The energy band alignment between atomic layer deposited (ALD) SiO2 and β-Ga2O3 (2¯01) is calculated using x-ray photoelectron spectroscopy and electrical measurement of metal-oxide semiconductor capacitor structures. The valence band offset between SiO2 and Ga2O3 is found to be 0.43 eV. The bandgap of ALD SiO2 was determined to be 8.6 eV, which gives a large conduction band offset of 3.63 eV between SiO2 and Ga2O3. The large conduction band offset makes SiO2 an attractive gate dielectric for power devices.

Recently, wide bandgap β-Ga2O3 (Ga2O3) has received much attention as an attractive semiconductor for power electronics and UV detector applications due to its large bandgap of 4.6–4.9 eV.1–3 In addition, high bulk electron mobilities in Ga2O3 lead to a Baliga Figure of Merit (BFoM), which exceeds that of SiC and GaN,4,5 which makes it as an attractive choice for next generation of power semiconductor devices. Moreover, large area bulk crystals of Ga2O3 can be grown using scalable crystal growth technologies.6–10 Both doped and semi-insulating bulk crystals are available commercially.4,5,11,12 Epitaxial growth of Ga2O3 by molecular beam epitaxy (MBE)4,13,14 and ion implantation doping technology15 has also been reported. All these factors make Ga2O3 a strong candidate for next generation power electronics. Ga2O3 metal-oxide semiconductor field effect devices (MOSFETs) with high breakdown voltages, large ON/OFF ratios, and high temperature operation have been recently demonstrated.5,15–17 These devices use atomic layer deposited (ALD) Al2O3 as a gate barrier. The conduction band offset between Al2O3 and Ga2O3 has been determined to be 1.5–1.7 eV.18,19 A higher conduction band offset is preferred to reduce thermal leakages during high temperature operation of power devices. However, the large bandgap of Ga2O3 limits the choice of gate dielectrics. In addition to be used as gate barrier, dielectrics are also used for passivation and electric field profiling by field plates in power semiconductor devices.

Silicon dioxide (SiO2) is an attractive gate barrier material for Ga2O3 due to its large bandgap of ∼9 eV.20 However, there is no report of the band parameters between SiO2 and Ga2O3. In this letter, we report the band alignment between ALD SiO2 and β-Ga2O3 (2¯01). Silicon dioxide deposited by ALD has great potential to serve as a gate dielectric in Ga2O3 based power device because of the expected large conduction band offset at the interface of SiO2/Ga2O3 and also because of its large critical breakdown electric field (∼10 MV/cm (Ref. 21)). In this work, the conduction band offset of ALD-SiO2/Ga2O3 hetero-junction was characterized using X-ray photoelectron spectroscopy (XPS) as well as the tunneling current through metal-oxide-semiconductor capacitors (MOSCAPs).

β-Ga2O3 (2¯01) crystals studied here was grown at Tamura Corporation with an n-type doping (Sn doped) density of ∼9×1018/cm3. A surface root mean square (rms) roughness of 0.13 nm was measured by atomic force microscopy (AFM) on these wafers. For XPS characterization, a thin layer (∼3 nm) of SiO2 was deposited on Ga2O3 by plasma-enhanced ALD in an Oxford FLEXAL system at 300 °C with trisdimethylaminosilane (3DMAS) and O2 plasma at 250 W. Standard solvent degreasing procedure was used to clean the wafers before SiO2 deposition. For band offset calculation by XPS, core level spectra of Si in bulk SiO2 are necessary. We use 40 nm thick SiO2 on Si as the bulk standard. For electrical studies, a 40 nm layer of SiO2 was deposited on Ga2O3. The growth rate was calibrated on silicon wafers to be 0.71 Å/cycle. For electrical characterization, the MOSCAP structure, shown in Figure 1, was fabricated. First, the top Ti/Au electrodes were defined by standard photolithography and lift-off technique. Next, the silicon oxide and Ga2O3 were etched by CF4/Ar based reactive ion etching. And finally, the bottom Ti/Au electrodes are defined. The sample was then annealed at 300 °C for 1 h to reduce the contact resistance.17 

FIG. 1.

Cross-section schematic of Ga2O3 MOSCAP device.

FIG. 1.

Cross-section schematic of Ga2O3 MOSCAP device.

Close modal

XPS measurements were performed using a Physical Electronic PHI VersaProbe 5000 equipped with a hemispherical energy analyzer. A monochromic Al Kα X-ray source (1486.6 eV) was operated at 25.3 W and 15 kV. The energy of the analyzer was operated at a pass energy of 117.5 eV for survey acquisitions and 23.50 or 11.75 eV for high-resolution acquisitions. The energy resolution was 0.025 eV for high resolution spectra or 1.0 eV for survey spectra. The operating pressure of XPS was <4×106 Pa (3.0×108 Torr) and the background pressure was <1×106 Pa (7.5×109 Torr). Dual charge neutralization was utilized to reduce the effects of charging on the acquired signal. Binding energies were calibrated by setting the CHx peak in the C 1s envelope at 284.8 eV to correct for charging effects.18,22 However, the valence band offset (VBO) measurement is not sensitive to charging effects.

Figures 2(a)–2(c) show the XPS spectra for 40 nm SiO2/Si, β-Ga2O3, and SiO2 (∼3 nm)/β-Ga2O3 structure. XPS results were curve fitted with a Gauss-Lorentzian band type with Shirley background subtraction23–25 with curve fitting limits as follows: binding energy ±0.4 eV, FWHM ±0.2 eV, and % Gauss = 92%. The core level (Ga 2p3/2) spectra on bare Ga2O3 shows a single peak (1118.4 eV) corresponding to Ga-O bond.18 The valence band maxima (VBM) were found by the linear extrapolation of the valence band states,18,24,25 the VBM of Ga2O3 was found to be 3.69 eV above the Fermi level (EF), as shown in Figure 2(b). For the SiO2 (3 nm)/Ga2O3 sample, in addition to the Si-O bonds, the XPS spectrum shows the Ga 3p peaks from underneath the ALD-SiO2 layer. Next, the valence band offset was calculated by the following equation:18,26

where the subscripts indicate the XPS peak and the superscripts indicate the sample. From the measured XPS profiles, (ESi2pSiO2/Ga2O3EGa2p3/2SiO2/Ga2O3), (EGa2p3/2Ga2O3EVBMGa2O3), and (ESi2pSiO2EVBMSiO2) are −1015.51, 1114.72, and 98.78 eV, respectively, which gives a valence band offset (ΔEV) of 0.43 eV. Figures 3(a) and 3(b) show core level and the loss structure of O 1s on SiO2/Si and bare Ga2O3 samples. From the loss peak, the band gap of Ga2O3 and SiO2 was found to be 4.54 eV and 8.6 eV,18,27 respectively. The conduction band offset is calculated by

where ΔEg is the band gap difference between Ga2O3 and SiO2 and ΔEV is the calculated valence band offset. Taking the band gap difference to be 4.06 eV, a conduction band offset of 3.63 eV is calculated.

FIG. 2.

XPS spectra used to calculate valence band offset. (a) Si 2p peak and valence band maximum acquired from 40 nm SiO2/Si. (b) Ga 2p3/2 peak and valence band maximum acquired from bare Ga2O3. (c) Ga 2p3/2 peak and Si 2p peak obtained from SiO2 (3 nm)/Ga2O3 heterostructure. Ga 3p3/2 and 3p1/2 peaks were also observed through SiO2 layer as shown in (c). The VBO was calculated as 0.43 eV.

FIG. 2.

XPS spectra used to calculate valence band offset. (a) Si 2p peak and valence band maximum acquired from 40 nm SiO2/Si. (b) Ga 2p3/2 peak and valence band maximum acquired from bare Ga2O3. (c) Ga 2p3/2 peak and Si 2p peak obtained from SiO2 (3 nm)/Ga2O3 heterostructure. Ga 3p3/2 and 3p1/2 peaks were also observed through SiO2 layer as shown in (c). The VBO was calculated as 0.43 eV.

Close modal
FIG. 3.

O 1s peaks obtained from (a) 40 nm SiO2/Si and (b) bare Ga2O3 to determine bandgap of SiO2 and Ga2O3. Inset of (a) and (b) shows the corresponding loss structure. The bandgap for SiO2 and Ga2O3 is 8.6 eV and 4.54 eV, respectively.

FIG. 3.

O 1s peaks obtained from (a) 40 nm SiO2/Si and (b) bare Ga2O3 to determine bandgap of SiO2 and Ga2O3. Inset of (a) and (b) shows the corresponding loss structure. The bandgap for SiO2 and Ga2O3 is 8.6 eV and 4.54 eV, respectively.

Close modal

In addition to XPS measurement, electrical characterization of the MOSCAPs was also carried out to calculate the conduction band offset between SiO2 and Ga2O3. C-V characteristic of MOS diode is shown in Figure 4. Both first derivative of C-V and flatband capacitance method indicate a flatband voltage about 9.7 V, suggesting the existence of negative surface charge between SiO2 and n-Ga2O3. An electron density of 9.7 × 1018 cm−3 for n-Ga2O3 was estimated using differential capacitance-voltage profile technique,28 which is given by

where n(W) is the carrier density, W is the depth from surface of metal and oxide, Ks is relative permittivity of the channel, and A is the area of contact. The current-voltage characteristics of the MOSCAP are shown in Figure 5. The current in the reverse bias is negligibly small (not shown), while in the forward direction current remains low till 50 V then rises rapidly due to Fowler-Nordheim (F-N) tunneling. Destructive breakdown of the MOSCAPs was observed at ∼60 V both in the forward and the reverse bias conditions. We extract the conduction band offset from the F-N tunneling current in forward bias.18 When sufficient forward bias is applied F-N tunneling takes place29 as indicated in the inset of Figure 5. The F-N tunneling current which depends on ΔEc is given by:18,29

where J is the current density, q is the electron charge, h is the Plank's constant, m0 is the free electron mass, and mox is the electron effective mass in oxide. Eox is the electric field strength in the oxide, which can be calculated easily if the diode is in the strong accumulation region at large forward bias voltages. ΔEC was extracted from the slope (S) of ln(J/Eox2) vs. (1/Eox), as shown in Figure 6. The slope of this curve in the F-N tunneling regime is measured which is given by18 

Using the calculated slope (S) and assuming an electron effective mass of SiO2 is 0.4m0,29,30 the ΔEC was calculated to be 3.76 eV, which is close to the result obtained from XPS. Taking bandgap of SiO2 and Ga2O3, band offset obtained from XPS, doping density as 9×1018/cm3, barrier height between Ti and SiO2 as 3.34 eV, a calculated band diagram at zero bias is shown in Figure 7.

FIG. 4.

C-V profile of Ti/SiO2/β-Ga2O3 diode. Inset shows carrier density profile derived from C-V profile. The extracted doping density is 9.7×1018/cm3.

FIG. 4.

C-V profile of Ti/SiO2/β-Ga2O3 diode. Inset shows carrier density profile derived from C-V profile. The extracted doping density is 9.7×1018/cm3.

Close modal
FIG. 5.

I-V characteristic of MOSCAP at forward bias. The F-N tunneling region is indicated. Inset shows a schematic band-diagram, which enables F-N tunneling.

FIG. 5.

I-V characteristic of MOSCAP at forward bias. The F-N tunneling region is indicated. Inset shows a schematic band-diagram, which enables F-N tunneling.

Close modal
FIG. 6.

ln(J/Eox2) vs. 1/Eox plot derived from forward I-V plot in Fig. 5. In the F-N tunneling region, as indicated by the line, the measured slope is 3.1488×1010.

FIG. 6.

ln(J/Eox2) vs. 1/Eox plot derived from forward I-V plot in Fig. 5. In the F-N tunneling region, as indicated by the line, the measured slope is 3.1488×1010.

Close modal
FIG. 7.

Calculated band diagram of MOSCAP device at zero bias. The bandgap and conduction band offset were extracted by XPS. Both I-V characteristics and XPS show similar conduction band offset, ∼3.7 eV.

FIG. 7.

Calculated band diagram of MOSCAP device at zero bias. The bandgap and conduction band offset were extracted by XPS. Both I-V characteristics and XPS show similar conduction band offset, ∼3.7 eV.

Close modal

In summary, we evaluated the band alignment between ALD SiO2 and n-doped β-Ga2O3 (2¯01) by XPS and electrical measurements. The conduction band offset is determined to be 3.63 and 3.76 eV by XPS and electrical measurements, respectively. The large conduction band offset is useful for power devices, especially for high temperature operation. However, the dielectric constant of SiO2 is lower than Al2O3, which reduces the equivalent oxide thickness (EOT). A composite gate dielectric stack with thin interfacial SiO2 layer and thicker Al2O3 layer can be used to obtain both large conduction band offset and lower EOT.

This work was partly supported by an ONR Grant (No. N000141310214) monitored by Dr. Paul A. Maki. A portion of this work was performed in the University at Buffalo Electrical Engineering Cleanroom and the Materials Characterization Laboratory, part of the university's Shared Instrumentation Laboratories. The authors would like to thank Dr. Brian Thibeault for the deposition of the SiO2 films.

1.
H.
Tippins
,
Phys. Rev.
140
(
1A
),
A316
A319
(
1965
).
2.
M.
Orita
,
H.
Ohta
,
M.
Hirano
, and
H.
Hosono
,
Appl. Phys. Lett.
77
(
25
),
4166
(
2000
).
3.
H.
He
,
R.
Orlando
,
M.
Blanco
,
R.
Pandey
,
E.
Amzallag
,
I.
Baraille
, and
M.
Rérat
,
Phys. Rev. B
74
(
19
),
195123
(
2006
).
4.
K.
Sasaki
,
A.
Kuramata
,
T.
Masui
,
E. G.
Víllora
,
K.
Shimamura
, and
S.
Yamakoshi
,
Appl. Phys. Express
5
(
3
),
035502
(
2012
).
5.
M.
Higashiwaki
,
K.
Sasaki
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
Appl. Phys. Lett.
100
(
1
),
013504
(
2012
).
6.
E. G.
Víllora
,
K.
Shimamura
,
Y.
Yoshikawa
,
K.
Aoki
, and
N.
Ichinose
,
J. Cryst. Growth
270
(
3–4
),
420
426
(
2004
).
7.
Z.
Galazka
,
R.
Uecker
,
K.
Irmscher
,
M.
Albrecht
,
D.
Klimm
,
M.
Pietsch
,
M.
Brützam
,
R.
Bertram
,
S.
Ganschow
, and
R.
Fornari
,
Cryst. Res. Technol.
45
(
12
),
1229
1236
(
2010
).
8.
Y.
Tomm
,
P.
Reiche
,
D.
Klimm
, and
T.
Fukuda
,
J. Cryst. Growth
220
,
510
(
2000
).
9.
N.
Ueda
,
H.
Hosono
,
R.
Waseda
, and
H.
Kawazoe
,
Appl. Phys. Lett.
70
(
26
),
3561
(
1997
).
10.
J.
Zhang
,
B.
Li
,
C.
Xia
,
G.
Pei
,
Q.
Deng
,
Z.
Yang
,
W.
Xu
,
H.
Shi
,
F.
Wu
,
Y.
Wu
, and
J.
Xu
,
J. Phys. Chem. Solids
67
(
12
),
2448
2451
(
2006
).
11.
E. G.
Víllora
,
K.
Shimamura
,
Y.
Yoshikawa
,
T.
Ujiie
, and
K.
Aoki
,
Appl. Phys. Lett.
92
(
20
),
202120
(
2008
).
12.
S.
Ohira
,
N.
Suzuki
,
N.
Arai
,
M.
Tanaka
,
T.
Sugawara
,
K.
Nakajima
, and
T.
Shishido
,
Thin Solid Films
516
(
17
),
5763
5767
(
2008
).
13.
T.
Oshima
,
T.
Okuno
, and
S.
Fujita
,
Jpn. J. Appl. Phys., Part 1
46
(
11
),
7217
7220
(
2007
).
14.
E. G.
Víllora
,
K.
Shimamura
,
K.
Kitamura
, and
K.
Aoki
,
Appl. Phys. Lett.
88
(
3
),
031105
(
2006
).
15.
M.
Higashiwaki
,
K.
Sasaki
,
T.
Kamimura
,
M.
Hoi Wong
,
D.
Krishnamurthy
,
A.
Kuramata
,
T.
Masui
, and
S.
Yamakoshi
,
Appl. Phys. Lett.
103
(
12
),
123511
(
2013
).
16.
M.
Higashiwaki
,
K.
Sasaki
,
M. H.
Wong
,
T.
Kamimura
,
D.
Krishenamurthy
,
A.
Kuramata
,
T.
Masui
, and
S.
Tamakoshi
,
Int. Electron Devices Meet.
2013
,
28.7.1
.
17.
W. S.
Hwang
,
A.
Verma
,
H.
Peelaers
,
V.
Protasenko
,
S.
Rouvimov
,
H.
(Grace) Xing
,
A.
Seabaugh
,
W.
Haensch
,
C.
Van de Walle
,
Z.
Galazka
,
M.
Albrecht
,
R.
Fornari
, and
D.
Jena
,
Appl. Phys. Lett.
104
,
203111
(
2014
).
18.
T.
Kamimura
,
K.
Sasaki
,
M.
Hoi Wong
,
D.
Krishnamurthy
,
A.
Kuramata
,
T.
Masui
,
S.
Yamakoshi
, and
M.
Higashiwaki
,
Appl. Phys. Lett.
104
(
19
),
192104
(
2014
).
19.
T. H.
Hung
,
K.
Sasaki
,
A.
Kuramata
,
D. N.
Nath
,
P.
Sung Park
,
C.
Polchinski
, and
S.
Rajan
,
Appl. Phys. Lett.
104
(
16
),
162106
(
2014
).
20.
J.
Robertson
,
J. Vac. Sci. Technol., B
18
(
3
),
1785
(
2000
).
21.
S.-J.
Won
,
S.
Suh
,
M.
Soo Huh
, and
H.
Joon Kim
,
IEEE Electron Device Lett.
31
(
8
),
857
(
2010
).
22.
C. F.
Shih
,
N. C.
Chen
,
P. H.
Chang
, and
K. S.
Liu
,
Jpn. J. Appl. Phys., Part 1
44
(
11
),
7892
7895
(
2005
).
23.
A.
Eisenhardt
,
G.
Eichapfel
,
M.
Himmerlich
,
A.
Knübel
,
T.
Passow
,
C.
Wang
,
F.
Benkhelifa
,
R.
Aidam
, and
S.
Krischok
,
Phys. Status Solidi C
9
(
3–4
),
685
688
(
2012
).
24.
P.
King
,
T.
Veal
,
C.
Kendrick
,
L.
Bailey
,
S.
Durbin
, and
C.
McConville
,
Phys. Rev. B
78
(
3
),
033308
(
2008
).
25.
P. D. C.
King
,
T. D.
Veal
,
P. H.
Jefferson
,
C. F.
McConville
,
T.
Wang
,
P. J.
Parbrook
,
H.
Lu
, and
W. J.
Schaff
,
Appl. Phys. Lett.
90
(
13
),
132105
(
2007
).
26.
E.
Kraut
,
R.
Grant
,
J.
Waldrop
, and
S.
Kowalczyk
,
Phys. Rev. Lett.
44
(
24
),
1620
1623
(
1980
).
27.
M. L.
Huang
,
Y. C.
Chang
,
C. H.
Chang
,
T. D.
Lin
,
J.
Kwo
,
T. B.
Wu
, and
M.
Hong
,
Appl. Phys. Lett.
89
(
1
),
012903
(
2006
).
28.
D. K.
Schroder
,
Semiconductor Material and Device Characterization
(
Wiley-Interscience
,
2006
).
29.
M.
Lenzlinger
,
J. Appl. Phys.
40
(
1
),
278
(
1969
).
30.
B.
Brar
,
G. D.
Wilk
, and
A. C.
Seabaugh
,
Appl. Phys. Lett.
69
(
18
),
2728
(
1996
).