Atomic layer deposition of NiO films on Si(100) using cyclopentadienyl-type compounds and ozone as precursors

Article abstract of DOI:10.1149/1.2965456

NiO films were grown on Si(100) by atomic layer deposition using Ni(Cp)₂ (Cp=cyclopentadienyl, C₅H₅) or Ni (EtCp)₂ [EtCp=ethylcyclopentadienyl, (C₂H₅) (C₅H₄)] and ozone in the 150-300°C temperature range. The growth temperature dependence of structure, electronic density, and impurity levels for the prepared NiO films was studied using X-ray reflectivity, X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, time of flight-secondary ion mass spectroscopy (ToF-SIMS), and transmission electron microscopy. The behavior of films deposited using Ni (Cp)₂ and Ni (EtCp)₂ is compared and discussed. NiO films with good stoichiometry and low amounts of contaminants are obtained at a growth temperature (Tg) of 250°C or above. At a fixed Tg, the growth rate for NiO films deposited using Ni (Cp)₂ is higher than the one of films deposited using Ni (EtCp)₂. Furthermore, the growth rate for NiO deposited using Ni(Cp)₂ at Tg =150°C is 0.32 nmcycle and decreases substantially in films deposited at higher temperatures. The electronic density of NiO films deposited at 300°C is close to the one of bulk NiO (1.83 e- A3). According to XRD and FTIR results, films deposited at Tg ≤ 200°C have a simple cubic polycrystalline structure. Impurities in NiO films decrease with increasing Tg, as detected by ToF-SIMS.

Full text of DOI:10.1149/1.2965456

Journal of The Electrochemical Society, 155 ͑10͒ H807-H811 ͑2008͒
H807
0013-4651/2008/155͑10͒/H807/5/$23.00 © The Electrochemical Society
Atomic Layer Deposition of NiO Films on Si(100) Using
Cyclopentadienyl-Type Compounds and Ozone as Precursors
H. L. Lu,a G. Scarel,a C. Wiemer,a M. Perego,a S. Spiga,a M. Fanciulli,a,b,z and
G. Paviac
aCNR-INFM MDM National Laboratory, 20041 Agrate Brianza (MI), Italy
bDipartimento di Scienza dei Materiali, Università degli Studi Milano-Bicocca, 20125 Milano, Italy
cSTMicroelectronics, 20041 Agrate Brianza (MI), Italy
NiO films were grown on Si͑100͒ by atomic layer deposition using Ni͑Cp͒ ͑Cp = cyclopentadienyl, C5H5͒ or Ni͑EtCp͒
2
2
͓EtCp = ethylcyclopentadienyl, ͑C2H5͒͑C5H4͔͒ and ozone in the 150–300°C temperature range. The growth temperature depen-
dence of structure, electronic density, and impurity levels for the prepared NiO films was studied using X-ray reflectivity, X-ray
diffraction ͑XRD͒, Fourier transform infrared ͑FTIR͒ spectroscopy, time of flight–secondary ion mass spectroscopy ͑ToF-SIMS͒,
and transmission electron microscopy. The behavior of films deposited using Ni͑Cp͒ and Ni͑EtCp͒ is compared and discussed.
2
2
NiO films with good stoichiometry and low amounts of contaminants are obtained at a growth temperature ͑Tg͒ of 250°C or
above. At a fixed Tg, the growth rate for NiO films deposited using Ni͑Cp͒ is higher than the one of films deposited using
2
Ni͑EtCp͒2. Furthermore, the growth rate for NiO deposited using Ni͑Cp͒ at Tg = 150°C is 0.32 nm/cycle and decreases sub-
2
stantially in films deposited at higher temperatures. The electronic density of NiO films deposited at 300°C is close to the one of
bulk NiO ͑1.83 e−/Å3͒. According to XRD and FTIR results, films deposited at Tg ജ 200°C have a simple cubic polycrystalline
structure. Impurities in NiO films decrease with increasing Tg, as detected by ToF-SIMS.
© 2008 The Electrochemical Society. ͓DOI: 10.1149/1.2965456͔ All rights reserved.
Manuscript submitted February 25, 2008; revised manuscript received June 30, 2008. Published August 19, 2008.
Nickel oxide ͑NiO͒ has been extensively studied for technologi-
cal applications due to its excellent chemical stability as well as its
optical, electrical, and magnetic properties.1-3 NiO is of great inter-
est because it represents a good example of p-type semiconducting
oxide, which can be used as transparent electrode and window
coating.4,5 Because of its magnetic properties, NiO is also an attrac-
tive material for the implementation of high-density magnetic stor-
age technologies.2,6 Currently, substantial work is devoted to the
investigation of the resistive switching properties of NiO that may
be exploited in nonvolatile resistance random access memory
devices.6-8 In addition, NiO films have been employed in various
other applications, such as antiferromagnetic layers in spin valve
structures,9,10 in electrochromic devices,11,12 and as functional layers
for chemical sensors.13 Thin NiO films have been grown using vari-
ous physical and chemical vapor deposition techniques, such as
sputtering,14,15 sol-gel process,16,17 pulsed laser deposition,18 spray
pyrolysis,19 chemical vapor deposition,20-22 and atomic layer depo-
sition ͑ALD͒.23-25 Among these growth techniques, ALD is recog-
nized as one of the most promising. The self-limiting nature in ALD
process enables the growth of high-quality thin films with precise
thickness control, excellent uniformity, and superior step
coverage.26-28
cently. The thermal stability of Cp precursors is limited, but it can be
improved through a wide variety of different substituents. In addi-
tion, Ni͑Cp͒ ͑Cp = cyclopentadienyl, C5H5͒ and H2O were re-
2
ported by Chae et al. to serve as precursors to fabricate NiO layers,
which were then reduced to metallic Ni layers using atomic hydro-
gen at 165°C.25
In the present work, thin NiO films were grown by ALD from
two different Ni-containing Cp-type precursors, namely, Ni͑Cp͒
2
and Ni͑EtCp͒ ͓EtCp = ethylcyclopentadienyl, ͑C2H5͒͑C5H4͔͒, and
2
O3 on HF-last Si substrates. The effect of growth temperature on
growth rate, film structure, and composition was investigated using
X-ray reflectivity ͑XRR͒, X-ray diffraction ͑XRD͒, Fourier trans-
form infrared spectroscopy ͑FTIR͒, time of flight–secondary ion
mass spectroscopy ͑ToF-SIMS͒, and transmission electron micros-
copy ͑TEM͒. The two precursors are compared in terms of purity,
smoothness, and growth rate of the corresponding NiO layer.
Experimental
NiO thin films were deposited on n-type Si͑100͒ in an ALD
F-120 ASM-Microchemistry Ltd. reactor. The Si substrates were
first cleaned using RCA standard clean
2 ͑SC-2͒ solutions
To our knowledge, however, there are only a few reports on ALD
of NiO. Utriainen et al. reported an ALD process for NiO in which
Ni͑acac͒ ͑acac = acetylacetonato͒, Ni͑dmg͒ ͑dmg = dimethyl-
͑HCl:H2O2:H2O = 1:1:5͒ at 85°C for 10 min and then dipped into a
dilute HF solution ͑HF:H2O = 1:50͒ at room temperature to pro-
2
2
duce H-terminated surfaces. Inside the ALD reactor, the Ni͑Cp͒
glyoximato͒ and Ni͑apo͒ ͑apo = 2, amio-pent-2-en-4-onato͒ were
2
2
and Ni͑EtCp͒ were evaporated from an open boat at source tem-
used as metal precursors with various oxygen sources, O2, H2O, O3,
etc.23 All these Ni-precursors showed poor gas-phase stability, lead-
ing to negligible reactivity. Furthermore, the impurity levels, such as
2
peratures of 40 and 80°C, respectively. The oxygen source ͑O3͒ was
generated from O2 ͑99%͒ in an ozone generator ͑Applied Science
and Technology Inc., AX8200͒ and fed into the reactor at a flux of
395 sccm and at a concentration of 175 g/cm3. Alternate pulsing of
precursors was carried out by inert gas and using pure nitrogen
͑99.9%͒ as carrier and purging gas. The self-limiting reaction and
growth rate for the Ni͑Cp͒2/O3 process were obtained with pulse
carbon and hydrogen, were not discussed in detail. Ni͑dmamp͒
2
͑dmamp = 1-dimethylamino-2-methyl-2-propanolate͒ has also been
adopted as a Ni-precursor for ALD of NiO together with water as
oxygen source.24 The ALD window for this precursor was found to
be between 90 and 150°C. Recently, cyclopentadienyl-type ͑Cp-͒
compounds were introduced as promising ALD precursors to deposit
high-quality metal oxide films.29,30 The Cp compounds are generally
volatile and highly reactive and, thus, suitable as ALD precursors
with low deposition temperatures and high growth rates. For ex-
ample, ALD processes utilizing Cp precursors for HfO2,31-34
times 8.0 and 10.0 s for Ni͑Cp͒ and O3, respectively, while purge
2
time was held constant at 9.0 s.40 For Ni͑EtCp͒2/O3, the best self-
limiting reaction and growth rate were obtained using the following
sequence: 5.0 s pulse of Ni͑EtCp͒2, followed by 4.0 s of nitrogen
purge, 7.0 s of O3, and finally, 4.0 s of nitrogen purge. The growth
temperature was varied from 150 to 300°C. Such a low-growth tem-
perature range is chosen for the Ni͑Cp͒2/O3 and the Ni͑EtCp͒2/O3
processes to verify whether a remarkable decrease of growth rate
with growth temperature takes place. This kind of phenomenon was
39
ZrO2,35,36 Y2O3,37 Gd2O3,38 and Lu2O3 have been developed re-
z E-mail: marco.fanciulli@mdm.infm.it
Downloaded on 2014-09-29 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

H808
Journal of The Electrochemical Society, 155 ͑10͒ H807-H811 ͑2008͒
Table I. Summary of the structural data of the NiO films extracted from XRR and GIXRD. The thickness of film grown with Ni„Cp…2ÕO3 at
150°C was measured using spectroscopic ellipsometry. (The error of the grain size, fixing microstrain to zero, is 5%.)
Tg
͑°C͒
Thickness
͑Ϯ0.1 nm͒
Growth rate
͑nm/cycle͒
Electronic density
Surface roughness
Grain size
͑nm͒
Precursor
͑Ϯ0.05 e−/Å3͒
͑Ϯ0.1 nm͒
Ni͑Cp͒
150
200
250
300
150
200
250
300
51.5
40.0
19.3
14.0
22.4
13.6
9.8
0.32
0.25
0.12
0.08
0.09
0.05
0.04
0.05
0.92
1.01
1.38
1.83
0.70
1.19–1.44
1.39–1.83
1.56–1.84
10.4
4.5
3.5
1.2
4.5
1.5
1.1
1.2
5.8
6.5
8.3
9.8
—
7.9
8.6
7.7
2
Ni͑Cp͒
2
Ni͑Cp͒
Ni͑Cp͒
Ni͑EtCp͒
Ni͑EtCp͒
Ni͑EtCp͒
Ni͑EtCp͒
2
2
2
2
2
2
12.9
found in Al2O3 and Lu-silicate films deposited in an ALD process
employing O3 and H-containing metal precursors, and the results
were reported in Ref. 41 and 42.
Ni͑EtCp͒ compared to Ni͑Cp͒2. Niinistö et al. have also observed
2
similar trends in growth rate for Y2O3 films grown using Cp-based
compounds.37
Thickness, roughness, and density of NiO films were determined
using XRR. The XRR data were acquired with an Italstructures
HRD3000 system with an angular step of 0.01° in the incidence
angle. In order to correctly simulate the XRR spectrum of film
XRR was also employed to determine electronic density and sur-
face roughness of the deposited layers. The measured XRR spectra
of thin films deposited using Ni͑Cp͒ and Ni͑EtCp͒ together with
fitted curves are shown in Fig. 2. The fitting results are summarized
in Table I. To obtain good quality fittings of the XRR spectra of
2
2
grown with Ni͑Cp͒ at 150°C, a preliminary check of the thickness
2
range was performed using spectroscopic ellipsometry. Film crystal
structure was evaluated by grazing incidence XRD ͑GIXRD͒ and
transmission FTIR spectroscopy. The FTIR spectra were collected in
transmission using a Bruker IFS66v/s spectrometer in vacuum
͑0.01 mbar͒ both at normal incidence ͑␪ = 0°͒ and at grazing inci-
dence ͑␪ = 70°͒ to acquire both transverse optical ͑TO͒ mode and
longitudinal optical ͑LO͒ mode ͑Berreman effect͒.43 Film growth
parameters and characteristics are summarized in Table I.
Impurities depth profiles were obtained by ToF-SIMS measure-
ments performed on a dual beam ION-TOF IV instrument using
25 keV Ga+ ions for analysis and 0.5 keV Cs+ ions for sputtering.
The data were acquired in negative polarity rastering the Cs+ ion
beam on a 200 ϫ 200 ␮m2 area and the Ga+ ion beam on a 50
ϫ 50 ␮m2 area. No depth scale-calibration was performed.
films deposited using Ni͑Cp͒ and O3 at Tg = 200–300°C, one has
2
to divide NiO into two distinct layers with quite different densities.
It means that a NiO top layer with low density is required for the
fitting. This top layer allows the simulation of a non-Gaussian elec-
tronic density profile at the sample surface/air interface, which
might be related to both a non-Gaussian distribution of the surface
roughness and a density gradient that might be caused by chemical
nonuniformity. The electronic density value for the film grown at
150°C is much lower than the one for bulk NiO ͑1.83 e−/Å3͒,44
suggesting that the grown film is not pure NiO. As Tg increases to
300°C, the electronic densities for the grown films become gradu-
ally closer to the one of bulk NiO. In particular, it can be seen that
the electronic density of film deposited at 300°C is 1.83 e−/Å3. The
low density of film grown at low Tg suggests the films are rich in
carbon contaminants or contain voids. Moreover, the film root-
mean-square ͑rms͒ surface roughness is observed to decrease with
increasing Tg.
TEM investigations were performed with a TECNAI F30
equipped with an energy filter. Bright-field images were acquired
using only elastically scattered electrons, thus overcoming chro-
matic aberration effects of the lenses. Compositional maps were
acquired using three-window electron spectroscopy imaging ͑ESI͒.
The electronic densities of films grown from the Ni͑EtCp͒2/O3
process are not homogeneous throughout the film cross section. In-
deed, a single-layer model cannot provide good fitting of the experi-
Results and Discussion
The growth rate of thin NiO films was determined by film thick-
ness, which was measured using XRR. Figure 1 shows the growth
rate of thin films deposited using Ni͑Cp͒ or Ni͑EtCp͒ as metal
2
2
precursor and O3 as oxygen source as a function of growth tempera-
ture. The growth rate ͑0.32 nm/cycle͒ in the Ni͑Cp͒2/O3 process is
quite high at 150°C. As Tg increases from 150 to 300°C, a signifi-
cant decrease in growth rate from 0.32 to 0.08 nm/cycle is ob-
served. For the films deposited from Ni͑EtCp͒ and O3, a similar
2
trend can be found in the 150–250°C range, although it is not as
evident as in the case of the Ni͑Cp͒2/O3 process. This behavior,
which will be explained later, is very similar to the one reported for
Al2O341 and Lu silicate42 deposited by ALD from Al͑CH3͒3/O3 and
͓͑Me3Si͒2N͔3Lu/O3, respectively. It should be noted that for the
Ni͑EtCp͒2/O3 process the growth rate at 300°C was slightly higher
than the one at 250°C. This could be a consequence of the partial
decomposition of Ni͑EtCp͒2. In addition, as shown in Table I, the
growth rate of the prepared films was lower when Ni͑EtCp͒2 instead
of Ni͑Cp͒ was used as Ni precursor at a fixed Tg. One should
2
consider that the larger size of the former precursor molecule com-
pared to the latter one could lead to steric hindrance on the substrate
surface, thus decreasing the growth rate for films deposited using
Figure 1. ͑Color online͒ Effect of growth temperature on the film growth
rate, determined from film thickness measured by XRR. Squares: Ni͑Cp͒2,
circles: Ni͑EtCp͒2.
Downloaded on 2014-09-29 to IP 129.81.226.78 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, 155 ͑10͒ H807-H811 ͑2008͒
H809
Figure 2. ͑Color online͒ XRR spectra ͑dots͒ and simulations ͑lines͒ of films
grown using ͑a͒ Ni͑Cp͒ and ͑b͒ Ni͑EtCp͒ as Ni precursors in the tempera-
2
2
ture range of 150–300°C. Diamonds: 150°C, triangles: 200°C, squares:
250°C, and circles: 300°C.
Figure 3. ͑Color online͒ XRD patterns ͑dots͒ and simulations ͑lines͒ of films
deposited from ͑a͒ Ni͑Cp͒2/O3 and ͑b͒ Ni͑EtCp͒2/O3. Diamonds: 150°C,
triangles: 200°C, squares: 250°C, circles: 300°C. The simulation of a NiO
powder with lattice parameter equal to 4.16 Å, including the peak indexing
and the background bumps due to the Si substrate, are also shown.
mental XRR spectra. An additional layer between the NiO films and
the Si͑100͒ should therefore be introduced. The fitting results show
that the electronic densities of the additional layer are lower than the
one of bulk NiO. Also, the electronic density of the film grown at
150°C is only 0.7 e−/Å3. In addition, the rms surface roughness for
the Ni͑EtCp͒2/O3 processed films deposited at Tg = 200–300°C
varies in the 1.1–1.5 nm range. However, the corresponding value is
observed to significantly increase to 4.5 nm in the film grown at
Tg = 150°C.
͑Table I͒. A higher value of D corresponds either to a bigger domain
size or to an enhanced degree of crystallinity. Within the explored
Tg, D is always lower than the film thickness. For films grown with
Ni͑Cp͒ D increases with increasing Tg whereas for films grown
2
using Ni͑EtCp͒ D does not significantly vary for Tg in the
2
200–300°C range. No evidence of metallic Ni is revealed in any of
the analyzed diffraction patterns.20
XRD was employed to investigate the dependence of film crys-
tallinity on Tg. Figures 3a and b show XRD patterns and correspond-
ing Rietveld refinement of films deposited between 150 and 300°C
using Ni͑Cp͒ and Ni͑EtCp͒ as Ni sources, respectively. For a
FTIR spectroscopy was used to determine the chemical compo-
sition, structure, and corresponding variations with Tg, of the layers.
Figure 4 shows the FTIR spectra of NiO films deposited using
Ni͑Cp͒ ͑a͒ and Ni͑EtCp͒ ͑b͒ in the Tg range from 150 to 300°C.
2
2
given growth temperature, films grown using both Ni precursors
have similar XRD patterns. At Tg = 150°C, only the film grown
2
2
In the case of films deposited using Ni͑Cp͒ and O3, the NiO film
2
with Ni͑Cp͒ is partially crystallized, whereas the one grown with
2
deposited at 300°C exhibits a broad absorption band corresponding
to a TO mode at 396 cm−1, which is in good agreement with the
wavenumber expected for simple cubic NiO.45 The same TO mode
of the films deposited below 300°C is also visible, but its intensity
drastically decreases as Tg decreases. Moreover, the wavenumber of
the TO mode of films deposited at Tg Ͻ 300°C is slightly shifted
toward higher wavenumbers compared to those of the bulk oxide.
The absorption band near the frequency corresponding to a LO
mode9,45 at 583 cm−1 is revealed only in the FTIR spectrum ac-
Ni͑EtCp͒2 is found to be amorphous. The lattice parameter extracted
from the Rietveld refinement varies in the 4.16–4.21 Å range, with
higher values being measured for films grown at lower temperature.
These values are in good agreement with the value reported for bulk
NiO ͑4.18 Å͒.44 Neither texture nor anisotropic strain distribution is
evidenced from Rietveld refinement of all these grazing incidence
diffraction data. Fixing microstrain at zero, an average value of the
diffraction domain size ͑D͒ can be extracted from the refinement
Downloaded on 2014-09-29 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

H810
Journal of The Electrochemical Society, 155 ͑10͒ H807-H811 ͑2008͒
Figure 5. ͑Color online͒ ToF-SIMS depth profiles of thin NiO films grown
using ͑a͒ Ni͑Cp͒2/O3 and ͑b͒ Ni͑EtCp͒2/O3 at Tg = 300°C. Upper triangles:
NiO, circles: Si, squares: C, down triangles: Ni, and diamonds: SiO3.
Ni͑EtCp͒2/O3 between 150 and 300°C. It can be clearly seen that
the carbon level decreases significantly with increasing Tg. A similar
trend has also been observed in samples grown using Ni͑Cp͒ and
2
O3 ͑not shown͒. It is important to note that, in films deposited using
both Ni precursors in the 150–250°C Tg range, the carbon signal is
almost three orders of magnitude higher than the one detected in
samples grown at 300°C. This phenomenon might be ascribed to
incomplete reactions between the precursors and the growing sur-
face, and/or readsorption of by-products. This trend is in good
agreement with the increase in film electronic density with an in-
creasing Tg detected using XRR. In addition, the carbon content in
the Ni͑EtCp͒2-based film grown at Tg = 300°C is slightly higher
than the one in the Ni͑Cp͒2-based film. The spike in carbon signal at
the sample surface and at the NiO/Si interface in the
Ni͑EtCp͒2-based film grown at 300°C is attributed to surface con-
tamination during air exposure.
Figure 4. ͑Color online͒ FTIR spectra acquired at ␪ = 70° of thin films
deposited from ͑a͒ Ni͑Cp͒2/O3 and ͑b͒ Ni͑EtCp͒2/O3 in the 150–300°C
temperature range.
To investigate the nature of the low-density films deposited at the
lowest Tg, TEM analysis and ESI elemental mapping ͑Fig. 7͒ of
quired at ␪ = 70° of the film deposited at 300°C, again the one with
electronic density closer to the one characteristic of NiO. For the
films grown using Ni͑EtCp͒ and O3 at Tg = 150 and 300°C were
2
performed. As shown in Fig. 7a, the TEM image of the film depos-
ited at 150°C exhibits regions with significantly different contrast
attributed to the presence of microvoids in the films. These voids
samples prepared using Ni͑EtCp͒ as Ni precursor, the spectra are
2
similar to those obtained for the films deposited using Ni͑Cp͒2.
However, the TO mode in the NiO film deposited at 250°C is almost
identical to the one of the film deposited at 300°C. Moreover, this
TO mode disappears in the films deposited at 150°C, indicating that
this film is amorphous. The typical peaks arising from SiO2 at
ϳ1107 cm−1 are visible in all samples, suggesting the formation of
a silicon oxide interfacial layer between NiO and Si͑100͒. The ab-
sorption bands assigned to the C v O stretching modes close to
1600 cm−1 for NiO films deposited using Ni͑Cp͒ and Ni͑EtCp͒
2
2
are detectable in films deposited at Tg Ͻ 250°C and 200°C, respec-
tively.
Depth profiles were determined by ToF-SIMS. Figures 5a and b
show representative ToF-SIMS depth profiles of the NiO films
grown using Ni͑Cp͒ and Ni͑EtCp͒ as Ni precursors at 300°C,
2
2
respectively. In the film deposited using Ni͑Cp͒2 and O3, a relatively
uniform NiO profile is obtained. In contrast, a decrease of the NiO
signal is observed near the surface region of the Ni͑EtCp͒2-based
films. The SiO3 signal in Fig. 5a has a peak at the film/substrate
interface, indicating the formation of a SiOx interfacial layer. Fur-
thermore, the Si levels in the NiO film obtained using the
Ni͑EtCp͒2/O3 process are higher than those measured in the film
grown using Ni͑Cp͒2/O3, especially near the film/substrate inter-
face.
Figure 6. ͑Color online͒ Carbon ToF-SIMS depth profiles of thin films
grown using Ni͑EtCp͒ and O3 in the 150–300°C temperature range. Upper
2
Figure 6 shows carbon depth profiles of the films deposited using
triangles: NiO, circles: Si, and squares: C.
Downloaded on 2014-09-29 to IP 129.81.226.78 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, 155 ͑10͒ H807-H811 ͑2008͒
H811
Figure 7. ͑Color online͒ TEM images of
thin films deposited using Ni͑EtCp͒ and
2
O3 at Tg: ͑a͒ 150°C, ͑b͒ 300°C. ESI map
͑c͒ of thin film deposited using Ni͑EtCp͒
and O3 at Tg = 300°C.
2
4. H. Sato, T. Minami, S. Takata, and T. Yamada, Thin Solid Films, 236, 27 ͑1993͒.
5. C. M. Lampert, Mater. Today, 7, 28 ͑2004͒.
account for the extremely low electronic densities measured in this
film by XRR ͑Table I͒. Similar phenomena were detected in Al2O3
films deposited using trimethylaluminum and O3.41 One possible
contribution to void generation is the high carbon concentration de-
tected in this sample ͑see Fig. 5 and 6͒. Another cause of void
generation could be the low dissociation rate of O3 at 150°C, which
6. M.-J. Lee, S. Seo, D.-C. Kim, S.-E. Ahn, D. H. Seo, I.-K. Yoo, I.-G. Baek, D.-S.
Kim, I.-S. Byun, S.-H. Kim, et al., Adv. Mater. (Weinheim, Ger.), 19, 73 ͑2007͒.
7. S. Seo, M. J. Lee, D. H. Seo, E. J. Jeoung, D.-S. Suh, Y. S. Joung, I. K. Yoo, I. R.
Hwang, S. H. Kim, I. S. Byun, et al., Appl. Phys. Lett., 85, 5655 ͑2004͒.
8. Y.-H. You, B.-S. So, J.-H. Hwang, W. Cho, S. S. Lee, T.-M. Chung, C. G. Kim, and
K.-S. An, Appl. Phys. Lett., 89, 222105 ͑2006͒.
9. Y. Hamakawa, H. Hoshiya, T. Kawabe, Y. Suzuki, R. Arai, K. Nakamoto, M.
Fuyama, and Y. Sugita, IEEE Trans. Magn., 32, 149 ͑1996͒.
10. H. J. M. Swagten, G. J. Strijkers, P. J. H. Bloemen, M. M. H. Willekens, and W. J.
M. de Jonge, Phys. Rev. B, 53, 9108 ͑1996͒.
11. G. A. Niklasson and C. G. Granqvist, J. Mater. Chem., 17, 127 ͑2007͒.
12. E. L. Miller and R. E. Rocheleau, J. Electrochem. Soc., 144, 1995 ͑1997͒.
13. K. Arshak, O. Korostynska, and F. Fahim, Sensors, 3, 176 ͑2003͒.
14. K.-S. Ahn, Y.-C. Nah, and Y.-E. Sung, Appl. Surf. Sci., 199, 259 ͑2002͒.
15. F. F. Ferreira, M. H. Tabacniks, M. C. A. Fantini, I. C. Faria, and A. Gorenstein,
Solid State Ionics, 86–88, 971 ͑1996͒.
16. Y. R. Park and K. J. Kim, J. Cryst. Growth, 258, 380 ͑2003͒.
17. R. C. Korošec and P. Bukovec, Acta Chim. Slov., 53, 136 ͑2006͒.
18. Y. Kakehi, S. Nakao, K. Satoh, and T. Kusaka, J. Cryst. Growth, 237–239, 591
͑2002͒.
19. P. Puspharajah, S. Radhakrishna, and A. K. Arof, J. Mater. Sci., 32, 3001 ͑1997͒.
20. J.-K. Kang and S.-W. Rhee, Thin Solid Films, 391, 57 ͑2001͒.
21. W.-C. Yeh and M. Matsumura, Jpn. J. Appl. Phys., Part 1, 36, 6884 ͑1997͒.
22. K. Sugimoto, M. Seto, S. Tanaka, and N. Hara, J. Electrochem. Soc., 140, 1586
͑1993͒.
23. M. Utriainen, M. Kröger-Laukkanen, and L. Niinistö, Mater. Sci. Eng., B, 54, 98
͑1998͒.
24. T. S. Yang, W. Cho, M. Kim, K.-S. An, T.-M. Chung, C. G. Kim, and Y. Kim, J.
Vac. Sci. Technol. A, 23, 1238 ͑2005͒.
25. J. Chae, H.-S. Park, and S.-W. Kang, Electrochem. Solid-State Lett., 5, C64 ͑2002͒.
26. T. Suntola, Thin Solid Films, 216, 84 ͑1992͒.
27. M. Leskelä and M. Ritala, Thin Solid Films, 409, 138 ͑2002͒.
28. R. L. Puurunen, J. Appl. Phys., 97, 121301 ͑2005͒.
29. M. Putkonen and L. Niinistö, Top. Organomet. Chem., 9, 125 ͑2005͒.
30. J. W. Elam, A. B. F. Martinson, M. J. Pellin, and J. T. Hupp, Chem. Mater., 18,
3571 ͑2006͒.
31. J. Niinistö, M. Putkonen, L. Niinistö, S. L. Stoll, K. Kukli, T. Sajavaara, M. Ritala,
and M. Leskelä, J. Mater. Chem., 15, 2271 ͑2005͒.
32. K. Kukli, J. Niinistö, A. Tamm, J. Lu, M. Ritala, M. Leskelä, M. Putkonen, L.
Niinistö, F. Song, P. Williams, et al., Microelectron. Eng., 84, 2010 ͑2007͒.
33. J. Niinistö, M. Putkonen, L. Niinistö, F. Song, P. Williams, P. N. Heys, and R.
Odedra, Chem. Mater., 19, 3319 ͑2007͒.
leaves many incompletely reacted Ni͑EtCp͒ ligands in the fabri-
2
cated film. Figure 7b shows that the film grown at 300°C is compact
and composed of three layers. Right above the Si substrate, a light
layer of ϳ2.6 nm thick is thought to be interfacial SiOx, which also
can be further confirmed by the ESI map ͑Fig. 7c͒. The top layer is
pure NiO with a thickness of 11.4 nm, in which small grains of
crystalline phase can be seen. A gray band of ϳ3.9 nm thick, be-
tween the NiO layer and the SiOx one, is identified as an interme-
diate layer. This layer is related to the additional layer necessary to
simulate the XRR data. However, the intermediate layer cannot be
easily distinguished in the ESI maps, indicating that its chemical
composition could be similar to the one of the main layer, just with
a different density.
Conclusions
In conclusion, thin NiO films grown by ALD using Ni͑Cp͒ or
2
Ni͑EtCp͒ as Ni precursor and O3 as the oxygen source were inves-
2
tigated. This study was focused on the influence of growth tempera-
ture on the structural and chemical properties of the prepared films.
At a fixed temperature, the growth rate of films deposited using
Ni͑Cp͒2/O3 is higher than the one of films deposited using
Ni͑EtCp͒2/O3. XRD and FTIR analyses show that films deposited at
Tg ജ 200°C have a cubic polycrystalline structure. The electronic
density of NiO films deposited at 300°C is close to the one of bulk
NiO, especially for the films grown using Ni͑Cp͒2. Compared to the
films deposited using Ni͑EtCp͒2/O3, the films grown with Ni͑Cp͒
2
have a lower carbon content. These results reveal that Ni͑Cp͒ is a
2
more suitable precursor than Ni͑EtCp͒ for the growth of NiO films
2
by ALD, both in terms of growth rate, compositional, and structure
quality. The low density at low Tg probably results from carbon
contaminants and/or voids in the films, as confirmed by ToF-SIMS
and TEM measurements.
34. J. Niinistö, M. Putkonen, L. Niinistö, K. Arstila, T. Saiavaara, J. Lu, M. Ritala, and
M. Leskelä, J. Electrochem. Soc., 153, F39 ͑2006͒.
35. J. Niinistö, M. Putkonen, L. Niinistö, K. Kukli, M. Ritala, and M. Leskelä, J. Appl.
Phys., 95, 84 ͑2004͒.
36. J. Niinistö, A. Rahtu, M. Putkonen, M. Ritala, M. Leskelä, and L. Niinistö, Lang-
muir, 21, 7321 ͑2005͒.
Acknowledgments
37. J. Niinistö, M. Putkonen, and L. Niinistö, Chem. Mater., 16, 2953 ͑2004͒.
38. J. Niinistö, N. Petrova, M. Putkonen, L. Niinistö, K. Arstila, and T. Sajavaara, J.
Cryst. Growth, 285, 191 ͑2005͒.
This work was supported by the European Project EMMA–
Emerging Materials for Mass-Storage Architectures ͑grant no. FP6-
033751͒.
39. G. Scarel, E. Bonera, C. Wiemer, G. Tallarida, S. Spiga, and M. Fanciulli, Appl.
Phys. Lett., 85, 630 ͑2004͒.
40. E. Cianci and M. Fanciulli, Personal communication.
41. S. D. Elliott, G. Scarel, C. Wiemer, and M. Fanciulli, Chem. Mater., 18, 3764
͑2006͒.
National Laboratory MDM assisted in meeting the publication costs of
this article.
42. G. Scarel, C. Wiemer, M. Fanciulli, I. L. Fedushkin, G. K. Fukin, G. A. Dom-
rachev, Y. Lebedinskii, A. Zenkevich, and G. Pavia, Z. Anorg. Allg. Chem., 633,
2097 ͑2007͒.
43. D. W. Berreman, Phys. Rev., 130, 2193 ͑1963͒.
44. ICSD, Inorganic Crystal Structure Database, file no. 61324 ͑2006͒.
References
1. R. J. Powell and W. E. Spicer, Phys. Rev. B, 2, 2182 ͑1970͒.
2. R. H. Kodama, S. A. Makhlouf, and A. E. Berkowitz, Phys. Rev. Lett., 79, 1393
͑1997͒.
3. H. Yamane and M. Kobayashi, J. Appl. Phys., 83, 4862 ͑1998͒.
45. S. Mochizuki and M. Satoh, Phys. Status Solidi B, 106, 667 ͑1981͒.
Downloaded on 2014-09-29 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).