Surface nanostructures based on tantalum and aluminum oxides

Article abstract of DOI:10.1134/S0036024410020172

Generalized results from investigations into the processes of ormation of tantalum and aluminum oxide nanostructures and their multilayered systems produced by the method of molecular lamination on a (100) silicon surface and aluminum are reported. Conditions for the layer-by-layer growth of oxide structures and multilayered low-dimensional systems with alternating zones of the said oxides are determined. Dielectric characteristics of the synthesized nanostructures are evaluated. Pleiades Publishing, Ltd., 2010.

Full text of DOI:10.1134/S0036024410020172

ISSN 0036ꢀ0244, Russian Journal of Physical Chemistry A, 2010, Vol. 84, No. 2, pp. 261–266. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © Yu.K. Ezhovskii, 2010, published in Zhurnal Fizicheskoi Khimii, 2010, Vol. 84, No. 2, pp. 314–320.
PHYSICAL CHEMISTRY OF NANOCLUSTERS
AND NANOMATERIALS
Surface Nanostructures Based on Tantalum
and Aluminum Oxides
Yu. K. Ezhovskii
St. Petersburg State Technological Institute (Technical University), St. Petersburg, 190013 Russia
eꢀmail: office@tu.spb.ru
Received January 11, 2009
Abstract—Generalized results from investigations into the processes of formation of tantalum and aluminum
oxide nanostructures and their multilayered systems produced by the method of molecular lamination on a
(100) silicon surface and aluminum are reported. Conditions for the layerꢀbyꢀlayer growth of oxide structures
and multilayered lowꢀdimensional systems with alternating zones of the said oxides are determined. Dielecꢀ
tric characteristics of the synthesized nanostructures are evaluated.
DOI: 10.1134/S0036024410020172
INTRODUCTION
and aluminum oxides and multilayered systems of
them deposited by the ML method on a silicon surface
and aluminum films on silicon.
Progress in the chemical nanotechnology of subꢀ
micronꢀscale structures involving surface chemical
reactions that form the basis of the atomic layer depoꢀ
sition (ALD technology), better known in Russia as
the molecular lamination (ML) method [1–3],
requires the solving of a number of problems related to
establishing the properties of nanostructures and reguꢀ
lar features of their formation. One major problem is
determining the conditions for the mechanism of their
layerꢀbyꢀlayer synthesis, excluding the stage of threeꢀ
dimensional nuclei formation and ensuring the matrix
synthesis of lowꢀdimensional structures with the posꢀ
sibility of verifying composition and thickness at the
monolayer level. It should be noted that chemical
methods for the synthesis of lowꢀdimensional objects
are an integral part of a vigorously progressing chemiꢀ
cal nanotechnology for the production of new materiꢀ
als having different end uses, and of nanoelectronic
systems in particular. Employing the gasꢀphase feed of
reagents and being selfꢀorganizing in character, such
processes allow entire batches of goods to be treated at
once, making them more profitable.
EXPERIMENTAL
The nanolayers were synthesized on freshly etched
silicon (KEFꢀ7.5, (100) orientation) and vacuumꢀ
sputtered aluminum films on silicon ~0.2 m thick.
The process was implemented in a vacuum flowꢀ
through installation (with residual gas pressure of no
more than 10 Pa), ensuring interaction between the
metal chloride vapor and surface hydroxyl groups and
subsequent hydrolysis with water vapor for recovery of
the hydroxyl coating.
μ
–
1
The reactions for the silicon matrix can be written
in the form
(≡Si–OH) + MCl
m n
(I)
II)
(
≡
Si–O–) MCl +
n – m
m
HCl,
m
(
≡
Si–O–) MCl
m
+ (n – m
– m
)Н О
2
n
(
(
≡
Si–O–) M(ОН)
m
+ (n – m)HCl.
– m
n
Successful application of ALD technology in tackꢀ Repeated reactions (I) and (II) and the intermediate
ling certain problems in electronics has been reported removal of excess reagents and reaction products led
in publications abroad, e.g., [4–6].
to formation of an oxide layer of the required thickness
), which was proportional, as was found in [2, 3], to
the number of the treatment cycles ( ):
(
d
Submicron layers of aluminum and tantalum
oxides are of practical significance as they have high
dielectric characteristics. Multilayered nanostructures
based on these materials ought to possess special propꢀ
erties such as dielectric superlattices, the characterisꢀ
tics of which would depend on the composition, the
sequence of layers, and the distribution of potential.
N
d
=
d0N
,
(1)
where d0 is the proportionality factor characterizing
the structure of the synthesized layer and indicating
average thickness of the film in a single cycle of treatꢀ
ment with one component or another.
This paper describes the generalized results of
A comparison of d0, now called “the layer growth
investigations into synthesis processes and the basic parameter,” and the metal–oxygen interlayer spacing
dielectric characteristics of nanolayers of tantalum in the oxide structure allows us to assess the level of
261

262
EZHOVSKII
surface filling and is one criterion for determining the stages of growth) subjected to magnetic modulation,
layer formation mechanism. Of greatest interest is a and parameters and were measured to within
Δ
Ψ
lamination mechanism that can operate under the
nonequilibrium conditions of reactions (I) and (II)
upon the limiting hydroxylation of the surface and sufꢀ
ficient activity of the OH groups. However, hydroxyl
groups of monocrystalline silicon with an oxide <3 nm
±
0.01'.
Composition of the ultrathin layers was monitored
according to the XPS data, using NPꢀ5950A (Al
K
α
radiation, EK = 1486 eV) and SERꢀ1 (Mg
K radiaꢀ
α
α
thick exhibit weak protonꢀdonor properties in surface tion, EK = 1253 eV) spectrometers. The energy was
α
reactions [7].
measured against the С1s1/2 carbon standard with
The same is true of the aluminum matrix surface.
Although the hydroxyl groups in the surface layer of
the aluminum oxide have a sufficiently high protonꢀ
donor capacity [8], the electronꢀsaturated core of the
metal matrix suppresses the inductive effect that facilꢀ
itates protonization of the hydroxyl groups. We thereꢀ
fore proposed using triethylamine (TEA) [9] to actiꢀ
vate surface reactions. Possessing high protonꢀaccepꢀ
tor properties, this reagent can perform a number of
functions simultaneously: it can stabilize hydroxyl
Est = 285.0 eV. At these energies, the probe penetrated
the surface layer to a maximum depth of 8 nm [12].
The energy lines of the elements under study (Es) were
identified using data from [13, 14], while layer compoꢀ
sition was determined from the relationship [12]
I /I = n σ E /n σ E ,
(2)
are the intensities of the examined lines
of the atoms under study; and are the relative
1
2
1
1
K
2
2
K
1
2
where 1 and 2
I I
σ
σ
1
2
coating with additional hydrogen bonds, activate a crossꢀsections of ionization for the studied levels;
n
1
reaction by forming an intermediate complex, and and n2 are the concentrations of the studied atoms;
bind liberated hydrogen chloride, thereby ensuring
completeness of the reactions. In synthesizing triethyꢀ
lamine, which is not inclined toward nucleophilic subꢀ
and EK1 and EK2 are the kinetic energies of knockedꢀ
Fs, where Fs is the elecꢀ
EKα
on electrons ( K Es
E = – –
stitution reactions, we can add water vapor and tron work function of the spectrometer material).
(
H O + N(C H5)3) simultaneously, and the complex Spectra were computerꢀprocessed using a special proꢀ
2
2
formed on the surface (e.g.,
facilitates protonization of the hydroxyl groups.
≡Si–OHꢀꢀꢀN(C H )
)
gram with Gaussian approximation of the curves.
Layer composition was also identified from ellipsoꢀ
metric observation of the refraction index determined
by the Holmes method [11].
2
5 3
Since the topology of a hydroxyl coating on silicon
with a thin oxide layer at Ts > 450 K is capable of
ensuring the chemical bonds of a halogenide with not
Kinetic characteristics of the metal halogenide and
more than two hydroxyl groups [7], the chemisorption water were subjected to ellipsometric analysis so as to
of a halogenide (e.g., TaCl5 with TEA) can be visualꢀ establish conditions for the limiting filling of the surꢀ
ized in the form
face with groups synthesized by reactions (I) and (II)
at a given synthesis temperature and vapor pressure.
The evaluation criterion was the dependence of the
growth parameter (the thickness of the formed oxide,
reduced to one treatment cycle) on the time of contact
between the reagent and the substrate. An analysis of
this dependence at different temperatures of the subꢀ
strate and different pressures of the halogenide showed
III) (Fig. 1) that within the range of the studied reagent
vapor pressures, the surface filling kinetics depended
essentially on the substrate temperature only.
H...N(C H )
H...N(C H )
2
5 3
2
5 3
–
Cl
Cl
–
Cl
Cl
–
Si O
–
Si O
–
+
Ta Cl
Cl
Ta Cl
–
–
–
–
Si O
Si O
–
Cl
Cl
Cl
H...N(C H )
H...N(C H )
2 5 3
2
5 3
(
–
–
Si O
Cl
–
Ta Cl + 2N(C H ) HCl
.
2
5 3
–
–
Si O
Cl
–
Halogenide chemisorption was of a similar characꢀ
This scheme depicts only the stoichiometry of the surꢀ ter for both oxides (Fig. 1). At Ts < 423 K and a haloꢀ
face reaction rather than the coordination of the metal genide vapor pressure
in the oxide being formed.
p = 0.5–10 Pa, parameter d0
increased continuously, pointing to polymolecular
adsorption of the halogenide. When the substrate temꢀ
perature was higher, the surface was saturated with
oxide groups (Fig. 1). This suggested the process was
of a selfꢀlimiting nature. Considering these data, the
time of contact between vapors of halogenide and
The thickness of the synthesized nanolayers was
determined from ellipsometric measurements of
polarization parameters
as an approximation of the Drude–Tronston oneꢀ
layer model [10]. The parameters and were meaꢀ
sured on an ellipsometer in the PQSA [11] arrangeꢀ
ment with a fixed compensator. The source of linearly
polarized light was an LGꢀ75 laser with an emission
wavelength of 632.8 nm. To improve measurement
Δ
and
Ψ
; it was also calculated
Δ
Ψ
water at
p 1.3 Pa and ~10 Pa, respectively, was chosen
≈
to be 30 s for AlCl3 and 60 s for TаCl5. These times
were used in all our subsequent experiments.
Analysis of the growth dynamics of the Al O and
2
3
accuracy, the light beam was in some cases (the initial Та О5 layers (Fig. 2) demonstrated that at Ts = const,
2
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 2
2010

SURFACE NANOSTRUCTURES BASED ON TANTALUM AND ALUMINUM OXIDES
263
d0, nm
1
1'
0
.6
.4
3
2
0
2
'
'
0.2
3
0
20
40
70
, s
τ
Fig. 1. Kinetics of surface filling with components during formation of Al O3 (1–3) and Ta O5 (1'–3') at substrate temperatures
2
2
Ts = 403 (1, 1'), 523 (2, 2'), and 603 K (3, 3'); p = 1.3 Pa.
the dependences d = f
(
N
) followed Eq. (1) over the layers synthesized in the absence of TEA. In this case,
studied range of
tions were of a sustained nature, and that equal
amounts of product were formed in each reaction
N
. This indicated the surface reacꢀ halogenide chemisorption was of an expressed activaꢀ
tion character at Ts > 453 K and layer growth could be
effected only at Ts > 550 K.
cycle. In the case of aluminum oxide, (Fig. 2,
value of d0 was in better agreement with the calculated
data if the surface reactions were assumed to involve the activation barrier, extending the temperature range
1) the
The use of TEA in the synthesis process eliminated
the Al Cl6 dimer dominant in the composition of the of layer growth almost by 100 K, into the lowꢀtemperꢀ
2
halogenide vapor and having an orientation that led to ature region (Fig. 3). At higher synthesis temperatures,
the formation of two aluminumꢀoxygen layers:
the decline in d0 was due to partial dehydroxylation of
–
–
–
Si
O
Cl
Cl
–
Si OH
+
Al Cl
2
Al
Al + H O
6
2
d, nm
–
1
–
Si OH
–
–
Si
O
Cl
Cl
(IV)
2
–
–
–
Si
O
O
OH OH
Al
–
Si
O
O
O
OH
t
16
Al
Al
Al
–
–
–
–
Si
OH OH
Si
O
OH
Scheme (IV) does not show the formation, analoꢀ
gous to reaction (III), of an intermediate complex
with thriethylamine; neither is the actual coordination
of aluminum in the forming oxide layer reflected.
12
Analysis of the series of dependences
d = f(N),
obtained for the Al O and Та О layers at different
2
3
2
5
temperatures of the substrate, allowed us to establish
how this factor influenced the layer growth parameter
8
4
(
Fig. 3). The mechanism of layer formation and the
conditions for layerꢀbyꢀlayer growth can be inferred
from the dependence d0 Ts . The large values of d0
= f( )
at low temperatures and the reduction in this parameꢀ
ter with rising Ts suggested polymolecular sorption of
the reagents at Ts < 450 K and their interaction in the
adsorbed layer, followed by the formation of hydrated
oxide. At
T > 450 K, the value of d0 approached the
double metalꢀoxide interlayer spacing for the given
oxide. These conditions could therefore be characterꢀ
ized as the layer growth conditions. For comparison,
0
20
40
60
N
Fig. 2. Growth dynamics of Al O3 (1) and Ta O (2) films
2 5
on a silicon matrix at Ts = 523 K with TEA.
2
Fig. 3 (
1'
and
2') depicts d0
=
f
(Ts)
for Та О and Al O3
2
5
2
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84 No. 2 2010

264
EZHOVSKII
are given in Table 1. This table also presents data
marked with asterisks) for layers that were subjected
d0, nm
(
–
2
to thermal treatment in a vacuum (10 Pa) at
T =
0
.6
773 K for 1 h. According to these data, the oxide nanoꢀ
structures remained virtually unhydrated at Ts > 523 K
and their composition corresponded to Та О and
2
5
1
Al O3, respectively. The high oxygen concentration
2
*
was most probably due to hydration of the oxides at
low synthesis temperatures and the small contribution
from the oxygen of the matrix oxide layer at high synꢀ
0
0
.4
1
'
*
2
thesis temperatures (mainly Al O3).
2
*
*
*
*
* *
IR spectroscopic analysis by the method of multiꢀ
ple disturbed total internal reflection (MDTIR) of
aluminum and tantalum oxide layers synthesized on
silicon elements at Ts = 553 K confirmed the XPS
data. One could, for example, clearly see the absorpꢀ
*
.2
2
'
0
4
tion bands characteristic of Si–O–Al bonds in alumiꢀ
00
450
500
550
600
Ts, K
–1
num silicates with
7
ν
= 1130 cm and those with
ν
=
–1
44 and 835 cm , typical of AlꢀO bonds with sixꢀ and
fourꢀcoordination aluminum [15].
Electron diffraction analysis of the structure of synꢀ
thesized oxide films more than 10 nm thick showed
Fig. 3. Effect of synthesis temperature on the growth
parameter of Al O 1, 1') and Ta O (2, 2') layers on siliꢀ
con with (1, 2) and without (1', 2') exchange activator (triꢀ
(
2
3
2 5
ethylamine).
that Та О5 layers were amorphous over the temperaꢀ
2
ture interval. Films of Al O3 deposited at Ts < 450 K
2
were amorphous too, while
them at high temperatures.
α
ꢀAl O3 was observed in
2
the surface and the formation of twoꢀdimensional
island structures.
The similar tension for vapors of the tantalum and
Composition of the synthesized products was studꢀ aluminum halogenides, along with the similar temperꢀ
ied by the methods of IR and XPS spectroscopy for ature conditions for layerꢀbyꢀlayer growth of their
films more than 10 nm thick. According to the XPS oxide systems (Fig. 3), considerably simplified syntheꢀ
data, the lines characteristic of halogenides (
2p levels) were absent at energies of 75–120 eV, while the need to heat the entire system of halogenide vapor
2
s
ꢀ, and sis of multilayered structures on their basis, obviating
+
feed to the reactor to the temperature of the haloꢀ
genide source.
Since synthesis of aluminum oxide layers by treatꢀ
ment of the surface of the tantalum oxide layer with
aluminum halogenide vapor might involve the substiꢀ
tution reaction
lines typical of 1s levels of nitrogen of the NR4 group
were absent near energies of 401–402 eV [12]. TEA
was then fully desorbed during the synthesis process.
Peaks with energies of 100–150 eV, which are characꢀ
teristic of 2
for either Та О or Al O layers with
s
and 2
p
levels of silicon, were not detected
> 5 nm, indicatꢀ
d
2
5
2
3
Ta O + AlХ
2
Al O + TaХ ,
2
(V)
ing layer continuity at this thickness.
5
3
3
5
the thermal effects were estimated and the equilibrium
constant (Kf, T) was calculated by the Temkin–
Schwarmann method, using equations for the isobaric
The oxygen–metal ratio in the oxide layer at differꢀ
ent synthesis temperatures was found from the area of
the peaks of oxygen ( b
respective metals ( b = 74.6 eV for Al , Eb = 118.8 eV potential
for Al , and Eb = 26.8 eV for Ta ). The relevant results
E = 530–534 eV) and the
E
Δ
= –
GT = 2.3RT logKf, T and ΔGT° ΔHT°
2p
2s
4f
°
TΔST
.
Calculations performed for AlCl3 and AlBr3 demꢀ
onstrated (Table 2) that the substitution reaction can
be eliminated at Ts < 470 K for AlBr3 and at Ts < 600 K
for AlCl3. With this in mind, multilayered systems were
synthesized using tantalum and aluminum chlorides at
Ts = 473 K. This corresponded to the conditions of
layerꢀbyꢀlayer growth for each component (Fig. 3).
Table 1. Oxygen–metal ratio in aluminum and tantalum
oxide layers at different synthesis temperatures
Ts, K
[O]/[Al]
[O]/[Ta]
423
473
523
553
573
773*
4.0
2.3
1.8
1.5
1.6
1.5
4.6
3.0
2.6
2.5
2.5
2.5
Ellipsometric analysis of the growth dynamics of
multilayered nanostructures [16] revealed that linear
dependence
tic of the oxide (
Al O3) (Fig. 2) persisted within layers of each oxide,
d
=
f
≈
(
N
)
with a parameter d0 characterisꢀ
0.3 nm for Та О5 and 0.45 nm for
2
2
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 2
2010

SURFACE NANOSTRUCTURES BASED ON TANTALUM AND ALUMINUM OXIDES
265
suggesting that the mechanism of their layerꢀbyꢀlayer
formation was maintained.
I
/
I0
,
rel. units
(a)
1
.0
The composition and the distribution profile of eleꢀ
ments were studied according to synthesized structure
by the methods of XPS and Auger spectroscopy, taking
twoꢀlayer Si–Al O –Та О and Si–Та О –Al O sysꢀ
4
1
2
3
2
5
2
5
2
3
tems with layers up to 20 nm thick each. The XPS
spectra did not exhibit the energy maxima characterisꢀ
tic of chlorine and silicon, testifying to the completeꢀ
ness of surface reactions and the continuity of the
deposited layers at the said thickness.
0.5
2
3
Lines with Eb = 118.9 and 74.8 eV (corresponding
to
2s and 2p levels of Al) and a line with Eb = 26.8 eV
(
typical of 4f levels of Ta) were clearly identified in the
XPS spectra.
(
b)
Examination of the Auger spectra from layerꢀbyꢀ
layer etching with argon ions (etching rate ~0.5 nm/min)
showed that the distribution of Ta and Al in the nanoꢀ
1.0
0.5
1
4
structured systems with
alternation of their oxide layers during synthesis
Fig. 4). It should be noted that, in the case of multiꢀ
d > 10 nm corresponded to
(
layered nanostructures synthesized with aluminum
bromide at Ts = 573 K, there was no clearly defined
zonal distribution of Ta and Al but rather a mixture of
oxides. This confirmed the validity of assessing synꢀ
thesis conditions from changes in the isobaric potenꢀ
tial of the system. A signal from the silicon matrix
appeared only after a 40ꢀmin etching period; at the
given etching rate, this was in good agreement with the
total thickness of the oxide nanostructure.
3
2
0
20
40
60
80
, min
τ
Electrophysical parameters were measured on
nanostructures ~60 nm thick synthesized on alumiꢀ
num films. These nanostructures were synthesized
with TEA and were formed according to analogous
Fig. 4. Intensities of Auger peaks of (
num, ( ) tantalum, and ( ) silicon during layerꢀbyꢀlayer
etching of Si–Ta O –Al O (a) and Si–Al O –Ta O (b)
1) oxygen, (2) alumiꢀ
3
4
3
2
5
2
2
3
2 5
nanostructures.
laws. The effect of highꢀtemperature (T = 873 K) therꢀ
mal treatment was studied, taking structures with a silꢀ
icon matrix as samples. Vacuumꢀevaporated nickel noteworthy contribution from the diffusion processes
films served as electrodes.
(
T
= 873 K) only at
The conductivity (
d < 5 nm.
–1
–1
Zonal distribution of the metal in the oxide nanoꢀ
σ
,
Ω
cm ) of the multilayered
–12
–14
structure (Fig. 4) ought to result in the properties of nanostructures changed from 10 in Та О5 to 10 in
2
the nanostructure differing from those of the oxide Al O3 [16]. At di
,
dj > 5 nm it was described well by the
mixture. The most characteristic parameter is the perꢀ relationship for an equivalent circuit if the multilayꢀ
mittivity ( ). This was determined by measuring the
2
ε
capacitance of the system on samples having a total
thickness of nearly 60 nm. Both the thickness of each
layer di (Al O ) and dj (Та О ) and their alternation
°
Table 2. Calculated ΔG (kJ/mol) and Кf,T for reaction (5)
T
2
3
2
5
were different in the test samples. A comparison of
experimental values of the permittivity of multilayered
systems with values calculated in terms of the Lanꢀ
dau–Lifshitz model for a statistical mixture of dielecꢀ
trics [17] showed a considerable difference in all the
cases (Table 3). Only when the samples were heat
Δ
G
°
Кf,T
Тs, К
AlBr3
AlСl3
AlBr3
AlСl3
400
420
450
10.01
4.55
18.2
16.41
13.73
11.94
9.25
–1.35
–0.39
–0.81
0.36
–2.38
–2.04
–1.60
–1.32
–0.95
–0.45
2.61
0.70
treated for two hours (
T
= 873 K) did values of
ε
approach those calculated for a mixture of dielectrics,
with the best match observed in samples with a thin
470
–3.26
–4.60
–6.32
–10.51
500
550
600
0.48
layer (di
,
dj 5 nm) of at least one of the system comꢀ
≅
4.77
0.60
ponents (Table 3, Nos. 1, 3, and 7). This suggested a
zonal distribution of the oxides in the dielectric and a
–0.3
0.92
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 2
2010

266
EZHOVSKII
Table 3. Permittivity (
ε
) of multilayered Al O –Ta O nanostructures of different compositions
3
2
2 5
Alternation
of oxide layers
No.
di, nm
dj, nm
ε
1
ε
2
ε
3
1
2
3
4
5
6
7
5
25
10
20
10
10
5
5
25
5
аtаtаtаtаtаt
аt
13.6
13.7
12.2
12.2
15.8
18.1
17.7
14.8
14.8
13.0
13.0
16.9
19.35
18.7
14.8
14.1
12.8
12.7
16.2
18.5
18.7
tаtаtаt
tаtа
10
20
50
20
аtаt
аt
аtаtа
Note: d = Al O layer thickness; d = Ta O layer thickness; a, t = alternation of oxide layers, a = aluminum oxide layer, t = tantalum oxide
i
2
3
j
2 5
layer; ε , ε , and ε = experimental, calculated, and experimental (following thermal treatment) values.
1
2
3
ered structure was viewed as a chain of seriesꢀconꢀ
nected resistors with the conductivities and j:
2. A. L. Egorov, Yu. K. Ezhovskii, and S. I. Kol’tsov,
Zh. Fiz. Khim. 59, 683 (1985).
σ
σ
i
3
. G. V. Anikeev, Yu. K. Ezhovskii, and S. I. Kol’tsov,
Izv. Akad. Nauk SSSR, Neorg. Mater. 24, 619 (1988).
1
/σ =
(1/σ ) +
(1/σ ).
j
∑
(3)
i
∑
4
. E. Granneman, Solid State Technol., No. 5, 92 (2000).
. J. Gelatos, L. Chen, H. Chung, and R. Thakur, Solid
State Technol. 2, 44 (2003).
The relationship (3) was not fulfilled if at least one
component of the structure was <5 nm thick. This was
most likely due to manifestation of the charge transfer
tunneling mechanism, and to conditions at the interꢀ
faces of the multilayered structure.
5
6
. R. L. Puurunen, J. Appl. Phys. 97, 121301 (2005).
7
. Yu. K. Ezhovskii and P. M. Vainshtein, Zh. Fiz. Khim.
7
1
, 2222 (1997) [Russ. J. Phys. Chem. A 71, 2009
(
1997)].
CONCLUSIONS
8
. Yu. K. Ezhovskii and A. I. Klyuikov, Zh. Fiz. Khim. 72
,
On the whole, the regular features established for
9
08 (1998) [Russ. J. Phys. Chem. A 72, 804 (1998)].
formation of nanostructures of Та О and Al O3 oxides
2
5
2
9. Yu. K. Ezhovskii, Zh. Fiz. Khim. 61, 1598 (1987).
and their multilayered systems reflect a general tenꢀ
dency that the temperature factor influences the 10. V. K. Gromov and S. I. Kol’tsov, in Ellipsometry:
growth mechanism of MLꢀdeposited layers. It was
A Method of Studying Surface, Collected Vol., Ed. by
A. V. Rzhanov (Nauka, Novosibirsk, 1983), p. 73
in Russian].
shown in particular that formation of Al O and Та О5
2
3
2
[
layers by alternate chemisorption of vapors of metal
halogenide and water can be accomplished in three
ways: reaction between components in a polymolecuꢀ
lar adsorbed layer and the formation of hydrated
oxides; successive buildup of monomolecular layers
1
1. V. K. Gromov, Introduction to Ellipsometry (Lening.
Gos. Univ., Leningrad, 1986) [in Russian].
1
2. V. I. Nefedov and V. T. Cherepin, Physical Methods of
Studying the Surface of Solids (Nauka, Moscow, 1983)
(
layerꢀbyꢀlayer growth); and formation and further
[in Russian].
development of twoꢀdimensional island structures.
The use of TEA as the exchange activator in the synꢀ 13. V. I. Nefedov, Xꢀray Electron Spectroscopy of Chemical
Compounds, Handbook (Khimiya, Moscow, 1984)
thesis process intensifies halogenide chemisorption,
[in Russian].
stabilizes the process of oxide layer formation, and
extends the temperature range of layerꢀbyꢀlayer
growth. Evaluation of basic dielectric characteristics
of the synthesized nanostructures showed that their
parameters allow use of the molecular lamination
method for the synthesis of dielectric structures
1
4. Handbook of Xꢀray Photoelectron Spectroscopy (Perkinꢀ
Elmer Corp. Phys. Electron. Division, Flying Cloud
Drive, Edom Prairie, Minnesota, 1978).
1
5. A. A. Tsyganenko, P. P. Mardilovich, G. M. Lysenko,
and A. I. Trokhimets, in Photonics Achievements, Iss. 9,
Collected Vol. (Lening. Gos. Univ., Leningrad, 1987),
p. 28 [in Russian].
(
including multilayered structures) in submicron eleꢀ
ments of microꢀ and nanoscale electronic systems.
1
6. Yu. K. Ezhovskii and A. I. Klusevich, Fiz. Tverd. Tela
45, 2099 (2003) [Phys. Solid State 45, 2203 (2003)].
REFERENCES
1
7. V. B. Lazarev, V. G. Krasov, and I. S. Shaplygin, Electriꢀ
cal Conductivity of Oxide Systems and Film Structures
(Nauka, Moscow, 1979) [in Russian].
1
. V. B. Aleskovskii, Vestn. Akad. Nauk SSSR
6, 48
(1975).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 2
2010