Surface passivation of c-Si by atmospheric pressure chemical vapor deposition of Al2O3

Article abstract of DOI:10.1063/1.4718596

Atmospheric pressure chemical vapor deposition of Al₂O ₃ is shown to provide excellent passivation of crystalline silicon surfaces. Surface passivation, permittivity, and refractive index are investigated before and after annealing for deposition temperatures between 330 and 520 °C. Deposition temperatures >440 °C result in the best passivation, due to both a large negative fixed charge density (～2 × 10¹²cm^-2) and a relatively low interface defect density (～1 × 10¹¹eV^-1cm^-2), with or without an anneal. The influence of deposition temperature on film properties is found to persist after subsequent heat treatment. Correlations between surface passivation properties and the permittivity are discussed.

Full text of DOI:10.1063/1.4718596

Surface passivation of c-Si by atmospheric pressure chemical vapor deposition of
Al2O3
Lachlan E. Black and Keith R. McIntosh
Citation: Applied Physics Letters 100, 202107 (2012); doi: 10.1063/1.4718596
View online: http://dx.doi.org/10.1063/1.4718596
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/100/20?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Drastic reduction in the surface recombination velocity of crystalline silicon passivated with catalytic chemical
vapor deposited SiNx films by introducing phosphorous catalytic-doped layer
J. Appl. Phys. 116, 044510 (2014); 10.1063/1.4891237
Reaction kinetics during the thermal activation of the silicon surface passivation with atomic layer deposited
Al2O3
Appl. Phys. Lett. 104, 061606 (2014); 10.1063/1.4865901
Si surface passivation by Al2O3 thin films deposited using a low thermal budget atomic layer deposition process
Appl. Phys. Lett. 102, 131603 (2013); 10.1063/1.4800541
Chemical and structural study of electrically passivating Al2O3/Si interfaces prepared by atomic layer deposition
J. Vac. Sci. Technol. A 30, 04D106 (2012); 10.1116/1.4704601
Interface engineering for the passivation of c-Si with O3-based atomic layer deposited AlOx for solar cell
application
Appl. Phys. Lett. 100, 143901 (2012); 10.1063/1.3701280
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29

APPLIED PHYSICS LETTERS 100, 202107 (2012)
Surface passivation of c-Si by atmospheric pressure chemical vapor
deposition of Al2O3
Lachlan E. Black1,a) and Keith R. McIntosh2
1Centre for Sustainable Energy Systems, College of Engineering and Computer Science, Australian National
University, Canberra ACT 0200, Australia
2PV Lighthouse, Coledale NSW 2515, Australia
(Received 18 November 2011; accepted 29 April 2012; published online 16 May 2012)
Atmospheric pressure chemical vapor deposition of Al2O3 is shown to provide excellent
passivation of crystalline silicon surfaces. Surface passivation, permittivity, and refractive index
are investigated before and after annealing for deposition temperatures between 330 and 520 ꢀC.
Deposition temperatures >440 ꢀC result in the best passivation, due to both a large negative
fixed charge density (ꢁ2 ꢂ 1012 cmꢃ2
) and a relatively low interface defect density
(ꢁ1 ꢂ 1011 eVꢃ1 cmꢃ2), with or without an anneal. The influence of deposition temperature on film
properties is found to persist after subsequent heat treatment. Correlations between surface
C
V
passivation properties and the permittivity are discussed. 2012 American Institute of Physics.
[http://dx.doi.org/10.1063/1.4718596]
Silicon surface passivation by aluminum oxide (Al2O3)
is currently a subject of significant research within the photo-
voltaic industry. The distinctive feature of this material is its
high density of negative fixed charge, which combined with
a relatively low interface defect density has been shown to
provide excellent passivation of silicon surfaces.1–10 In par-
ticular, it is able to effectively passivate both diffused and
undiffused p-type surfaces, a task for which conventional
positively charged silicon nitride (SiNx) passivation layers
have generally been found to be unsuited11,12 (although it
should be noted that good passivation of boron emitters with
SiNx has recently been observed in some cases following a
firing step13). An effective means of passivating such surfa-
ces is necessary both to overcome efficiency limitations
imposed by the use of conventional Al-diffused “back sur-
face fields,” and to enable the adoption of cells based on
n-type silicon, with its attendant advantages.14
sure, self-limiting, or continuous reactions), and chemistry
(precursors) may result in fundamentally different film struc-
tures and properties, and passivation may only be achieved
under a limited range of process conditions.
Despite the interest in high-throughput alternatives to
ALD, one comparatively simple high-throughput technique
has so far been neglected, namely atmospheric pressure
chemical vapor deposition (APCVD). This is ironic, given
that APCVD was used in the first work on surface passiva-
tion with Al2O3, that of Hezel and Jaeger in 1989.24 While
they achieved promising results, their work was not followed
up until 2006, and then with ALD.5,6 In this paper, we inves-
tigate the surface recombination velocity, interface defect
density, and fixed charge of APCVD Al2O3 on silicon surfa-
ces, as a function of deposition temperature and annealing,
using photoconductance and capacitance-voltage (C-V) tech-
niques. It is found that Al2O3 deposited by APCVD can pro-
vide excellent surface passivation of crystalline silicon
surfaces, comparable to the best results of PECVD and ALD.
Furthermore, we demonstrate the dependence of passivation
properties on deposition temperature and subsequent anneal-
ing and show how these properties are correlated with the
permittivity of the films.
Al2O3 for surface passivation has to date been mostly
deposited by conventional atomic layer deposition (ALD), a
technique that allows for very fine control of film thickness
and a high degree of conformality, but which does not
provide the high throughput necessary for application in the
photovoltaic industry. Recent work has, therefore, focused
on alternative high-throughput deposition techniques,
such as plasma-enhanced chemical vapor deposition
(PECVD),8–10,15,16 reactive sputtering,10,17 and high-rate
spatial ALD.10,18,19 Promising results have been achieved
with all of these methods, although sputtered films still lag
behind PECVD and ALD in terms of surface passivation.
Other deposition techniques such as sol-gel20,21 and evapora-
tion22 have so far shown only marginal passivation potential.
It is evident that passivation depends sensitively on interfa-
cial rather than bulk film properties,23 and that these may be
strongly dependent on the method used to synthesize the
films. Even for techniques based on similar principles, such
as ALD and PECVD, differences in activation method (ther-
mal or plasma), deposition regime (temperature and pres-
Al2O3 films were deposited using an inline APCVD belt
furnace system manufactured by SierraTherm Production
Furnaces Inc. A detailed description of this system is given
elsewhere.25 Triethyldialuminum tri-(sec-butoxide) (TEDA-
TSB)26 was used as the chemical precursor. Like trimethyla-
luminum (TMA), the precursor generally used in ALD and
PECVD deposition of Al2O3, this material forms a low-
viscosity liquid at room temperature, but unlike TMA it has
the practical advantage of being non-pyrophoric and, there-
fore, easier to handle. TEDA-TSB was reacted with water
vapor at substrate temperatures between 330 and 520 ꢀC to
form Al2O3. Substrates were passed under the injector head
on a moving belt at a speed of 12 in. per minute. All films
deposited at a given temperature were deposited in the same
run. Post-deposition annealing was performed in a quartz
tube furnace under various ambients at either 400 or 425 ꢀC.
a)Electronic mail: lachlan.black@anu.edu.au.
C
V
0003-6951/2012/100(20)/202107/5/$30.00
100, 202107-1
2012 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29

202107-2
L. E. Black and K. R. McIntosh
Appl. Phys. Lett. 100, 202107 (2012)
Lifetime measurements were performed on 325 lm
thick 0.8 X cm p-type h100i FZ-Si wafers. The wafers were
etched in TMAH to remove saw damage, cleaned by the
RCA (Radio Corporation of America) procedure, and dif-
fused with phosphorus to getter iron impurities. The wafers
were then etched in HCl to remove the phosphorus diffu-
sions, cleaned by the RCA procedure, and given an HF dip
and de-ionized (DI) water rinse to remove the native oxide
just prior to the deposition of APCVD Al2O3. The Al2O3
was deposited on the front surface and then the rear surface
immediately after.
The samples prepared for capacitance-voltage measure-
ments were fabricated from 8 X cm p-type h100i Cz-Si
wafers and 1 X cm n-type h100i FZ-Si wafers. The front
surfaces of these wafers were polished. The wafers were
cleaned by the RCA procedure, including an HF dip and DI
water rinse, just prior to the deposition of APCVD Al2O3. Al
dot contacts were thermally evaporated on the front side
through a shadow mask, while GaIn paste was applied to the
rear side to form an ohmic contact. For annealed samples,
annealing occurred prior to contact deposition.
FIG. 1. seff and Seff,UL as a function of Al2O3 deposition temperature.
Values are presented as-deposited and after various successive 30 min
post-deposition anneals. Results from two samples, taken from different
wafers, are shown at each deposition temperature and annealing step. Lines
are plotted between mean values at each temperature.
Carrier lifetime was measured using a Sinton Labs
WCT-120 photoconductance tool operated in either transient
or generalized quasi-steady-state mode. The upper limit of
the effective surface recombination velocity Seff,UL was cal-
culated according to
tion measurements (Filmetrics F20 Thin Film Measurement
System) with the extracted values of n and k. The reported
values are averages derived from multiple measurements
over the area of each sample. Standard deviation was always
less than 5% and generally less than 3%. The dielectric con-
stant j was calculated from this thickness and the average in-
sulator capacitance determined from both high-frequency
(1 MHz) and quasi-static C-V measurements on the same
samples.
ꢀ
ꢁ
W
1
1
Seff;UL
¼
ꢃ
;
2
seff sbulk;intrinsic
where W is the sample thickness, seff is the measured effec-
tive minority carrier lifetime at an excess carrier concentra-
tion of 1015 cmꢃ3, and sbulk,intrinsic is the Auger-limited
intrinsic bulk lifetime determined using the parameterization
of Kerr and Cuevas.27
Figure 1 shows effective minority carrier lifetime as a
function of deposition temperature prior to annealing (blue
circles) and after successive anneals. The lifetime increases
markedly with increasing deposition temperature over the
range 330–440 ꢀC, and plateaus at 440–520 ꢀC. Subsequent
annealing results in increased lifetime, with the relative
increase being greater for lower deposition temperatures.
Whether the annealing ambient is important is unclear from
this exploratory study and will be the subject of further
work. It is interesting to note that the influence of deposition
temperature on lifetime is reduced, but not eliminated by
subsequent anneals. This suggests that a significantly larger
thermal budget is needed to achieve the same level of surface
passivation when applied after rather than during deposition.
A minimum Seff,UL of 10 cm/s was achieved at a deposi-
tion temperature of 440 ꢀC, after annealing. This is the best
value of surface recombination velocity reported to date for
Al2O3 on such heavily doped substrates. It is comparable to
the best values reported for Al2O3 from thermal and spatial
ALD (both 4 cm/s) and PECVD (5 cm/s) for less heavily
doped 1.3 X cm p-type samples,10 and much better than the
value of 55 cm/s reported for sputtered films on the same
substrates.17 For comparison, Seff,UL values of 5 cm/s and
12 cm/s have been reported as exemplary values for alnealed
SiO2 and SiNx, respectively, on 1 X cm p-type sub-
strates.32,33 We note that we have measured lower values of
High frequency and quasi-static capacitance-voltage
measurements were performed using an HP 4284 A Precision
LCR Meter and HP 4140B Picoammeter/DC Voltage
Source. Capacitance measurements were corrected for para-
sitic series resistance and inductance, leakage currents, and
dielectric dispersion. Gate area was determined on an indi-
vidual basis using optical microscopy. The method of Ghi-
baudo et al.28 was used to determine the insulator
capacitance, while the flatband voltage was determined from
the offset of the measured and ideal high-frequency curves
in depletion, after correction for interface trap stretchout.29
The midgap interface defect density Dit was extracted using
the combined high-frequency/quasi-static capacitance
method.30 Fixed charge Qf was calculated by assuming all
charge to be located at the Si–Al2O3 interface, using work
function values of WAl ¼ 4.23 V and WSi(100) ¼ 4.71 V for
the aluminum gate and silicon substrate, respectively.31
A
more detailed description of the analysis procedure used is in
preparation for a separate publication.
The Al2O3 refractive index was determined as a function
of deposition temperature by fitting a combination of polar-
ized multi-angle reflectance and fixed-angle spectral ellips-
ometry measurements with optical modeling software. Film
thickness for individual capacitance-voltage samples was
then determined by fitting normal incidence spectral reflec-
S
eff,UL than those reported here with the same APCVD Al2O3
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29

202107-3
L. E. Black and K. R. McIntosh
Appl. Phys. Lett. 100, 202107 (2012)
FIG. 3. Fixed charge density Qf/q as a function of film deposition tempera-
ture, both as-deposited and after annealing at 400 ꢀC in N2 for 30 min. Val-
ues are shown both for p and n-type samples. As-deposited and annealed
values are for separate samples with films deposited in the same run. Lines
are plotted between mean values for p- and n-type samples at each
temperature.
FIG. 2. Midgap interface defect density Dit as a function of film deposition
temperature, both as-deposited and after annealing at 400 ꢀC in N2 for
30 min. As-deposited and annealed values are for separate samples with
films deposited in the same run. Lines are plotted between mean values at
each temperature.
films on different substrates, and these are reported else-
where,25 where they are compared with the results of other
techniques.
ence in apparent Qf at flatband for n- and p-type samples at
lower temperatures is consistent with the effect of interface
trapped charge Qit, which scales directly with Dit and will
most likely be positive for p-type samples and negative for
n-type.29 The true value of Qf at these temperatures will,
therefore, be somewhere between the measured p- and
n-type values.
The C-V measurements show how the deposition tem-
perature and the post-deposition anneal influence the inter-
face states and the charge density, which in turn determine
surface recombination. Figure 2 shows the interface defect
density and Figure 3 the fixed charge density for films depos-
ited at temperatures between 330–520 ꢀC, both before and
after annealing in N2 at 400 ꢀC for 30 min. Interface defect
density decreases with increasing deposition temperature,
falling sharply between 410–440 ꢀC and leveling out at
ꢁ1.4 ꢂ 1011 eVꢃ1 cmꢃ2 above 470 ꢀC. Annealing results in a
reduction in defect density across all deposition tempera-
tures, although Dit remains higher for films deposited at
lower temperatures. For films deposited at temperatures
above 440 ꢀC, annealing has only a minor effect on Dit,
reducing it to ꢁ1.2 ꢂ 1011 eVꢃ1 cmꢃ2 for temperatures above
470 ꢀC.
The extracted values of Dit and charge compare closely
to those reported in the literature for ALD and PECVD,
which are typically on the order of ꢁ1 ꢂ 1011 eVꢃ1 cmꢃ2 for
Dit and 1012–1013 cmꢃ2 for charge.1–3,7–9,15,16,34 They are
similar to the results reported recently by Dingemans et al.7
of Dit ¼ 1 ꢂ 1011 eVꢃ1, and Qf /q ¼ ꢃ2.4 ꢂ 1012 cmꢃ2 for
thermal ALD films annealed at the same temperature. Nota-
bly, however, such excellent values are achieved here even
without any post-deposition annealing. The values of Dit are
the lowest reported for as-deposited Al2O3. Indeed, the life-
time results confirm that there is little need to anneal the
films to gain excellent surface passivation, so long as deposi-
tion occurs above 440 ꢀC.
These trends for Dit correlate directly to those for Seff,UL
,
but the passivation is also influenced by the charge density in
the films. Figure 3 shows that the charge density is zero,
within error, for films deposited at 355 ꢀC or below, and
increases sharply at higher temperatures, appearing to pla-
teau at ꢃ1.8 to ꢃ2.2 ꢂ 1012 cmꢃ2 above 410 ꢀC. Following
annealing, the trend is reversed, with films deposited at lower
temperatures having higher charge, with the apparent excep-
tion of the films deposited at the lowest temperature, 330 ꢀC.
Above 440 ꢀC, charge is actually reduced somewhat by
annealing. It is interesting to note that the value of Qf is very
similar for substrates of both polarities when Dit is small.
This allows us to be confident in the work function values
used in the extraction and indicates that the charge in the
film is independent of the substrate doping. The large differ-
It is desirable to link changes in Dit and Qf to structural
changes in the film, and these may be elucidated by studying
other film properties. Figure 4 shows Al2O3 film thickness,
refractive index n, and both high-frequency and static dielec-
tric constant j as a function of deposition temperature before
and after annealing. Film thickness varies between 13 and
25 nm and is generally thicker for lower temperatures. No
change in thickness is observed with annealing. Previous
studies have observed no thickness dependence of Al2O3
passivation for thicknesses >ꢁ10 nm, so this variation is
unlikely to be the cause of the differences in passivation
properties observed here.23 Refractive index is practically
constant with temperature, with an average value of n ¼ 1.62
at a wavelength of 632 nm. Since refractive index is related
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29

202107-4
L. E. Black and K. R. McIntosh
Appl. Phys. Lett. 100, 202107 (2012)
where an Seff,UL of 13 cm/s was attained before annealing
and 10 cm/s after. It was shown that this temperature pro-
vides the lowest Dit, both before and after annealing. It was
also shown that there is little charge within the films when
deposited at <355 ꢀC and that it increases with temperature
to up to 2.2 ꢂ 1012 cmꢃ2 above 410 ꢀC, but increases above
this value for lower temperatures after the anneal. Seff,UL, Dit
and charge were found to have a dependence on deposition
temperature that persisted even after annealing. Changes in
these values were observed to be linked to variations of the
permittivity, suggesting that the reduction of Dit at higher
temperatures is linked to lattice restructuring in the film, and
that there is a relationship between the generation of nega-
tive charge centers and the reduction of permittivity during
deposition and annealing.
The authors would like to thank Kenneth M. Provancha
and James N. Cotsell for their invaluable technical
assistance.
1B. Hoex, S. B. S. Heil, E. Langereis, M. C. M. van de Sanden, and W. M.
M. Kessels, Appl. Phys. Lett. 89, 042112 (2006).
FIG. 4. Al2O3 film thickness, refractive index n at a wavelength of 632 nm,
and both high-frequency (1 MHz) and static dielectric constants j as a func-
tion of Al2O3 deposition temperature. The dielectric constant is shown
before and after annealing. For the thickness and refractive index, the mea-
surement error is generally smaller than the symbol size. The relative error
for j is estimated to be approximately 3%, most of which is due to system-
atic error in the extraction of the insulator capacitance.
2G. Agostinelli, A. Delabie, P. Vitanov, Z. Alexieva, H. F. W. Dekkers, S.
De Wolf, and G. Beaucarne, Sol. Energy Mater. Sol. Cells 90, 3438
(2006).
3B. Hoex, J. Schmidt, P. Pohl, M. C. M. van de Sanden, and W. M. M.
Kessels, J. Appl. Phys. 104, 044903 (2008).
4G. Dingemans, P. Engelhart, R. Seguin, F. Einsele, B. Hoex, M. C. M. van
de Sanden, and W. M. M. Kessels, J. Appl. Phys 106, 114907 (2009).
5G. Dingemans, R. Seguin, P. Engelhart, M. C. M. van de Sanden, and W.
M. M. Kessels, Phys. Status Solidi RRL 4, 10 (2010).
6J. Schmidt, B. Veith, F. Werner, D. Zielke, and R. Brendel, in Proceedings
of the 35th IEEE PVSC, Honolulu, USA, 20–25 June 2010 (IEEE, 2010),
pp. 885–890.
to density, this suggests that density also does not vary sig-
nificantly with temperature. The reduction in thickness at
higher temperatures is, therefore, due to a lower deposition
rate, which is consistent with the presence of parasitic gas-
phase reactions.35 In contrast, the dielectric constant varies
strongly with deposition temperature, increasing from a min-
imum of 7.0 at 385 ꢀC to 8.9 at 520 ꢀC (taking the values at
1 MHz). This is most likely due to a gradual change in the
bonding configuration within the amorphous films, resulting
in a larger contribution of lattice polarization.36 We note that
this increase in permittivity is correlated with the reduction
of Dit. The sharp increase of the dielectric constant below
355 ꢀC is likely to be related to significantly increased hydro-
gen incorporation in these films (probably in the form of OH
groups),37 resulting in a strong dipolar contribution.38 This
transition appears to coincide with the appearance of signifi-
cant negative charge in the films. It is notable that the magni-
tude of the post-annealing reduction in permittivity of the
films deposited at the lowest two temperatures is correlated
with the change in fixed charge density of the same samples
(Figure 3). This could be explained if desorption of hydrogen
during annealing (and consequent reduction of the dipolar
contribution to permittivity) is linked to the formation of
negative charge centers (possibly oxygen interstitials39). The
reason why a smaller reduction is observed for the lowest
temperature film is unclear, though it is possible that its
greater thickness may restrict hydrogen desorption. Further
investigation is needed to establish these relationships more
firmly.
7G. Dingemans, N. M. Terlinden, D. Pierreux, H. B. Profijt, M. C. M. van
de Sanden, and W. M. M. Kessels, Electrochem. Solid State 14, H1
(2011).
8G. Dingemans, M. C. M. van de Sanden, and W. M. M. Kessels, Electro-
chem. Solid State 13, H76 (2010).
9G. Dingemans, P. Engelhart, R. Seguin, M. M. Mandoc, M. C. M. van de
Sanden, and W. M. M. Kessels, in Proceedings of the 35th IEEE PVSC,
Honolulu, USA, 20–25 June 2010 (IEEE, 2010), pp. 3118–3121.
10J. Schmidt, F. Werner, B. Veith, D. Zielke, R. Bock, V. Tiba, P. Poodt, F.
Roozeboom, A. Li, A. Cuevas, and R. Brendel, in Proceedings of the 25th
EUPVSEC, Valencia, Spain, 6–10 September 2010 (WIP, 2010), pp.
1130–1133.
11S. Dauwe, L. Mittelsta¨dt, A. Metz, and R. Hezel, Prog. Photovolt: Res.
Appl. 10, 271 (2002).
12M. J. Kerr, Ph.D. dissertation, Australian National University, Canberra,
2002.
13F. Chen, I. Romijn, A. Weeber, J. Tan, B. Hallam, and J. Cotter, in
Proceedings of the 22nd EUPVSEC, Milan, Italy, 3–7 September 2007
(WIP, 2007), pp. 1326–1331.
14D. Macdonald and L. J. Geerligs, Appl. Phys. Lett. 85, 4061 (2004).
15P. Saint-Cast, D. Kania, M. Hofmann, J. Benick, J. Rentsch, and R. Preu,
Appl. Phys. Lett. 95, 151502 (2009).
16S. Miyajima, J. Irikawa, A. Yamada, and M. Konagai, Appl. Phys. Express
3, 012301 (2010).
17T. T. Li and A. Cuevas, Phys. Status Solidi RRL 3, 160 (2009).
18P. Poodt, A. Lankhorst, F. Roozeboom, K. Spee, D. Maas, and A. Ver-
meer, Adv. Mater. 22, 3564 (2010).
19G. Dingemans and W. M. M. Kessels, in Proceedings of the 25th
EUPVSEC, Valencia, Spain, 6–10 September 2010 (WIP, 2010), pp.
1083–1090.
20P. Vitanov, A. Harizanova, T. Ivanova, and T. Dimitrova, Thin Solid
Films 517, 6327 (2009).
21H. Xiao, C. Zhou, X. Cao, W. Wang, L. Zhao, H. Li, and H. Diao, Chin.
Phys. Lett. 26, 088102 (2009).
Our results show that excellent passivation can be
achieved with Al2O3 using a relatively simple deposition
technique, APCVD. It is evident that the optimal deposition
temperature for the precursors used lies at 440–520 ꢀC,
22M. Y. Seo, E. N. Cho, C. E. Kim, P. Moon, and I. Yun, in Proceedings of
the 3rd INEC, Hong Kong, China, 3–8 January 2010 (IEEE, 2010),
pp. 238–239.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29

202107-5
L. E. Black and K. R. McIntosh
Appl. Phys. Lett. 100, 202107 (2012)
23F. Werner, B. Veith, D. Zielke, L. Ku¨hnemund, C. Tegenkamp, M. Seibt,
R. Brendel, and J. Schmidt, J. Appl. Phys. 109, 113701 (2011).
24R. Hezel and K. Jaeger, J. Electrochem. Soc. 136, 518 (1989).
25L. E. Black, K. M. Provancha, and K. R. McIntosh, in Proceedings of the.
26th EUPVSEC, Hamburg, Germany, 5–9 September 2011 (WIP, 2011),
pp. 1120–1124.
30R. Castagne´ and A. Vappaille, Surf. Sci. 28, 157 (1971).
31H. Kawano, Prog. Surf. Sci. 83, 1 (2008).
32M. J. Kerr and A. Cuevas, Semicond. Sci. Technol. 17, 35 (2002).
33M. J. Kerr and A. Cuevas, Semicond. Sci. Technol. 17, 166 (2002).
34B. Hoex, J. J. H. Gielis, M. C. M. van de Sanden, and W. M. M. Kessels,
J. Appl. Phys. 104, 113703 (2008).
26R. G. Gordon, K. Kramer, and X. Liu, Mater. Res. Soc. Symp. Proc. 446,
383 (1997).
35E. J. Kim and W. N. Gill, J. Cryst. Growth 140, 315 (1994).
36H. Momida, T. Hamada, and Y. Takagi, Phys. Rev. 73, 054108 (2006).
37A. N. Gleizes, C. Vahlas, M. Sovar, D. Same´lor, and M. Lafont, Chem.
Vap. Deposition 13, 23 (2007).
27M. Kerr and A. Cuevas, J. Appl. Phys. 91, 2473 (2002).
28G. Ghibaudo, S. Bruye`re, T. Devoivre, B. DeSalvo, and E. Vincent, IEEE
Trans. Semicond. Manuf. 13, 152 (2000).
38K. Murase, Jpn. J. Appl. Phys. 33, 1385 (1994).
29E. H. Nicollian and J. R. Brews, MOS (Metal Oxide Semiconductor)
Physics and Technology (Wiley, New York, 1982), pp. 477–486.
39J. R. Weber, A. Janotti, and C. G. Van de Walle, J. Appl. Phys. 109,
033715 (2011).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 158.42.28.33
On: Thu, 18 Dec 2014 10:05:29