22nd International Symposium on Plasma Chemistry
July 5-10, 2015; Antwerp, Belgium
Low temperature plasma-assisted atomic layer deposition of silicon nitride
moisture permeation barrier layers
A.-M. Andringa1, A. Perrotta1, K. de Peuter1, H.C.M. Knoops1,2, W.M.M. Kessels1 and M. Creatore1
1
Department of Applied Physics, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, the
Netherlands
2
Oxford Instruments Plasma Technology, North End, Bristol BS49 4AP, U.K.
Abstract: Silicon nitride layers have been deposited by means of low temperature plasmaassisted atomic layer deposition. The calcium test pointed out the excellent barrier quality
of the layers, in the order of 10-6 g/m2day for a layer thickness of 10 - 40 nm. The
microstructure of the layers was investigated by means of ellipsometric porosimetry, which
pointed out the absence of accessible pores larger than 0.3 nm in diameter, thus confirming
the excellent results provided by the calcium test.
Keywords: plasma-assisted atomic layer deposition, permeation barriers, silicon nitride
1. Introduction
Emerging technologies such as OLED lighting and
displays, organic thin film transistors and (flexible) thin
film solar cells have become very attractive due to their
performance in combination with large-area processing
and flexibility. However, in order for these devices to
have a long lifetime and high reliability, there are a
number of challenges to be met. One of them is the
efficient encapsulation of the device against moisture
permeation. To ensure a lifetime of 10 years for OLEDs,
an encapsulation layer is required with a water vapor
transmission rate (WVTR) of 10-6 g/m2/day. For flexible
devices developed on polymer substrates, there is also
need for large area, low temperature processed (generally
below 120 °C) thin-film barrier layers. Barrier layer
solutions commonly adopt inorganic thin films deposited
by sputtering, plasma-enhanced chemical vapor deposited
(PECVD) and (plasma-assisted) atomic layer deposition
(ALD). Generally, it can be concluded from literature
studies that thinner ALD layers are found to outperform
thicker PECVD or sputtered layers, in terms of barrier
performance [1].
Water permeation has been universally found to occur
unhindered through local pinholes/defects present in the
inorganic layer. Therefore, a more effective approach
consists in developing multi-layers of inorganic barrier
layers and polymer-like interlayers [2]. These latter have
been shown to provide a defect decoupling effect, thereby
reducing the number of substrate-induced pinholes in the
subsequent barrier layer. Such approach can lead to than
three orders of magnitude less permeable to water and
oxygen than a single inorganic layer. Next to the local
defects, water permeation occurs also through the (bulk)
nano-porosity of the inorganic barrier [3]. Previously, we
have reported on the correlation between the intrinsic
barrier properties and the pore size and relative pore
content of PECVD and plasma-assisted ALD inorganic
(e.g.SiO 2 and Al 2 O 3 ) barriers [4].
O-20-6
In this contribution, we investigate the moisture
permeation barrier properties of plasma-assisted ALD
SiN x layers. Deposition of high-quality SiN x by ALD has
turned out to be very challenging, especially in
controlling the impurity levels in the layer, when
operating at substrate temperatures compatible with
thermally-sensitive substrates. Using a recently, PA-ALD
process in our group developed, SiN x is grown at a
substrate temperature between 80 and 200 °C by
alternating the precursor SiH 2 (NHtBu) 2 (BTBAS) and the
reactant N 2 plasma.
2. Experimental
The PA-ALD SiN x depositions were carried out using
an Oxford Instruments FlexALTM reactor, described in
detail elsewhere [5]. SiH 2 (NHtBu) 2 (BTBAS, purity
≥98.5%, Air Products Inc.) was used as the precursor and
held at a bubbler temperature of 50 °C. The reactor is
equipped with a remote inductively coupled (ICP) plasma
generator, which was operated at 600 W at 13.56 MHz. A
mixture of N 2 (100 sccm, purity 99.999%) and Ar (200
sccm) was used to generate the plasma. The depositions
were performed on c-Si substrates, with a thin native
oxide (SiO 2 ) layer. 100 nm thick SiN x layers were
deposited at 80, 120, 160 and 200 °C, according to the
recipe outlined in Figure 1. The film thickness and
optical properties of the layers were obtained by SE, using
a J.A. Woollam Co. M-2000F ellipsometer over a
wavelength range of 400-1000 nm. The chemical
composition and stoichiometry of the SiN x films were
investigated with X-ray photoelectron spectroscopy
(XPS), using a Thermo Scientific K-Alpha spectrometer
with a monochromatic Al Κα X-ray source (hν = 1486.6
eV). Further chemical characterization of the SiN x films
was carried out by means of Fourier-transform infrared
spectroscopy (FTIR) with a Bruker Tensor 27
spectrophotometer. The open porosity of the layers was
investigated by means of ellipsometric porosimetry. This
1
Normalized absorption (a.u.)
technique combines the sensitivity of spectroscopic
ellipsometry with the isothermal adsorption studies of
probe molecules which can infiltrate in the open porosity
of the later as well as develop multi-layer adsorption on
its surface [4,6]. Three solvents were used as adsorptives:
water (kinetic diameter d H2O ~ 0.3 nm), ethanol (d C2H5OH
~ 0.4 nm) and toluene (d C6H5CH3 ~ 0.6 nm).
The intrinsic barrier performance of the (10, 20, 40 nm
thick) SiN x films deposited at 120C was evaluated by
means of the calcium test, in an environmental chamber at
20 °C and 50% relative humidity (RH).
SiNx ALD
80 °C
120 °C
160 °C
200 °C
νas Si-N-Si
νs Si-O-Si
νSi-H
νas Si-O-Si
νSi-C
δN-H
4000
δSi-N
νC=O
νN-H
3000
2000
1000
-1
Wavenumber (cm )
Fig. 2. FT-IR spectra of SiN x layers at different substrate
temperatures.
Fig. 1. Atomic layer deposition recipe of SiN x films.
3. Results and Discussion
The SiN x growth per cycle (GPC), thin film refractive
index (n) and chemical composition in the bulk of the
layer are reported in Table 1. The lower GPC at higher
deposition temperature is a result of the lower impurity
content (carbon) due to improved ligand removal and,
therefore, higher film density, at higher temperature. This
conclusion is supported by the trends in refractive index
and chemical composition.
Table I. GPC, refractive index and chemical composition
for the layers at different substrate temperature.
T
(°C)
80
120
160
200
GPC
(Å/cycle)
Refr. index
(633 nm)
N/Si
0.44 ±
0.02
0.33
0.26
0.24
1.802 ±
0.002
1.839
1.884
1.904
1.91 ±
0.1
1.69
1.64
1.60
XPS
[C]
(at%)
[O]
(at%)
14 ± 1
2±1
9
6
5
3
3
2
The chemical composition of the SiN x layers deposited
at the indicated substrate temperatures was investigated
with FTIR. Typical FTIR spectra are shown in Error!
Reference source not found., reporting the characteristic
(hydrogenated) SiN x absorption bands. The Si-N-Si
stretching bands continue to grow stronger with
increasing deposition temperature, and at the same time,
all C, O and H contamination related absorption peaks
decrease in intensity, in agreement with the XPS data.
2
Figure 3a shows the adsorption/desorption isotherms
for the 100 nm thick SiN x layers deposited at 80, 120, 160
and 200 °C, when water is adopted as adsorptive. The
variation of the optical thickness of the SiN x layer
combined with the optical thickness of the adsorptive
layer (n SiNx *d SiNx + n solvent *d multilayer ), which accounts for
the total adsorbed quantity, has been reported as a
function of P l /P sat . The adsorption isotherm is of type II
(non-porous) [7]. The observed small increase at P l /P sat
<0.2 is attributed to adsorptive monolayer formation,
followed by multilayer formation that grows stronger
closer to P l /P sat is 1. This unrestricted multilayer growth
is typical for non-porous layers and depends on the SiN x
deposition temperature and the solvent used. For ethanol
and toluene as adsorptives, the same conclusions are
drawn.
In order to distinguish between the two contributions to
the optical thickness, i.e. the refractive index and
thickness, the refractive index of the SiN x layer is
separately presented in Error! Reference source not
found.. The refractive index of the SiN x layers remains
constant upon increasing the water vapor pressure,
indicating that no infiltration of the probing molecules in
the SiN x layer occurs. Taking into account the size of the
smallest probe molecule, i.e. water, it can be concluded
that no micropores larger than 0.3 nm are present.
Similarly, no open porosity permeable to ethanol and
toluene is detected.
O-20-6
4. Conclusions
Silicon nitride moisture permeation barriers have been
fabricated using low temperature plasma-assisted ALD.
The deposited films have been characterized in terms of
refractive index and chemical composition by
spectroscopic ellipsometry (SE), XPS and FTIR,
respectively. The calcium test has been performed to
determine the intrinsic water vapor transmission rate
(WVTR). Intrinsic WVTRs in the order of 10-6 g/m2/day
for 10-40 nm thick silicon nitride layers indicate an
excellent barrier quality. The microstructure of the layers
was studied by ellipsometric porosimetry (EP), using
water (d= 0.3 nm), ethanol (d = 0.4 nm) and toluene (d =
0.6 nm) as probe adsorptive molecules. Adsorption
isotherms have been recorded by ellipsometric
porosimetry. Irrespective of the tested deposition
conditions (deposition temperature in the range 80120°C), no uptake of the probe in the layer is observed,
indicating the absence of accessible pores larger than 0.3
nm in diameter. The isotherms follow a type II behavior.
These results demonstrate the non-porous nature of the
silicon nitride films and confirm the excellent results in
terms of intrinsic WVTR values.
Fig.3. a) Water adsorption isotherms for SiN x layers
deposited at the indicated temperatures. The change in
optical thickness of the SiN x and the adsorptive
multilayer (n SiNx *d SiNx + n solvent *d multilayer ) is presented as
a function of P l /P sat (P l /P sat is the ratio of the partial
pressure to the saturation vapor pressure) and the
isotherms are offset for clarity. b) Refractive index as a
function of P l /P sat during adsorption of water (a) on SiN x
samples. The standard deviation of n is 0.0002.
The main effect of increasing P l /P sat is therefore an
increase in multilayer thickness of the probe molecule on
top of the non-porous barrier layer (not shown here): the
shapes of the thickness adsorption isotherms are
obviously similar to the overall adsorption isotherm of
type II in Fig. 3a.
In conclusions, EP measurements on the PA-ALD SiN x
layers showed that no nanopores larger than 0.3 nm are
present. Previously, it has been demonstrated for oxides
that absence of porosity larger than 0.3 nm leads to
excellent moisture barriers with a WVTR in the range of
10-6 g/m2/day. Therefore, we tested the intrinsic barrier
properties of the SiN x layers with the Ca test.
The intrinsic WVTR of the SiN x barrier films has been
found in the range of 10-6 g/m2/day range. Strikingly, the
measured intrinsic WVTR is, within the error margin of
2⋅10-6 g/m2day, not dependent on the SiN x layer
thickness. Apparently, only 10 nm of PA-ALD SiN x are
sufficient to have a closed continuous layer on the Ca
plate.
O-20-6
5. Acknowledgments
The authors would like to thank Pieter Klaassen and
Marc Kuilder (Philips Research, High Tech Campus,
Eindhoven, The Netherlands) for the Ca test. This work is
supported by NanoNextNL, a micro and nanotechnology
programme of the Dutch ministry of economic affairs,
agriculture and innovation (EL&I) and 130 partners.
6. References
[1] W. Keuning, P. van der Weijer, H. Lifka, W.M.M.
Kessels, M. Creatore, J.Vac.Sci.Technol., 30, 01A131
(2012)
[2] G. Graff, R. Williford, P. Burrows, J.Appl.Phys., 96,
1840 (2004).
[3] J. Affinito, D. Hilliard, 47th S.V.C. Annual Tech. Conf.
Proc., 563 (2004).
[4] A. Perrotta, E. van Beekum, G. Aresta, A. Jagia, W.
Keuning, M.C.M. van de Sanden, W.M.M. Kessels, M.
Creatore, Microporous and Mesoporous Mat., 188,
163171 (2014).
[5] S. Heil, J. van Hemmen, C. Hodson, N. Singh, J.
Klootwijk, F. Roozeboom, M.C.M. van de Sanden,
W.M.M. Kessels, J. Vac. Sci. Technol. A, 25, 1357
(2007).
[6] S. Eslava, M. Baklanov, C. Kirschhock, F. Iacopi, S.
Aldea, K. Maex, J. Martens, Langmuir, 23, 12811 (2007).
[7] K.S.W. Sing, D.H. Everett, L. Moscou, R.A. Pierotti,
J. Rouquerol, T. Siemieniewska, Pure and Applied
Chemistry, 56, 603 (1985).
3