High ion density dry etching of bulk single-crystal β-Ga2O3 was carried out as a function of source power (100–800 W), chuck power (15–400 W), and frequency (13.56 or 40 MHz) in inductively coupled plasma (ICP) systems using Cl2/Ar or BCl3/Ar discharges. The highest etch rate achieved was ∼1300 Å min−1 using 800 W ICP source power and 200 W chuck power (13.56 MHz) with either Cl2/Ar or BCl3/Ar. This is still a comfortably practical set of conditions, where resist reticulation does not occur because of the effective He backside cooling of the sample in the tool and the avoidance of overly high powers in systems capable of 2000 W of source power. The etching is ion-assisted and produces anisotropic pattern transfer. The etched surface may become oxygen-deficient under strong ion-bombardment conditions. Schottky diodes fabricated on these surfaces show increased ideality factors (increasing from 1.00 to 1.29 for high power conditions) and reduced barrier heights (1.1 on reference diodes to 0.86 eV for etched surfaces). This electrically active damage is dependent on ion energy and flux during the etching. An obvious strategy is to reduce plasma powers toward the end of an etch sequence to reduce the disruption to the Ga2O3 surface.

There is a strong need to develop high resolution pattern transfer processes for monoclinic Ga2O3 because of its emerging role in high power electronics, solar-blind ultraviolet (UV) photodetectors for fire detection and military surveillance, and also gas sensors.1–12 Ga2O3 remains transparent well into the UV, which allows its use as a transparent conducting oxide in this region of the spectrum.13–18 To be solar-blind, the deep-UV photodetectors need a cut-off wavelength of 280 nm. With its large direct band gap of ∼4.9 eV, Ga2O3 also has a very high theoretical breakdown electric field (∼8 MV/cm).19–28 This means it is of interest for high power electronics in industrial and military applications. A commonly used comparative parameter is the Baliga figure-of-merit for power electronics. This number is almost four times higher for Ga2O3 compared to GaN.2,6,9,19 Reported values of the experimental breakdown field in Ga2O3 have reached 3.8 MV/cm, which is higher than the bulk critical field strengths of the more mature GaN and SiC technologies.22 It is expected that as control of crystal growth and device processing modules improves, the experimental values will move even closer to the theoretical maximum.1,2

A variety of electronic devices have been reported for Ga2O3, including diode rectifiers, metal-semiconductor field-effect transistors, depletion-mode metal-oxide-semiconductor field-effect transistors, and finfets19–26 fabricated on either bulk or thin films. In addition, transistors fabricated on exfoliated Ga2O3 nanobelts have been shown to exhibit well-behaved dc characteristics such as saturation and transconductance,10,27,28 with associated large values of on-off ratios and stable operation in air.

High quality bulk Ga2O3 is now commercially available from several sources1,14,15 and n-type epistructures are also coming onto the market in limited quantities. There are also significant efforts worldwide to grow more complex epistructures, including heterostructures in the AlInGaOx system,18 and thus, this materials system is poised to make rapid advances in the various device technologies mentioned earlier. However, needs at the moment include the availability of high quality patterning, doping, and contacting processes that exist for the more mature semiconductors. Etching is needed for intradevice isolation or for exposing layers for subsequent contacting. While numerous wet etchants have been reported for Ga2O3, including HNO3/HCl, H2SO4, H3PO4, and hydrofluoric acid (HF)-based solutions,29–32 little is known about its dry etching characteristics and the associated mechanisms and effects on the optical properties of the material.33–35 Hogan et al.33 reported etch rates with Cl2/BCl3 under conventional reactive ion etching conditions and BCl3, BCl3/SF6, and CF4/O2 under inductively coupled plasma (ICP) conditions. The etch rates were generally low and the resultant surfaces slightly rougher in comparison to those obtained with most compound conducting oxides. Shah and Bhattacharya35 found no significant temperature dependence of etch rate in BCl3/Ar or Cl2/Ar over the range 25–200 °C. Plasma-induced damage in Ga2O3 is found to increase the conductivity of the near surface and lead to improved n-type Ohmic contact resistivities, as was the case in the early days of GaN technology.

In this paper, we report on the dry etching characteristics of high quality bulk Ga2O3 using ICP discharges of either Cl2/Ar or BCl3/Ar and measure the etch rates, near-surface stoichiometry, and electrically active damage as a function of source and chuck power and also chuck power frequency to understand the nature of the etch mechanism and residual damage. The etching is ion-enhanced and leads to a reduction of Schottky barrier height in diodes fabricated on the etched surfaces.

The starting sample was a bulk β-phase Ga2O3 single crystal with (-201) surface orientation (Tamura Corporation, Japan) grown by the edge-defined film-fed growth method.1 Hall effect measurements showed the sample was unintentionally n-type with an electron concentration of ∼3 × 1017 cm−3. Full-area back Ohmic contacts were created using Ti/Au (20 nm/80 nm) deposited by e-beam evaporation. Photoresist masked and unmasked samples were exposed to 15Cl2/5Ar or 15BCl3/5Ar discharges (where the numbers represent the respective gas flows in standard cubic centimeters per minute) in a Plasma-Therm Versaline ICP reactor. The plasma chemistries were chosen based on the initial reports in the literature and the fact that BCl3 is a Lewis acid that is effective in reducing oxides. The Cl2/Ar mixture is a common one for Ga-containing semiconductors such as GaAs and GaN.36 Since Ga2O3 is strongly bonded, ion-assisted etching through Ar ion bombardment is expected to be necessary to achieve practical etch rates We used three different configurations that involved either different rf chuck biasing frequency (13.56 or 40 MHz) or chuck diameter (6 or 10 in., which effectively varied the dc self-bias at a fixed power). The 2 MHz power applied to the 2000 W, three-turn ICP source was varied from 100 to 800 W, while the rf (13.56 or 40 MHz) chuck power was varied from 15 to 400 W (each configuration uses a 600 W source). This gave us maximum flexibility in developing high and low etch rate processes and examining the effect of each on near-surface electrical damage and stoichiometry. The source power controls the ion density while the latter controls ion energy. To give some idea of dc self-biases on the sample electrode over the rf power range investigated, this varied from −102 to −820 V over the range of conditions used for the Cl2/Ar etching. For constant 400 W ICP, it increased from −146 V at 50 W rf power to 820 V at 400 W rf power. For a constant 200 W rf power, it decreased from −710 V at 100 W ICP and 200 W rf power to −295 V at 800 W ICP and 200 W rf power. Also, for what we will term the low etch rate case, the DC bias voltage was about −102 V for 150 W ICP and 15 W rf power. The electrode setups use mechanical clamping on the periphery of the wafer, with the region between the electrode and wafer back side being enveloped with He (maintaining at least 3.5 Torr pressure) sealed behind the wafer to get good heat transfer between the wafer and electrode even at high bias process. Then the heat exchanger temperature of 25 °C can be maintained efficiently on the sample helping to prevent any resist reticulation.

Etch rates were obtained from stylus profilometry measurements. Schottky contacts were prepared on the front sides of the unmasked samples by e-beam deposited contacts Ni/Au (20 nm/80 nm) through a stencil mask. The forward and reverse current–voltage (I–V)36 characteristics were recorded at 25 °C using an Agilent 4145B parameter analyzer. The near-surface stoichiometry was obtained from Auger electron spectroscopy (AES) measurements, which was completed at EAG Laboratories, while the anisotropy of etched features were examined by scanning electron microscopy (SEM). The AES system was a Physical Electronics 670 Scanning Auger Microprobe. The electron beam conditions were 10 keV, 15 nA beam current at 30° from sample normal. For depth profiling, the ion beam conditions were 4 keV Ar+, 0.8 μA, (3 mm)2 raster. Figure 1 shows a schematic of the completed diodes.

Fig. 1.

(Color online) Schematics of Ni/Au Schottky diode on the β-Ga2O3 substrate.

Fig. 1.

(Color online) Schematics of Ni/Au Schottky diode on the β-Ga2O3 substrate.

Close modal

Figure 2(a) shows the Ga2O3 etch rate as a function of ICP source power for the three different configurations and two plasma chemistries. The etch rate increases with source power as the ion density and reactive neutral density increases and reaches a maximum value of ∼1300 Ǻ min−1 using 800 W ICP source power and 200 W chuck power (13.56 MHz) with both Cl2/Ar and BCl3/Ar. Shah and Bhattarchya35 reported the highest dry etch rate to date of 1440 Å min−1 using BCl3/Ar on bulk Ga2O3, with a selectivity of 2.7 over the SiNx mask. The use of the BCl3/Ar in our studies also increases the dc self-bias at a given set of rf and ICP powers. As ICP source power increases, the dc self-bias decreases but the ion density increases and since the etching is ion-assisted, this leads to higher rates. The dc self-biases as a function of source power are shown in Fig. 2(b). Note that the 40 MHz rf biasing on the electrode produces very low self-biases and the lowest etch rates. The high rate conditions provide a practical parameter space for mesa formation, where depths up to 1 μm are needed and also does not require such high powers that photoresist reticulation becomes an issue. The fact that BCl3-based discharges under some conditions can provide higher etch rates than Cl2-based has been discussed previously33,35 and may be related to the role of both BCl3 reacting with the oxygen in the Ga2O3 and BCl2+ ions providing sputtering, while in chlorine-based discharges, the atomic and molecular chlorine species are less effective at removing oxygen.

Fig. 2.

(Color online) (a) β-Ga2O3 etch rate as a function of ICP power at constant 200 W rf power (13.56 MHz) for the three different chamber configurations and the two plasma chemistries. (b) dc self-biases under these conditions.

Fig. 2.

(Color online) (a) β-Ga2O3 etch rate as a function of ICP power at constant 200 W rf power (13.56 MHz) for the three different chamber configurations and the two plasma chemistries. (b) dc self-biases under these conditions.

Close modal

To demonstrate the effect of rf chuck power on etch rate, Fig. 3(a) shows values as a function of chuck power in the different electrode configurations and the two plasma chemistries. We can obtain a controllable etch rate above 500 Å min−1 even at low dc self-biases, with the latter shown in Fig. 3(b). Note that the BCl3/Ar produces higher self-biases, presumably because of a lower dissociation efficiency compared to Cl2. The threshold ion energy for the initiation of etching was around 75 eV for pure Ar using the Steinbruchl model.37 Figure 4 shows the Ga2O3 etch rate in both types of plasma chemistry as a function of the substrate bias, Vb. The x-axis is plotted as the square root of the average ion energy, which is the plasma potential of ∼25 V minus the dc self-bias. The Steinbruchl37 model is a commonly accepted one for an etching process occurring by ion-enhanced sputtering in a collision-cascade process and predicts the etch rate will be proportional to E0.5-ETH0.5, where E is the ion energy and ETH is the threshold energy.37 Therefore, a plot of etch rate versus E0.5 should be a straight line with an x-intercept equal to ETH. As mentioned above, in the case of pure Ar, the value of ETH was ∼75 eV. The BCl3/Ar data for 13.56 MHz rf power would indicate negative activation energy, but this is an artifact of the complexity of the ion energy distribution in that chemistry, as reported in detail previously.38–40 The Cl2/Ar exhibits a threshold energy of ∼56 eV, lower than for the pure physical sputtering mode. The fact that Cl2/Ar exhibits an ion-assisted etch mechanism is consistent with the moderate vapor pressure for the expected group III etch product, GaCl3 [1 Torr at 48 °C, melting temperature 78 °C (Ref. 41)] and the high bond strength of Ga2O3 (Ref. 42). To form the etch product, the Ga–O bonds must first be broken by ion bombardment, consistent with the low etch rates for these chemistries. The trends are consistent with the process being an ion-assisted mechanism, which is the case for many compound semiconductors and oxides. We did not have enough data on the 40 MHz rf biasing to draw meaningful conclusions.

Fig. 3.

(Color online) (a) β-Ga2O3 etch rate as a function of rf power at constant 400 W ICP power for the three different chamber configurations and the two plasma chemistries (b) dc self-biases under these conditions.

Fig. 3.

(Color online) (a) β-Ga2O3 etch rate as a function of rf power at constant 400 W ICP power for the three different chamber configurations and the two plasma chemistries (b) dc self-biases under these conditions.

Close modal
Fig. 4.

(Color online) Etch rate of Ga2O3 in BCl3/Ar or Cl2/Ar plasmas as a function of the average ion kinetic energy (plasma potential of +25 V minus the measured dc bias voltage) for the three different chamber configurations.

Fig. 4.

(Color online) Etch rate of Ga2O3 in BCl3/Ar or Cl2/Ar plasmas as a function of the average ion kinetic energy (plasma potential of +25 V minus the measured dc bias voltage) for the three different chamber configurations.

Close modal

One characteristic of ion-assisted etch processes is their high degree of anisotropy because of absence of purely chemical etching component that would lead to undercutting of patterned features. Figure 5 shows SEM micrographs of features etched into the bulk Ga2O3 using a Cl2/Ar discharge and an SiO2 mask, which is still in place on the sample. After etching about 3 μm in the case at the top of Fig. 5, the sidewalls still have good verticality (as shown in (a)) while the surface (b) shows a small amount of roughening for an etch depth of over 2 μm. Further optimization of etch conditions will almost certainly improve the morphology further.

Fig. 5.

(Color online) SEM micrographs of bulk Ga2O3 dry etched in an ICP discharge of Cl2/Ar using an SiO2 mask, which is still in place. The image at top (a) shows the anisotropic nature of the etching, while the image at bottom (b) shows the surface morphology after an etch depth of over 2 μm..

Fig. 5.

(Color online) SEM micrographs of bulk Ga2O3 dry etched in an ICP discharge of Cl2/Ar using an SiO2 mask, which is still in place. The image at top (a) shows the anisotropic nature of the etching, while the image at bottom (b) shows the surface morphology after an etch depth of over 2 μm..

Close modal

Figure 6 shows the reverse I–V characteristics of diodes fabricated on the etched surfaces using either the 13.56 (a) or 40 MHz (b) rf chuck biasing conditions with BCl3/Ar discharges. The unetched reference diodes showed reverse breakdown voltages of order 50 V and this was decreased by the presence of ion-induced damage and nonstoichiometry in the diodes on the dry etched surfaces. The extent of the damage depended on the self-bias, as expected, since the trap states are likely to be point defect complexes. Zhang et al.43 found five different deep trap states in as-grown Ga2O3, ranging in depth from EC–0.62 eV to EV+0.42 eV.

Fig. 6.

(Color online) Reverse I-V characteristics of diodes fabricated on the etched surfaces using either the 13.56 (top) or 40 MHz (bottom) rf chuck biasing conditions with BCl3/Ar discharges.

Fig. 6.

(Color online) Reverse I-V characteristics of diodes fabricated on the etched surfaces using either the 13.56 (top) or 40 MHz (bottom) rf chuck biasing conditions with BCl3/Ar discharges.

Close modal

The Schottky barrier heights and ideality factors were extracted from the forward I–V characteristics of the type shown in Fig. 7 for diodes etched in BCl3 discharges with 40 MHz rf chuck biasing. We obtained this type of data for all the etch configurations and fit the forward I–V characteristics to the relation for the thermionic emission over a barrier37 

J F = A * T 2 exp ( e ϕ b k T ) exp ( e V n k T ) ,
(1)

where J is the current density, A* is the Richardson's constant for n-GaN, T the absolute temperature, e the electronic charge, ϕb the barrier height, k Boltzmann's constant, n the ideality factor, and V the applied voltage. We did this for the high and low etch rate conditions for each of the three reactor configurations, and the results are shown in Fig. 8. The Schottky diodes fabricated on these surfaces show increased ideality factors (increasing from 1.00 to a worst-case of 1.29 for high power conditions) and reduced barrier heights (1.1 on reference diodes to a worst-case of 0.86 eV for etched surfaces). Note that the amount of degradation in the barrier height from the reference values is a function of etch time under low etch rate conditions, indicating that longer times allow the damage to accumulate under these conditions and also that the frequency of the rf chuck bias plays a strong role through its influence on chuck dc self-bias and hence incident ion energy. Table I also shows a summary of the etch rates, barrier heights, and ideality factors for Ga2O3 etched in the three different configurations of the etch tools.

Fig. 7.

(Color online) Forward I-V characteristics for diodes etched in BCl3/Ar discharges with 40 MHz rf chuck biasing.

Fig. 7.

(Color online) Forward I-V characteristics for diodes etched in BCl3/Ar discharges with 40 MHz rf chuck biasing.

Close modal
Fig. 8.

(Color online) Schottky barrier height and ideality factor for diodes etched at high and low etch rate conditions shown in Table I.

Fig. 8.

(Color online) Schottky barrier height and ideality factor for diodes etched at high and low etch rate conditions shown in Table I.

Close modal
Table I.

Summary of etch rates, barrier heights, and ideality factors as a function of ICP power and also rf chuck power and frequency. The etch rate data for high rate, low rate A, and low rate B was obtained for 5, 10, and 25 min, respectively.

Sample/plasma chemistry Rf power frequency (MHz) Rf power (W) ICP power (W) DC self-bias (−V) Etch rate (Å/min) Schottky barrier height (eV) Ideality factor
Reference    1.20  1.00 
High etch rate-BCl3/Ar  13.56 MHz  200  400  480  692  0.86  1.20 
Low etch rate A-BCl3/Ar  13.56 MHz  15  150  102  121  1.01  1.06 
Low etch rate B-BCl3/Ar  13.56 MHz  15  150  102  121  1.02  1.08 
High etch rate-BCl3/Ar  40 MHz  200  400  19  302  0.97  1.20 
Low etch rate A-BCl3/Ar  40 MHz  15  250  27  1.10  1.17 
Low etch rate B-BCl3/Ar  40 MHz  15  250  27  1.04  1.29 
High etch rate-Cl2/Ar  13.56 MHz  400 (small electrode)  400  547  173  0.89  1.69 
Low etch rate A-Cl2/Ar  13.56 MHz  15 (small electrode)  100  76  15  1.02  1.11 
Low etch rate B-Cl2/Ar  13.56 MHz  15 (small electrode)  100  76  15  1.01  1.15 
Sample/plasma chemistry Rf power frequency (MHz) Rf power (W) ICP power (W) DC self-bias (−V) Etch rate (Å/min) Schottky barrier height (eV) Ideality factor
Reference    1.20  1.00 
High etch rate-BCl3/Ar  13.56 MHz  200  400  480  692  0.86  1.20 
Low etch rate A-BCl3/Ar  13.56 MHz  15  150  102  121  1.01  1.06 
Low etch rate B-BCl3/Ar  13.56 MHz  15  150  102  121  1.02  1.08 
High etch rate-BCl3/Ar  40 MHz  200  400  19  302  0.97  1.20 
Low etch rate A-BCl3/Ar  40 MHz  15  250  27  1.10  1.17 
Low etch rate B-BCl3/Ar  40 MHz  15  250  27  1.04  1.29 
High etch rate-Cl2/Ar  13.56 MHz  400 (small electrode)  400  547  173  0.89  1.69 
Low etch rate A-Cl2/Ar  13.56 MHz  15 (small electrode)  100  76  15  1.02  1.11 
Low etch rate B-Cl2/Ar  13.56 MHz  15 (small electrode)  100  76  15  1.01  1.15 

Note that the usual procedures for eliminating the near-surface damage region of dry etched semiconductors include either annealing or a wet etch clean-up.44–47 The wet etching under these conditions in wide bandgap materials can often be self-limiting because only the damaged region etches at a significant rate. Such annealing and shallow wet etch processes still need to be developed for Ga2O3.

The near-surface composition of the samples was measured by AES depth profiling. Figure 9 shows the data for the low (a) and high (b) etch rate conditions of rows 2 and 3 from Table I. Within the experimental error, there was no significant change in O/Ga ratio in this region relative to the unetched control samples (Fig. 10), indicating that the disruption to the surface is not sufficient to be detected by chemical analysis techniques but is present to a level that it changes the electrical properties (Schottky barrier height and ideality factor). In our experience, this indicative of a nearly optimized process, i.e., the damage is at a small enough level that it should be relatively simple to use the annealing or wet-etch clean-up steps to remove this affected region. When nitride or oxide semiconductors are dry etched under nonoptimized conditions, it is common to note a preferential loss of the nitrogen or oxygen from the surface, leading in general to highly conducting layers.48 

Fig. 9.

(Color online) AES depth profiles of Ga2O3 surfaces before and after either low rate (a) or high rate (b) etching in BCl3/Ar discharges. The etch conditions correspond to rows 2 and 3 of Table I.

Fig. 9.

(Color online) AES depth profiles of Ga2O3 surfaces before and after either low rate (a) or high rate (b) etching in BCl3/Ar discharges. The etch conditions correspond to rows 2 and 3 of Table I.

Close modal
Fig. 10.

(Color online) O/Ga ratio obtained from AES depth profiles of Ga2O3 near surface region after etching in BCl3/Ar discharges. The etch conditions correspond to rows 2 and 3 of Table I.

Fig. 10.

(Color online) O/Ga ratio obtained from AES depth profiles of Ga2O3 near surface region after etching in BCl3/Ar discharges. The etch conditions correspond to rows 2 and 3 of Table I.

Close modal

In conclusion, we have investigated the etch mechanism for Ga2O3 in plasma chemistries of BCl3/Ar and Cl2/Ar. For both chemistries, the etch rate increases with ion energy as predicted from an ion-assisted chemical sputtering process. Both chemistries give similar maximum rates under ICP conditions. Both high rate and low rate processes have been developed for applications that need either higher rates for deep mesa formation or finer control for adjustment of threshold voltage in gate-recessed transistor structures. Anisotropic features can be created to depths of several microns with these processes and simple masks. The near-surface stoichiometry changes as a result of dry etching are enough to reduce the barrier height and increase the ideality factors on Schottky diodes fabricated on these surfaces, but are not significant enough to be detected by AES.

The project or effort depicted was also sponsored by the Department of the Defense, Defense Threat Reduction Agency, HDTRA1-17-1-011, monitored by Jacob Calkins. The content of the information does not necessarily reflect the position or the policy of the federal government, and no official endorsement should be inferred. Part of the work at Tamura was supported by “The research and development project for innovation technique of energy conservation” of the New Energy and Industrial Technology Development Organization (NEDO), Japan. The authors also thank Kohei Sasaki from Tamura Corporation for fruitful discussions.

1.
A.
Kuramata
,
K.
Koshi
,
S.
Watanabe
,
Y.
Yamaoka
,
T.
Masui
, and
S.
Yamakoshi
,
Jpn. J. Appl. Phys., Part 1
55
,
1202A2
(
2016
).
2.
M. J.
Tadjer
 et al,
J. Electron. Mater.
45
,
2031
(
2016
).
3.
M.
Higashiwaki
 et al,
Appl. Phys. Lett.
108
,
133503
(
2016
).
4.
K.
Sasaki
,
M.
Higashiwaki
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
Appl. Phys. Express
6
,
086502
(
2013
).
5.
Z.
Guo
 et al,
Appl. Phys. Lett.
106
,
111909
(
2015
).
6.
M.
Higashiwaki
,
K.
Sasaki
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
Phys. Status Solidi B
211
,
21
(
2014
).
7.
O.
Ueda
,
N.
Ikenaga
,
K.
Koshi
,
K.
Iizuka
,
A.
Kuramata
,
K.
Hanada
,
T.
Moribayashi
,
S.
Yamakoshi
, and
M.
Kasu
,
Jpn. J. Appl. Phys., Part 1
55
,
1202BD
(
2016
).
8.
S. I.
Stepanov
,
V. I.
Nikolaev
,
V. E.
Bougrov
, and
A. E.
Romanov
,
Rev. Adv. Mater. Sci.
44
,
63
(
2016
); available at http://www.ipme.ru/e-journals/RAMS/no_14416/06_14416_stepanov.html.
9.
M. J.
Tadjer
,
N. A.
Mahadik
,
V. D.
Wheeler
,
E. R.
Glaser
,
L.
Ruppalt
,
A. D.
Koehler
,
K. D.
Hobart
,
C. R.
Eddy
, Jr.
, and
F. J.
Kub
,
ECS J. Solid State Sci. Technol.
5
,
468
(
2016
).
10.
J.
Kim
,
S.
Oh
,
M.
Mastro
, and
J.
Kim
,
Phys. Chem. Chem. Phys.
18
,
15760
(
2016
).
11.
R.
Pandeeswari
and
B. G.
Jeyaprakash
,
Sens. Actuators, B
195
,
206
(
2014
).
12.
A. M.
Armstrong
,
M. H.
Crawford
,
A.
Jayawardena
,
A.
Ahyi
, and
S.
Dhar
,
J. Appl. Phys.
119
,
103102
(
2016
).
13.
M. H.
Wong
,
K.
Sasaki
,
A.
Kuramata
,
S.
Yamakoshi
, and
M.
Higashiwaki
,
Appl. Phys. Lett.
106
,
032105
(
2015
).
14.
K.
Hoshikawa
,
E.
Ohba
,
T.
Kobayashi
,
J.
Yanagisawa
,
C.
Miyagawa
, and
Y.
Nakamura
,
J. Cryst. Growth
447
,
36
(
2016
).
15.
Z.
Galazka
 et al,
ECS J. Solid State Sci. Technol.
6
,
Q3007
(
2017
).
16.
M.
Baldini
,
M.
Albrecht
,
A.
Fiedler
,
K.
Irmscher
,
R.
Schewski
, and
G.
Wagner
,
ECS J. Solid State Sci. Technol.
6
,
Q3040
(
2017
).
17.
S.
Rafique
,
L.
Han
,
M. J.
Tadjer
,
J. A.
Freitas
, Jr.
,
N. A.
Mahadik
, and
H.
Zhao
,
Appl. Phys. Lett.
108
,
182105
(
2016
).
18.
S. W.
Kaun
,
F.
Wu
, and
J. S.
Speck
,
J. Vac. Sci. Technol., A
33
,
041508
(
2015
).
19.
M.
Higashiwaki
,
K.
Sasaki
,
H.
Murakami
,
Y.
Kumagai
,
A.
Koukitu
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakosh
,
Semicond. Sci. Technol.
31
,
034001
(
2016
).
20.
M.
Higashiwaki
,
K.
Sasaki
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
Appl. Phys. Lett.
100
,
013504
(
2012
).
21.
K. D.
Chabak
 et al,
Appl. Phys. Lett.
109
,
213501
(
2016
).
22.
A. J.
Green
 et al,
IEEE Electron Device Lett.
37
,
902
(
2016
).
23.
M. H.
Wong
,
K.
Sasaki
,
A.
Kuramata
,
S.
Yamakoshi
, and
M.
Higashiwaki
,
IEEE Electron Device Lett.
37
,
212
(
2016
).
24.
K.
Sasaki
,
M.
Higashiwaki
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
IEEE Electron Device Lett.
34
,
493
(
2013
).
25.
J.
Yang
,
S.
Ahn
,
F.
Ren
,
R.
Khanna
,
K.
Bevlin
,
D.
Geerpuram
,
S. J.
Pearton
, and
A.
Kuramata
,
Appl. Phys. Lett.
110
,
142101
(
2017
).
26.
S.
Oh
,
G.
Yang
, and
J.
Kim
,
ECS J. Solid State Sci. Technol.
6
,
Q3022
(
2017
).
27.
W. S.
Hwang
 et al,
Appl. Phys. Lett.
104
,
203111
(
2014
).
28.
S.
Ahn
,
F.
Ren
,
J.
Kim
,
S.
Oh
,
J.
Kim
,
M. A.
Mastro
, and
S. J.
Pearton
,
Appl. Phys. Lett.
109
,
062102
(
2016
).
29.
S.
Ohira
and
N.
Arai
,
Phys. Status Solidi C
5
,
3116
(
2008
).
30.
T.
Oshima
,
T.
Okuno
,
N.
Arai
,
Y.
Kobayashi
, and
S.
Fujita
,
Jpn. J. Appl. Phys., Part 1
48
,
040208
(
2009
).
31.
M.
Passlack
 et al,
J. Appl. Phys.
77
,
686
(
1995
).
32.
F.
Ren
,
M.
Hong
,
J. P.
Mannaerts
,
J. R.
Lothian
, and
A. Y.
Cho
,
J. Electrochem. Soc.
144
,
L239
(
1997
).
33.
J. E.
Hogan
,
S. W.
Kaun
,
E.
Ahmadi
,
Y.
Oshima
, and
J. S.
Speck
,
Semicond. Sci. Technol.
31
,
065006
(
2016
).
34.
H.
Liang
,
Y.
Chen
,
X.
Xia
,
C.
Zhang
,
R.
Shen
,
Y.
Liu
,
Y.
Luo
, and
G.
Du
,
Mater. Sci. Semicond. Proc.
39
,
582
(
2015
).
35.
A. P.
Shah
and
A.
Bhattacharya
, “ICP-RIE etching of β-Ga2O3: Comprehensive investigation of plasma chemistry and temperature,”
J. Vac. Sci. Technol., A
(submitted).
36.
R. J.
Shul
,
M. C.
Lovejoy
,
A. G.
Baca
,
J. C.
Zolper
,
D. J.
Rieger
,
M. J.
Hafich
,
R. F.
Corless
, and
C. B.
Vartuli
,
J. Vac. Sci. Technol., A
13
,
912
(
1995
).
37.
D. K.
Schroder
,
Semiconductor Material and Device Characterization
(
Wiley
,
New York
,
1990
).
38.
C.
Steinbruchel
,
Appl. Phys. Lett.
55
,
1960
(
1989
).
39.
K.
Pelhos
,
V. M.
Donnelly
,
A.
Kornblit
,
M. L.
Green
,
R. B.
Van Dover
,
L.
Manchanda
, and
E.
Bower
,
J. Vac. Sci. Technol., A
19
,
1361
(
2001
).
40.
C. C.
Chang
,
K. V.
Guinn
,
V. M.
Donnelly
, and
I. P.
Herman
,
J. Vac. Sci. Technol., A
13
,
1970
(
1995
).
41.
K.
Ip
,
K.
Baik
,
M. E.
Overberg
,
E. S.
Lambers
,
Y. W.
Heo
,
D. P.
Norton
,
S. J.
Pearton
,
F.
Ren
, and
J. M.
Zavada
,
Appl. Phys. Lett.
81
,
3546
(
2002
).
42.
C.
Bernard
,
C.
Chatillon
,
A.
Ait-Hou
,
R.
Hillel
,
Y.
Monteil
, and
J.
Bouix
,
J. Chem. Thermodyn.
20
,
129
(
1988
).
43.
Z.
Zhang
,
E.
Farzana
,
A. R.
Arehart
, and
S. A.
Ringel
,
Appl. Phys. Lett.
108
,
052105
(
2016
).
44.
A. G.
Baca
,
A. M.
Armstrong
,
A. A.
Allerman
,
E. A.
Douglas
,
C. A.
Sanchez
,
M. P.
King
,
M. E.
Coltrin
,
T. R.
Fortune
, and
R. J.
Kaplar
,
Appl. Phys. Lett.
109
,
033509
(
2016
).
45.
X. A.
Cao
,
S. J.
Pearton
,
A. P.
Zhang
,
G. T.
Dang
,
F.
Ren
,
R. J.
Shul
,
L.
Zhang
,
R.
Hickman
, and
J. M.
Van Hove
,
Appl. Phys. Lett.
75
,
2569
(
1999
).
46.
E. A.
Douglas
,
C. A.
Sanchez
,
R. J.
Kaplar
,
A. A.
Allerman
, and
A. G.
Baca
,
J. Vac. Sci. Technol., A
35
,
021305
(
2017
).
47.
X. A.
Cao
 et al,
Appl. Phys. Lett.
75
,
232
(
1999
).
48.
See, for example,
S. W.
Pang
,
Handbook of Advanced Plasma Processing Techniques
, edited by
R. J.
Shul
and
S. J.
Pearton
(
Springer
,
Berlin
,
2000
), Chap. 8.