Gas jet assisted vapor deposition of yttria stabilized zirconia
D. D. Hass and H. N. G. Wadleya兲
Department of Materials Science and Engineering, School of Engineering and Applied Science,
University of Virginia, Charlottesville, Virginia 22903
共Received 29 August 2008; accepted 26 January 2009; published 27 February 2009兲
A gas jet assisted electron beam evaporation process for synthesizing yttria stabilized zirconia
共YSZ兲 coatings has recently been reported. The process uses a rarefied inert gas jet to entrain and
transport vapor to a substrate. The gas jet enables the lateral spreading of the flux to be controlled
and large fractions of the vapor to be deposited on samples of relatively small size. When the gas
pressure is high, coatings grown at 1050 ° C and below have a columnar structure and a high pore
fraction. The total pore volume fraction, the morphology of the inter- and intracolumn pores and the
coating texture are all observed to be a strong function of the gas pressure in the chamber with
increasing chamber pressure leading to larger intercolumnar pore spacings, wider pores, a higher
total pore volume fraction, and a reduction in the coating texture. A direct simulation Monte Carlo
simulation approach has been used to investigate vapor transport for the various gas pressures
explored in this study. The simulation indicates that as the gas pressure increases, binary scattering
events between the vapor and background gas broaden the vapor molecule incidence angle
distribution. This intensifies flux shadowing and results in the incorporation of voids in the coating.
Increasing the gas pressure also results in a rapid increase in the vapor phase nucleation of YSZ
clusters. This observation coincides with a transition from a 关200兴 textured columnar morphology at
moderate pressures to a nanogranular structure with no texture and a very high nanoscopic pore
volume fraction at high pressures. © 2009 American Vacuum Society. 关DOI: 10.1116/1.3085725兴
I. INTRODUCTION
Yttria stabilized zirconia 共YSZ兲 thermal barrier coatings
共TBCs兲 are widely employed in the thermal protection systems used to enhance the high temperature life of hot section
components in gas turbine engines.1,2 These superalloy components are protected from high temperature oxidation by a
coating system consisting of a metallic bond coat based upon
platinum modified nickel aluminides3 or MCrAlY alloys,4 a
thermally grown 共ideally pure ␣-alumina兲 oxide 共TGO兲
layer5 and the YSZ thermal protection layer. Yttria partially
stabilized zirconia containing about 7 wt % Y2O3 共7YSZ兲 is
usually chosen as the thermal barrier material because of its
low thermal conductivity,6 high melting point,7 thermal expansion coefficient close to that of the substrate metal,8 and
its thermochemical compatibility with the TGO layer that it
contacts during service. These 7YSZ coatings are usually
applied to hot section engine components by either by air
plasma spray9 共APS兲 or electron beam physical vapor deposition 共EB-PVD兲.10 However other gas jet processes are also
being explored including electron beam evaporation combined with gas jet enabled directed vapor deposition11 and a
solution precursor plasma spray 共SPPS兲 process.12 The performance of a TBC coating is measured by its ability to
reduce the temperature of the TGO layer 共thereby reducing
its growth rate兲, its resistance to spallation during thermal
cycling and erosion/foreign object damage by small 共or
a兲
Author to whom correspondence should be addressed; electronic mail:
[email protected]
404
J. Vac. Sci. Technol. A 27„2…, Mar/Apr 2009
larger兲 particle impacts and its susceptibility to reaction with
calcium aluminum magnesium silicates that accumulate
upon the coating surface during use.
When thermal transport through the TBC occurs by conduction, the coating thermal resistance is controlled by its
thermal conductivity. The TBC layer thermal conductivity
depends upon that of the material used to make the layer and
the volume fraction and morphology of the pores within the
layer.13 In APS coatings, repeated molten ceramic droplet
impingement, spreading and rapid solidification leads to the
formation of long pores parallel to the substrate. The orientation of these pores is highly effective at impeding the flow
of the heat through the coating and the through thickness
thermal conductivity of 7YSZ coatings made this is very low
共0.8– 1.0 W / mK at 25 ° C兲.14 However, pores in this orientation provide little compliance to accommodate in the plane
differential strains created between the coating and substrate
by their thermal expansion mismatch. The coatings made by
this approach are therefore susceptible to spallation during
thermal cycling. A recent variant of the APS deposition processes has resulted in dense vertically cracked structures
which are more resistance to thermal cycling. The SPPS process also results in structures with mixtures of pore shapes
and orientations15 and those containing voids and cracks in
the through thickness direction are claimed to offer significant increases in thermal cycle lifetimes.16 An ultrafast variant of the plasma deposition technique has also been recently
proposed for TBC deposition.17 In this approach some of the
ceramic particles are evaporated during their travel within
the plasma and vapor condensation, gas phase nucleated
nanocluster impacts and liquid droplet impingement occur
0734-2101/2009/27„2…/404/11/$25.00
©2009 American Vacuum Society
404
405
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
simultaneously resulting in a structure of low conductivity.
The cyclic spallation life of this novel coating has not been
reported.
TBC coatings with the best thermal cycling life are usually made by the EB-PVD process.18 The 7YSZ source material is vaporized by impingement of an electron beam upon
a source ingot and transported under low pressure
共⬍10−3 Pa兲 conditions to a heated air foil where condensation occurs. By rotating the air foil in the vapor plume, a
porous coating can be produced at deposition temperatures in
the 1000– 1050 ° C range. These vapor deposited YSZ coatings have highly 关200兴 textured, columnar microstructures
with primary intercolumnar pores oriented in the coating
growth direction. These help to accommodate the 共in-plane兲
coefficient of thermal expansion mismatch between the coating and the metal substrate thereby enhancing the cyclic
spallation life. However, these intercolumnar pores are orientated perpendicular to the plane of the coating and therefore only modestly reduce the coatings thermal
conductivity.19 Finer scale 共initially feathery兲 pores are also
present within the columns and are much more effective at
impeding the heat flow through the coating. The through
thickness thermal conductivity of EB-PVD YSZ coatings is
still significantly higher than that of the APS coatings 共it lies
in the 1.5– 1.9 W / mK range at 25 ° C兲 共Ref. 20兲 but can be
varied by modifying the rotation rate and substrate temperature which effect the pore fraction and morphology.14 Recent
modeling has begun to establish the causal linkages between
the pore morphology and performance in these systems21 and
has led to the identification of novel zig-zag 共or herringbone兲
morphologies that seek to achieve both in plane compliance
and low through thickness thermal conductivity.22 These
studies indicate that control of the pore morphology is essential if optimal coating performance is to be achieved.23
Recent experimental and atomistic modeling studies indicate that the morphology in vapor deposited coatings is affected by many processing variables. These include the vapor atoms kinetic energy,24,25 its incidence angle,26,27 the
substrate’s temperature,28,29 the deposition rate,28,30 the background pressure,28,31,32 vapor plume composition,33 and the
substrate’s roughness.34 The experimental studies clearly
show that porous, columnar morphologies are associated
with low kinetic energy, oblique vapor incidence angles, low
substrate temperatures, high deposition rates, rough substrates, and high chamber pressures.24,26,28,34,35 Atomistic
models generally support these observations36 and link pore
formation to flux shadowing under conditions of restricted
adatom mobility.37
Several groups have attempted to use atomistic simulations to investigate the relationships between the pore volume fraction and morphology and the conditions used to
grow a vapor deposited film. The most fundamental approach based upon density functional theory is too computationally constrained for the problems encountered in even
thin film growth. Instead, the interatomic bonding has been
coarse grained and represented by an interatomic potential.
Potentials for many materials have been proposed,38 but
JVST A - Vacuum, Surfaces, and Films
405
none as yet have reliably addressed the interatomic forces
that control the structure and properties of yttria stabilized
zirconia. Molecular dynamics methods have been utilized
with some of the potentials proposed for metallic systems to
investigate pore control.39 Unfortunately computational constraints force the use of greatly accelerated deposition rates
for the simulations and an underestimation of the contribution of diffusional processes upon the atomic assembly of a
coating.
Gilmer et al. pioneered the use of a kinetic Monte Carlo
modeling approach to vapor deposition that does not require
acceleration of the deposition rate.40 In this approach, the
activation barriers for diffusion of the species of interest are
precalculated 共for example, by density functional theory or
by molecular statics using interatomic potentials兲 and then
the system of interest is assembled atom by atom allowing
diffusion to occur at time steps defined by the smallest activation energy diffusive event. A two-dimensional version of
the approach has been used to explore the multidiffusion
path assembly of thin nickel films using activation barriers
obtained from an embedded atom potential for nickel.41 The
approach successfully predicted the pore structures and
trends in pore fraction observed experimentally in metal systems. Cho et al. extended the approach to investigate the
growth of columnar structures in coatings that were rotated
in a vapor plume and showed that local high points on the
surface shadowed the incident flux and this controlled the
growth of the both the primary and feathery pore structures
observed in TBC coatings.42 These atomistic modeling approaches reveal that the pore fraction and morphology can be
controlled by the molecular weight and incidence angle distribution of the depositing flux, the diffusion rate on the
growth surface and the initial surface upon which deposition
occurs.
The low chamber pressures 共⬍10−4 Pa兲 used to facilitate
electron beam propagation in the EB-PVD process results in
collisionless vapor transport to the substrate. When the substrate is far from the vapor source, this collisionless transport
results in a narrow vapor angle of incidence distribution
upon a flat, substrate positioned perpendicular to the flux.
Very few oblique atom impacts with the substrate occur and
shadowing is therefore reduced. While the vapor atoms kinetic energy is low 共⬃0.2 eV兲 the porous, columnar morphology is only achieved when the flux incidence angle distribution is broadened to promote flux shadowing. In the EBPVD process, this is normally accomplished by rotating the
substrate. This rotation is also utilized to ensure uniform
coating of air foil components. The ability to exert greater
control of the pore shape and distribution during the vapor
deposition of TBCs is limited by limitations to the design of
current EB-PVD systems which provide little ability to control the incidence angle of the vapor other than by sample
rotation.
There are other ways of manipulating the angle of incidence distribution of the vapor atoms/molecules. One approach is to entrain the flux in a supersonic gas jet that is
directed to the substrate.43 Within the jet, the pressure can be
406
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
high 共10– 100 Pa兲 and the mean free path is greatly reduced
so that many vapor phase collisions occur. For example, the
mean free path of zirconia molecules in helium is predicted
by binary collision theory to decrease from 10 m to 1 mm
when the background pressure is increased from
10−3 to 10 Pa.44 Multiple scattering events during the transport of vapor both widens the angular incidence distribution
and thermalizes the flux. However, increasing the pressure
also increases the frequency of three body collisions and
therefore the probability of nucleating vapor phase clusters.
An electron beam directed vapor deposition 共DVD兲 process has been proposed as a means for depositing zirconia
based coatings at high pressure 共10– 100 Pa兲.45 Preliminary
studies indicated that YSZ layers with porous, columnar
morphologies could be deposited onto stationary substrates
using this technique.46 Subsequent experiments have shown
that TBCs with very high pore fractions and very low thermal conductivities 共comparable to APS coatings兲 can be
made by combining gas phase scattering with substrate rotation. However, a detailed study of the dependence of the
coating morphology on pressure has not been conducted.
Here, the morphology, texture and density of YSZ films deposited over a range of pressures is systematically investigated. Altering the chamber pressure is observed to greatly
affect the characteristics of the coatings. A direct simulation
Monte Carlo 共DSMC兲 approach is used to estimate the flow
field characteristics for the range of pressures used in the
experiments. These flow fields are then used to simulate multiple scattering of the vapor flux and to deduce the angle of
incidence distribution of the vapor flux at the stationary substrate. This allows the mechanisms responsible for the large
changes in the coatings pore morphology and microstructure
to be identified.
II. EXPERIMENT
A. Deposition process
In a DVD process, material is vaporized using a high
voltage/medium power 共60 kV/ 10 kW兲 axial e-beam gun.
By incorporating differential pumping of the gun column and
a very small 共⬃3 mm diameter兲 e-beam exit opening into the
process chamber the gun can function in a high pressure
environment 共up to 200 Pa in helium兲. Figure 1 shows a
schematic illustration of the process.47 The gas jet was created by supersonic expansion through a nozzle near the site
of evaporation. This was achieved by maintaining an
upstream/downstream pressure ratio in excess of two. The
gas jet entrained the vapor and transported it to the substrate.
The gas jet was introduced into the chamber via a 1.27 cm
nozzle opening. In the present experiments an upstream/
processing chamber pressure ratio of between 2.5 and 10.0
was used. The process chamber pressure was varied from
13.3 to 106.4 Pa 共Table I兲. The substrates onto which deposition occurred were 25.4 mm diameter IN100 buttons with a
nickel aluminide bond coat 共supplied by GE Aircraft兲. The
substrate surface had a root mean square 共rms兲 roughness of
405.10 nm and a mean roughness 共Ra兲 of 106.66 nm.48 StaJ. Vac. Sci. Technol. A, Vol. 27, No. 2, Mar/Apr 2009
406
FIG. 1. Experimental setup used for the deposition of partially stabilized
zirconia 共containing 7 wt % yttria兲 on platinum modified nickel aluminied
coated IN100 substrates at a temperature of 1050 ° C.
tionary substrates were positioned 12.0 cm from the nozzle
共x-direction兲 and 9.5 cm from the source. The height
共y-direction兲 of the substrate was set so that the bottom of the
substrate was level with the top of the crucible. This resulted
in the center point of the substrate being slightly lower than
the nozzle center point and maximized the deposition efficiency for many of the process conditions used.48
7YSZ vapor was created by evaporation from a porous
7YSZ source rod 共Transtech, Inc.兲. To compensate for oxygen lost from the source during evaporation, a heliumoxygen 共3.0 vol % oxygen兲 carrier gas was used to create the
gas jet. The substrate was heated to 1050 ° C using a resistive
heater placed behind the substrate. The substrate temperature
was monitored using a thermocouple inserted in a drilled
hole centered with respect to the substrate thickness. This
limited errors from convective cooling of the thermocouple
by the carrier gas flow and by radiant heating of the thermocouple from by the heater. Evaporation was allowed to continue until coating thicknesses of between 50 and 100 ␮m
were achieved. The deposition rate was between 2.0 and
6.0 ␮m min−1.
TABLE I. DVD process conditions employed.
Chamber
pressure
共Pa兲
Pressure
ratio
Gas
flow
共slm兲
Deposition
rate
共␮m / min兲
13.3
26.6
39.9
53.2
66.5
79.8
93.1
10.0
10.0
6.6
5.0
4.0
3.4
2.5
5.0
8.0
8.0
8.0
8.0
8.0
8.0
2.0
3.8
5.8
7.7
6.9
4.9
2.6
407
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
407
FIG. 2. Schematic illustration defining the parameters used to quantify the
morphology of the porosity found in 7 YSZ coatings.
B. Coating characterization
The coating density was measured directly from the substrates weight gain and dimensional changes. For many of
the conditions investigated, elongated, intergrowth column
pores were observed that extended from the substrate to the
coating surface. Finer, intracolumnar porosity was also observed. Figure 2 schematically illustrates the coating morphology. The pore morphology has been characterized by
determining the primary pore width, ␳, and the intercolumnar
spacing, ␧ on mechanically polished cross sections of the
coatings. All pore parameters were measured within
6.35 mm of the center of the substrate. The pore spacing was
measured at the midplane of the coating. The pore width was
measured at three positions: 10 ␮m below the coating surface, at the coating midpoint, and 10 ␮m above the substrate.
For the measurement of the pore width it was recognized that
the polished surface resulted in an expanded width due to the
range of angles that the pores intersected the polished cross
section. These angles varied from 0° to 70°. The error was
corrected by multiplying the measured value of the spacing
by a correction factor of 1.22. This corresponds to an intersection angle of 35° 共the midpoint between two extreme values of 0° and 70°兲. The coating texture was determined by
x-ray diffraction through comparison of the peak intensities
with those of a randomly orientated sample. Error bars of
one standard deviation are shown on all graphs of data.
III. RESULTS
Figure 3 shows the cross-section 共a兲 and surface views
关共b兲 and 共c兲兴 of a coating deposited at the lowest investigated
JVST A - Vacuum, Surfaces, and Films
FIG. 3. SEM micrographs showing 共A兲 the cross section and 共B兲 the top
surface of a 7 YSZ coating deposited using a chamber pressure of 13.3 Pa
and a pressure ratio of 10.0. Note the intercolumnar pores that extend from
the substrate to the coating surface.
pressure 共13.3 Pa兲. This and all other coatings deposited had
a columnar structure. The primary growth diameters did not
generally vary with thickness. The columns were approximately parallel sided and not highly tapered like those seen
in EB-PVD samples.13 Wide, intercolumnar pores up to
5 ␮m in width defined the primary columns. These extended
from the substrate to the coatings outer surface, Fig. 4共A兲.
The primary columns had a fractal structure with each column subdivided by smaller intracolumnar pores.
The intercolumnar pore width, the intercolumnar pore
spacing, and the total pore volume fraction of the yttria stabilized zirconia coatings all varied with chamber pressure. In
Fig. 4, the effect of chamber pressure on the intercolumnar
pore spacing and width is shown. The pore spacing steadily
increased with chamber pressure. The pore width increased
with chamber pressure until a maximum width was observed
at an intermediate pressure. Further pressure increases led to
a reduction in pore width.
The total pore volume fraction 共deduced from the macroscopic density data兲 exhibited a large increase 共from
2 to 50 vol %兲 as the pressure was increased from
13.3 to 106.4 Pa, Fig. 5. The intercolumnar pore volume
fraction is also shown. The largest intercolumnar pore vol-
408
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
408
FIG. 5. Plot showing the measured change in 共A兲 the total pore volume
fraction and 共B兲 the intercolumnar pore volume fraction in YSZ coatings
deposited using a range of chamber pressures 共13.3– 106.4 Pa兲. Note the
dramatic increase in pore volume fraction as the chamber pressure increased. The difference between the curves in 共A兲 and 共B兲 is a measure of
the intracolumnar pore volume fraction.
the chamber pressure was increased. A randomly orientated,
nanocrystalline structure resulted at the highest pressure
共106.4 Pa兲.
IV. DIRECT SIMULATION MONTE CARLO
SIMULATIONS
FIG. 4. Plot showing the change in 共A兲 the intercolumnar pore width and 共B兲
the intercolumnar pore spacing with chamber pressure. Note the systematic
increase in both parameters as the chamber pressure increased.
ume fraction occurred at the lowest investigated pressure
共13.3 Pa兲 due primarily to the close spacing of the intercolumnar pores in this condition. The difference between the
total pore volume and the intercolumnar pore volume gives a
measure of the intracolumnar pore volume. This systematically increased with pressure. The pore volume fraction data
are summarized in Table II. The intracolumnar pore volume
fraction data corresponded well with the observed coating
morphology. At the lowest pressure 共13.3 Pa兲 the columns
appeared relatively dense. As the pressure increased the column sides developed a feathery appearance. At the highest
growth pressure 共106.4 Pa兲 the columns had a granular appearance and contained fewer intercolumnar pores, Fig. 6.
The coating texture was examined by x-ray diffraction.
The peak intensities normalized by the 兵111其 peak intensity
are given in Table III and plotted versus pressure in Fig. 7.
Coatings deposited at the lowest pressure 共13.3 Pa兲 had a
共200兲 preferred orientation. This 共200兲 texture disappeared as
J. Vac. Sci. Technol. A, Vol. 27, No. 2, Mar/Apr 2009
A DSMC technique was used to simulate binary collisions
during transport of the vapor to the substrate thus estimate
the vapor incidence angle distribution 共LAD兲.49 A two-step
modeling approach was used. In step 1 the gas jet supersonic
expansion and its interaction with a flat substrate was analyzed. The code output was the gas jet flow field in the region near the substrate. This was used in step 2 to analyze
the interaction of the gas jet with the vapor flux and to estimate the incidence angle distribution at the substrate. Details
can be found in Ref. 48.
Representative simulations of the gas jet velocity field
components in the axial and radial directions during the su-
TABLE II. Pore parameters for each pressure condition.
Chamber
pressure
共Pa兲
Interclumnar
pore
spacing 共␮m兲
Intercolumnar
pore
width
共␮m兲
Intercolumnar
pore
volume
共%兲
Total
pore
volume
共%兲
13.3
26.6
39.9
53.2
66.5
79.8
93.1
3.3
11.8
15.5
24.9
48.4
69.0
133
0.4
0.5
0.9
1.4
1.9
1.9
1.4
12.1
4.2
5.8
5.6
3.9
2.7
1.1
¯
7.0
¯
31.5
¯
38.5
45.6
409
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
409
FIG. 7. XRD data showing the change in the relative intensity of the 共200兲
peak with respect to the 共111兲 peak. High 共200兲 peak intensities indicate the
presence of a 共200兲 preferred orientation. Note the increased 共200兲 peak
intensity as the chamber pressure was decreased.
FIG. 6. High magnification SEM micrograph showing cross sections of zirconia coatings deposited using 共A兲 a chamber pressure of 13.3 Pa and a
pressure ratio of 10.0 and 共B兲 a chamber pressure of 106.6 Pa and a pressure
ratio of 2.5. Note the smoother and denser appearance of the low pressure
condition and the porous, granular appearance of the high pressure case.
The IAD at a given location on the substrate 共near to the
region where porosity measurements were performed兲 was
estimated for the various process conditions. The results, Fig.
9, show that many of the molecules impinged the substrate at
oblique angles of incidence. To characterize changes in the
distribution, two parameters were defined and are given in
Table IV: 共i兲The peak maximum angle, ␪m, defined as the
incident angle at which the maximum intensity, Io, occurred
and 共ii兲 the peak dispersion width, Pw, defined as the width
of the distribution at 0.5 Io.
In Fig. 10, Pw is plotted as a function of chamber pressure. As the chamber pressure increased Pw broadened. ␪m
personic expansion are shown in Fig. 8. In the axial direction
the maximum velocity decreased as the chamber pressure
increased, Table IV. This was the result of a decrease in the
pressure ratio with higher chamber pressure. The interaction
of the gas with the substrate resulted in the formation of a
strong radial component to the velocity 共i.e., the wall jet
velocity兲 near the substrate. The maximum wall jet velocity
is also given in Table IV.
TABLE III. Relative peak intensities from XRD patterns of the yttria stabilized zirconia films deposited with different chamber pressures.
Chamber
pressure 共Torr兲
13.6
26.6
39.9
53.2
66.5
Random
I兵111其
I兵200其
I兵220其
I兵311其
I兵400其
1
1
1
1
1
1
0.56
0.32
0.25
0.30
0.26
0.20
0.55
0.37
0.38
0.43
0.47
0.56
0.53
0.33
0.34
0.36
0.37
0.38
0.23
0.10
0.07
0.09
0.09
0.07
JVST A - Vacuum, Surfaces, and Films
FIG. 8. DSMC simulations showing the speed of a helium-3.0% oxygen
carrier gas in 共A兲 the axial direction and 共B兲 the radial direction. The axial
component was reduced in the region near the substrate and significant
radial component developed 共i.e., the wall jet兲.
410
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
410
TABLE IV. DSMC simulation results for the gas jet expansion and IAD distribution.
Chamber
pressure
共Torr兲
Maximum
velocity
Z-dir 共m/s兲
Wall jet
velocity
共m/s兲
Average
IAD
共deg兲
Peak
maximum, Pm
共deg兲
Peak width,
Pw 共at 0.5Pm兲
共deg兲
13.3
26.6
39.9
53.2
66.5
79.8
1455
1635
1482
1300
1270
1175
147
255
200
167
126
94
16.5
18.0
16.7
9.5
5.9
4.4
18
22
15
14
14
5
61
68
92
92
101
101
was observed to decrease from −18° at the lowest pressure
共13.3 Pa兲 to 0° at the highest pressure 共106.4 Pa兲.
V. HIGH PRESSURE VAPOR DEPOSITION
When vapor deposition is performed using chamber pressures greater than ⬃0.1 Pa, the mean free path for vapor
atoms collisions with the background gas becomes less than
the typical source-to-substrate distance 共30– 50 cm兲. These
collisions can alter the coating morphology in two ways. The
first is by altering flux shadowing mechanisms. Binary collisions scatter the vapor onto various trajectories resulting in
changes to the IAD of the depositing flux. This controls
shadowing phenomena. The second is the nucleation and
growth of multiple atom clusters. The impingement of clusters on the substrate has been observed to affect the coating
growth process.50,51 These mechanisms are discussed below
in detail.
A. Flux shadowing effects
Flux shadowing results in porosity when oblique atom
arrivals are “shadowed” by fast growing grains or surface
asperities. This creates local reductions in the density of va-
FIG. 9. Change in the IAD distribution with chamber pressure is given.
Higher chamber pressures resulted in broader distributions and peak positions close to the substrate normal.
J. Vac. Sci. Technol. A, Vol. 27, No. 2, Mar/Apr 2009
por flux that impinges on the substrate. A reduced local
growth rate results in the formation of pores in the coating.
The degree to which this occurs depends upon the IAD and
the adatom surface mobility. A high fraction of oblique adatom arrivals and a low surface mobility promote porosity. A
high surface mobility will compensate the local flux depletion by surface diffusion.
An increased chamber pressure during DVD deposition
alters the mean free path of vapor molecules as well as the
velocity distribution of the background gas. This combination controls the distribution of incidence angles of the depositing molecules. Atomistic models have revealed that
changes to these distributions will result in changes to the
coating microstructure.36
B. Cluster formation
The advent of binary and ternary collisions during vapor
transport will, in some cases, promote gas phase clustering.
The formation and ensuing deposition of clusters is important as they can affect the pore morphology and texture of
vapor deposited films.50,51 The effect of a cluster impact depends strongly on its impact energy and size. High energy
cluster impacts lead to denser coating morphologies and
FIG. 10. Change in the peak dispersion width with chamber pressure is
given. The anticipated effect of cluster formation on Pw is schematically
shown.
411
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
lower energy cluster result in porous, granular coatings with
equiaxed grain structures.52 To further explore the probability of cluster formation in a DVD environment a kinetic
approach given by Bensen53 is considered.
Generally speaking, three body collisions are required to
form clusters in the vapor phase in order to conserve the laws
of momentum and energy for elementary reactions.54 Either
three vapor species or two vapor species and a background
gas atom or molecule may facilitate a reaction. The reaction
kinetics is therefore a strong function of the density of the
vapor species as it controls the likelihood that a third body
impacts the transition state of a metastable dimer. The vapor
density is a related to the evaporation rate, the system geometry and the operating chamber pressure.
For the current study, it is believed that the probability of
cluster formation is greatly affected by the presence of molecular evaporation and a “reactive” carrier gas 共containing
oxygen兲. Thus, instead of single atoms, ZrO and ZrO2 molecules are the predominate species leaving the vapor source.
Oxygen molecules are also present in the carrier gas stream.
The presence of these molecules alters the cluster formation
kinetics. Examples of binary collisions that could enable the
formation of stable clusters are given below:
ZrO + O2 ↔ ZrO2 + O*,
共Step I兲,
共1兲
ZrO + ZrO2 ↔ 2ZrO + O*,
共Step IIA兲,
共2兲
ZrO2 + ZrO2 ↔ 2ZrO + O2,
共Step IIB兲,
共3兲
2ZrO + ZrO2 ↔ 3ZrO + O*,
共Step III兲,
共4兲
共N − 1兲ZrO + ZrO2 ↔ 共N兲ZrO + O*,
共Step IV兲.
共5兲
The oxygen atoms 共O*兲 provide one mechanism to remove the latent heat of fusion. Helium/ cluster collisions
provide a second mechanism. The important result is that
three-body collisions are not necessarily required to nucleate
clusters. As a result, the nucleation of clusters is not considered rate limiting for the process conditions used here and
only the subsequent cluster growth from continued collision
events with the vapor molecules, oxygen molecules, and the
helium carrier gas need be considered.
Cluster growth can be estimated using
冉
␴aAN + ␴aA
dN ZAN−A
=
= 关A兴␲
dT 关AN兴
2
冊冉 冊
2
8kT
␲␮
TABLE V. Input values for the cluster size calculations.
Chamber
pressure
共Torr兲
Average gas
jet axial
velocity 共m/s兲
Source to
substrate
distance 共m兲
Time of flight,
␶ 共␮s兲
Average ZrO2
density 共# / m3兲
13.3
26.6
39.9
53.2
66.5
79.8
106.4
833.46
999.88
890.11
781.29
719.23
637.29
278.14
0.095
0.095
0.095
0.095
0.095
0.095
0.095
113.98
95.01
106.72
121.95
132.08
149.06
341.55
2.4⫻ 10−18
3.0⫻ 10−18
4.0⫻ 10−18
6.0⫻ 10−18
9.0⫻ 10−18
1.2⫻ 10−19
1.2⫻ 10−19
Nav =
1
␶
冕
␶
N共t兲d␶ ,
0
where ␶ is the time of flight of the vapor species.
Using this approach, the effect of chamber pressure on the
cluster size was determined. Estimates of the key parameters
are given in Table V. The zirconia concentration in the vapor
will increase with chamber pressure due to increased vapor
focusing by the gas jet. Visualization studies are used to
observe these effects.48 Values for ␶ are determined from
DSMC calculations. The effect of pressure on the cluster size
for a range of ␴aA values is shown in Fig. 11. The value of
␴aA is not precisely known for ZrO and ZrO2 molecules and
will increase as the cluster grows. Thus, an estimated range
of ␴aA values is given in Fig. 11. The increased zirconia
concentration and longer time of flight that results when the
pressure is increased results in a higher probability of large
clusters 共clusters greater than 100 molecules in size appear
possible兲. The effect of these impacts on the growth and
morphology of the yttria stabilized zirconia coatings is considered below.
1/2
,
where 关A兴 is the volume concentration of the vapor species,
␴aAN is the hard sphere diameter of the cluster, ␴a is the hard
sphere diameter of the vapor species, ␬ is the Boltzman constant, T is the gas temperature, and ␮ is the reduced mass of
collision, defined by 1 / ␮ = 1 / mAN + 1 / mA. Note that cluster
growth is largely dependent on the vapor species concentration. The average size of the cluster can then be determined
using
JVST A - Vacuum, Surfaces, and Films
411
FIG. 11. Estimate of the cluster size for a range of zirconia hard sphere
diameters. Increased chamber pressures and larger hard sphere diameters
result in larger cluster impacts.
412
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
VI. DISCUSSION
Large changes to the morphology of the zirconia coatings
resulted as the chamber pressure was increased during deposition. These changes included 共i兲 a larger intercolumnar pore
spacing, 共ii兲 wider intercolumnar pores, 共iii兲 an increased
volume fraction of fine scaled, intracolumnar pores, and 共iv兲
a disappearance of the texture typically observed in coatings
of this type. The large multifaceted changes observed indicate a fundamental transition in the growth mechanisms of
these coatings. These changes are related to changes in the
flux shadowing and cluster deposition mechanisms discussed
above.
A. Columnar pore structures
Columnar porosity is often present in coatings assembled
from the vapor phase due to differences in the growth rates
of different crystal surfaces and flux shadowing. The preferential growth of surface nucleated 7 YSZ crystallites orientated with their fastest growing directions towards the vapor
source rapidly raises these crystals above others and they
harvest a majority of the flux.55 This then results in a
strongly 共200兲 textured, columnar coating morphology. Here,
such a 共200兲 texture was observed, but only in the coatings
deposited at the lowest pressure 共13.3 Pa兲. The absence of
texture in the higher pressure process conditions indicates
that competitive growth of this type is not a dominate growth
mechanism for those high pressure conditions. Recent kinetic Monte Carlo simulations indicate that the intercolumnar pores observed here nucleate at surface asperities on the
rough substrate and rapidly grow due to flux shadowing
mechanisms.56 The role of substrate asperities is important as
many of the intercolumnar pores are eliminated when coating occurs onto atomistically flat substrates.48
Atomistic simulations indicate that pore nucleation at surface asperities depends upon the asperity size, the surrounding surface geometry and the IAD distribution.56 As the IAD
distribution becomes broader and more asymmetric, smaller
asperity dimensions will nucleate pores 共i.e., the IAD distribution can alter the population of nucleation sites on the
substrate兲. Pores nucleate when regions on the substrate are
depleted in incident vapor flux with respect to neighboring
regions. The depletion is limited by the diffusion of adatoms
across the growth surface into flux depleted regions. Thus,
factors that affect the adatom surface mobility 共i.e., the vapor
species translation energy, the latent heat of condensation
release, and molecular weight together with the substrate
temperature, deposition rate, and surface topology兲 will also
play a role.
B. Nanoscopic structures
Interestingly, intercolumnar pores did not form as readily
in coatings as the pressure was increased. This is surprising
as DSMC simulations indicate that a higher pressure increased Pw due to a higher probability of repeated binary
collision scattering of vapor atoms and the promotion of oblique vapor incident angles. The resulting increase in flux
J. Vac. Sci. Technol. A, Vol. 27, No. 2, Mar/Apr 2009
412
shadowing should ease the nucleation of intercolumnar pores
and increase the number of intercolumnar pore nucleation
sites.36 The fact that this was not observed suggests that a
different growth process must be occuring.
The most likely possibility is the formation of the gas
phase clusters. When clusters impact on a substrate their affect on the coating growth process is determined by their
kinetic energy and the cohesion energies of the substrate.
Several scenarios exist, including implantation of the cluster
into the substrate, dissociation upon impact, or the formation
of mounds on the surface. For the present case, the clusters
have a low kinetic energy per atom and low energy cluster
impacts are resulting in small mounds on the surface are then
expected. When the cluster size is small, the mounds will
affect the adatom surface mobility. For an adatom to diffuse
over a mound the activation barrier for an atom jump over a
ledge 共i.e., the Schwoebel barrier兲 must be overcome. This
barrier is much greater than for hopping over a flat surface
and thus, surface diffusion will be reduced. As the frequency
of cluster impacts and the cluster size increase, the process is
no longer governed by the surface adatom diffusion, but increasingly by the successive cluster impacts.
The nucleation of intercolumnar pores will be affected in
two ways by the transition from atomistic to cluster deposition. First, as the size of the clusters increase they will begin
to approach the size of some of the surface asperities. This
will limit shadowing induced pore nucleation, as shown
schematically in Fig. 12. Second, DSMC simulations reveal
that the IAD peak width 共Pw兲 increases with pressure if atomistic deposition is predominant. However, as the cluster
size increases with pressure, the rate at which Pw 共for the
combined cluster and vapor flux兲 increases with pressure will
be reduced and eventually decrease. The trajectories of clusters increasingly higher mass are increasingly less effected
by collisions with light helium atoms. As the volume fraction
and size of the clusters increases 共with increasing pressure兲 a
greater fraction of material will impact the substrate at near
normal incidence. Thus, impacts onto a rough substrate are
not as likely to be shadowed by surface asperities and intercolumnar pore nucleation is partially eliminated. The result
is a reduction in the number of pore nucleation sites on the
substrate and increased intercolumnar pore spacing.
The transition to a cluster dominated deposition regime
will also alter the pore size. Kinetic Monte Carlo simulations
have indicated that the pore width is linked to both the IAD
and the size and geometry of asperities 共or discontinuities兲
on the substrate.56 Broad IAD distributions result in large
pore widths due to the increase in the area of the substrate
that is flux depleted, Fig. 13. This relationship is observed by
noting that both the IAD peak width 共see the projected line in
Fig. 10 that assumes cluster formation兲 and the pore width
关see Fig. 4共B兲兴 increase until an intermediate pressure
共106.4 Pa兲 is reached after which both values decrease. This
is further complicated by the observation that the pore spacing decreases with increased pressure and thus, the number
of pore nucleation sites is also reduced. Whether a specific
substrate geometry will nucleate a pore is determined by the
413
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
413
FIG. 13. Schematic illustration showing the effect of the IAD on the width of
intercolumnar pores.
FIG. 12. Schematic illustration showing 共A兲 atomistic deposition onto a
rough substrate resulting in the nucleation of intercolumnar porosity and 共B兲
cluster deposition onto a rough substrate. The clusters shown are of the same
size scale as the substrate roughness leading to the inability to result in
shadowing induced pore nucleation.
IAD and the size of frequency of cluster impacts. For example, a region on the substrate that nucleates an intercolumnar pore at low pressure may not necessarily nucleate a pore
at higher pressure and vice versa. This is due to changes in
the IAD that result when the chamber pressure is altered
and/or the onset of cluster formation and deposition. Pores
that nucleate at intermediate pressures are observed to have a
largest width. This is because large asperities 共such as bond
coat grain boundaries兲 that result in the widest pores are only
observed to nucleate pores at intermediate pressures when
Pw is large. The thinnest pores are gradually eliminated with
increased pressure presumably because of an increased frequency of cluster impacts or loss of the crystallographic texture of the columns in the coating. The disappearance of the
thinnest pores together with the formation of wider pores at
the bond coat grain boundaries results in the observed increase in the average pore width with pressure until Pw begins to decrease due to the increased frequency of cluster
impacts.
While critical for TBC applications, the intercolumnar
pore volume fraction typically represents only 5%–10% of
JVST A - Vacuum, Surfaces, and Films
the porosity in these coatings. The total pore volume fraction
was often much higher 共up to 50%兲 and dramatically increased with chamber pressure. Kinetic Monte Carlo
simulations56 indicate that the level of IAD variation observed here cannot account for the large increase in porosity.
Other parameters that could affect the intercolumnar pore
fraction 共such as the vapor species translation energy and the
various contributions an adatom’s surface mobility兲 are also
not anticipated to be significantly changed by increasing the
chamber pressure since the translation energy of the incident
species is low 共⬍0.2 eV per atom兲 in all cases and the adatom surface mobility is primarily controlled by the substrate
temperature. However, low energy cluster impacts, such as
those shown schematically in Fig. 12共B兲 do lead to the formation of nanoscale porosity in the coating by a process
analogous to that formed between the much larger “splats” in
the APS process. Molecular dynamics simulations of cluster
impacts57 have shown that nanoscale porosity does form between impacting clusters. This phenomenon becomes important as the pressure is increased and cluster impacts become
more frequent.
VII. CONCLUSIONS
Zirconia coatings have been deposited using a directed
vapor deposition processing approach. Large effects effects
of deposition pressure upon the microstructure and pore morphology were observed. Relatively dense coatings, with textured columns were found at the lowest chamber pressures
共13.3 Pa兲. As the chamber pressure was increased, wide intercolumnar pores were observed. The spacing and width of
these pores increased with pressure. At the highest pressure
共106.4 Pa兲 large pore volume fractions were measured 共up to
50%兲. However, this porosity consisted almost entirely of
fine scaled 共nanoscopic兲 intracolumnar pores. The morpho-
414
D. D. Hass and H. N. G. Wadley: Gas jet assisted vapor deposition of yttria stabilized zirconia
logical changes have been linked to changes in the incidence
angle distribution of the depositing and the increased occurrence of low energy cluster impacts at the highest pressure.
ACKNOWLEDGMENTS
The authors are grateful to Timothy Bartel at Sandia National Laboratories for use of his ICARUS DSMC code. This
research was supported by the Office of Naval Research
共Grant No. N0001400-0147兲 under the program direction of
Steve Fishman and David Shifler.
1
NRC Report: Coatings for High Temperature Structure Materials, edited
by R. V. Hillary 共National Academy, Washington, D.C., 1996兲.
2
J. W. Fairbanks and R. J. Hecht, Mater. Sci. Eng. 88, 321 共1987兲.
3
I. T. Spitsberg, D. R. Mumm, and A. G. Evans, Mater. Sci. Eng., A 394,
176 共2005兲.
4
J. R. Nicholly, MRS Bull. 28, 659 共2003兲.
5
V. K. Tolpygo, D. R. Clarke, and K. S. Murphy, Surf. Coat. Technol.
146–147, 124 共2001兲.
6
P. Morrell and R. Taylor, High Temp. - High Press. 17, 79 共1985兲.
7
Handbook Chem and Physics 共CRC, Boston, 1990兲.
8
R. L. Jones, Report No. NRL/MR/6170-96-7841, 1996.
9
O. Yamamoto, Electrochim. Acta 45, 2423 共2000兲.
10
D. J. Wortman, B. A. Nagaraj, and E. C. Duderstadt, Mater. Sci. Eng., A
121, 443 共1989兲.
11
H. Zhao, Fengling Yu, T. D. Bennett, and Haydn Wadley, Acta Mater. 54,
5195 共2006兲.
12
P. R. Strutt, B. H. Kear, and R. Boland, U.S. Patent No. 6025024 共2000兲.
13
O. Unal, J. Am. Ceram. Soc. 77, 984 共1994兲.
14
S. M. Meier, D. M. Nissley, K. D. Sheffler, and T. A. Cruse, J. Eng. Gas.
Turbine Power Trans. of the ASME 114, 258 共1992兲.
15
Liangde Xie, Xinging Ma, E. H. Jordan, Nitin P. Padture, D. T. Xiao, and
M. Gell, J. Mater. Sci. 39, 1639 共2004兲.
16
M. Gell, Liangde Xie, Xinging Ma, Eric Jordan, and Nitin P. Padature,
Surf. Coat. Technol. 177–178, 97 共2004兲.
17
H. Huang, K. Eguchi, M. Kambara, and T. Yoshida, J. Therm. Spray
Technol. 15, 83 共2006兲.
18
N. Padture, M. Gell, and E. Jordan, Science 296, 280 共2002兲.
19
S. M. Meier and D. K. Gupta, J. Eng. Gas. Turbine Power Trans. of the
ASME 116, 250 共1994兲.
20
C. Levi, Curr. Opin. Solid State Mater. Sci. 8, 77 共2004兲.
21
S. Gu, T. J. Lu, D. D. Hass, and H. N. G. Wadley, Acta Mater. 49, 2539
共2001兲.
22
D. Hass, A. J. Slifka, and H. N. G. Wadley, Acta Mater. 49, 973 共2001兲.
23
U. Schultz, B. Sarukan, K. Fritscher, and C. Leyens, Int. J. Appl. Ceram.
Technol. 1, 302 共2004兲.
24
X. W. Zhou, R. A. Johnson, and H. N. G. Wadley, Acta Mater. 45, 4441
共1997兲.
25
K. H. Müller, Surf. Sci. 184, L375 共1987兲.
26
S. M. Rossnagel, C. Nichols, S. Hamaguchi, D. Ruzic, and R. Turkot, J.
Vac. Sci. Technol. B 14, 1819 共1996兲.
27
J. G. W. van de Waterbeemd and G. W. van Oosterhout, Philips Res. Rep.
22, 375 共1967兲.
28
J. A. Thornton, J. Vac. Sci. Technol. 12, 830 共1975兲.
J. Vac. Sci. Technol. A, Vol. 27, No. 2, Mar/Apr 2009
29
414
K. Kennedy, Proceedings of the AVS Vacuum Metallurgical Conference,
Los Angeles, 1968 共unpublished兲, p. 195.
30
K. H. Müller, J. Appl. Phys. 58, 2573 共1985兲.
31
J. J. Cuomo, J. M. E. Harper, C. R. Guarnieri, D. S. Yee, L. J. Attanasio,
J. Angilello, C. T. Wu, and R. H. Hammond, J. Vac. Sci. Technol. 20, 349
共1982兲.
32
T. C. Huang, G. Lim, F. Parmigiani, and E. Kay, J. Vac. Sci. Technol. A
3, 2161 共1985兲.
33
B. V. Movchan and A. V. Demchishin, Phys. Met. Metallogr. 28, 83
共1969兲.
34
D. V. Rigney, R. Viguie, D. J. Wortman, and D. W. Skelly, NASA Conf.
Publ. 3312, 135 共1995兲.
35
K. Robbie, L. J. Friedrich, S. K. Drew, T. Smy, and M. J. Brett, J. Vac.
Sci. Technol. A 13, 1032 共1995兲.
36
Y. G. Yang, R. A. Johnson, and H. N. G. Wadley, Acta Mater. 45, 1455
共1997兲.
37
R. A. Dugdale, Trans. Inst. Met. Finish. 54, 61 共1976兲.
38
M. Finnis, Prog. Mater. Sci. 49, 1 共2004兲.
39
X. W. Zhou and H. N. G. Wadley, Surf. Sci. 487, 159 共2001兲.
40
G. Gilmer, H. Huang, T. dela Rubia, and F. Bauman, Thin Solid Films
365, 189 共2000兲.
41
X. W. Zhou, R. A. Johnson, and H. N. G. Wadley, Acta Mater. 45, 4441
共1997兲.
42
J. Cho, S. G. Terry, R. LeSar, and C. G. Levi, Mater. Sci. Eng., A 391,
390 共2005兲.
43
J. F. Groves, H. N. G. Wadley, A. P. Ritenour, D. D. Hass, and P. L.
Ratnaparkhi, Proceedings of Electron Beam Melting and Refining State of
the Art, Bakish Materials Corp., Englewood, NJ, 1997 共unpublished兲, p.
46; D. D. Hass, P. A. Parrish, and H. N. G. Wadley, J. Vac. Sci. Technol.
A 16, 3396 共1998兲.
44
G. A. Bird, Molecular Gas Dynamics and the Direct Simulation of Gas
Flows 共Clarendon, New York, 1994兲.
45
D. D. Hass, A. J. Slifka, and H. N. G. Wadley, Acta Mater. 49, 973
共2001兲.
46
Hengbei Zhao, Fengling Yu, Ted D. Bennett, and Haydn N. G. Wadley,
Acta Mater. 54, 5195 共2006兲.
47
J. F. Groves and H. N. G. Wadley, Composites, Part B 28B, 57 共1997兲.
48
D. D. Hass, Ph.D. thesis University of Virginia, 2001.
49
T. J. Bartel and S. J. Plimpton, AIAA Paper No. 92-2860 共1992兲.
50
L. Bardotti, P. Jenson, A. Hoareau, M. Treilleux, B. Cabaud, A. Perez,
and F. Cadete Santos Aires, Surf. Sci. 367, 276 共1996兲.
51
P. Jenson, J. F. Roux, L. Bardotti, B. Cabaud, M. Treilleux, and A. Hoareau, Mater. Sci. Eng., B 60, 51 共1999兲.
52
A. M. Dongare, D. D. Hass, and L. V. Zhigilei, Proceedings of the International International Symposium on Clusters and Nano-Assemblies:
Physical and Biological Systems, 10–13 November 2003, Richmond, Virginia, USA 共unpublished兲.
53
S. W. Bensen, Foundations of Chemical Kinetics 共McGraw-Hill, New
York, 1960兲, Chap. 12.
54
S. H. Bauer and D. J. Frurip, J. Phys. Chem. 81, 1015 共1977兲.
55
U. Schulz, S. G. Terry, and C. G. Levi, Mater. Sci. Eng., A 360, 319
共2003兲.
56
D. D. Hass, Y. Y. Yang, and H. N. G. Wadley, J. Porous Mater. 共to be
published兲.
57
R. S. Averback, M. Ghaly, and H. Zhu, Radiat. Eff. Defects Solids, 130–
131, 211 共1994兲.