD to investigate the material’s stability to protonexposure. The current during the proton irradiations wasaround two orders of magnitude larger than the pre-irradia-tion current because electron–hole pairs were generated bythe proton irradiations and thermal agitation. The measuredcurrent during the proton irradiations oscillated due to theirradiated area being shifted by the XY-axis scanning system.We evaluated the device degradation by using the current atconcave points, at which the proton beam was the farthestfrom the samples. The depletion width (Wd) of a SBD isobtained by using W V V ,d eN dkTe2 sD= - -e ( ) where εs(= 10.4ε0) is the dielectric constant of GaN,23) ND (∼[Si])is the donor concentration, Vd (= 0.95 V) is the built-inpotential,24) V is the vias, k is the Boltzmann constant, and Tis the absolute temperature. Wd of the GaN SBD is estimatedto be 110 nm at V = −20 V.The dependence of the current at the reverse bias of 20 V onthe proton fluence (fpe) was estimated by assuming thatthe proton flux per unit time was constant. As shown inFig. 2(a), the reverse current showed little degradation atfpe < 5 × 1014 neq cm−2 and was close to those of thereported GaN devices.7,13) The reverse current graduallydecreased with increasing fpe due to the increased number ofproton-induced defects. The degraded reverse current was fittedusing the equation I I 1 ,p pe0 n1a f= + -( ) 25) where I0 is thecurrent at the start of the irradiation, αp is a fitting parameterdefined as the damage constant, and n is another fittingparameter. According to Rose-Barnes theory, n approaches 1/3 when the current is dominated by the space charge recombi-nation, while n approaches 2/3 when the current is dominatedby diffusion.26) The experimental data were fitted using αp= 2.5 × 10−16 cm2 and n = 0.33, which is close to 1/3 due tothe high density of recombination centers. A positive currentappeared under reverse bias at a fpe over 6 × 1015 cm−2. Weconsider that excess carriers that were generated by the protonsand thermal agitation might have remained during the real-timeI–V measurement. Hall-effect and I–V measurements should beconducted after cooling the sample in an area with less radiationto clarify the reason of the positive currents.(a) (b)Fig. 1. (a) Real-time current measurement system of n-GaN SBD with 1 mm diameter Ni contact during 70 MeV proton irradiation. (b) Current–voltagecharacteristics of n-GaN SBD before and after 70 MeV proton irradiation.064001-2 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.The relation between the normalized output value ofvarious GaN devices and the displacement damage dose(Dd), which is equal to NIEL times fp, is shown in Fig. 2(b).The plotted values include the current transition of GaN high-electron mobility transistors (HEMTs),27–30) the currenttransition of the GaN PND,31) and intensity transition of GaNLEDs.32) The measured current during the proton irradiationsshould be higher than the current after the proton irradiationand cooling because the measured current includes thecarriers generated by radioactivated surrounding objects andthermal agitation. However, the fitting curve was close to thereported current transitions of the GaN HEMTs and PNDs,indicating that radiation damage is dominant on the degrada-tion transition of the GaN SBD. NIEL increases withdecreasing proton energy due to the increased electronic-stopping component; SR-NIEL web calculators gave 2.3 ×10−2 MeV cm2 g−1 for 2.5 MeV, 3.9 × 10−3 MeV cm2 g−1for 50MeV, and 3.5 × 10−3 MeV cm2 g−1 for 70MeV.33)The current of the GaN SBD fell by 18% after protonirradiations with Dd = 1012 MeV g−1. Thus, we concludedthat the performance of the GaN devices was stable againstproton irradiations with Dd < ∼1012 MeV g−1. The fittingcurve showed a slightly smaller degradation compared withthe relation between normalized intensity and fp for a GaNLED irradiated with 50MeV protons, for which αp = 6.6 ×10−16 cm2 and n = 0.32 were used.32) This smallerdegradation may have resulted from the formation of com-plexes between the proton-induced defects and donor/ac-ceptor dopants in the GaN LED and PND.27–32) We need toconduct further investigations to elucidate the relationbetween the degradation and the structure of GaN devices.3. Exposure of GaN PND to Xe ions3.1. Experimental procedureWe used a GaN homoepitaxial layer grown on a 2 inchn-type GaN substrate with a threading dislocation density of∼106 cm−2 and [Si] = 2 × 1018 cm−3 by metal-organic CVD(SCIOCS Co., Ltd.). As schematically shown in Fig. 3(a), thePND consisted of the following layers from the top: 500 nmp-GaN ([Mg] = 1 × 1018 cm−3)/10 μm n−-GaN ([Si] = 1 ×1016 cm−3)/2 μm n-GaN ([Si] = 1 × 1018 cm−3) layers with a50 nm p+-GaN ([Mg] = 2 × 1020 cm−3) contact layer. Aftercutting the sample to the size of 5 mm × 4 mm, we fabricatedthe GaN PND. A 20 nm Ti/100 nm Al/10 nm Ni/50 nm Auelectrode was evaporated on the backside of the sample byusing electron-beam deposition, followed by thermal an-nealing at 800 °C for 1 min in a nitrogen ambient. After25 nm Ni/25 nm Au electrodes were deposited, thermalannealing was carried out at 500 °C for 10 min in an oxygenambient.34) I–V and capacitance-voltage (C–V ) characteris-tics were measured using a parameter analyzer. The samplewas mounted on a package with silver paste. Anodeelectrodes were bonded to each pin with 40 μm thick Al wire.We used a xenon beam with a momentum of 400MeV/nprovided by the heavy-ion medical accelerator in Chiba(HIMAC) located at the National Institute of RadiologicalScience. The GaN PND was irradiated in the air through acollimator at 1 meter from the beam duct by a 132Xe beamwith a size of 3–5 mmf. During irradiation, the xenon-ionfluence (fxe) was maintained to 5 × 106 particles cm−2 per3.3 s pulse, which was monitored by a scintillator counter.The total xenon-ion fluence over the course of 256 min was3 × 1010 cm−2. The ion LET, given by ,dEdxrwas calculatedusing an SRIM simulation. From Fig. 3, the average LET inthe GaN PND was estimated to be 6.3 MeV cm2 mg−1. Asingle-event particle effect occurs when the xenon beamstrikes the GaN layers, resulting in transient signals. Thecharge per unit length created by an ionization is given bye ,dQdx EdEdxeh= where Eeh (=9.59 eV for GaN) is the meanelectron–hole pair creation energy.35) The charge created bya 400MeV n−1 xenon ion is estimated to be 65 fC μm−1 forGaN. An amplifier-shaper-discriminator IC was used to readthe detection signal. The number of charges detected by theGaN PND was calculated from the value of the analog-to-digital converter (ADC) peak. The detector capacitanceincluding the coaxial cable was 22 pF. The signal wasmeasured at RT in air by an oscilloscope (MDO4104C).3.2. Degradation transition of GaN PNDThe current density–voltage (J–V ) characteristics of the GaNPND with an anode diameter of 1 mm are shown in Fig. 4(a).(a) (b)Fig. 2. (a) Reverse current of n-GaN SBD at reverse bias of 20 V during 70 MeV proton irradiation. (b) Relation between displacement-damage dose andnormalized current of the n-GaN SBD fitted using Rose-Barnes theory (solid line), output intensity of GaN LED (dashed line),30) current of GaN PND(circle),29) and drain current of GaN HEMT (squares).25–28)064001-3 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.The GaN PND was fabricated without a mesa structure tominimize the effect of process damage on the electricalcharacteristics. The apparent turn-on voltage was 3.1 V. Aforward current of 100 mA cm−2 was measured at 3.7 V. Ronwas 6.8 Ωcm2 for 100 mA cm−2. The on/off current ratio was∼103. The large leakage current of ∼10−4 A cm−2 isattributed to the high concentration of threading dislocations,∼106 cm−2, and the large device size having a 1 mmdiameter and no mesa structure.The fxe dependence of the change in displacement-damage-induced leakage current normalized to volume forthe GaN PND is shown in Fig. 4(b). Here, the reverse biaswas 7 V to protect the circuit from the large leakage current.Neither single-event burnout nor catastrophic failure thatwould have increased the leakage current were observed atthis reverse bias. The increase in reverse current results fromelectron–hole pairs generated at damage-induced defects. Thevariation in leakage current is proportional to fxe, i.e.,IV c xesa f=D 36) where Vs is the sensitive volume corre-sponding to the junction area times the detector thickness andαc is a fitting parameter called the current-related damagerate. αc was 1.7 × 10−11 A cm−1, which is five orders ofmagnitude larger than that (10−17– 10−16 A cm−1) of asilicon detector for a 1 MeV neq flux.36) A precise determina-tion of αc for a GaN detector will require a reduction in thelateral capacitance by using a guard ring.3.3. Signal degradation of GaN PNDA single event, as shown by the ADC double peaks inFig. 5(a), was observed at a reverse bias of 7 V. The ADCpeak of 250 may be derived from the leakage current to theelectrode of adjacent devices with different anode diameters.The ADC peak of 427 is evaluated to be 136 fC, whichcorresponds to 20% of the expectation for fully depletion ofthe n−-GaN layer. When the concentration of the generatedcarriers is smaller than the doping concentration, the totalcollected charges consist of carriers collected from thedepletion region as drift current and carriers collected fromthe non-depletion region as diffusion current. In this low-injection charge collection model, the CCE is calculatedfrom ⎛⎝⎞⎠x xCCE d e d ,EWEx WDEx10ddddddx WdL0 ò ò= + - - 37) whereE0 ( xdDEx0ddò= ) is the total energy which the incidentparticle gives to fully depleted GaN, D (∼10 μm) is thethickness of the GaN epilayer, and L (∼250 nm) is thediffusion length of hole in n-GaN, using SRIM simulation.(a) (b)Fig. 4. (a) J–V characteristics of 1 mm diameter GaN PND. (b) Fluence dependence of leakage current for GaN PND at reverse bias of 7 V during400 MeV n−1 xenon-ion irradiation.Fig. 3. Schematic cross-section of vertical GaN PND and linear energy transfer (LET) of xenon ion with 400 MeV n−1 in GaN at the sample depth andSRIM calculation of electronic stopping power of 52400 MeV xenon ions for GaN.064001-4 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.Wd of the GaN PND is estimated to be 729 nm at V = −7 V.Expected CCE of the GaN PND is calculated to be 9.8%,which is smaller than the experimental result. This discre-pancy may be attributed to the carriers collected fromextended drift region as drift current at a low bias.38)The dependence of the normalized collected charges on Ddis shown in Fig. 5(b) for the GaN PND and Si detector. SR-NIEL web calculators indicated that NIEL increases withdecreasing xenon-ion energy; 4.2 × 101 MeV cm2 g−1 inGaN and 4.6 × 101 MeV cm2 g−1 in Si for 400MeV n−1.After xenon-ion irradiation with a fluence of 2 × 1010 cm−2,the charges collected by the Si detector and GaN PND werereduced to 80% and 92%, respectively, of their pre irradiationvalues. We consider that xenon-ion irradiation increases thecarrier scattering and reduces the carrier lifetime due to thecreated trap states,39–41) resulting in signal degradation. Theexperimental data of the collected charges were fitted byQ Q 1 ,q xe0 a f= -( ) where Q0 is the charge collected at thestart of the irradiation, and αp is a fitting parameter defined asthe damage constant. αq was 1 × 10−13 cm2 for the GaNPND and 4 × 10−13 cm2 for the Si detector. This indicatesthat the GaN PND was more tolerant to the radiation than theSi detector due to the high displacement energy. Thedegradation of the GaN PND, despite it having channelwith a high-quality epilayer, was close to that of the GaNSBD caused by the proton irradiation, indicating that thesignal reduction is due to the displacement damage. We thusfound that irradiations with a Dd > ∼1012 MeV g−1dramatically degraded the performance of the GaNdevices.3.4. Signal detection using GaN PND with mesastructureA GaN PND of the design in Fig. 3 was fabricated with amesa termination of 600 nm depth to reduce the leakagecurrent. The mesa termination was fabricated by reactive-ionetching at an inductively coupled plasma power of 150Wwith Cl2/BCl3 mixing gas of 20/50 sccm under 5 Pa for 2 minafter deposition of a 50 nm thick Ni mask. The J–Vcharacteristics of the GaN PND with an anode diameter of200 μm is shown in Fig. 6(a). The apparent turn-on voltagewas 3.2 V, as expected from the bandgap energy of GaN. TheGaN PND exhibited good rectifying behavior with a forwardcurrent of 100 A cm−2 at 3.7 V. The leakage current wasdramatically reduced by the mesa isolation and small anodecontact,42) resulting in an on/off current ratio of ∼109. Ronwas 5 mΩcm2 for 100 A cm−2, which is higher than the otherreports, 0.1–2.3 mΩcm2.43–48) Ron would be reduced byoptimizing the fabrication process of the p-type GaN contact.The forward bias J–V characteristics were analyzed using therelation J J e 1 ,0eVnkT= -( ) where J0 is the saturation currentdensity, e is the electron charge, k is the Boltzmann constant,and T is the absolute temperature. n of the GaN PND was 2 atvoltages between 2.4 and 3.2 V, because of the Shockley–Read–Hall recombination current.49,50) The avalanche break-down voltage (VB) of an ideal planar junction under thepunch-through condition is calculated to be 1.5 kV from therelation V E W ,e N N WB c 2 sD A2= -e-( ) where W is the thicknessof the n−-GaN drift layer, ND–NA the effective donorconcentration, and Ec the critical electric field (=2.4 MVcm−1 for ND–NA = 1 × 1016 cm−3).51) VB measured without(a) (b)Fig. 5. (a) Outputs from 1 mm diameter GaN PND for xenon-ion irradiation. (b) Displacement-damage dose dependence of normalized corrected charges for1 mm diameter GaN PND (circle) and Si detector (square) at reverse bias of 7 V during 400 MeV n−1 xenon-ion irradiation.(a) (b)Fig. 6. (a) RT J–V characteristics of 200 μm diameter GaN PND. (b) Outputs from 200 μm diameter GaN PND for xenon-ion irradiation.064001-5 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.fluorinert was 444 V, which is much smaller than thecalculated value. Surface passivation, good edge termination,and a field-plate structure would increase the breakdownvoltage.Signals from xenon ions were detected at a reverse bias of50 V by using the GaN PND with the mesa structure. Wd ofthe GaN PND is estimated to be 1.53 μm at V = −50 V.Expected CCE of the GaN PND is calculated to be 17%. Asshown in Fig. 6(b), a single event was observed, and a singleADC peak was obtained. The ADC peak value of 760corresponded to 242 fC, which is 37% of what would beexpected for full depletion of the n−-GaN layer. This meansthat the reverse bias is still small. The collected chargeswould be further increased by increasing a reverse bias.4. ConclusionsWe investigated the material stability of a vertical GaN SBDduring 70MeV proton irradiation by making real-time mea-surements of the reverse current. The current of the GaN SBDfell by 18% after the proton irradiation with Dd =1012MeV g−1. The reverse current gradually decreased withincreasing proton fluence. We also investigated signal degra-dation in vertical GaN-on-GaN PND during 400MeV n−1xenon-ion irradiation. A single event was observed, showingthe collected charge of 240 fC at a reverse bias of 7 V.Although the signal gradually decreased with increasingxenon-ion fluence, the degradation of the collected chargesin the GaN PND was smaller than that of an Si detector. Wefound that the performance of the GaN devices remainedstable against irradiations with displacement-damage doses<∼1012 MeV g−1.AcknowledgmentsThis work was financially supported by the TIA “Kakehashi”collaborative research program, a project for Tsukuba in-dustry-university collaboration promotion, and the Muratascience foundation. Part of the experiments was performedunder the Research Project with Heavy Ions at NIRS-HIMAC, program No. 21H455. The irradiation for assessingthe radiation damage was performed at CYRIC. The authorsalso wish to thank the accelerator staff at HIMAC andCYRIC for supplying the excellent beams used in thiswork. The device fabrication was carried out at the Nano-Processing Facility at the National Institute of AdvancedIndustrial Science and Technology and open facility of theUniversity of Tsukuba.ORCID iDsHironori Okumura https://orcid.org/0000-0002-5464-9169Tadaaki Isobe https://orcid.org/0000-0001-5163-030X1) G. F. Knoll, Radiation Detection and Measurement (Wiley, New Jersey,1980) 4th ed.2) G. Lindstrom, M. Moll, and E. Fretwurst, Nucl. Instrum. Methods Phys.Res., Sect. A 426, 1 (1999).3) R. J. Tapper, Rep. Prog. Phys. 63, 1273 (2000).4) D. C. Look, Mater. Sci. Eng. B80, 383 (2001).5) A. Owens and A. Peacock, Nucl. Instrum. Methods Phys. Res., Sect. A 531,18 (2004).6) F. Nava, G. Bertuccio, A. Cavallini, and E. Vittone, Meas. Sci. Technol. 19,102001 (2008).7) J. Wang, P. Mulligan, L. Brillson, and L. R. Cao, Appl. Phys. Rev. 2,031102 (2015).8) M. Sugiura, M. Kushimoto, T. Mitsunari, K. Yamashita, Y. Honda,H. Amano, Y. Inoue, H. Mimura, T. Aoki, and T. Nakano, Jpn. J. Appl.Phys. 55, 05FJ02 (2016).9) Q. Xu, P. Mulligan, J. Wang, W. Chuirazzi, and L. Cao, Nucl. Instrum.Methods Phys. Res., Sect. A 849, 11 (2017).10) L. Zhou, Z. Lu, J. Wu, H. Jiang, L. Chen, X. Ouyang, and K. M. Lau, IEEEElectron Device Lett. 40, 1044 (2019).11) A. Sandupatla, S. Arulkumaran, N. G. Ing, S. Nitta, J. Kennedy, andH. Amano, Micromachines 11, 519 (2020).12) J. R. Srour, C. J. Marshall, and P. W. Marshall, IEEE Trans. Nucl. Sci. 50,653 (2003).13) S. J. Pearton, R. Ren, E. Patrick, M. E. Law, and A. Y. Polyakov, ECS J.Solid State. Sci. Technol. 5, Q35 (2016).14) M. P. King et al., IEEE Trans. Nucl. Sci. 62, 2912 (2015).15) K. C. Collins, A. M. Armstrong, A. A. Allerman, G. Vizkelethy, S. B. VanDeusen, F. Leonard, and A. A. Talin, J. Appl. Phys. 122, 235705 (2017).16) K. Nakamura, Y. Arai, M. Hagihara, K. Hanagaki, K. Hara, R. Hori,M. Hirose, Y. Ikegami, O. Jinnouchi, and S. Kamada, J. Instl. 10, C06008(2015).17) (https://rd50.web.cern.ch/NIEL/default.html) (accessed August 15, 2000).18) Y. Cao, R. Chu, R. Li, M. Chen, R. Chang, and B. Hughes, Appl. Phys. Lett.108, 062103 (2016).19) T. Maeda, M. Okada, M. Ueno, Y. Yamamoto, T. Kimoto, M. Horita, andJ. Suda, Appl. Phys. Lett. 10, 051002 (2017).20) G. Lindstorm, Nucl. Instrum. Methods Phys. Res., Sect. A 512, 30 (2003).21) Z. Zhang, A. R. Arehart, E. Cinkllic, J. Chen, E. X. Zhang, D. M. Fleetwood,R. D. Schrimpf, B. McSkimming, J. S. Speck, and S. A. Ringel, Appl. Phys.Lett. 103, 042102 (2013).22) J. L. Lyons and C. G. Van de Walle, npj Comput. Mater. 3, 1 (2017).23) M. J. Kane, M. J. Uren, D. J. Wallis, P. J. Wright, D. E. J. Soley, A. J. Simons,and T. Martin, Semicond. Sci. Technol. 26, 085006 (2011).24) T. Maeda, M. Okada, M. Ueno, Y. Yamamoto, T. Kimoto, M. Horita, andJ. Suda, Appl. Phys. Exp. 10, 051002 (2017).25) B. H. Rose and C. E. Barnes, J. Appl. Phys. 53, 1772 (1982).26) K. Saarinen et al., Phys. Rev. Lett. 79, 3030 (1997).27) A. Karmarkar, B. Jun, D. M. Fleetwood, R. D. Schrimpf, R. A. Weller,B. D. White, L. J. Brillson, and U. K. Mishra, IEEE Trans. Nucl. Sci. 51,3801 (2004).28) L. Lv, J. G. Ma, Y. R. Cao, J. C. Zhang, W. Zhang, L. Li, S. R. Xu,X. H. Ma, X. T. Ren, and Y. Hao, Microelectron. Reliab. 51, 2168(2011).29) B. D. Weaver, P. A. Martin, J. B. Boos, and C. D. Cress, IEEE Trans. Nucl.Sci. 59, 3077 (2012).30) Y. Xi et al., J. Vac. Sci. Technol. B 32, 012201 (2014).31) G. Vizkelethy, M. P. King, O. Aktas, I. C. Kizilyalli, and R. J. Kaplar, Nucl.Instrum. Methods Phys. Res., Sect. B 404, 264 (2017).32) S. M. Khanna et al., IEEE Trans. Nucl. Sci. 51, 2729 (2004).33) (http://Sr-niel.org/) (accessed June 15, 2023).34) H. Okumura, D. Martin, and N. Grandjean, Appl. Phys. Lett. 109, 252101(2016).35) E. B. Yakimov, A. Y. Polyakov, I. V. Shchemerov, N. B. Smirnov,A. A. Vasilev, P. S. Vergeles, E. E. Yakimov, A. V. Chernykh, F. Ren,and S. J. Pearon, Appl. Phys. Lett. 118, 202106 (2021).36) M. Moll, “Radiation Damage in Silicon Particle Detectors,” Ph.D. ThesisHumburg University (1999).37) M. B. H. Breese, J. Appl. Phys. 74, 3789 (1993).38) N. Iwamoto, S. Onoda, T. Makino, T. Oshima, K. Kojima, A. Koizumi,K. Uchida, and S. Nozaki, IEEE Trans. Nucl. Sci. 58, 305 (2011).39) V. V. Ursaki, I. M. Tiginyanu, O. Volciuc, V. Popa, V. A. Skuratov, andH. Morkoc, Appl. Phys. Lett. 90, 161908 (2007).40) L. M. Zhang, C. H. Zhang, L. Q. Zhang, X. J. Jia, T. D. Ma, Y. Song,Y. T. Yang, B. S. Li, and Y. F. Jin, Nucl. Instrum. Methods Phys. Res., Sect.B 269, 1782 (2011).41) P. N. M. Ngoepe, W. E. Meyer, F. D. Auret, E. Omotoso, T. T. Hlatshwayo,V. A. Skuratov, and M. Diale, Nucl. Instrum. Methods Phys. Res., Sect. B409, 69 (2017).42) H. Okumura, Y. Ogawara, M. Togawa, M. Miyahara, J. Nishinaga,M. Imura, and T. Isobe, Ext. Abstr. Int. Conf. Solid State Devices andMaterials, 2021, p. 199.43) I. C. Kizilyalli, A. P. Edwards, H. Nie, D. Bour, T. Prunty, and D. Disney,IEEE Electron Device Lett. 35, 247 (2014).44) K. Nomoto, B. Song, Z. Hu, M. Zhu, M. Qi, N. Kaneda, T. Mishima,T. Nakamura, D. Jena, and H. G. Xing, IEEE Electron Device Lett. 37, 161(2016).064001-6 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.https://orcid.org/0000-0002-5464-9169https://orcid.org/0000-0002-5464-9169https://orcid.org/0000-0002-5464-9169https://orcid.org/0000-0001-5163-030Xhttps://orcid.org/0000-0001-5163-030Xhttps://orcid.org/0000-0001-5163-030Xhttps://doi.org/10.1088/0034-4885/63/8/203https://doi.org/10.1016/j.nima.2004.05.071https://doi.org/10.1016/j.nima.2004.05.071https://doi.org/10.1088/0957-0233/19/10/102001https://doi.org/10.1088/0957-0233/19/10/102001https://doi.org/10.1063/1.4929913https://doi.org/10.1063/1.4929913https://doi.org/10.7567/JJAP.55.05FJ02https://doi.org/10.7567/JJAP.55.05FJ02https://doi.org/10.1016/j.nima.2016.12.061https://doi.org/10.1016/j.nima.2016.12.061https://doi.org/10.3390/mi11050519https://doi.org/10.1149/2.0251602jsshttps://doi.org/10.1149/2.0251602jsshttps://doi.org/10.1063/1.5006814https://rd50.web.cern.ch/NIEL/default.htmlhttps://doi.org/10.1063/1.4941814https://doi.org/10.1063/1.4941814https://doi.org/10.1063/1.4816423https://doi.org/10.1063/1.4816423https://doi.org/10.1088/0268-1242/26/8/085006https://doi.org/10.7567/APEX.10.051002https://doi.org/10.1063/1.331649https://doi.org/10.1103/PhysRevLett.79.3030https://doi.org/10.1016/j.microrel.2011.04.022https://doi.org/10.1016/j.microrel.2011.04.022https://doi.org/10.1116/1.4836577https://doi.org/10.1016/j.nimb.2016.11.031https://doi.org/10.1016/j.nimb.2016.11.031http://Sr-niel.org/https://doi.org/10.1063/1.354471https://doi.org/10.1063/1.2723076https://doi.org/10.1016/j.nimb.2011.04.118https://doi.org/10.1016/j.nimb.2011.04.118https://doi.org/10.1016/j.nimb.2017.05.032https://doi.org/10.1016/j.nimb.2017.05.03245) S. Usami et al., Appl. Phys. Lett. 112, 182106 (2018).46) J. Wang, L. Cao, J. Xie, E. Beam, R. McCarthy, C. Youtsey, and P. Fay,Appl. Phys. Lett. 113, 023502 (2018).47) H. Lee, Y. Zhang, Z. Chen, M. W. Rahman, H. Zhao, and S. Rajan, IEEETrans. Electron Devices 67, 3553 (2020).48) H. Fu et al., IEEE Electron Device Lett. 41, 127 (2020).49) Z. Hu, K. Nomoto, B. Song, M. Zhu, M. Qi, M. Pen, X. Cao, V. Protasenko,D. Jena, and H. G. Xing, Appl. Phys. Lett. 107, 243501 (2015).50) T. Maeda, T. Narita, H. Ueda, M. Kanechika, T. Uesugi, T. Kachi,T. Kimoto, M. Horita, and J. Suda, Jpn. J. Appl. Phys. 58, SCCB14 (2019).51) T. Maeda, T. Narita, S. Yamada, T. Kachi, T. Kimoto, M. Horita, andJ. Suda, IEEE Electron Device Lett. 43, 96 (2021).064001-7 © 2023 The Japan Society of Applied PhysicsJpn. J. Appl. Phys. 62, 064001 (2023) H. Okumura et al.https://doi.org/10.1063/1.5035267https://doi.org/10.7567/1347-4065/ab07ad 1. Introduction 2. Prolonged exposure of GaN SBD to protons 2.1. Experimental procedure 2.2. Electrical properties of GaN SBD 2.3. Degradation transition of GaN SBD 3. Exposure of GaN PND to Xe ions 3.1. Experimental procedure 3.2. Degradation transition of GaN PND 3.3. Signal degradation of GaN PND 3.4. Signal detection using GaN PND with mesa structure 4. Conclusions Acknowledgments A6