Cyclic chemical-vapor-deposited TiO2/Al2O3 film using trimethyl aluminum, tetrakis(diethylamino)titanium, and O 2

Article abstract of DOI:10.1149/1.2744136

Titanium aluminum oxide films have been studied as potential alternative gate dielectrics. However, most studies have focused on sputtered films. In this study, we demonstrate that a combination of tetrakis(diethylamino)titanium, trimethyl aluminum, oxygen, and cyclic chemical vapor deposition (CVD) is a promising approach for laminated Ti O2 Al2 O3 films with low impurities and high thermal stability even at low temperatures. The growth of the films is carried out in a cold-wall CVD chamber at 300°C and 0.7 Torr. Our studies show that the properties of Ti O2 improve with the addition of even a few percent of Al2 O3. X-ray diffraction analyses indicate that as-deposited Ti O2 Al2 O3 films have amorphous structure. Upon annealing as-deposited films in Ar at 700°C for 5 min, Ti O2 Al2 O3 films maintain their amorphous structure, while pure Ti O2 films crystallize at these conditions. Atomic force microscopy shows that the surfaces of Ti O2 Al2 O3 films are smoother than those of Ti O2 films deposited at the same conditions. Even though annealing increases the roughness of the Ti O2 Al2 O3 films, film roughness is still significantly lower than that of as-deposited Ti O2 films. Moreover, Rutherford backscattering spectroscopy and X-ray photoelectron spectroscopy show that there is no detectable formation of interfacial silicon oxide and negligible carbon impurity in as-deposited Ti O2 Al2 O3 films.

Full text of DOI:10.1149/1.2744136

Journal of The Electrochemical Society, 154 ͑8͒ G177-G182 ͑2007͒
G177
0013-4651/2007/154͑8͒/G177/6/$20.00 © The Electrochemical Society
Cyclic Chemical-Vapor-Deposited TiO2/Al2O3 Film Using
Trimethyl Aluminum, Tetrakis(diethylamino)titanium, and O2
,z
Xuemei Songa and Christos G. Takoudisa,b,
*
aDepartment of Chemical Engineering and bDepartment of Bioengineering, University of Illinois at
Chicago, Chicago, Illinois 60607, USA
Titanium aluminum oxide films have been studied as potential alternative gate dielectrics. However, most studies have focused on
sputtered films. In this study, we demonstrate that a combination of tetrakis͑diethylamino͒titanium, trimethyl aluminum, oxygen,
and cyclic chemical vapor deposition ͑CVD͒ is a promising approach for laminated TiO2/Al2O3 films with low impurities and high
thermal stability even at low temperatures. The growth of the films is carried out in a cold-wall CVD chamber at 300°C and
0.7 Torr. Our studies show that the properties of TiO2 improve with the addition of even a few percent of Al2O3. X-ray diffraction
analyses indicate that as-deposited TiO2/Al2O3 films have amorphous structure. Upon annealing as-deposited films in Ar at 700°C
for 5 min, TiO2/Al2O3 films maintain their amorphous structure, while pure TiO2 films crystallize at these conditions. Atomic
force microscopy shows that the surfaces of TiO2/Al2O3 films are smoother than those of TiO2 films deposited at the same
conditions. Even though annealing increases the roughness of the TiO2/Al2O3 films, film roughness is still significantly lower than
that of as-deposited TiO2 films. Moreover, Rutherford backscattering spectroscopy and X-ray photoelectron spectroscopy show
that there is no detectable formation of interfacial silicon oxide and negligible carbon impurity in as-deposited TiO2/Al2O3 films.
© 2007 The Electrochemical Society. ͓DOI: 10.1149/1.2744136͔ All rights reserved.
Manuscript submitted January 26, 2007; revised manuscript received April 6, 2007. Available electronically June 7, 2007.
The continuous decrease of transistor feature sizes and related
concerns of high tunneling leakage current and low gate capacitance
of ultrathin SiO2 have been driving the development of alternative
dielectric materials with higher permittivity.1 Alumina ͑Al2O3͒ is a
promising gate dielectric material. It is stable on Si and remains
amorphous up to high temperatures. Also, it has the largest bandgap
͑8.8 eV͒ next to SiO2 and has high band offset with respect to Si
͑2.8 eV for conduction band offset and 4.9 eV for valence band
offset͒.1 However, its dielectric constant, k, ͑in the range of 8–10͒
makes it a relatively short-term solution for industry needs.1 TiO2
films have attracted a lot of attention due to their higher dielectric
constant, up to 80.2 However, issues such as several stable oxidation
states,1 low bandgap ͑3.0–3.5 eV͒,3 low crystallization temperature
͑around 300–400°C͒,4 and instability on Si5 induce high leakage
current and impede the application of TiO2 as gate dielectric.
Recently, many studies focus on heterogeneous materials, espe-
cially silicates and aluminates,6 which may modify film properties
and overcome the limitations of TiO2 films. Among these composite
films, titanium aluminum oxide thin films have emerged as one of
the promising alternative high-k materials. It has been reported that
alloying titanium oxide and aluminum oxide yields TiAlO materials
titanium tetrachloride ͑TiCl4͒, and CO2/H2 at 450°C have been re-
ported to have a dielectric constant between 28 and 75.12 However,
the use of halide precursor, i.e., TiCl4, may result in corrosive Cl
contamination. ALD Al2O3–TiO2 nanolaminates ͑11% Al2O3͒ with
a dielectric constant of 45 have been grown using trimethyl alumi-
num ͑TMA͒, tetraisopropyl titanium ͑TTIP͒, and water at 250°C.13
Kim and Yun13 have reported that the addition of Al2O3 inhibits
crystallization of as-deposited ALD TiO2 films without further study
on phase transition temperature, which is one of the important char-
acteristics that determines the applicability of the material, because
high temperature is used in the integration processes of complemen-
tary metal oxide semiconductor ͑CMOS͒ devices.
In this study, cyclic CVD TiO2/Al2O3 films were deposited on
Si͑100͒ using tetrakis͑diethylamino͒titanium ͑TDEAT͒, TMA, and
oxygen at 300°C. TMA is a promising Al-containing precursor and
has been extensively studied for CVD or ALD Al2O3 films. Amide-
based titanium precursors, i.e., TDEAT, have been used mainly for
titanium nitride deposition. Amide-based metal precursors are prom-
ising for low-temperature film deposition due to their high reactivity,
which induces lower impurity incorporation even at low deposition
temperatures.14-17 We have demonstrated that TDEAT is a promising
precursor for TiO2 film deposition.18 Here, we further explore it for
TiO2/Al2O3 film deposition. Moreover, cyclic CVD is utilized in
this study; this not only allows easy control of the Ti/Al ratio in the
films but also lowers impurity incorporation by reducing possible
gas-phase reactions when compared to that in traditional CVD pro-
cesses. The focus of this study is to obtain high-k TiO2/Al2O3 films
at low deposition temperatures with high crystallization tempera-
tures, low impurity levels, and negligible interfacial reactions, which
can lead to low leakage current.
7,8
that retain a high permittivity close to that of TiO2 and also have
excellent thermal stability like that of Al2O3.8,9 In addition, the high
oxygen affinity of Al2O3 decreases the reduction reaction of Ti4+ to
Ti3+ and the formation of oxygen vacancies.8 The addition of Al2O3
also enhances the adhesion of TiO2 to the substrate and the film does
not display an orientation polarization under an electric field.10
Crystallization retardation due to the alloying of Al2O3 into TiO2
formed by radio-frequency magnetron sputtering7,10 and oxidation
of sputtered TiAl11 has also been reported.
Even though the studies mentioned above have shown encourag-
ing results, most titanium aluminum oxide films were deposited us-
ing sputtering methods. The disadvantage of a sputter process is the
directional nature of the process, which does not satisfy conformal-
ity requirements. On the contrary, chemical vapor deposition ͑CVD͒
or atomic layer deposition ͑ALD͒, which provide better step cover-
age, are promising methods in integration processes of nonplanar
topographies. The few studies on the CVD and ALD titanium alu-
minum oxide films have shown encouraging results in obtaining
high-permittivity films. Amorphous low-pressure CVD ͑LPCVD͒
TiAlOx films from a mixture of aluminum tri-sec-butoxide ͑ATSB͒,
Experimental
TDEAT and TMA precursors were introduced into the reactor by
bubbling Ar carrier gas at 20 and 15 sccm, respectively. Due to the
low vapor pressure of TDEAT, its bubbler was maintained at 40°C
and 80–320 Torr, depending on the desired flow rate of TDEAT. The
TMA bubbler was kept at 760 Torr and 25°C. In order to prevent
the condensation of the precursors, the temperatures of TMA and
TDEAT lines were maintained at 50 and 90°C, respectively. The
flow rates of oxygen and Ar purge gas were maintained constant at
10 and 30 sccm, respectively. The experiments were carried out in a
custom-designed cold-wall CVD chamber.19 The base pressure was
0.03 Torr. The substrate holder was heated resistively and controlled
with a proportional-integral-differential controller. The substrate
temperature was 300°C and the chamber pressure was 0.7 Torr.
*
Electrochemical Society Active Member.
z E-mail: takoudis@uic.edu
Downloaded on 2015-03-15 to IP 129.161.206.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G178
Journal of The Electrochemical Society, 154 ͑8͒ G177-G182 ͑2007͒
One-side polished Si͑100͒ wafers were treated with
H2SO4/H2O2 ͑4:1͒ for 15 min to remove organic contamination and
with 49% HF for 10 s to remove the native oxide. This was followed
by deionized ͑DI͒-water rinse and N2 purge dry. Then the sample
was immediately loaded into the reactor and continuously purged
with Ar for 1 h at 0.7 Torr. Next, the sample temperature was
ramped to 300°C within 2 min. This was followed by alternating
exposures of the sample to TMA ͑30 s͒, O2 ͑30 s͒, TDEAT ͑30 s͒,
and O2 ͑30 s͒, with Ar purging ͑1 min͒ in between reactant expo-
sures at the operating pressure of interest. This process is similar to
the ALD process which minimizes undesired reactions and particle
formation in the gas phase. First, Al2O3 was deposited to increase
the stability of the deposited films to Si. In order to study phase
transitions and interfacial reactions, films were rapid thermal an-
nealed in Ar for 5 min at a temperature between 500 and 900°C at
1 atm in a preheated quartz horizontal furnace ͑Lindberg Blue STF͒.
The furnace was kept in Ar ambient during wafer loading. The tem-
perature in the furnace can be maintained within 2°C.
The thickness of TiO2/Al2O3 film was measured using spectro-
scopic ellipsometry ͑J. A. Woollam Co. M-44͒. The surface rough-
ness of both as-deposited and annealed films was analyzed using
atomic force microscopy ͑AFM, Digital Instruments Dimension
3100͒ using the tapping mode. The bonding states and atomic com-
position of the samples were investigated using X-ray photoelectron
spectroscopy ͑XPS, Kratos Axis ULTRA and Kratos AXIS-165͒.
The data was calibrated with respect to the C 1s peak at 285 eV.
Peak fitting was accomplished using Gaussian–Lorentzian functions
and Shirley background subtraction. X-ray diffraction ͑XRD, Philips
X’pert͒ was utilized to identify the phase and structure of the film.
The analysis was characterized using Cu K␣ radiation ͑wavelength
rX-ray = 0.1542 nm͒ at a glancing incident angle of 1°. The X-ray
generator was operated at 45 kV and 40 mA. Rutherford back-
scattering ͑RBS͒ spectroscopy with a General Ionex Tandetron ac-
celerator employing 2 MeV He+ particles was used to study the bulk
composition of the films. The simulation was done using RUMP
software.
Figure 1. Thickness of TiO2/Al2O3 films as a function of cycle number
͓sequential expose of TMA ͑30 s͒, oxygen ͑30 s͒, TDEAT ͑30 s͒, and oxy-
gen ͑30 s͒, with Ar purging ͑1 min͒ in between each reactant feed͔; TDEAT
bubbler pressure is 80 ͑triangles͒, 160 ͑hollow circles͒, and 240 Torr ͑solid
circles͒. The films were deposited at a substrate temperature of 300°C and
chamber pressure of 0.7 Torr.
Si substrate; the TiO2 film is rougher and its rms is about 0.99 nm.
Typically, TiO2 films are rough. It has been reported that TiO2 films
deposited using TTIP have rms of more than 4 nm.21,22 The prob-
able difficulty with TTIP is that its activation energy is very high
and such a TiO2 film is usually deposited at temperatures above
300°C.5 It was reported that LPCVD TiO2 films using TTIP as
precursor are crystallized at deposition temperatures as low as
320°C.5,23 Generally, when TiO2 is deposited using TTIP, the co-
lumnar growth structure is known as a prime factor which causes the
roughness of the film.8,9,24 In our study, LPCVD TiO2 deposited
using TDEAT maintains amorphous structure up to 500°C, as evi-
denced by XRD ͑data not shown here͒. It is likely that the amor-
phous structure of TiO2 deposited using TDEAT results in a rela-
tively flat surface, which may make TDEAT an effective precursor
for TiO2 deposition.
Results and Discussion
By incorporating Al2O3 into TiO2, the roughness of the film is
improved dramatically as shown in Fig. 3. Addition of Al2O3 into
TiO2 ͑7:3 Ti/Al atomic ratio͒ decreases the RMS of the film from
0.99 to 0.16 nm, which is close to that of Al2O3 film. The appear-
The growth rate of laminated TiO2/Al2O3 films was studied at
various TDEAT feed flow rates. One layer of Al2O3 and one layer of
TiO2 typically constitute one deposition cycle. The Ti/Al ratio in
TiO2/Al2O3 film was controlled by varying the thickness of the
TiO2 layer while keeping the thickness of the Al2O3 layer constant.
This was accomplished by altering the TDEAT bubbler pressure
while maintaining all other conditions constant. Because the flow
rate of the carrier gas, Ar, is constant and TDEAT vapor pressure is
constant at constant temperature, lowering the TDEAT bubbler pres-
sure increases the ratio of TDEAT partial pressure to total bubbler
pressure, which results in higher TDEAT flow rate and, in turn, in
higher TiO2 deposition rate. Figure 1 illustrates the deposition rate
of TiO2/Al2O3 films as a function of number of deposition cycles at
various TDEAT bubbler pressures. The linear increase of the growth
rate with increasing deposition cycle number indicates that, at a
given TDEAT bubbler pressure, the thickness increase resulting
from each cycle is constant.
Because the deposition rate of Al2O3 is held constant at
0.54 nm/cycle ͑Fig. 2a͒, the thickness ratio of TiO2/Al2O3 at each
cycle as a function of TDEAT bubbler pressure can be derived and it
is as shown in Fig. 2b. Thus, by altering the TDEAT bubbler pres-
sure, laminated TiO2/Al2O3 films with controllable Ti/Al ratio can
be deposited.
The surface roughness of both as-deposited and annealed films is
important because it affects shifts in electronic energy levels and
degrades electrical characteristics.20 Figure 3 shows AFM images of
surface roughness of as-deposited TiO2, laminated TiO2/Al2O3, and
Al2O3 films on Si substrates. The images shown are from 2
ϫ 2 ␮m scan areas. The Al2O3 film is flat and its root-mean-square
͑rms͒ roughness is only 0.11 nm, which is comparable to that of the
Figure 2. ͑a͒ Al2O3 growth rate as a function of cycle number and ͑b͒
calculated thickness ratio of TiO2/Al2O3 as a function of TDEAT bubbler
pressure ͑other conditions are constant͒. The growth rate of A2O3 is constant
at 0.54 nm in each cycle. The films were deposited at a substrate temperature
of 300°C and chamber pressure of 0.7 Torr.
Downloaded on 2015-03-15 to IP 129.161.206.73 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, 154 ͑8͒ G177-G182 ͑2007͒
G179
Figure 3. ͑Color online͒ AFM surface roughness of as-deposited ͑a͒ 9 nm thick Al2O3 film ͑rms roughness is 0.11 nm͒, ͑b͒ 10 nm thick TiO2/Al2O3 film ͑7:3
Ti/Al atomic ratio, rms roughness is 0.16 nm͒, ͑c͒ 8 nm thick TiO2/Al2O3 film ͑9:1 Ti/Al atomic ratio, rms roughness is 0.50 nm͒, and ͑d͒ 7 nm thick TiO2 ͑rms
roughness is 0.99 nm͒ on Si substrate at 300°C and 0.7 Torr.
ganic precursor is used. Carbon impurity could increase the film
crystallization temperature, but it degrades the electrical character-
istics of the gate dielectric layer by increasing the interfacial and
bulk trap densities. In this study, RBS and XPS were utilized for
studies on impurities. Figure 6 shows RBS spectra of as-deposited
TiO2, Al2O3, and TiO2/Al2O3 films. The RBS spectra reveal that all
films have no detectable amount of carbon or nitrogen. XPS analy-
ses of these films ͑Fig. 7͒, after sputtering to remove the surface
contamination, corroborate the RBS results. Less than 1.5% carbon
and nitrogen were detected, which is close to the XPS detection
limit. This suggests that the deposited films have negligible carbon
contamination and the high crystallization temperature of the
LPCVD TiO2/Al2O3 films is not attributed to the incorporation of
carbon impurities.
ance of the TiO2 surface is also different from that of the films with
Al2O3. The surface of pure TiO2 is a dense narrow spike in shape,
while the TiO2 alloyed with Al2O3 is more mountainlike.
The roughness of TiO2/Al2O3 films ͑9:1 Ti/Al atomic ratio͒ as a
function of annealing temperature is shown in Fig. 4. The as-
deposited and annealed samples originated from the same 55 nm
TiO2/Al2O3 film. The AFM scan area was 1 ϫ 1 ␮m. The samples
annealed at temperatures of 500 and 700°C have a smooth surface
similar to that of the as-deposited sample with a rms value below
1 nm, which suggests that the film is stable up to 700°C. However,
when the annealing temperature is increased to 900°C, the film
shows apparent nucleation with larger grain size and its roughness
increases to more than 2 times that of as-deposited films. This sharp
roughness change indicates phase transformation.
The two important factors which may contribute to the high crys-
tallization temperature and the low impurity incorporation in the
films are the choice of the precursors and the deposition method.
TiO2 films have been deposited via hydrolysis. It has been found
Even though there are studies on the crystallization inhibition of
TiO2 films by Al2O3 incorporation,7,10,11 very few focus on CVD or
ALD films. Kim and Yun have studied laminated ALD TiAlO films
but did not report annealing studies.13 In this study, XRD was uti-
lized to study the morphology of both as-deposited and annealed
films. All as-deposited TiO2/Al2O3 and TiO2 films are amorphous.
Upon annealing in Ar at 700°C for 5 min, the XRD patterns shown
in Fig. 5 reveal that the 64 nm thick TiO2 film is crystallized. The
peaks are well interpreted to be a mixture of anatase and little rutile
phase.7,25 However, at the same annealing conditions, the 50 nm
thick TiO2/Al2O3 film ͑9:1 Ti/Al atomic ratio͒ remains amorphous.
This film crystallizes when the annealing temperature is increased to
900°C and only anatase phase reflections appear in the XRD spec-
trum. These results corroborate AFM results that suggest
TiO2/Al2O3 films maintain amorphous structure at temperatures up
to 700°C, which is comparable with some LPCVD Al2O3 films
reported in the literature.13
that TiO2 formed using titanium tetra-n-butoxide ͓Ti͑OBu͒ ͔ is
4
more likely to crystallize than that formed using other precursors
͑e.g., TDEAT͒. It was reported that Et2NH formed and incorporated
in the film during the hydrolysis of TDEAT inhibits the crystalliza-
tion of TiO2.25 However, that is not likely the case in our study,
because negligible amounts of nitrogen impurity were found. It
seems that the precursor itself plays an important role in the film
structure. Because amide-based metal precursors have been reported
to be promising due to their high reactivity and lower impurity
incorporation,16,17 TDEAT, as an amino Ti-containing precursor,
should result in a lower level of impurity incorporation than that
obtained using an alkoxide-based precursor, i.e., TTIP, even at low
process temperatures. Also, it has been reported that CH4, which
may be a major product in the TMA–H2O reaction and may partici-
pate actively in the carbon incorporation mechanism, can be re-
High deposition rate in a low-temperature CVD process likely
introduces impurities in films, especially carbon, when a metallor-
Downloaded on 2015-03-15 to IP 129.161.206.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G180
Journal of The Electrochemical Society, 154 ͑8͒ G177-G182 ͑2007͒
Figure 4. ͑Color online͒ AFM surface roughness of 55 nm thick TiO2/Al2O3 film with 9:1 Ti/Al atomic ratio: ͑a͒ as-deposited film at 300°C and 0.7 Torr has
rms roughness of 0.53 nm and ͑b–d͒ Ar-annealed at 500, 700, and 900°C for 5 min with rms roughness of 0.70, 0.72, and 1.17 nm, respectively.
moved as CO2 when oxygen is used as oxidizer.26 Furthermore, the
method, in which excess reactants in the gas phase are purged out
and the reaction takes place only on the substrate surface. Conse-
quently, lower impurity levels are obtained.
low impurity level can also be due to the ALD-like cyclic deposition
method used. In our earlier studies on LPCVD TiO2 films using
TDEAT, about 3 atom % carbon and 2 atom % nitrogen impurities
were found in the films.27 In this study, we use the cyclic deposition
The thermal stability of TiO2/Al2O3 film on Si was investigated
through postthermal annealing at temperatures of 500 and 700°C in
Ar for 5 min. The interface chemistry was analyzed using XPS.
Figure 8 shows the XPS spectra of Si 2p obtained from 6 nm thick
as-deposited and annealed films. The bonding energy at 99.1 eV is
attributed to Si 2p from the substrate.28,29 There is no signal from
Sin+ ͑n Ͼ 0͒ in the as-deposited film; such a signal starts appearing
in the annealed film. The atomic composition calculated from XPS
analyses suggests that the as-deposited TiO2/Al2O3 film has about
10% excess oxygen. After annealing in inert environment at tem-
peratures of 500 and 700°C, the film appears almost stoichiometric,
with less than 2% oxygen vacancies. This change in oxygen content,
from oxygen-rich in the as-deposited film to almost stoichiometric
in the annealed film, suggests that excess oxygen may facilitate the
interfacial reactions and induce SiOx formation at increased anneal-
ing temperatures.
Conclusions
Laminated TiO2/Al2O3 films with a controllable Ti/Al ratio have
been successfully deposited with TDEAT, TMA, and oxygen using
cyclic LPCVD at low temperatures. AFM analyses revealed that the
roughness of TiO2 film decreases dramatically with the addition of
even a few percent of Al2O3. As-deposited films are found to have
negligible amounts of carbon and nitrogen impurities as probed us-
ing both XPS and RBS. XRD shows that TiO2/Al2O3 film with 9:1
Ti/Al atomic ratio maintains an amorphous structure up to 700°C,
while TiO2 film crystallizes at the same conditions. There was no
interfacial SiOx detected in the as-deposited films, but it was formed
Figure 5. XRD diffraction patterns of Ar-annealed 64 nm thick TiO2 film
and 50 nm thick TiO2/Al2O3 film ͑9:1 Ti/Al atomic ratio͒. Annealing time
was 5 min. The analysis by XRD was carried out using Cu K␣ radiation
͑wavelength rX-ray = 0.1542 nm͒ at a glancing angle of 1°.
Downloaded on 2015-03-15 to IP 129.161.206.73 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, 154 ͑8͒ G177-G182 ͑2007͒
G181
Figure 7. C 1s and N 1s XP spectra of TiO2/Al2O3 film at 90° take-off angle
collected before Ar+ sputtering and 5 min after Ar+ sputtering, indicating less
than 1.5% carbon and nitrogen incorporation in the film.
Figure 8. XP spectra of Si 2p collected at 90°C take-off angle for a 6 nm
thick TiO2/Al2O3 ͑9:1 Ti/Al atomic ratio͒ film deposited at 300°C and
0.7 Torr: as-deposited film and after annealing in Ar for 5 min at 500°C and
700°C.
Acknowledgments
Figure 6. RBS spectra of as-deposited TiO2, Al2O3, and TiO2/Al2O3 ͑3:7
Ti/Al atomic ratio͒ films at 300°C and 0.7 Torr. 2.0 MeV He+ particle was
used for the analysis. RBS revealed that there is no detectable carbon and
nitrogen incorporation in the film.
The authors would like to thank Akzo Nobel for the kind supply
of TMA precursor. We are also grateful to Dr. Scott MacLaren, Dr.
Rick Haasch, Dr. Mauro Sardela, and Dr. Doug Jeffers for the AFM,
XPS, XRD, and RBS analyses, respectively, carried out at the Cen-
ter for Microanalysis of Materials, University of Illinois at Urbana-
Champaign. Partial funding by the National Science Foundation
͑NSF͒ is also gratefully acknowledged.
upon annealing in Ar at high temperatures. The laminated cyclic
CVD TiO2/Al2O3 films using TDEAT, TMA, and O2 at low tem-
peratures can be promising high-k structures for gate dielectric ap-
plications. Electrical characterization of such structures remains to
be done so that their full potential can be investigated.
University of Illinois at Chicago assisted in meeting the publication costs
of this article.
References
1. G. D. Wilk, R. M. Wallace, and J. M. Anthony, J. Appl. Phys., 89, 5243 ͑2001͒.
Downloaded on 2015-03-15 to IP 129.161.206.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

G182
Journal of The Electrochemical Society, 154 ͑8͒ G177-G182 ͑2007͒
2. C. C. Fulton, G. Lucovsky, and R. J. Nemanich, J. Vac. Sci. Technol. B, 20, 1726
͑2002͒.
3. K. Kukli, J. Ihanus, M. Ritala, and M. Leskela, Appl. Phys. Lett., 68, 3737 ͑1996͒.
4. A. Erbil, W. Braun, B. S. Kwak, B. J. Wilkens, L. A. Boatner, and J. D. Budai, J.
Cryst. Growth, 124, 684 ͑1992͒.
5. C. J. Taylor, D. C. Gilmer, D. G. Colombo, G. D. Wilk, S. A. Campbell, J. Roberts,
and W. L. Gladfelter, J. Am. Chem. Soc., 121, 5220 ͑1999͒.
6. A. S. R. B. Clark-Phelps, L. Cleveland, T. E. Seidel, and O. Sneh, Mater. Res. Soc.
Symp. Proc., 607, K.2.2.1 ͑2001͒.
7. D.-H. Kuo, K.-H. Tzeng, and C.-H. Chien, J. Vac. Sci. Technol. A, 21, 1996
͑2003͒.
Technol. A, 22, 1175 ͑2004͒.
17. K. A. Son, A. Y. Mao, B. Y. Kim, F. Liu, E. D. Pylant, D. A. Hess, J. M. White, D.
L. Kwong, D. A. Roberts, and R. N. Vrtis, in Papers from the 44th National
Symposium of the AVS, Vol. 16, p. 1670, San Jose, CA ͑1998͒.
18. X. Song and C. G. Takoudis, Presented in the 52nd AVS International Symposium,
AVS, Boston, MA ͑2005͒.
19. A. Deshpande, Ph.D. Thesis, Department of Chemical Engineering, University of
Illinois at Chicago, Chicago, IL ͑2005͒.
20. P. Young-Bae, K. H. Ahn, and P. Dong-Wha, J. Mater. Sci. Lett., V22, 1325
͑2003͒.
21. M.-K. Lee, J.-J. Huang, and T.-S. Wu, Semicond. Sci. Technol., 20, 519 ͑2005͒.
22. W. C. Hyun and C. D. Jin, Yoop Hakhoechi, 35, 88 ͑1998͒.
23. S. A. Campbell, H. S. Kim, D. C. Gilmer, B. He, T. Ma, and W. L. Gladfelter, IBM
J. Res. Dev., 43, 383 ͑1999͒.
8. D.-H. Kuo and K.-H. Tzeng, Thin Solid Films, 460, 327 ͑2004͒.
9. N. V. Nguyen, J.-P. Han, J. Y. Kim, E. Wilcox, Y. J. Cho, W. Zhu, Z. Luo, and T.
P. Ma, AIP Conf. Proc., 683, 181 ͑2003͒.
10. D. H. Kuo and K. H. Tzeng, Thin Solid Films, 420–421, 497 ͑2002͒.
11. O. Auciello, W. Fan, B. Kabius, S. Saha, J. A. Carlisle, R. P. H. Chang, C. Lopez,
E. A. Irene, and R. A. Baragiola, Appl. Phys. Lett., 86, 042904 ͑2005͒.
12. D.-H. Kuo and C.-N. Shueh, J. Non-Cryst. Solids, 336, 120 ͑2004͒.
13. Y. S. Kim and S. Jin Yun, J. Cryst. Growth, 274, 585 ͑2005͒.
14. G. K. Dalapati, S. Chatterjee, S. K. Samanta, and C. K. Maiti, Appl. Surf. Sci., 210,
249 ͑2003͒.
15. M. N. L. Niinistö, J. Päiväsaari, J. Niinistö, M. Putkonen, and M. Nieminen, Phys.
Status Solidi A, 201, 1443 ͑2004͒.
16. Y. Senzaki, S. Park, H. Chatham, L. Bartholomew, and W. Nieveen, J. Vac. Sci.
24. Y. Leprince-Wang and K. Yu-Zhang, Surf. Coat. Technol., 140, 155 ͑2001͒.
25. I. Yoshizumi, H. Hideyasu, and S. Yoshiyuki, J. Sol-Gel Sci. Technol., V19, 365
͑2000͒.
26. R. C. Abhijit, Ph.D. Thesis, Department of Chemical Engineering, University of
Illinois at Chicago, Chicago, IL ͑2003͒.
27. X. Song and C. G. Takoudis, J. Vac. Sci. Technol. A, In press.
28. T. M. Klein, D. Niu, W. S. Epling, W. Li, D. M. Maher, C. C. Hobbs, R. I. Hegde,
I. J. R. Baumvol, and G. N. Parsons, Appl. Phys. Lett., 75, 4001 ͑1999͒.
29. W. R. Salaneck, R. Bergman, J. E. Sundgren, A. Rockett, T. Motooka, and J. E.
Greene, Surf. Sci., 198, 461 ͑1988͒.
Downloaded on 2015-03-15 to IP 129.161.206.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).