Upload
others
View
9
Download
0
Embed Size (px)
Citation preview
Chemical bonding, optical constants, and electrical resistivity of sputter-depositedgallium oxide thin filmsC. V. Ramana, E. J. Rubio, C. D. Barraza, A. Miranda Gallardo, Samantha McPeak, Sushma Kotru, and J. T.
Grant
Citation: Journal of Applied Physics 115, 043508 (2014); doi: 10.1063/1.4862186 View online: http://dx.doi.org/10.1063/1.4862186 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/115/4?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Effect of etching with cysteamine assisted phosphoric acid on gallium nitride surface oxide formation J. Appl. Phys. 114, 064907 (2013); 10.1063/1.4817899 Compositional study of vacuum annealed Al doped ZnO thin films obtained by RF magnetron sputtering J. Vac. Sci. Technol. A 29, 051514 (2011); 10.1116/1.3624787 Optical properties of amorphous Ga As 1 x N x film sputtering with different N 2 partial pressures J. Vac. Sci. Technol. A 24, 1714 (2006); 10.1116/1.2217977 Influence of sputtering parameter on the optical and electrical properties of zinc-doped indium oxide thin films J. Vac. Sci. Technol. A 23, 1187 (2005); 10.1116/1.1924473 Optical behavior of reactive sputtered carbon nitride films J. Appl. Phys. 91, 2812 (2002); 10.1063/1.1446224
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
Chemical bonding, optical constants, and electrical resistivityof sputter-deposited gallium oxide thin films
C. V. Ramana,1,a) E. J. Rubio,1 C. D. Barraza,1 A. Miranda Gallardo,1 Samantha McPeak,2
Sushma Kotru,2,3 and J. T. Grant41Department of Mechanical Engineering, University of Texas at El Paso, El Paso, Texas 79968, USA2Department of Electrical and Computer Engineering, The University of Alabama, Tuscaloosa,Alabama 35487, USA3Materials for Information and Technology, The University of Alabama, Tuscaloosa, Alabama 35487, USA4Research Institute, University of Dayton, Dayton, Ohio 45469, USA
(Received 9 October 2013; accepted 1 January 2014; published online 24 January 2014)
Gallium oxide (Ga2O3) thin films were made by sputter deposition employing a Ga2O3 ceramic
target for sputtering. The depositions were made over a wide range of substrate temperatures (Ts),
from 25 to 600 �C. The effect of Ts on the chemical bonding, surface morphological characteristics,
optical constants, and electrical properties of the grown films was evaluated using X-ray
photoelectron spectroscopy (XPS), atomic force microscopy (AFM), spectroscopic ellipsometry
(SE), and four-point probe measurements. XPS analyses indicate the binding energies (BE) of the
Ga 2p doublet, i.e., the Ga 2p3/2 and Ga 2p1/2 peaks, are located at 1118.0 and 1145.0 eV,
respectively, characterizing gallium in its highest chemical oxidation state (Ga3þ) in the grown
films. The core level XPS spectra of O 1s indicate that the peak is centered at a BE� 531 eV, which
is also characteristic of Ga-O bonds in the Ga2O3 phase. The granular morphology of the
nanocrystalline Ga2O3 films was evident from AFM measurements, which also indicate that the
surface roughness of the films increases from 0.5 nm to 3.0 nm with increasing Ts. The SE analyses
indicate that the index of refraction (n) of Ga2O3 films increases with increasing Ts due to improved
structural quality and packing density of the films. The n(k) of all the Ga2O3 films follows the
Cauchy’s dispersion relation. The room temperature electrical resistivity was high (�200 X-cm) for
amorphous Ga2O3 films grown at Ts¼RT-300 �C and decreased to �1 X-cm for nanocrystalline
Ga2O3 films grown at Ts� 500–600 �C. A correlation between growth conditions, microstructure,
optical constants, and electrical properties of Ga2O3 films is derived. VC 2014 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4862186]
I. INTRODUCTION
Oxide dielectrics have been the subject of numerous
investigations for many years due to their possible device
integration in a wide range of technologies involving elec-
tronics, electro-optics, opto-electronics, and magneto-
electronics. Gallium oxide (Ga2O3), which is the stable oxide
of Ga, is one of the wide band gap oxides1–6 with potential
applications in luminescent phosphors,7 high temperature
sensors,8–11 antireflection coatings, and solar cells.10,12
Ga2O3 has been recognized as a deep-ultraviolet transparent
conducting oxide (UV–TCO),1–5,13 which makes the mate-
rial interesting for transparent electrode applications in
UV optoelectronics.6,14–16 While conventional transparent
oxides, such as In2O3, SnO2, ZnO, and ITO, are opaque in
the UV region due to their small band gaps (�3 eV), Ga2O3
exhibits a wide band gap (�5 eV) and high transparency in
the UV region.4,6,12–16
Gallium oxide can crystallize in five different crystal
structures: a, b, c, d, and e phases.17,18 Among these phase,
b-Ga2O3 is the most stable with a melting point of
1780 �C.10,17 The crystal structure of b-Ga2O3 is base
centered monoclinic (space group C2/m)17,18 where the oxy-
gen ions are in a distorted cubic packing arrangement and
the gallium ions are in distorted tetrahedral and octahedral
sites.18 The lattice parameters of b-Ga2O3 are
a¼ 12.2140(3) A, b¼ 3.0371(9) A, c¼ 5.7981(9) A, and
b¼ 103.83�.19–21 Despite great promise, studies on
b-Ga2O3 are rather scarce compared to other well explored
wide band gap oxides. For thin films grown using either
physical or chemical vapor deposition methods, properties
and phenomena of b-Ga2O3 films depend on the processing
conditions such as base pressure,9 growth temperature,22–24
reactive pressure (if any), deposition rate, and annealing
conditions. Therefore, the controlled growth and manipula-
tion of b-Ga2O3 at nanoscale dimensions has important
technological implications. Specifically, the ultramicrostruc-
ture in terms of the chemical valence state of Ga and the
surface chemistry of the grown films strongly influences the
electronic properties and hence their device applications.
Most importantly, determination of the index of refraction
(n) and its functional dependence on the wavelength is im-
portant for transparent oxides, such as Ga2O3 in this case.
Furthermore, for thin films and nanomaterials, the “n” val-
ues are sensitive to the microstructure; n is influenced by
various factors such as surface/interface structure, crystal
quality, packing density, lattice parameters, lattice strain,
a)Author to whom correspondence should be addressed. Email:
[email protected]. Tel.: 915-747-8690. Fax: 915-747-5015
0021-8979/2014/115(4)/043508/9/$30.00 VC 2014 AIP Publishing LLC115, 043508-1
JOURNAL OF APPLIED PHYSICS 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
defect structure, and chemical composition. On the other
hand, the index of refraction is the most important physical
quantity, which is connected to most of the optical and
electro-optic phenomena in matter and is useful for engi-
neering modern electronic and optical devices. The present
work was, therefore, performed on sputter-deposited Ga ox-
ide films made by sputter-deposition at various temperatures
in order to probe their surface chemistry, Ga-O bonding,
and optical constants. Combined X-ray photoelectron spec-
troscopy (XPS), atomic force microscopy (AFM), and spec-
troscopic ellipsometry (SE) measurements allowed us to
determine the chemical composition, morphology, and
optical constants of Ga2O3 films. The results obtained are
presented and the effects of growth temperature and micro-
structure on the index of refraction profiles and electrical
conductivity of Ga2O3 films are discussed in this paper.
II. EXPERIMENTS
A. Fabrication
Gallium oxide films were deposited onto silicon (Si)
(100) wafers by radio-frequency magnetron sputtering. All
the substrates were thoroughly cleaned using the RCA
(Radio Corporation of America) procedure. Briefly, the RCA
cleaning procedure removes organic, alkali ions, and heavy
metal contaminants present on the surface of the
substrate. The cleaning procedure has three major steps. (1)
Removal of insoluble organic contaminants using 5:1:1
H2O/H2O2/NH4OH solution. (2) Removal of ionic and heavy
metal atomic components using a solution of 6:1:1
H2O/H2O2/HCl solution. (3) Removal of native oxide by
buffered oxide etching solution. Finally, the substrates were
dried with nitrogen before introducing them into the vacuum
chamber, which was initially evacuated to a base pressure of
�10�6 Torr. A gallium oxide target (Plasmaterials Inc.) of
5 cm diameter and 99.999% purity was employed for sputter-
ing. The Ga2O3-target was placed on a 5 cm sputter gun,
placed at a distance of 8 cm from the substrate. A sputtering
power of 40 W was initially applied to the target while intro-
ducing high purity argon (Ar) into the chamber to ignite the
plasma. Once the plasma was ignited, the power was
increased to 100 W for reactive deposition. The flow of Ar
was controlled using an MKS mass flow meter. Before each
deposition, the Ga2O3-target was pre-sputtered for 10 min
with the shutter closed. The deposition was carried out for
30 min. The thickness of the films was �40 nm. The samples
were deposited at different temperatures in the range of
25–600 �C. Substrate rotation was maintained during the
entire deposition time to ensure uniform coverage on the
substrate surface.
B. Characterization
1. X-ray photoelectron spectroscopy
The chemical composition of the Ga2O3 films was
determined using X-XPS. The samples were mounted on a
stainless steel sample holder, held down with double-sided
carbon tape, and analyzed using a calibrated Physical
Electronics PHI 5700 XPS system with a monochromated
Al K-a (1486.7 eV) source operating at 350 W at
�1� 10�9 torr and an angle of emission of 45�. Survey and
multiplex (multiple high resolution regions) scans were
taken before and after sputtering with Ne ions at 4 keV for
5 min. Survey spectra were obtained over the binding
energy (BE) range of 0–1200 eV with an energy step of
0.80 eV, a total time of 0.58 s per data point, a dwell time
of 20 ms, and an analyzer pass energy of 187.85 eV.
Multiplex spectra were obtained with an energy step of
0.125 eV and analyzer pass energy of 29.35 eV. Peaks
scanned in multiplex mode were Ga 2p, Si 2p, O 1 s, and C
1 s with a total time per data point of 0.85 s, 1.7 s, 0.85 s,
and 1.7 s, respectively. The Ga 2p region was chosen
instead of the more commonly used Ga 3d region to avoid
interference of the Ga 3d peaks with the O 2 s peak.24,25
The chemical shifts of the Ga 2p and Ga 3d peaks are simi-
lar.26 Probing the Ga 2p region, therefore, also provides in-
formation on the Ga-O chemical bonding and Ga valence
state in the grown films. After data acquisition, survey spec-
tra were analyzed using CasaXPS V2.3.16 to obtain atomic
composition. Peak areas were calculated using Shirley back-
ground subtraction and concentrations were obtained using
sensitivity factors for the instrument.
2. Atomic force microscopy
Surface morphology of the Ga2O3 films was studied
employing AFM using a Veeco Multimode scanning probe
microscope with a Nanoscope V controller. AFM images
were acquired using the tapping mode at a scanning fre-
quency of 1.53 Hz. Silicon non-contact cantilevers
(Nanoandmore, USA) were used to acquire the tapping
mode images. These uncoated silicon cantilevers are
125 lm long, 30 lm wide, and 4 lm thick, with a radius of
curvature of �8 nm. The resonance frequency of the tips
was �320 kHz. Tip quality was qualitatively assessed by
the clarity and presence of artifacts in the acquired images.
The calibration standard used to calibrate the scanner con-
sisted of platinum-coated, 200 nm-tall silicon columns
spaced at 10 lm intervals on centers. The columns have a
length of 5 nm on a side. AFM images (512 scan lines and
512 pixels per scan line) were acquired with an integral
gain of approximately 2 and an amplitude setpoint of
300 mV. The drive amplitude varied between 30 and
180 mV while the proportional gain varied from 4 to 25.
The images were then subjected to a 3rd order flattening
procedure using the Bruker NanoScope Analysis V1.40
software to remove the non-linear background artifact intro-
duced by the piezo scanner. Following the flattening proce-
dure, the surface roughness was quantified over five areas
of 1 lm� 1 lm on the 5 lm scan size images.
3. Spectroscopic ellipsometry
The optical properties and surface/interface characteris-
tics were probed by SE. The measurements were made using
a W. A. Woollam VVASE spectroscopic ellipsometer oper-
ating in the wavelength range of 250–1350 nm with a step
size of 2 nm and at angles of incidence of 65�, 70�, and 75�,near the Brewster’s angle of silicon. The number of
043508-2 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
revolutions per measurement was set at 20. The ellipsometry
data analysis was performed using commercially available
WVASE32 software.27
4. Electrical resistivity
The four-point Van der Pauw method was used to mea-
sure the resistivity of the films.28 Four electrical contacts on
the sample were made using a two part conducting epoxy
and silver wires.
In this method, voltage and current are measured
using eight different combinations of the four contact
points on the sample. This means that in the first set of
measurements, current is applied to one pair of electrodes
and voltage is measured across the other two electrodes
and then the process is repeated for other contact configu-
rations. A current of 2 mA was applied to the sample using
a Keithley 220 programmable current source. A Keithley
196 sensitive digital voltmeter was used to read the volt-
age and a 196 DMM system was used to read the current
between two contacts. A Keithley 2001 switch was used to
switch between contacts. The meters were connected to a
computer and LABVIEW data acquisition system via the
IEEE-488 GPIB for automatic data recording. The final re-
sistivity was calculated using the Van der Pauw principle
taking into account the sample thickness, the shape factor,
and other limitations.28
III. RESULTS
A. Chemical bonding—x-ray photoelectronspectroscopy
Structural and surface morphology characterization of
Ga2O3 films was performed employing X-ray diffraction
(XRD) and scanning electron microscopy (SEM), and the
results were reported elsewhere.29 The chemical valence
state and surface chemistry of the grown Ga2O3 films was
analyzed by XPS. The XPS survey spectra recorded for
as-grown and Neþ sputtered Ga-oxide films are shown in
Fig. 1. The data are shown in the BE range of 0–1200 eV for
two representative samples grown at RT and 500 �C pre-
sented in the upper and lower panels, respectively (Fig. 1).
For the as-grown Ga-oxide films, all the spectral features,
except the C 1s level, are attributed to the constituent ele-
ment core levels or Auger lines. The C 1s signal is due to
hydrocarbons adsorbed on the film surface during air expo-
sure during transfer from the growth chamber to the XPS
instrument. It should be noted that the C 1s signal is com-
pletely absent in the XPS spectra of Neþ bombarded films
for both the samples shown. This observation confirms the
removal of hydrocarbon contamination from the surface. It is
clear from the XPS data that the spectral components related
to constituent elements become more intense as a result of
ion-bombardment of the film surfaces. The absence of a Si
substrate peak is due to the fact that the deposited Ga-oxide
FIG. 1. XPS survey scans of represen-
tative Ga2O3 films grown at different
substrate temperatures. The survey
scans obtained for films grown at RT
(upper panel) and 500 �C (lower panel)
are shown. The data shown are from
as-grown and Neþ sputtered samples.
The removal of surface carbon species
upon sputtering is evident.
043508-3 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
films cover the Si surface and are sufficiently thick to prevent
detection of photoelectrons from the substrate.
The core level photoelectron spectra of the Ga 2p doublet
are shown in Fig. 2. The high resolution XPS spectra are
shown for as-grown (left panel) and Neþ sputtered (right
panel) samples. The BE values of these Ga 2p3/2 and Ga 2p1/2
peaks are 1118.0 and 1145.0 eV, respectively; they are almost
constant and do not exhibit any change or shift as a function
of substrate temperature during growth. Furthermore, no
appreciable change is noted in the BE values of these peaks
in as-grown and Neþ sputtered samples. However, a very
small variation in full-width at half-maximum is noted for
these peaks in Neþ sputtered samples compared to as-grown
samples. The BE values of Ga 2p levels for metallic Ga are:
1117.0 and 1144.0 eV for Ga 2p3/2 and Ga 2p1/2, respectively,
and are indicated by the dashed lines in Fig. 2.24 Compared to
metallic gallium, the grown Ga-oxide films show a positive
shift in the BE (1.0 eV). It should be noted that the BE shift
occurs due to the redistribution of electronic charge around
the constituent atoms.30–32 Therefore, the differences in Ga
chemical bonding cause a BE shift that can be used to extract
information about the oxidation state of Ga in the grown
films. Based on these considerations, it can be concluded that
the chemical state of Ga in the films is in the highest valence
state (Ga3þ) without any of the reported intermediate valence
states.24,33
The core level photoelectron spectra of O 1s also
revealed interesting features about the surface chemistry of
the grown Ga-oxide films. The high resolution XPS spectra
of O 1s core levels are shown for as-grown (left panel) and
Neþ sputtered (right panel) samples in Fig. 3. The O 1s peak
is asymmetric in as-grown samples. While the main peak is
centered at a BE �531 eV, there is a small shoulder at a
BE� 532 eV. As shown in Fig. 4 for a representative sam-
ples grown at RT, the O 1s peak for as-grown Ga-oxide films
can be resolved into two components. The most intense peak
is located at 531.0 eV and the less intense peak is at
532.3 eV, representing the different chemistries of oxygen.
The less intense contribution at a BE of 532.3 eV disappears
in all the samples upon Neþ sputtering of the surface. With
Neþ sputtering of the surface, the O 1s peak changes into a
single, sharp peak centered at a BE of 531.0 eV. These fea-
tures can be understood as follows. It has been reported that
the O 1s peak in Ga-oxide occurs at a BE� 531 eV, which
FIG. 2. Ga 2p core level spectra of
Ga2O3 thin films grown at various sub-
strate temperatures. The core level
spectra obtained from as-grown and
Neþ sputtered samples are shown. The
vertical lines indicate the BE positions
for Ga 2p levels in gallium metal and
serve as reference to indicate the posi-
tive shift of the Ga 2p lines in the
grown Ga-oxide films.
FIG. 3. O 1s core level spectra of
Ga2O3 thin films grown at various sub-
strate temperatures. The core level
spectra obtained from as-grown and
Neþ sputtered samples are shown.
043508-4 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
characterizes Ga-O bonding with the highest oxidation
state of Ga (Ga3þ). The less intense contribution at a
BE� 532.3 eV can be attributed to either carbonyl (oxygen
bonded to carbon) or hydroxyl (oxygen bonded to hydrogen)
groups, which are adsorbed onto the surface.34,35 However,
it was noted in the survey XPS spectra (Fig. 1) that the C 1s
signal was present on all of the as-grown Ga-oxide samples.
The presence of C 1s disappeared upon Neþ sputtering of the
surface. Based on these observations, it is clear that the
minor contribution that appears in O 1s spectra is due to the
surface C-O or hydroxide bonds, which disappeared on sput-
tering away the few top-most surface layers.
B. Surface characteristics—atomic force microscopy
The AFM results for the Ga2O3 samples grown at
Ts� 500 �C are shown in Fig. 5. The topographic images are
shown along with the corresponding 3D surface images. It is
remarkable to note the evident nano-particulate morphology
the film surfaces, where the conically shaped grains are dis-
tributed uniformly all over the film surface. Samples grown
at Ts< 500 �C (not shown) are mostly amorphous in nature
and no crystalline grains are presented on the film surfaces.
This observation is in agreement with the XRD data, which
indicate that Ga2O3 films grown at Ts< 500 �C were amor-
phous in nature while those grown at Ts� 500 �C are
nano-crystalline. It is evident from Fig. 5 that the grain-size
increases with increasing growth temperature. The size
increases from �20 nm to �32 nm with increasing growth
temperature from 500 �C to 600 �C. The effect of growth
temperature is also reflected in the case of root-mean-square
surface roughness values, which increased from 1 nm to
�3 nm with increasing temperature from 500 to 600 �C. The
surface roughness values of the amorphous films were not
significant. For instance, the set of films grown at
Ts< 500 �C exhibit a surface roughness of <1 nm.
C. Optical constants—spectroscopic ellipsometry
Optical constants of the Ga2O3 films were primarily
probed by SE, which measures the relative changes in the
amplitude and phase of the linearly polarized monochro-
matic incident light upon oblique reflection from the sample
surface. The experimental parameters obtained by SE are the
angles W (azimuth) and D (phase change), which are related
to the microstructure and optical properties, defined by36–40
q ¼ Rp=Rs ¼ tan W exp iDð Þ; (1)
FIG. 4. Peak fitting analysis of the O 1s core-level peak in as-grown Ga-ox-
ide films fabricated at RT. It is evident that the peak can be resolved into
two separate components, which are located at 531.0 eV and 532.3 eV with
high and low intensities, respectively.
FIG. 5. AFM images of sputter depos-
ited Ga2O3 films; (a) sample grown at
500 �C, (b) sample grown at 600 �C.
043508-5 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
where Rp and Rs are the complex reflection coefficients of
the light polarized parallel and perpendicular to the plane of
incidence, respectively. The spectral dependencies of the
ellipsometric parameters, W and D, determined for Ga2O3
films grown at various temperatures are shown in Fig. 6. The
spectral dependencies of ellipsometric parameters W (azi-
muth) and D (phase change) can be fitted with appropriate
models to extract film thickness and the optical constants,
i.e., the refractive index (n) and extinction coefficient (k),
based on the best fit between experimental and simulated
spectra.36,37 The curves obtained for Ga2O3 films indicate
(Fig. 6) a reasonable agreement between the experimental
and simulation data. The Levenberg-Marquardt regression
algorithm was used for minimizing the mean-squared error
(MSE)36
MSE ¼ 1
2N �M
Xn
i¼1
Wexp: � Wcalc:ð Þrexp
Wi
( )224
þDexp: � Dcalc:
� �rexp
Di
( )2 #; (2)
where Wexp, Wcalc and Dexp, Dcalc are the measured (experi-
mental) and calculated ellipsometry functions, N is the num-
ber of measured W, D pairs, M is the number of fitted
parameters in the optical model, and r are standard devia-
tions of the experimental data points.
Extracting meaningful physical information from ellips-
ometry requires the construction of an optical model of the
sample, which generally accounts a number of distinct layers
with individual optical dispersions. Interfaces between these
layers are optical boundaries at which light is refracted and
reflected according to the Fresnel relations. The stack model
used to simulate the spectra for the purpose of determining
the optical constants of Ga2O3 films is schematically shown
in Fig. 7. The model contains, from top, the Ga2O3 film, the
SiO2 interface, and the Si substrate. The surface roughness
was also considered in order to accurately fit the experimen-
tal data. A Cauchy dispersion model for the refractive index
was employed to deduce film thickness, roughness values,
and refractive index (n) dispersion. The Cauchy dispersion
model is40
n ¼ n1 þA
k2þ B
k4: (3)
FIG. 6. The spectral dependence of W and D for Ga2O3 films grown at various temperatures. The experimental data obtained and modeling curves are shown.
FIG. 7. The stack model of the Ga2O3 sample constructed for ellipsometry
data analysis.
043508-6 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
The microstructure information, specifically, film thick-
ness, roughness, and interfacial oxide thickness, were
allowed to vary and determined from SE analysis. The varia-
tion of film thickness from 43 nm to 32 nm was noted as a
function of growth temperature. It is evident that the film
thickness is more or less constant for Ts¼ 25–400 �C at
which point a decreasing trend prevails. This can be under-
stood based on the fact that the films grown at Ts< 500 �Care amorphous while those grown at Ts� 500 �C were
nano-crystalline.29 The observed thickness decrease from
44 nm to 32 nm is, thus, due to Ga2O3 film crystallization
and densification with increasing substrate temperature.
The dispersion profiles of index of refraction (n) deter-
mined from SE data for Ga2O3 films are shown in Fig. 8.
The “n” dispersion curves also indicate a sharp increase at
shorter wavelengths corresponding to fundamental absorp-
tion of energy across the band gap. The range of n-values
and the observed behavior agree well with those reported in
the literature for electron beam deposited Ga2O3 films.15
However, the effect of growth temperature is evident in the
dispersion curves (Fig. 8), where there is an increase in “n”values with increasing Ts. While the increasing trend of “n”values with increasing Ts from very beginning is clear, the
dispersion curves of Ga2O3 films grown at higher tempera-
ture (500–600 �C) are distinctly separated from those of sam-
ples grown at lower temperatures, indicating their
characteristic optical quality difference.
D. Electrical resistivity
Figure 9 shows the room temperature resistivity of the
films grown at various deposition temperatures. Films grown
at lower temperatures are seen to exhibit higher resistivity,
the values being approximately 200–210 X-cm for films
grown up to 300 �C. A sharp decrease in resistivity is
observed for samples prepared at 400 �C. As the growth
temperature increases, the resistivity is seen to decrease fur-
ther. The sample grown at 600 �C exhibits the lowest resis-
tivity of 1 X-cm.
IV. DISCUSSION
The emphasis in this work is to establish the direct rela-
tionship between the index of refraction (“n”) and structural
quality of sputter-deposited Ga2O3 films grown under vari-
able growth temperature (Ts). It must be emphasized that the
optical properties are sensitive to the film microstructure.
Various factors such as surface/interface structure, grain
size, crystal quality, lattice parameters, defect structure, and
chemical composition influence the optical properties of ox-
ide films. The growth temperature, which is an important
thermodynamic parameter, plays an important role,
besides the reactive oxygen pressure, in deciding the micro-
structure as well as properties of materials resulting from
sputter-deposition. Therefore, the XPS, AFM, and SE results
along with electrical resistivity variation are considered in
order to derive the functional relationship between the sur-
face physics, morphology, and optical properties of Ga2O3
films.
The effect of Ts on the crystal structure, phase, grain
size, and band gap of Ga2O3 films is summarized in Table I.
The crystal structure, phase, and band gap were evaluated
in our earlier work.29 Briefly, Ga2O3 films grown at
Ts¼RT-400 �C were amorphous (a-Ga2O3). A substrate
temperature of 500 �C is critical to the onset of film crystalli-
zation with the b-Ga2O3 phase formation. Therefore, films
grown at Ts� 500 �C crystallize in the b-Ga2O3 phase. The
grain size determined varied from 15 to 35 nm with increas-
ing Ts. Now, we focus our attention on the XPS results to es-
tablish the surface chemistry of the Ga2O3 films. XPS
analyses indicate that the films are nearly stoichiometric and
Ga exists in its highest chemical valence state. The BE val-
ues of Ga 2p3/2 and Ga 2p1/2 peaks (1118.0 and 1145.0 eV,
respectively) and the BE shift from Ga0 (1 eV) characterizes
the Ga in its highest chemical valence state (Ga3þ) in all the
films, almost without any appreciable change, as a function
FIG. 8. The dispersion profiles of index of refraction of Ga2O3 films grown
at various Ts. The sharp increase in index of refraction at k’ 400 nm is due
to fundamental absorption across the band gap of the films. The distinct dis-
persion curves as a function of Ts indicate the differences in structural qual-
ity of Ga2O3 films.
FIG. 9. Electrical resistivity variation of Ga2O3 films with Ts. It can be noted
that the resistivity of amorphous Ga2O3 films is generally high and exhibits
a sharp decrease for nanocrystalline Ga2O3 films.
043508-7 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
of Ts. This is important from the view point of electronic
structure since reduced chemical valence states of Ga can
create defects, which in turn alter the optical and electrical
properties. Furthermore, the chemical stoichiometry analyses
made by XPS in this work, and RBS results reported else-
where29 are in good agreement with each other. Also, these
chemical studies rule out the possible compositional defects
or reduced Ga states and their influence on the optical and
electrical properties of Ga2O3 films.
Finally, the observed variation in the index of refraction
and electrical resistivity of Ga-oxide films can be explained
based on the microstructure variation of Ga2O3 films with
Ts. A simple model can be formulated to explain the
observed functional relationship between the optical con-
stants and Ts in Ga2O3 films. Evident from the SE results,
the dispersion of optical constants depend on the growth
temperature. XRD measurements demonstrated that the
Ga2O3 films are completely amorphous at Ts¼ 25–500 �C at
which point amorphous to nano-crystalline structural trans-
formation occurs.29 The observed increase in “n” values
when Ga2O3 films are grown at higher temperatures can,
therefore, be attributed to the improved structural order and
packing density of the films. Furthermore, while surface
roughening can induce scattering losses, AFM data evi-
denced the rms roughness of the films are not significant in
Ga2O3 films as a function of Ts. Usually, rough film surfaces
result in increased light-scattering losses at the interface.
Therefore, the critical temperature (500 �C) noted in the
XRD measurements seems to promote the structural order
and increases the mobility of ad-atoms to join together to
increase the packing density of the material in the films lead-
ing to observed increase in “n” values. Such behavior was
also noted in sputter-deposited Y2O3 films.39
The effect of microstructure was also reflected in the
electrical resistivity analysis of the Ga2O3 films. The electrical
resistivity usually becomes higher with grain size reduction
due to the increasing grain boundary volume and associated
impedance to the flow of charge carriers.41,42 If the crystallite
size is smaller than the electron mean free path, grain bound-
ary scattering dominates and hence the higher electrical resis-
tivity. Furthermore, the electrical resistivity is also very
sensitive to lattice imperfections in solids, such as vacancies
and dislocations, which are often present in nanocrystalline
materials.41 In addition to that, lattice strain and their distor-
tions can affect the charge mobility causing a decrease in con-
ductivity.41,42 The room-temperature electrical resistivity
variation with Ts observed for Ga2O3 films (Fig. 9) can be
explained taking these factors into consideration. Ga2O3 films
grown at Ts¼RT-300 �C are completely amorphous as evi-
denced in the XRD studies. The randomness or disordered
structure of the films, therefore, accounts for the observed
high electrical resistivity of amorphous Ga2O3 films. The re-
sistivity decrease with increasing Ts can be attributed to the
increasing crystalline nature of Ga2O3 films. The observed
sharp decrease in the electrical resistivity for Ga2O3 films
grown at higher temperatures seems to be a result of the com-
bined effects of the grain-size increase and the phase transfor-
mation from amorphous-Ga2O3 to b-Ga2O3 films. Thus, this
trend in resistivity is expected and can be attributed to better
crystallinity, the presence of larger grains and fewer grain
boundaries in the samples prepared at higher temperatures.
Furthermore, the electrical results are in agreement with the
results obtained from AFM and XRD studies.
V. CONCLUSIONS
Gallium oxide films were produced by sputter deposi-
tion onto Si(100) substrates. The chemical bonding in the
grown films was probed by XPS. The Ga 2p core level spec-
tra and analyses indicate that the Ga 2p3/2 and Ga 2p1/2 peaks
are located at BE values of 1118.0 and 1145.0 eV, respec-
tively, in all the grown Ga-oxide films. The positive shift in
the BE values (compared to elemental Ga) indicate they are
characteristic of gallium in the Ga3þ state. The O 1s core
level XPS spectra from as-grown samples exhibit the pres-
ence of surface carbonyl or hydroxyl groups, which could be
removed by ion sputtering of the top-most surface layers.
The core level XPS spectra of O 1s coupled with the Ga 2p
core level data indicate the presence of the Ga2O3 phase in
the films. AFM revealed the granular morphology of the
nanocrystalline Ga2O3 films. The index of refraction of
Ga2O3 films increases with increasing Ts while the n-disper-
sion obeys the Cauchy dispersion relation. The n-Ts direct
relationship is explained based on the structural transforma-
tion, improved structural order, and enhanced packing den-
sity of the Ga2O3 films with increasing growth temperature.
The room temperature electrical resistivity measurements
demonstrated the direct relationship between crystal struc-
ture and electrical properties. The resistivity was high (�200
X-cm) for amorphous Ga2O3 films and it sharply decreased
at the amorphous-to-crystalline film formation at higher Ts.
ACKNOWLEDGMENTS
This material was based on the work supported by the
Department of Energy under Award Number DE-PS26-
08NT00198-00. C.V.R. also acknowledges with pleasure the
support from the National Science Foundation (NSF) with
NSF-PREM Grant # DMR-1205302. One of us (A.M.G.) is
thankful to the Campus Office of Undergraduate Research
Initiative (COURI) at the University of Texas at El Paso for
the funding and opportunity for research experience.
1M. Orita, H. Ohta, M. Hirano, and H. Hosono, Appl. Phys. Lett. 77, 4166
(2000).
TABLE I. Summary of the effect of growth temperature on the crystal struc-
ture, phase, grain size, and band gap of Ga2O3 films.29
Growth
temp. (�C)
Crystal structure
and phase
Ga/O
ratio
Band gap
(eV)
Gran size
(nm)
25 Amorphous 1.6 5.17 …
200 Amorphous 1.6 5.02 …
300 Amorphous 1.5 5.01 14
400 Amorphous 1.5 4.98 16
500 Monoclinic (b-Phase) 1.5 4.99 20
600 Monoclinic (b-Phase) 1.5 4.96 35
043508-8 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11
2M. Mohamed, C. Janowitz, I. Unger, R. Manzke, Z. Galazka, R. Uecker,
R. Fornari, J. R. Weber, J. B. Varley, and C. G. Van de Walle, Appl. Phys.
Lett. 97, 211903 (2010).3J. Q. Hu, Q. Li, X. M. Meng, C. S. Lee, and S. T. Lee, J. Phys. Chem. B
106, 9536 (2002).4N. Ueda, H. Hosono, R. Waseda, and H. Kawazoe, Appl. Phys. Lett. 70,
3561 (1997).5T. C. Lovejoy, E. N. Yitamben, N. Shamir, J. Morales, E. G. Villora, K.
Shimamura, S. Zheng, F. S. Ohuchi, and M. A. Olmstead, Appl. Phys.
Lett. 94, 081906 (2009).6P. D. C. King, I. McKenzie, and T. D. Veal, Appl. Phys. Lett. 96, 062110
(2010).7T. Minami, T. Nakatani, and T. J. Miyata, J. Vac. Sci. Technol. A 18,
1234 (2000).8M. Fleischer and H. Meixner, Sens. Actuators B 26–27, 81 (1995).9M. Ogita, N. Saika, Y. Nakanishi, and Y. Hatanaka, Appl. Surf. Sci. 142,
188 (1999).10M. Bartic, C.-I. Baban, H. Suzuki, M. Ogita, and M. J. Isai, Am. Ceram.
Soc. 90, 2879 (2007).11H. Jiang, Y. Chen, Q. Zhou, Y. Su, H. Xiao, and L. Zhu, Mater. Chem.
Phys. 103, 14 (2007).12V. Bhosle, J. T. Prater, F. Yang, D. Burk, S. R. Forrest, and J. Narayan,
J. Appl. Phys. 102, 023501 (2007).13J. B. Varley, J. R. Weber, A. Janotti, and C. G. Van de Walle, J. Appl.
Phys. Lett. 97, 142106 (2010).14H. He, R. Orlando, M. A. Blanco, and R. Pandey, Phys. Rev. B 74, 195123
(2006).15M. Rebien, W. Henrion, M. Hong, J. P. Mannaerts, and M. Fleischer,
Appl. Phys. Lett. 81, 250 (2002).16Z. Ji, J. Du, J. Fan, and W. Wang, Opt. Mater. 28, 415 (2006).17F. Litimein, D. Rached, R. Khenata, and H. J. Baltache, J. Alloys. Compd.
488, 148 (2009).18R. Roy, V. G. Hill, and E. F. Obson, J. Am. Chem. Soc. 74, 719 (1952).19K. Takakura, D. Koga, H. Ohyama, J. M. Rafi, Y. Kayamoto, M.
Shibuya, H. Yamamoto, and J. Vanhellemont, Physica B 404, 4854
(2009).20C. Janowitz, V. Scherer, M. Mohamed, A. Krapf, H. Dwelk, R. Manzke,
Z. Galazka, R. Uecker, K. Irmscher, R. Fornari et al., New J. Phys. 13,
085014 (2011).21L. Binet, D. Gourier, and C. Minot, J. Solid State Chem. 113, 420 (1994).
22H. W. Kim and N. H. Kim, Mater. Sci. Eng. B 110, 34 (2004).23S. Ou, D. Wuu, Y. Fu, S. Liu, R. Horng, L. Liu, and Z. Feng, Mater.
Chem. Phys. 133, 700 (2012).24W. Priyantha, G. Radhakrishnan, R. Droopad, and M. Passlack, J. Cryst.
Growth 323, 103 (2011).25A. I. Serykh and M. D. Amiridis, Surf. Sci. 604, 1002 (2010).26J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of
X-ray Photoelectron Spectroscopy (Perkin-Elmer Corp., Eden Prairie,
MN, 1992).27J. A. Woollam Co., Inc., Guide to Using WVASE32 Spectroscopic
Ellipsometry Data Acquisition and Analysis Software, 2008.28L. J. Van der Pauw, “A method of measuring the resistivity and Hall coef-
ficient on lamellae of arbitrary shape,” Philips Tech. Rev. 20, 220 (1958).29S. Sampath Kumar, E. J. Rubio, M. Noor-A-Alam, G. Martinez, S.
Manandhar, V. Shutthanandan, S. Thevuthasan, and C. V. Ramana,
J. Phys. Chem. C 117, 4194 (2013).30V. V. Atuchin, L. D. Pokrovsky, O. Yu. Khyzhun, A. K. Sinelnichenko
and C. V. Ramana, J. Appl. Phys. 104, 033518 (2008).31V. V. Atuchin, A. V. Kalinkin, V. A. Kochubey, V. N. Kruchinin, R. S.
Vemuri, and C. V. Ramana, J. Vac. Sci. Technol. A 29, 021004 (2011).32B. Zheng, W. Hua, Y. Yue, and Z. Gao, J. Catal. 232, 143 (2005).33L. Kong, J. Ma, C. Luan, W. Mi, and Y. Lv, Thin Solid Films 520, 4270
(2012).34J. C. Dupin, D. Gonbeau, P. Vinater, and A. Levasseur, Phys. Chem.
Chem. Phys. 2, 1319 (2000).35H. Perron, J. Vandenborre, C. Domain, R. Drot, J. Roques, E. Simoni, J. J.
Ehrhardt, and H. Catalette, Surf. Sci. 601, 518 (2007).36G. E. Jellison, Jr., Thin Solid Films 313, 33 (1998).37H. Fujiwara, Spectroscopic Ellipsometry: Principles and Applications
(John Wiley & Sons Inc, 2007).38C. V. Ramana, S. Utsunomiya, R. C. Ewing, U. Becker, V. V. Atuchin, V.
Sh. Aliev, and V. N. Kruchinin, Appl. Phys. Lett. 92, 011917 (2008).39V. H. Mudavakkat, V. V. Atuchin, V. N. Kruchinin, A. Kayani, and C. V.
Ramana, Opt. Mater. 34, 893 (2012).40F. A. Jenkins and H. E. White, Fundamentals of Optics, 4th ed. (McGraw-
Hill, Inc., 1981).41R. S. Vemuri, K. Kamala Bharathi, S. K. Gullapalli, and C. V. Ramana,
ACS Appl. Mater. Inter. 2, 2623 (2010).42K. Ajai Gupta, V. Kumar, and N. Khare, Solid State Sci. 9, 817
(2007).
043508-9 Ramana et al. J. Appl. Phys. 115, 043508 (2014)
[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:
129.108.52.159 On: Tue, 01 Apr 2014 20:49:11