Properties of ZnO/Al2O3 alloy films grown using atomic layer deposition techniques

Article abstract of DOI:10.1149/1.1569481

By varying the ratio of the constituents, compound films can exhibit a widely tunable range of physical properties. Atomic layer deposition (ALD) techniques are based on sequential, self-limiting surface reactions and can grow compound films. In this study, ZnO/Al₂O₃ alloy films were prepared using ALD techniques. By adjusting the ALD pulse sequence, the ZnO/Al₂O₃ alloy film composition was varied from 0-100% ZnO. These ZnO/Al₂O₃ alloy films are expected to display varying properties because ZnO and Al₂O₃ have very differential physical properties. For example, ZnO is a conductor and Al₂O₃ is an insulator. The physical properties of the ZnO/Al₂O₃ alloys were explored using a variety of techniques. The growth rate, refractive index, composition, surface roughness, crystallinity, resistivity and density of the ZnO/Al₂O₃ alloy films could be continuously tuned over the full range of values defined by the pure oxides. The refractive index varied from n = 2.00 for pure ZnO to n = 1.64 for pure Al₂O₃. The resistivity could be tuned over 18 orders of magnitude from 10^-2 Ω cm for pure ZnO to 10¹⁶ Ω cm for pure Al₂O₃. Anomalies in surface roughness and density vs. Zn content were observed at a Zn content of ～76%. These anomalies were attributed, in part, to the etching of Zn by Al(CH₃)₃ during the ZnO/Al₂O₃ alloy film growth.

Full text of DOI:10.1149/1.1569481

Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
G339
0
013-4651/2003/150͑6͒/G339/9/$7.00 © The Electrochemical Society, Inc.
Properties of ZnOÕAl O Alloy Films Grown Using Atomic
2
3
Layer Deposition Techniques
a
b,
a,c,z
J. W. Elam, D. Routkevitch * and S. M. George
a
c
Department of Chemistry and Biochemistry and Department of Chemical Engineering, University
of Colorado, Boulder, Colorado 80309-0215, USA
b
Nanomaterials Research, LLC, Longmont, Colorado 80501, USA
By varying the ratio of the constituents, compound films can exhibit a widely tunable range of physical properties. Atomic layer
deposition ͑ALD͒ techniques are based on sequential, self-limiting surface reactions and can grow compound films. In this study,
ZnO/Al O alloy films were prepared using ALD techniques. By adjusting the ALD pulse sequence, the ZnO/Al O alloy film
2
3
2
3
composition was varied from 0-100% ZnO. These ZnO/Al O alloy films are expected to display varying properties because ZnO
2
3
and Al O have very different physical properties. For example, ZnO is a conductor and Al O3 is an insulator. The physical
2
3
2
properties of the ZnO/Al O alloys were explored using a variety of techniques. The growth rate, refractive index, composition,
2
3
surface roughness, crystallinity, resistivity and density of the ZnO/Al O3 alloy films could be continuously tuned over the full
2
range of values defined by the pure oxides. The refractive index varied from n ϭ 2.00 for pure ZnO to n ϭ 1.64 for pure Al O .
2
3
Ϫ2
16
The resistivity could be tuned over 18 orders of magnitude from 10 ⍀ cm for pure ZnO to 10 ⍀ cm for pure Al O .
2
3
Anomalies in surface roughness and density vs. Zn content were observed at a Zn content of ϳ76%. These anomalies were
attributed, in part, to the etching of Zn by Al͑CH3)3 during the ZnO/Al O alloy film growth.
2
3
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1569481͔ All rights reserved.
Manuscript submitted May 31, 2002; revised manuscript received December 16, 2002. Available electronically April 11, 2003.
The physical properties of compound films can be tuned by ad-
justing the relative proportions of the constituent materials. This
strategy has been employed to control numerous thin film properties
Al O films can be deposited with atomic layer control by repeating
these reactions in an ABAB... sequence.
2
3
ZnO and Al O exhibit very dissimilar physical properties as
2
3
1
2
3
such as refractive index, dielectric constant, lattice constant,
given in Table I. ZnO ALD films are conducting, crystalline, and
hardness, charge storage capacity,5 and surface roughness. The
ability to engineer films with specific physical characteristics im-
pacts a broad range of technologies. These technologies include in-
4
6
44-47
rough.
Al O ALD films are insulating, amorphous, and
Because of these differences, the physical properties of
2 3
48-52
smooth.
ZnO/Al O alloys may span a broad range of values. Fortunately,
5,7
3,8
9,10
2
3
tegrated circuits1, optoelectronics, gas sensors,
and thermal
1,12
optimal growth rates for ZnO ALD and Al2O3 ALD are achieved at
nearly the same deposition temperatures and reactant exposures.
These similar reaction conditions facilitate the growth of the
ZnO/Al2O3 alloy films.
barrier coatings.
Atomic layer deposition ͑ALD͒ is a useful technique for growing
compound films. ALD relies on a binary sequence of self-limiting
surface chemical reactions.1
3,14
These reactions typically occur be-
ZnO/Al O alloy films may have numerous applications.
2
3
tween a gas-phase precursor and a solid surface. Films can be de-
Aluminum-doped ZnO ͑Al:ZnO͒ may provide a low cost alternative
to indium tin oxide ͑ITO͒ as a transparent conducting material for
posited by repeating the binary reaction sequence in an ABAB...
fashion.1
3,14
Over the last 10-15 years, ALD methods have been
5
3
54
15
flat panel displays and solar cells. The ZnAl O spinel is a
developed to deposit numerous materials. The resulting films are
usually dense, pinhole-free, and extremely conformal to the under-
lying substrate. ALD methods also have been used previously
2 3
5
5-57
58
widely used catalyst
an important semiconducting material for gas sensors.
catalysts or gas sensors may result from depositing ZnAl O or ZnO
and catalytic support material and ZnO is
59,60
Novel
1
,5,16-23
to deposit compound films including nanolaminates
2
3
2
,3,7,8,24-34
35-37
38-43
alloys,
complex oxides,
and doped materials.
on nanoporous substrates. ZnO and Al2O3 are both high bandgap
materials and are transmissive to visible light. Therefore,
ZnO/Al2O3 alloy films may benefit optoelectronic applications by
allowing the control of refractive index and surface roughness.
ZnO/Al O alloys are ideal for exploring the structure-property
2
3
relationships of ALD compound films. Well-established ALD
methods exist for depositing both ZnO4
4-47
and Al O
48-52
films.
2
3
ZnO/Al O alloy films with tunable resistivity may also improve the
ZnO ALD is performed using alternating Zn͑CH CH ) and H O
2
3
2
3
2
2
exposures44-47
reliability of microelectromechanical ͑MEMS͒ devices by prevent-
ing static charge buildup.
This investigation explored the physical properties of the
͑
A͒ ZnOH* ϩ Zn͑CH CH ͒ → ZnOZn͑CH CH )* ϩ CH CH
2
3
2
2
3
3
3
ZnO/Al O alloy films. A set of ZnO/Al O alloy films was pre-
2
3
2
3
͓
1͔
pared where the percentage of ZnO cycles used to prepare the
ZnO/Al2O3 alloys was varied from 0–100%. Ellipsometry, stylus
profilometry, atomic emission spectroscopy, atomic force micros-
copy, X-ray diffraction, four-point probe, mercury probe, and in situ
quartz crystal microbalance measurements were performed to evalu-
ate the film properties. These measurements enabled the determina-
tion of the refractive index, surface roughness, crystallinity, resistiv-
ity, and density of the ZnO/Al O alloy films vs. film composition.
͑
B͒ Zn͑CH CH ͒* ϩ H O → ZnOH* ϩ CH CH
͓2͔
2
3
2
3
3
where the asterisks represent the surface species. By repeating these
reactions in an ABAB... sequence, ZnO films can be deposited with
atomic layer control. Al O3 ALD is performed using alternating
2
48-52
Al͑CH ) and H O exposures
3
3
2
2
3
The influence of the percentage of ZnO cycles on the thickness and
͑
A͒ AlOH* ϩ Al͑CH ͒ → AlOAl͑CH )* ϩ CH
͓3͔
͓4͔
3
3
3
2
4
composition of ZnO/Al O alloy films was examined in a previous
2
3
6
1
study.
͑
B͒ AlCH* ϩ H O → AlOH* ϩ CH
3
2
4
Experimental
Viscous flow reactor for atomic layer deposition.—The
*
Electrochemical Society Active Member.
E-mail: Steven.George@Colorado.edu
z
ZnO/Al O alloy films were prepared in a viscous flow ALD reactor
2
3
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G340
Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
back surface of the sensor crystal. This nitrogen flow prevented
deposition on the back side of the sensor. A Maxtek TM400 film
thickness monitor interfaced to a computer measured the QCM sen-
sor signals. The thickness monitor and interface allowed mass mea-
Table I. Physical properties of ZnO and Al O ALD films pre-
pared in a viscous flow ALD reactor at 177°C. Surface roughness
2
3
„
rms… is for films deposited using 600 ALD cycles.
2
Physical property
ZnO
Al O3
surements with a 0.375 ng/cm resolution at 10 Hz. Assuming a ZnO
2
3
density of 5.6 g/cm , this mass resolution equates to a thickness
8 ϫ 10Ϫ3
2.00
5.62
ϳ1016
1.64
2.91
Resistivity ͑⍀ cm͒
resolution for ZnO of 0.007 Å.
Refractive index
3
Density ͑g/cm ͒
Ex situ measurements.—Elemental analysis was performed on
the ZnO/Al O alloy films deposited on Si͑100͒ by dissolving some
Crystalline phase
RMS surface roughness ͑nm͒
ALD growth rate ͑Å/cycle͒
Hexagonal
2.16
Amorphous
0.25
2
3
of each film in a 5% hydrofluoric acid/water mixture. The resulting
solutions were analyzed by inductively coupled plasma ͑ICP͒-
atomic emission spectroscopy ͑AES͒. The AES investigations deter-
mined the Zn and Al concentrations in each film. These AES mea-
surements allowed the molar Zn content to be evaluated using
2.01
1.29
assembled using stainless steel components and conflat seals.62 The
sample substrates were placed in the flow tube with a 1.4 in. inside
diameter that was resistively heated to 177°C. A flow of ϳ200 sccm
of filtered, ultrahigh purity nitrogen at a velocity of ϳ2.5 m/s and a
pressure of ϳ1 Torr transported the gaseous reactants to the sub-
strates and swept the reaction products into a mechanical pump. The
flow reactor was equipped with three independent reactant channels
employing computer-controlled solenoid valves to inject the diethyl
Zinc film content ͑%͒ ϭ ͓Zn/͑Zn ϩ Al͒]•100%
͓5͔
where Zn and Al are the atomic concentrations of these elements.
Thickness measurements were performed using a Dektak 3 stylus
profilometer. Steps were created in the ZnO/Al O alloy films using
2
3
one of two methods. For the films containing higher Zn percentages,
a line of Shipley AZ5214E photoresist was applied on top of the
alloy film and cured at 100°C for 60 s. Subsequently, the unpro-
tected film was etched away in 10% nitric acid. The photoresist was
zinc ͑DEZ͒, trimethyl aluminum ͑TMA͒, and H O precursors into
2
the nitrogen carrier gas. These studies employed Akzo Nobel semi-
conductor grade DEZ and TMA and Fisher Optima purity H O.
2
ZnO ALD and Al O ALD were performed using alternating
2
3
then removed by sonicating the sample in acetone. The ZnO/Al O3
2
DEZ/H O and TMA/H O exposures, respectively, as given by Eq.
2
2
alloy films containing у25% Al were insoluble in 10% nitric acid.
Although these films could be dissolved in hydrofluoric acid, the
hydrofluoric acid also dissolved the photoresist. Consequently, a
spot of Shipley AZ5214E photoresist was placed on the Si͑100͒
substrates prior to film growth. After the deposition of these alloy
films, the photoresist was removed by rubbing the sample with an
acetone-soaked cotton swab.
1
-4. Figure 1 illustrates the reactant pulse sequence employed to
deposit a ZnO/Al O3 alloy film with a ZnO cycle percentage of
2
6
7%. The DEZ and H O pulses alternate and every third DEZ pulse
2
is substituted with a TMA pulse. Consequently, 67% of the metal
alkyl pulses are diethyl zinc pulses. Throughout these experiments,
the shortest possible pulse sequence was used to obtain the desired
ZnO cycle percentage. By interposing the TMA pulses as often as
possible while maintaining the desired ZnO cycle percentage, the
resulting ZnO/Al O alloy should be nearly homogenous.
Sheet resistances for the ZnO/Al O alloy films deposited on the
2
3
glass substrates were measured with a four-point probe. To compute
the resistivity from the sheet resistances, the film thicknesses were
determined using the stylus profiler. The alloy films containing
у30% Al were too resistive to allow the four-point probe measure-
ments. Consequently, current-voltage measurements were performed
on these films deposited on n-doped Si͑100͒ substrates using a mer-
2
3
The ALD ZnO/Al O alloy films were deposited on glass and
2
3
n-doped Si͑100͒ substrates. Reactant exposure times were 1 s and
purge times varied between 2-5 s after reactant exposures. Prior to
deposition, the glass and Si substrates were degreased using a 15
min dip in a piranha solution containing 70 mL H SO and 30 mL of
6
3
cury microprobe and a picoammeter.
2
4
3
0% H O2 in H O. Subsequently, the Si͑100͒ substrates were
The surface topography of the alloy films was measured using a
Thermomicroscopes Autoprobe CP atomic force microscope in in-
termittent contact mode. 1ϫ1 ␮m images were recorded and the
root mean squared ͑rms͒ roughness was evaluated for each image as
2
2
etched in a clean room grade, 5% HF/H O solution for 1 min to
remove the native SiO layer.
2
2
The ALD viscous flow reactor was equipped with a quartz crystal
microbalance ͑QCM͒ allowing in situ film growth measurements.
Polished QCM sensors from Colorado Crystal Corporation ͑part no.
CCAT1BK-1007-000͒ were mounted in a Maxtek BSH-150 bake-
able sensor head attached to a 2.75 in. conflat flange. The sensor
head was modified to provide a nitrogen flow of ϳ20 sccm over the
6
described previously. The thicknesses and refractive indexes for the
alloy films deposited on Si͑100͒ substrates were measured using a
Rudolph Research Auto EL ellipsometer at a wavelength of 632.8
nm. The crystallinity of the alloy films was evaluated with X-ray
diffraction ͑XRD͒.
Results
Thickness, composition, and phase.—A set of ZnO/Al O alloy
2
3
films was deposited on Si͑100͒ substrates using a total of 600 ALD
cycles. The ZnO cycle percentage was varied from 0–100%. Figure
2
shows the thicknesses measured for the alloy films using ellipsom-
etry ͑solid circles͒ and stylus profilometry ͑open triangles͒. The
ZnO/Al O alloy film thickness is relatively constant at ϳ740-800
2
3
Å for ZnO cycle percentages below ϳ80%. Subsequently, the film
thickness increases sharply to ϳ1120-1270 Å for films using a ZnO
cycle percentage of у95%.
The average growth rates for pure ZnO and pure Al O ALD in
2
3
the viscous flow ALD reactor are 2.01 and 1.29 Å/cycle,
6
1
respectively. Consequently, the ZnO/Al O alloy film thicknesses
2
3
can be predicted using the rule of mixtures formula:
Thickness ϭ 600 cycles•͓͑2.01 Å/Cycle͒%ZnO
ϩ ͑1.29 Å/cycle͒͑100 Ϫ %ZnO͔͒/100
Figure 1. Diagram of ALD pulse sequence for depositing ZnO/Al O alloy
film where the percentage of ZnO cycles is 67%.
2
3
͓6͔
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
G341
Figure 4. X-ray diffractograms for ZnO/Al O alloy films vs. Zn content.
2
3
Figure 2. Thickness after 600 AB cycles vs. percentage of ZnO cycles used
to grow the ZnO/Al O alloy film. The thickness was measured using ellip-
2
3
sometry ͑solid circles͒ and stylus profiling ͑open triangles͒.
the AES measurements using Eq. 5. The results are given by the
solid squares in Fig. 3.
The expected Zn content can be calculated using the rule of
mixtures formula
where % ZnO is the percentage of ZnO cycles. The dashed line in
Fig. 2 shows the thicknesses predicted by Eq. 6. The measured
thicknesses are substantially below the predicted values over most
of the range of ZnO cycle percentages.
Zn film content ͑%͒ ϭ ␳Zn%ZnO/͑␳Zn%ZnO
ϩ ␳ ͑100 Ϫ %ZnO͒͒•100
͓7͔
Al
Portions of the ZnO/Al O alloy films prepared on Si͑100͒ sub-
2
3
strates using 600 AB cycles at 177°C were dissolved in hydrofluoric
acid. The Zn and Al concentrations in the solutions were determined
using ICP-AES. The Zn content for each film was calculated from
In this equation, ␳ ϭ 8.34 ϫ 1014 atom/cm2 and ␳Al ϭ 4.44
ϫ 10 atom/cm . These are the Zn and Al atomic densities depos-
Zn
14
2
ited during each ALD cycle for the pure ZnO and Al O com-
2
3
6
1
pounds, respectively. For ZnO cycle percentages р85%, the mea-
sured Zn content falls below the values predicted by Eq. 7. The gray
shaded region in Fig. 3 indicates the range of ZnO cycle percentages
where etching was observed during QCM measurements.61 This
etching was evident as a mass loss coincident with each TMA ex-
posure during the ZnO/Al O alloy film growth.
2
3
Figure 4 shows the X-ray diffractograms for the ZnO/Al O al-
2
3
loy films deposited using 600 AB cycles at 177°C on Si͑100͒ sub-
strates. The strong peak at 2⌰ ϭ 33° and the small peak at 62° in
the diffractograms arise from the Si͑100͒ substrate. The accepted
locations of the hexagonal ZnO and cubic ZnAl O4 diffraction
2
peaks are indicated by the solid lines with open circles and open
squares, respectively. The top trace shows the diffractogram for the
pure ZnO film.
The peaks at 2⌰ ϭ 32°, 34°, 36°, and 56° in Fig. 4 are all
characteristic of hexagonal ZnO. The height of the ZnO peak at 56°
increases slightly when the Zn content decreases from 100 to 99%.
As the Zn content decreases from 99 to 89%, the ZnO features
reduce in intensity, broaden, and shift to larger angles. Diffraction
peaks from the cubic phase of the ZnAl O spinel structure are
2
4
absent from all of the diffractograms. An unidentified diffraction
peak appears at 40° for the alloy films containing 95 and 91% Zn.
Electrical and optical properties.—The solid circles in Fig. 5
plot the resistivities of the ZnO/Al O alloy films prepared using
2
3
600 AB cycles at 177°C. The resistivities were measured using four-
point probe techniques vs. the Zn content as determined from the
Ϫ3
AES analysis. The resistivity of the pure ZnO films is 8.1 ϫ 10
⍀
Figure 3. Zn content vs. percentage of ZnO cycles used to grow the
cm. The resistivity decreases for small additions of Al and reaches a
minimum value of 2.2 ϫ 10 ⍀ cm at 98% Zn. The resistivity then
Ϫ3
ZnO/Al O alloy film. The Zn content was determined using ICP-AES.
2
3
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G342
Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
Figure 5. Four-point probe measurements of the resistivity of ZnO/Al O3
alloy films vs. Zn content.
Figure 7. Resistivity of ZnO/Al2O3 alloy films measured using the four-
point probe and the mercury probe.
2
increases for lower Zn content. The ZnO/Al O alloy resistivity in-
creases to 1.0 ϫ 10 ⍀ cm for films containing 76% Zn.
The resistivities of the ZnO/Al O alloy films can be estimated
2
3
2
3
2
from the I-V curves in Fig. 6. Most of the I-V curves show a
pronounced nonlinearity and the current densities measured at nega-
tive voltages are much lower than at positive voltages. In contrast,
the I-V curves measured using the mercury probe for films with
տ75% Zn content deposited on glass were linear and nearly sym-
metric with respect to the y axis. These mercury probe measure-
ments also yielded resistivities similar to those obtained using the
four-point probe. This similarity indicates a negligible contact resis-
tance at the mercury-alloy film interface. These findings suggest that
the asymmetry observed for the films deposited on Si͑100͒ arises
from a rectifying junction at the Si-alloy interface. This junction
may introduce a contact resistance in series with the alloy film re-
sistance, R ϭ RC ϩ R where R is the total resistance, RC is the
The ZnO/Al O alloy films with р75% Zn were not sufficiently
2
3
conducting to allow four-point probe measurements. Current-voltage
measurements were performed on these alloy films using a mercury
microprobe and picoammeter. Figure 6 displays current voltage
curves for five ZnO/Al O alloy films containing 82, 70, 64, 34, and
2
3
2.0% Zn deposited on n-doped Si͑100͒ substrates using 600 AB
cycles at 177°C. Figure 6 shows that the current density at a given
voltage decreases with decreasing Zn percentage.
T
F
T
contact resistance, and R is the film resistance.
F
The slope of the linear portion of the I-V curve for the 82% alloy
3
film at Ͼ2 V in Fig. 6 yields R ϭ 2.7 ϫ 10 ⍀. The resistance of
T
the 82% alloy film can be calculated using R ϭ ␳d/A where
F
␳
ϭ 7.9 ⍀ cm is the resistivity measured using the four-point
Ϫ8
probe, d ϭ 1030 ϫ 10 cm is the film thickness, and A ϭ 4.32
ϫ 10 cm is the area of the mercury probe. This equation yields
Ϫ2
2
R ϭ 0.019 ⍀. The contact resistance can then be determined using
F
3
RC ϭ R Ϫ R ϳ 2.7 ϫ 10 ⍀. Assuming that RC remains con-
T
F
stant with Zn content, the resistivities of the alloy films can be
calculated from the linear portions of the I-V curves in Fig. 6 after
3
subtracting RC ϳ 2.7 ϫ 10 ⍀. The results are displayed in Fig. 7.
Figure 7 demonstrates that the resistivity increases nearly exponen-
1
6
tially with decreasing Zn content to 1.7 ϫ 10 ⍀ cm at 2.0% Zn.
Figure 8 presents the refractive index values for the ZnO/Al O3
2
alloy films deposited on Si͑100͒ using 600 AB cycles at 177°C.
These values were obtained using ellipsometry. The refractive index
decreases monotonically from n ϭ 2.00 at 100% Zn to n ϭ 1.64
for the pure Al O film. The solid square at 33% Zn marks the
2
3
location of the refractive index of n ϭ 1.78 expected for cubic
6
4
ZnAl O spinel.
2
4
Figure 6. Current-voltage curves measured using the mercury probe for
Surface roughness and density.—The ZnO/Al O alloy films de-
2 3
ZnO/Al O alloy films with 82, 70, 64, 34, and 2.0% Zn content.
posited on Si͑100͒ substrates using 600 AB cycles at 177°C were
2
3
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
G343
Figure 8. Refractive index of ZnO/Al O alloy films vs. Zn content mea-
2
3
sured using ellipsometry.
analyzed using AFM techniques. Figure 9 shows the topography for
ϫ 1 ␮m AFM images for alloy films with compositions of 0, 76,
1
and 100% Zn. The light-to-dark ranges for these images are 2.0, 26,
and 19 nm, respectively. Root mean squared ͑rms͒ roughness values
calculated from AFM images of ZnO/Al O3 alloy films with Zn
2
contents ranging from 0-100% are shown in Fig. 10. The rms rough-
ness of the pure Al O sample is 0.25 nm. In comparison, the rms
2
3
roughness of the uncoated Si͑100͒ substrate is ϳ0.12 nm. The rms
roughness increases with increasing Zn content to 2.2 nm at 100%
Zn. In addition, a distinct peak appears at 76% Zn where the surface
roughness increases to 3.8 nm.
Densities for the ZnO/Al O alloy films were calculated by com-
2
3
bining the profilometry thickness measurements with the quartz
crystal microbalance ͑QCM͒ measurements. The thickness deposited
per AB cycle was obtained using the profilometer thickness mea-
surements in Fig. 2. The mass deposited per ALD cycle was ob-
tained using the in situ QCM to measure the mass increase for each
of the ZnO/Al O alloys during ϳ100 ALD cycles. Figure 11 plots
Figure 9. AFM images of ZnO/Al2O3 alloy films containing 0 ͑2.0 nm z
scale͒, 76 ͑26 nm z scale͒, and 100% ͑19 nm z scale͒ Zn.
2
3
the mass deposition rate obtained from the in situ QCM measure-
ments as a function of the Zn content determined from the AES
analysis. The mass deposition rate of the pure Al O film was 37.6
2
3
2
ng/cm /cycle. The mass deposition rate increases with increasing Zn
content and was 113 ng/cm /cycle for the 100% Zn film.
2
ϭ 113 ϫ 10Ϫ9 g/cm /cycle and m
2
In this equation, m
ZnO
2
Al O
2 3
The solid circles and solid line in Fig. 12 show the ZnO/Al O
ϭ 37.6 ϫ 10Ϫ9 g/cm /cycle are the masses deposited per ALD
cycle for pure oxide films obtained from QCM measurements.
2
3
alloy film densities obtained by dividing the mass deposited per
ALD cycle by the thickness deposited per ALD cycle. The density
61
Ϫ8
Ϫ8
GZnO ϭ 2.01 ϫ 10
cm/cycle and GAl2O3 ϭ 1.29 ϫ 10
cm/
3
of the 100% ZnO film is 5.62 g/cm . The density decreases with
3
cycle are the thicknesses deposited per ALD cycle for the pure
oxides. The dotted line in Fig. 12 shows the rule of mixtures
decreasing Zn content to 2.91 g/cm for the pure Al O film. In
61
2
3
addition, there is a pronounced reduction in the density between 70
3
density predicted for the ZnO/Al2O3 alloy films using Eq. 8. The Zn
content is obtained from % ZnO using the relationship between Zn
content and % ZnO shown in Fig. 3. The measured densities are
significantly lower than the densities predicted by Eq. 8 over most of
the range of Zn content.
and 82% Zn The density reduces to 3.16 g/cm at 76% Zn. In com-
parison, the solid square at 33% Zn content marks the location of the
3
64
4
.58 g/cm density expected for cubic ZnAl O spinel.
2 4
The expected density for the ZnO/Al O alloy films can be cal-
2
3
culated using the rule of mixtures formula
The alloy film densities can also be predicted from the measured
refractive indexes in Fig. 8 using Lorentz-Lorenz analysis.6
5,66
The
mZnO% ZnO ϩ m
͑100 Ϫ % ZnO͒
͑100 Ϫ % ZnO͒
3
Lorentz-Lorenz analysis evaluates the refractive index of a com-
pound film from the individual molar refractivities, A, of the com-
ponents. The molar refractivity of a single component is given by
Al O
2
3
Density ϭ
͓8͔
GZnO% ZnO ϩ G
Al O
2
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G344
Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
Figure 11. Mass deposition rate of ZnO/Al O3 alloy films vs. Zn content
determined by in situ QCM measurements.
2
Figure 10. RMS surface roughness of ZnO/Al O alloy films vs. Zn content
determined by AFM measurements.
2
3
The smaller thicknesses are explained, in part by the nucleation
behavior that occurs when making the transition between ZnO and
2
w͑n Ϫ 1͒
6
1
A ϭ
͓9͔
Al O . Both ZnO ALD growth on Al O and Al-doped ZnO sur-
2
2
3
2
3
␳
͑n ϩ 2͒
faces and Al O ALD growth on ZnO and Zn-doped Al O surfaces
2
3
2
3
require numerous AB cycles to obtain the bulk ZnO and Al O ALD
growth rates.
2
3
where w is molecular weight, ␳ is density, and n is refractive index.
The molar refractivity of a ZnO/Al O alloy film is given by
2
3
The Zn content shown in Fig. 3 is also below the rule of mixtures
prediction given by Eq. 7 for most of the ZnO/Al O alloy films.
2
2
3
wFilm͑n Ϫ 1͒
The lower than expected Zn content is attributed to the etching of Zn
AFilm ϭ XZnOAZnO ϩ ͑1 Ϫ XZnO͒A
ϭ
Al O
2
3
2
61
␳Film͑n ϩ 2͒
by TMA. This etching is believed to occur by the reaction
͓10͔
In this expression, XZnO is the mole fraction of ZnO in the film and
wFilm and ␳Film are the average molecular weight and density of the
film, respectively.
The mole fraction of ZnO in the alloy film can be obtained from
XZnO ϭ Zn/͑Zn ϩ 2Al) using the Zn and Al concentrations in the
alloy films determined by the AES analysis. AZnO and A
were
3
Al O
2
determined from Eq. 9 using the refractive index and density mea-
sured for pure ZnO and Al O films, respectively. The molar refrac-
2
3
tivities were AZnO ϭ 7.24 and A
ϭ 12.62. The molecular
Al O
2
3
weight for the film, wFilm was calculated using XZnO and the molecu-
lar weights of ZnO and Al O .
2
3
Equation 10 can be rearranged to predict the density of the alloy
film as a function of the measured refractive indexes. The densities
predicted for the ZnO/Al O alloy films using the Lorentz-Lorenz
2
3
analysis are given by the dashed line in Fig. 12. The Lorentz-Lorenz
formula predicts the alloy densities reasonably well between 0–65%
Zn. However, Eq. 10 fails to account for the distinct reduction in
density at 76% Zn. Equation 10 also underestimates the densities of
the alloy films between 85-98% Zn.
Discussion
Thickness, composition, and phase.—Figure 2 shows that the
thicknesses measured for many of the ZnO/Al O alloy films are
2
3
significantly below the thicknesses predicted by the rule of mixtures
formula given by Eq. 6. The alloy films are thinner than the expec-
tations from the growth rates for pure ZnO and Al O ALD films.
Figure 12. Density vs. Zn content for ZnO/Al O alloy films determined by
QCM and stylus profiler measurements.
2
3
2
3
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
G345
ZnOH* ϩ Al͑CH ) → Al͑OH͒͑CH )* ϩ Zn͑CH )
͓11͔
the ellipsometric data using the known surface roughnesses pre-
sented in Fig. 10.
3
3
3
3 2
The refractive index measured for the 34% alloy film in Fig. 8 is
n ϭ 1.66. This value is significantly below the literature value of
n ϭ 1.78 for the ZnAl2O4 spinel. This large difference suggests that
the spinel phase does not form under the ALD conditions used to
prepare the ZnO/Al O alloy films. The X-ray diffractograms in Fig.
The gray shaded region in Fig. 3 indicates the range of ZnO cycle
percentages where etching was observed during QCM measure-
ments. This etching region coincides with the distinct reduction in
Zn content in Fig. 3 and alloy film thickness in Fig. 2 compared with
the rule of mixtures predictions.
2
3
4 also provide no evidence for the cubic ZnAl O phase. In addition,
2
4
Previous investigations have established that ZnO ALD grows in
Fig. 12 shows that the density of the 34% alloy film is well below
the hexagonal crystalline phase.4
5,47
In agreement with these previ-
3
the value of 4.58 g/cm expected for the ZnAl O spinel.
2
4
ous studies, the diffractogram for the film with 100% ZnO shown in
Fig. 4 displays strong peaks at 2⌰ ϭ 32° and 56° that are consis-
tent with hexagonal ZnO. The 2⌰ ϭ 56° peak height increases
slightly when the Al content increases to ϳ1%. This increase may
result from a slightly thicker film after 600 AB cycles at 177°C and
larger ZnO nanocrystals in the film. In support of this explanation,
the profilometry measurements in Fig. 2 show a 7% thickness in-
crease from 1185 to 1269 Å when the ZnO cycle percentage is
reduced from 100 to 97.5%.
Additional measurements could provide a more detailed under-
standing of the electrical and optical behaviors of the ZnO/Al O3
2
alloy films. For instance, optical reflectivity measurements could
probe the variation in bandgap with Zn content allowing an exami-
nation of the transition between semiconducting and insulating be-
havior. Capacitance measurements performed using the mercury
probe could determine the variation in permittivity of the
ZnO/Al O alloy films vs. the Zn content. The resistivity values in
2
3
Fig. 7 derived from the mercury probe measurements assume a con-
stant contact resistance at the interface between the Si substrate and
the alloy film. This assumption could be verified by performing
The ZnO diffraction peaks in Fig. 4 decrease in height and in-
crease in width with increasing Al content. This behavior suggests a
loss of crystallinity. The ZnO/Al O alloy films appear to be amor-
2
3
supplementary resistance measurements on ZnO/Al O3 alloy film
2
phous for Zn contents Շ81%. These observations are consistent
test structures employing vapor deposited Ti/Au electrodes. Ti/Au
electrodes form an Ohmic contact with a very low contact resistance
4
7
with the previous findings that ZnO ALD is crystalline and Al O
2
3
ALD is amorphous67 under these growth conditions. In the Zn con-
tent range of 100-89% where the alloy films display crystallinity, the
ZnO diffraction peaks shift to larger angles with increasing Al con-
tent. According to Bragg’s law, n␭ ϭ 2d sin ⌰, a decrease in lattice
spacing, d, will lead to an increase in the detected peak position, ⌰.
71
when vapor deposited onto Al-doped ZnO films.
Surface roughness and density.—Previous investigations have
shown that ZnO ALD films roughen considerably with increasing
6
thickness as a result of the growth of ZnO nanocrystals. In contrast,
2
ϩ
3ϩ
The Zn ionic radius is 0.72 Å and the Al ionic radius is 0.53 Å.
As predicted by Bragg’s law and the smaller ionic radius of Al
Al O ALD films deposit in an amorphous form and remain remark-
2
3
3
ϩ
6
,
ably smooth with increasing thickness. The surface roughness of a
particular ZnO film thickness can be made progressively smoother
by interleaving Al O layers in the ZnO film to increase the number
the ZnO XRD peaks shift to larger angles with increasing Al con-
tent.
2
3
6
of ZnO/Al O interfaces. The progressive smoothing of the film
2
3
Electrical and optical properties.—Figure 5 shows that the resis-
surface is accompanied by a simultaneous reduction in the intensity
of the ZnO diffraction peaks observed in grazing incidence XRD
tivity of the ZnO/Al O alloy film decreases when doped with small
2
3
amounts of Al. This phenomenon has been observed
72
measurements. These results suggest that the ZnO nanocrystal
previously.4
4,53,68
The improved conductivity of the ZnO films re-
growth is interrupted by the Al O layers. This interruption results
44,53,68
2ϩ
2
3
sults from the n-doping of ZnO by Al ions.
When a Zn ion
in smoother, more amorphous films.
in the ZnO lattice is replaced by an Al3 ion, the Al3ϩ contributes
an extra valence electron as a charge carrier. This theory is sup-
ported by Hall probe measurements of Al:ZnO ALD films that ex-
hibit an increase in charge carrier density for small Al
ϩ
In agreement with these previous results, Fig. 10 shows that the
ZnO/Al O alloy films become smoother with increasing Al content.
2
3
However, the prominent peak in rms roughness at 76% Zn content is
unexpected and may indicate the formation of a new crystalline
phase. Contrary to this expectation, the XRD measurements in Fig.
4
4,53,68
percentages.
The XRD results shown in Fig. 4 also reveal that the Al3 dop-
ants perturb the ZnO lattice. This perturbation creates scattering
sites that decrease the electron mobility and ultimately increase the
resistivity. In addition, excess Al doping can create nonconducting
ϩ
4
indicate that the ZnO/Al O alloy films are amorphous at 76% Zn.
2 3
An alternative explanation for the peak in rms roughness shown
in Fig. 10 can be found in the density measurements of Fig. 12.
Expanded views of these two data sets are superimposed in Fig. 13.
The open squares and dotted line plot the rms roughness values. The
solid circles and solid line display the density measurements. Figure
6
9,70
Al O clusters that do not produce electron donors.
When the
2
3
Zn content is Ͻ98%, Fig. 5 and 7 show that the resistivity of the
ZnO/Al O alloy film increases nearly exponentially with decreas-
2
3
1
3 clearly shows that the peak in rms roughness coincides with the
1
6
ing Zn content. The resistivity of the 2.0% Zn film is 1.7 ϫ 10
⍀
reduction in density. Furthermore, these two features are fairly well
bracketed within the range of compositions where etching was ob-
cm. This value is similar to the resistivity measured for pure Al O
2
3
ALD films.63 The presence of only a few percent of Zn in Al O has
61
2
3
served by the in situ QCM measurements. This range is indicated
a negligible effect on the resistivity of the Al O films.
2
3
by the dashed lines in Fig. 13.
The refractive index of the ZnO/Al O3 alloy films decreases
The increased roughness and decreased density may result from
2
monotonically and relatively smoothly from n ϭ 2.00 to n ϭ 1.64
with decreasing Zn content as shown in Fig. 8. The refractive in-
dexes measured for the pure ZnO and Al O ALD films agree well
the etching process. The etching of the ZnO/Al O alloy films by the
2
3
TMA pulses appears to remove selectively Zn atoms. This selective
removal results in a sudden drop in Zn content below the rule of
mixtures estimate shown in Fig. 3. The etching may occur by the
reaction given in Eq. 11. The removal of Zn atoms will leave voids
that roughen the film surface.
A line scan analysis of the AFM image for the 76% Zn alloy film
in Fig. 9 reveals peak-to-valley thickness changes that are Ͼ200 Å
within a lateral spacing of ϳ100 nm. These topographical changes
are too abrupt to be detected accurately by the stylus profilometer
probe tip. The profilometer tip will ride over these surface features
and yield thicknesses that are too large. These larger thicknesses
2
3
6
48
with previous measurements of n ϭ 1.95 for ZnO and n ϭ 1.65
for Al O ALD films. There is a rapid drop in refractive index
2
3
between 100–95% Zn. This reduction is not accompanied by a si-
multaneous reduction in density in Fig. 12. However, the crystallin-
ity decreases significantly for the alloy films between 100-95% Zn
content as shown in Fig. 4. These findings suggest that the refractive
indexes of these Zn-rich films are more affected by crystallinity than
density. The refractive index and thickness values determined using
the ellipsometer could be determined more accurately by modeling
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G346
Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
Acknowledgments
This work was funded by the Air Force Office of Scientific Re-
search ͑AFOSR͒ and Nanomaterials Research, LLC. Professor John
Drexler of the Department of Geological Sciences at the University
of Colorado performed the ICP-AES elemental analysis of the alloy
films. The authors thank Tom Seidel and Ofer Sneh of Genus, Inc.,
for donating equipment used to construct the viscous flow ALD
reactor.
University of Colorado assisted in meeting the publication costs of this
article.
References
1
2
3
4
. H. Kattelus, M. Ylilammi, J. Saarilahti, J. Antson, and S. Lindfors, Thin Solid
Films, 225, 296 ͑1993͒.
. M. Stromme, G. A. Niklasson, M. Ritala, M. Leskela, and K. Kukli, J. Appl. Phys.,
9
0, 4532 ͑2001͒.
. H. Fujiwara, T. Nabeta, and I. Shimizu, Jpn. J. Appl. Phys., Part 1, 33, 2474
1994͒.
. X. Chu, M. S. Wong, W. D. Sproul, and S. A. Barnett, J. Mater. Res., 14, 2500
1999͒.
͑
͑
5. K. Kukli, M. Ritala, and M. Leskela, J. Appl. Phys., 86, 5656 ͑1999͒.
6. J. W. Elam, Z. A. Sechrist, and S. M. George, Thin Solid Films, 414, 43 ͑2002͒.
7. M. Vehkamaki, T. Hatanpaa, T. Hanninen, M. Ritala, and M. Leskela, Electrochem.
Solid-State Lett., 2, 504 ͑1999͒.
8
9
. J. Ihanus, M. Ritala, M. Leskela, and E. Rauhala, Appl. Surf. Sci., 112, 154 ͑1997͒.
. H. Y. Bae and G. M. Choi, Sens. Actuators B, 55, 47 ͑1999͒.
Figure 13. Comparison of density and surface roughness ͑rms͒ of
ZnO/Al O alloy films vs. Zn content between 64-100% Zn.
1
0. J. H. Yu and G. M. Choi, Sens. Actuators B, 75, 56 ͑2001͒.
2
3
11. S. Sharafat, A. Kobayashi, V. Ogden, and N. M. Ghoniem, Vacuum, 59, 185 ͑2000͒.
1
1
1
1
2. F. Wang, L. Wang, and G. Liu, Surf. Eng., 17, 35 ͑2001͒.
3. T. Suntola and J. Hyvarinen, Annu. Rev. Mater. Sci., 15, 177 ͑1985͒.
4. S. M. George, A. W. Ott, and J. W. Klaus, J. Phys. Chem., 100, 13121 ͑1996͒.
5. M. Ritala and M. Leskela, in Handbook of Thin Film Materials, H. S. Nalwa,
Editor, Academic Press, San Diego, CA ͑2001͒.
would yield smaller densities. This explanation partially accounts
for the reduction in density at 76% Zn shown in Fig. 13.
16. M. Ritala, M. Leskela, L. Niinisto, T. Prohaska, G. Friedbacher, and M. Grasser-
bauer, Thin Solid Films, 249, 155 ͑1994͒.
The Lorentz-Lorenz density prediction is given by the dashed
line in Fig. 12. This prediction is based on the relationship between
1
1
7. K. Kukli, J. Ihanus, M. Ritala, and M. Leskela, Appl. Phys. Lett., 68, 3737 ͑1996͒.
8. J. M. Hartmann, M. Charleux, H. Mariette, and J. L. Rouviere, Appl. Surf. Sci., 112,
6
5,66
the density and refractive index as expressed by Eq. 9 and 10.
If
1
42 ͑1997͒.
19. M. Ishii, S. Iwai, H. Kawata, T. Ueki, and Y. Aoyagi, J. Cryst. Growth, 180, 15
1997͒.
0. H. Kumagai, K. Toyoda, K. Kobayashi, M. Obara, and Y. Iimura, Appl. Phys. Lett.,
0, 2338 ͑1997͒.
an artifact in the stylus profilometer measurements caused the mea-
sured densities to be low, the measured densities may be below the
Lorentz-Lorenz estimates. However, some of the measured densities
are larger than the densities obtained from the refractive indexes. In
particular, the measured density values are significantly higher than
the Lorentz-Lorenz estimates between ϳ85-99% Zn. These obser-
vations argue against the possible stylus profilometer artifact.
͑
2
2
7
1. H. Zhang, R. Solanki, B. Roberds, G. Bai, and I. Banerjee, J. Appl. Phys., 87, 1921
͑2000͒.
22. K. Kukli, M. Ritala, and M. Leskela, J. Electrochem. Soc., 148, F35 ͑2001͒.
2
2
3. H. Zhang and R. Solanki, J. Electrochem. Soc., 148, F63 ͑2001͒.
4. B. T. McDermott, K. G. Reid, N. A. El-Masry, S. M. Bedair, W. M. Duncan, X.
Yin, and F. H. Pollak, Appl. Phys. Lett., 56, 1172 ͑1990͒.
A detailed examination of the ZnO/Al O alloy film properties
2
3
was conducted between 70-82% Zn. This investigation revealed that
the reduction in density at ϳ76% Zn results from a combination of
an increase in measured thickness and a simultaneous drop in mass
deposition rate. This combination suggests that etching may both
lower the mass deposition rate and substantially increase the free
volume of the ZnO/Al O alloy films.
25. H. Kattelus, M. Ylilammi, J. Salmi, T. Ranta-Aho, E. Nykanen, and I. Suni, Mater.
Res. Soc. Symp. Proc., 284, 511 ͑1993͒.
2
6. H. Fujiwara, T. Nabeta, H. Kiryu, and I. Shimizu, Jpn. J. Appl. Phys., Part 1, 33,
381 ͑1994͒.
7. T. Asikainen, M. Ritala, and M. Leskela, J. Electrochem. Soc., 142, 3538 ͑1995͒.
4
2
28. K. S. Boutros, F. G. McIntosh, J. C. Roberts, S. M. Bedair, E. L. Piner, and N. A.
El-Masry, Appl. Phys. Lett., 67, 1856 ͑1995͒.
2
3
2
9. H. Seim, M. Nieminen, L. Niinisto, H. Fjellvag, and L.-S. Johansson, Appl. Surf.
Sci., 112, 243 ͑1997͒.
Conclusions
ZnO/Al O alloys were grown using ZnO ALD with
30. H. Seim, H. Molsa, M. Nieminen, H. Fjellvag, and L. Niinisto, J. Mater. Chem., 7,
49 ͑1997͒.
31. O. Nilsen, M. Peussa, H. Fjellvag, L. Niinisto, and A. Kjekshus, J. Mater. Chem.,
, 1781 ͑1999͒.
2. J. H. Song, E. D. Sim, K. S. Baek, and S. K. Chang, J. Cryst. Growth, 214, 460
2000͒.
4
2
3
9
Zn͑CH CH ) /H O and Al O ALD with Al͑CH ) /H O. The
2
3
2
2
2
3
3
3
2
3
composition of the ZnO/Al O alloy films was controlled by adjust-
2
3
͑
ing the relative number of ZnO ALD and Al O ALD reaction
33. M. Schuisky, K. Kukli, M. Ritala, A. Harsta, and M. Leskela, Chem. Vap. Deposi-
tion, 6, 139 ͑2000͒.
2
3
cycles in the pulse sequence. Alloys were grown with physical char-
acteristics that could be tuned over the full range of values defined
3
4. M. Juppo, P. Alen, M. Ritala, and M. Leskela, Chem. Vap. Deposition, 7, 211
2001͒.
͑
by pure ZnO and Al O . A variety of film properties was investi-
2
3
35. S. Haukka, M. Lindblad, and T. Suntola, Appl. Surf. Sci., 112, 23 ͑1997͒.
36. M. Nieminen, T. Sajavaara, E. Rauhala, M. Putkonen, and L. Niinisto, J. Mater.
Chem., 11, 2340 ͑2001͒.
gated vs. Zn content including the film growth rate, refractive index,
surface roughness, crystallinity, resistivity, and density. The refrac-
tive indexes varied from n ϭ 2.00 for pure ZnO to n ϭ 1.64 for
pure Al O . The resistivities of the ZnO/Al O alloy films ranged
3
3
7. A. Rahtu, T. Hanninen, and M. Ritala, J. Phys. IV, 11, 923 ͑2001͒.
8. T. Suntola and A. Jorma, U.S. Pat. 4,058,430 ͑1977͒.
39. H. Viirola and L. Niinisto, Thin Solid Films, 251, 127 ͑1994͒.
40. W. Kong, J. Fogarty, and R. Solanki, Appl. Phys. Lett., 65, 670 ͑1994͒.
2
3
2
3
Ϫ3
16
from ϳ10 ⍀ cm for the 98% Zn film to ϳ10 ⍀ cm for the 2.0%
alloy film. This remarkable span of over 18 orders of magnitude for
the resistivity may be useful in many thin-film electrical applica-
tions. Anomalies in rms surface roughness and density were ob-
served for the alloy films with a Zn content of ϳ76%. These anoma-
4
4
4
1. K.-E. Elers, M. Ritala, M. Leskela, and E. Rauhala, Appl. Surf. Sci., 82Õ83, 468
͑
1994͒.
2. D. Jung, M. Leonard, N. E. El-Masry, and S. M. Bedair, J. Electron. Mater., 24, 75
͑1995͒.
3. E. B. Yousfi, B. Weinberger, F. Donsanti, P. Cowache, and D. Lincot, Thin Solid
Films, 387, 29 ͑2001͒.
lies may be partially explained by the etching of Zn by Al͑CH3)
3
44. V. Lujala, J. Skarp, M. Tammenmaa, and T. Suntola, Appl. Surf. Sci., 82Õ83, 34
during the ZnO/Al O alloy film growth.
͑1994͒.
2
3
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑6͒ G339-G347 ͑2003͒
G347
4
4
4
4
5. A. Yamada, B. Sang, and M. Konagai, Appl. Surf. Sci., 112, 216 ͑1997͒.
6. A. W. Ott and R. P. H. Chang, Mater. Chem. Phys., 58, 132 ͑1999͒.
7. E. B. Yousfi, J. Fouache, and D. Lincot, Appl. Phys. Lett., 153, 223 ͑2000͒.
8. A. W. Ott, J. W. Klaus, J. M. Johnson, and S. M. George, Thin Solid Films, 292,
60. J. Q. Xu, Y. A. Shun, Q. Y. Pan, and J. H. Qin, Sens. Actuators B, 66, 161 ͑2000͒.
61. J. W. Elam and S. M. George, Chem. Mater., 15, 1020 ͑2003͒.
62. J. W. Elam, M. D. Groner, and S. M. George, Rev. Sci. Instrum., 73, 2981 ͑2002͒.
63. M. D. Groner, J. W. Elam, F. H. Fabreguette, and S. M. George, Thin Solid Films,
1
35 ͑1997͒.
4
13, 186 ͑2002͒.
4
5
9. G. S. Higashi and C. G. Fleming, Appl. Phys. Lett., 55, 1963 ͑1989͒.
0. M. Ritala, H. Saloniemi, M. Leskela, T. Prohaska, G. Friedbacher, and M. Grass-
erbauer, Thin Solid Films, 286, 54 ͑1996͒.
6
6
4. CRC Handbook of Chemistry and Physics, 63rd ed., R. C. Weast and M. J. Astle,
Editors, CRC Press, Inc., Boca Raton, FL ͑1982-1983͒.
5. C. J. F. Bottcher, Theory of Electric Polarization, Elsevier Publishing, Amsterdam
5
5
1. P. Ericsson, S. Bengtsson, and J. Skarp, Microelectron. Eng., 36, 91 ͑1997͒.
2. R. Matero, A. Rahtu, M. Ritala, M. Leskela, and T. Sajavaara, Thin Solid Films,
͑
1952͒.
6
6
6. M. Born and E. Wolf, Principles of Optics, Pergamon Press, Oxford ͑1975͒.
7. A. Paranjpe, S. Gopinath, T. Omstead, and R. Bubber, J. Electrochem. Soc., 148,
G465 ͑2001͒.
3
68, 1 ͑2000͒.
3. H. Sato, T. Minami, S. Takata, T. Miyata, and M. Ishii, Thin Solid Films, 236, 14
1993͒.
5
5
͑
6
8. H. Kim, A. Pique, J. S. Horwitz, H. Murata, Z. H. Kafafi, C. M. Gilmore, and D. B.
Chrisey, Thin Solid Films, 377-378, 798 ͑2000͒.
4. B. Canava, J.-F. Guillemoles, E.-B. Yousfi, P. Cowache, H. Kerber, A. Loeffl, H.-W.
Schock, M. Powalla, D. Hariskos, and D. Lincot, Thin Solid Films, 361-362, 187
6
7
9. J. Hu and R. G. Gordon, J. Appl. Phys., 71, 880 ͑1992͒.
0. A. F. Aktaruzzaman, G. L. Sharma, and L. K. Malhotra, Thin Solid Films, 198, 67
͑2000͒.
5
5
5
5
5. T. K. Shioyama, U.S. Pat. 4,260,845 ͑1981͒.
6. R. L. Cobb, U.S. Pat. 4,568,784 ͑1986͒.
7. B. M. Welch, U.S. Pat. 4,692,430 ͑1987͒.
8. J. Wrzyszcz, M. Zawadzki, J. Trawczynski, H. Grabowska, and W. Mista, Appl.
Catal. A, 210, 263 ͑2001͒.
͑
1991͒.
71. H.-K. Kim, S.-H. Han, T.-Y. Seong, and W.-K. Choi, J. Electrochem. Soc., 148,
G114 ͑2001͒.
72. J. M. Jensen, A. B. Oelkers, R. Toivola, D. C. Johnson, J. W. Elam, and S. M.
George, Chem. Mater., 14, 2276 ͑2002͒.
5
9. J. Q. Xu, Q. Y. Pan, Y. A. Shun, and Z. Z. Tian, Sens. Actuators B, 66, 277 ͑2000͒.
Downloaded on 2014-08-12 to IP 130.63.180.147 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).