Surface and Coatings Technology 151 152 (2002) 412417

Improved sealing treatments for thick thermal barrier coatings

S. Ahmaniemi*, P. Vuoristo, T. Mantyla
Institute of Materials Science, Tampere University of Technology, P.O. Box 589, 33101 Tampere, Finland

Abstract

Zirconia-based 8Y2O3 ZrO2, 22MgO ZrO2 and 25CeO2 2.5Y2O3 ZrO2 thick thermal barrier coatings (1000 mm) were
studied with different sealing methods for diesel engine and gas turbine applications. The aim of the sealing procedure was to
improve the hot corrosion-resistance and mechanical properties of porous, thick thermal barrier coatings (TTBC). The surface of
the coatings was sealed with three different methods: (1) laser glazing; (2) an aluminium phosphate sealing treatment; and (3)
detonation gun spraying of a dense top coating on the TTBC. Sealant penetration into the coating and the coating microstructure
were determined by scanning electron microscopyyenergy-dispersive spectrometry (SEMyEDS) and optical microscopy. Coatings
were characterised by X-ray diffraction (XRD), microhardness and porosity measurements. The thickness of the densified top
layers in all cases was 50 400 mm. XRD analysis showed some minor phase changes and reaction products caused by the
phosphate-based sealing treatment and some crystal orientation and phase changes in laser-glazed coatings. The porosity of the
outer layer of the sealed coating decreased in all cases, which led to increased microhardness values. 2002 Elsevier Science
B.V. All rights reserved.

Keywords: Thermal barrier coating; Diesel engine; Phosphate sealing; Laser-glazing

1. Introduction overcoatings w811x. In this study, the microstructure of

Plasma-sprayed zirconia coatings are widely used as
thermal barrier coatings in gas-turbine hot section com-
ponents. From the early 1980s, there have been many
investigations into the application of these coatings in
diesel engines as well w14x. The conditions in a diesel-
engine combustion chamber differ considerably from
those in gas-turbine hot sections. Thermal and mechan-
ical loads and hot corrosion conditions set very demand-
ing requirements for thermal barrier coatings in both
cases, but long-term coating durability is a more difficult
problem in the diesel engine. The major TTBC failure
mechanism that causes coating spallation in gas turbines
is bond coat oxidation, while hot corrosion, thermal
cycling and mechanical loading are presumed to cause
more coating failures in diesel engines. There have been
several attempts to solve these problems by sealing the
zirconia-based thermal barrier coating. Sealing of open
porosity in TTBC coating can be carried out by liquid
metal impregnation w5x, laser glazing w6x, hot isostatic
pressing (HIP) w7x, a solgel process, or by thin CVD
* Corresponding author. Tel.: q358-3-365 2760.
E-mail address: samppa.ahmaniemi@tut.fi (S. Ahmaniemi).
sealed TTBCs was characterised. The goal was to find
methods to produce a dense top layer for TTBCs. In
this paper, we compare three separate sealing methods
and discuss their respective advantages and drawbacks.
2. Experimental procedure
2.1. Sample preparation
Three different TTBC coatings were sprayed with
plasma spray equipment (Plasma-Technik A3000S, Sulz-
er Metco AG, Wohlen, Switzerland). The thickness of
the bond coat was ;100 mm and the top coat layers
;1000 mm. The main spray parameters and powder
information are listed in Table 1. Coating specimens
were sprayed on AISI4142 steel substrates, with a height
of 5 mm and a diameter of 25 mm.
2.2. Laser-glazing
Coatings were laser-glazed using a 4-kW continuous-
wave fibre-coupled HAAS HL4006D lamp-pumped
Nd:YAG laser (HAAS laser GmbH, Schramberg, Ger-
many). The main laser-glazing parameters are presented

0257-8972/02/$ - see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 2 5 7 - 8 9 7 2 0 1 . 0 1 6 4 3 - 7
 S. Ahmaniemi et al. / Surface and Coatings Technology 151 152 (2002) 412417 413

Table 1
The main spray parameters used in APS process and the abbreviations for the coatings

Praxair ZRO Metco 205 NS Praxair ZRO 103 AMDRY 962

113y114
Coating abbreviation 8Y 25C 22M BC
Powder composition 8Y2O3 ZrO2 25CeO2 22.5MgO ZrO2 Ni22Cr10Al1Y
2.5Y2O3 ZrO2
Powder type AyS HOSP FyC Atomised
Particle size wmmx 11 125 16 90 10 75 56 106

Argon flow rate wlyminx 35 35 35 55

Hydrogen flow rate wlyminx 12 12 12 9.5
Current wAx 600 600 600 600
Voltage wVx 71.0 70.0 70.3 70.5

Table 2
The main spray parameters used in laser-glazing and the abbreviations for the coatings

8Y2O3 25CeO2 22.5MgO

ZrO2 2.5Y2O3 ZrO2
ZrO2
Coating abbreviation 8YL 25CL 22ML
Laser power wkWx 3.5 3.0 3.5
Surface speed wmmyminx 4500 4000 4500
Surface distance from the mirror wmmx 80 80 80
Laser-beam specific energy density wJymm2x 4.7 4.5 4.7

Table 3
The main spray parameters used in the D-gun process and abbreviations for the coatings

H.C Starck A706.072 H.C Starck A727.054 H.C Starck A840.1

Coating abbreviation 8YqDCr2O3 8YqD8Y 8YqDZrSiO4
Composition Cr2O3 8Y2O3 ZrO2 65ZrO2 35SiO2
Powder type AyS AyS FyC
Particle size wmmx 10 38 10 45 16 62
Powder phase structure a-Cr2O3 t9qm-ZrO2 t-ZrSiO4
Acetylene flow rate wlyminx 12 12 12
Oxygen flow rate wlyminx 25 21 21
Air flow rate wlyminx 11 11 11

in Table 2. Two parallel 10-mm-wide tracks, with a 1-
mm overlap, were used to produce a 19-mm-wide glazed
zone.
2.3. Aluminium phosphate sealing treatment
8Y and 25C coatings were sealed with Al(OH)3
(85%) H3PO4 solution diluted with 20 wt.% deionised
water. The ratio of Al(OH)3 y(85%) H3PO4 was 1:4.2
by weight, which corresponds to a PyAl molar ratio of
approximately 3. The solution was mixed and slightly
heated with a magnetic stirrer until it became clear. The
22M coating was sealed with orthophosphoric acid
(85%) H3PO4. Basic sealing tests showed that (85%)
H3PO4 sealant is more effective than Al(OH)3 (85%)
H3PO4 solution for the 22M coating. With both coatings,
the sealant was spread onto the surface just before the
heat treatment. Heat treatment was performed at 300 8C
for 4 h.
2.4. Detonation gun-sprayed dense top coatings
8Y2O3 ZrO2, Cr2O3 and ZrSiO4 layers were sprayed
onto the thick 8Y coatings with detonation gun equip-
ment (Perun-P, Paton Electric Welding Institute, Kiev,
Ukraine) in order to obtain as dense a top coating as
possible. Finer powder was used in the detonation spray
process than in APS, because of the lower heating
power. Spray parameters and powder information are
listed in Table 3.
2.5. Characterisation
The coating microstructure was determined by optical
microscopy (OM; Model Versamet 3, Union Co, Japan)
414 S. Ahmaniemi et al. / Surface and Coatings Technology 151 152 (2002) 412417
Fig. 1. Fracture surface of the laser-glazed 8Y2O3 ZrO2 coating.
and scanning electron microscopy (SEM; Model XL-
30, Philips, Eindhoven, Netherlands). Energy-dispersive
spectrometry (EDS; Model DX-4, EDAX International,
New Jersey, USA) was used for elemental analysis in
SEM studies. Polished micro-sections and fracture sur-
faces were prepared for microscopy studies. The coating
phase structure was characterised by X-ray diffracto-
metry (XRD; Siemens D500, Karlsruhe, Germany)
using CuKa radiation (scan step 0.028, step time 1.2 s).
XRD analysis for the sealed coatings was carried out
after grinding of a 1020-mm layer away from the
surface, because reaction products on the coating surface
normally differ considerably from those below the sur-
face. Coating microhardness (HV0.3 ) profiles were deter-
mined with a microhardness tester (Shimadzu, Kyoto,
Japan) from coating cross-sections at 150200-mm
intervals. Results are presented as mean values of three
separate measurements taken at each depth.
3. Results and discussion
3.1. Laser-glazed coatings
The laser-melted region was highly densified in 8YL,
25CL and 22ML coatings, but some vertical cracks were
detected, especially from the melted zone. The depth of
the melted zone was 50150 mm and hardness values
increased from a base level of 600700 up to 1100
1250 HV0.3. Melting had occurred in quite a uniform
layer in 8YL and 25CL coatings, but in the magnesia-
stabilised coating, the thickness of the melted layer
varied a great deal. This difference probably results from
the different laser-beam absorption of the coating mate-
rials. 8YL coatings had a transparent and glass-like
surface because of the better laser-beam absorption and
melting. The surface structure of the 25CL coating was
smooth, but not glass-like. In contrast, the surface of
the 22ML coating was quite rough and inhomogeneous.
The colour of the yttria-stabilised zirconia coatings
changed from light grey to transparent yellow due to
the laser-glazing process. The colour of the 25C coating
changed from yellow to black. The white colour of the
22M sample did not change after the laser treatment.
SEM studies showed a columnarydendritic structure for
the laser-glazed zone. The structure of the 8YL coating
is presented in Fig. 1.
From the fracture surface of the coating, some voids
could be detected in the lower region of the melted
zone, marked by an arrow in Fig. 1. Large voids
probably developed from coating porosity during the
melting process. A columnar structure in the 25C and
22ML coatings was not clearly detectable from the
fracture surface cross-sections, but backscattered elec-
tron images taken from the polished micro-sections
showed the columnarydendritic nature of the properly
glazed zone; see Figs. 2 and 3.
Some small-sized pores could be found in the glazed
zones, but the pore size distributions were smaller
compared to the 8YL coating. In the laser-glazed 8Y
and 25C coatings, the phase structure of the coating
surface was purely non-transformable tetragonal zirco-
nia, t9-ZrO2. In the as-sprayed 25C coating, the phase
structure was also t9-ZrO2 , but in the case of the 8Y
coating, a certain quantity of monoclinic (m-ZrO2) and
cubic (c-ZrO2) zirconia was detected. The formation of
non-transformable t9-ZrO2 indicates rapid solidification
in both plasma spraying and laser-glazing. The relative
intensity of the individual diffraction peaks greatly
changed due to the glazing process. This is due to the
columnar grain orientation caused by the laser-glazing
process. Detailed XRD studies of the 8YL coating for
the 2u region of 72768 showed that the lattice para-
meters of the glazed coating differed to a certain extent
Fig. 2. SEM backscattered electron image of the polished cross-sec-
tion of the laser-glazed 25CeO2 2.5Y2O3 ZrO2 coating.
 S. Ahmaniemi et al. / Surface and Coatings Technology 151 152 (2002) 412417
Fig. 3. SEM backscattered electron image of the polished cross-sec-
tion of the laser-glazed 22MgO ZrO2 coating.
from those of the as-sprayed coating. There was a wider
gap between the t9(400) and t9(004) peaks in the laser-
glazed coating. This could be an indication of the
slightly lower concentration of Y2O3 in the ZrO2 struc-
ture w12x. This might also be explained by the rearrange-
ment of Y2O3 in the ZrO2 structure during laser-melting.
In the 22ML coating, in addition to the c-ZrO2 structure,
some rhombohedral Mg2Zr5O12 could be detected by
XRD. EDS analysis showed that in the lighter region of
the melted zone, the concentration of MgO was ;8
wt.% and ;17 wt.% in the darker region. These regions
are marked with arrows in Fig. 3. The presence of the
rhombohedral Mg2Zr5O12 phase is evident in the XRD
diagram for 22MgOZrO2-based coatings (Fig. 4).
Fig. 4. XRD diagrams for 22MgO ZrO2 coating surfaces: (a) as-sprayed coating; (b) orthophosphoric acid-sealed coating; and (c) laser-glazed
coating; csc-ZrO2, psc-MgO (periclase), xsZrP2O7, zsMg2Zr5O12.
415
The vapour pressure of MgO in the molten state
during laser-glazing is much higher than that of ZrO2.
The decrease in the total amount of MgO in the coating
structure is a consequence of its vaporisation during the
laser-glazing process. Peak broadening of the c-ZrO2
phase in the 22ML coating did not appear, as in the as-
sprayed coating.
3.2. Phosphate-sealed coatings
Coating densification in the phosphate-sealed coatings
was clearly evident in optical microscopy studies. Phos-
phate-based sealants penetrated approximately 300400
mm into the coatings. The coating hardness increased
from 600700 up to 850950 HV0.3 due to phosphate
sealing. Phosphate phases were observed in SEMyEDS
studies of coating fracture surfaces using elemental
mapping. EDS elemental mapping results of the 8Y AP
coating are presented in Fig. 5.
Our earlier study w13x showed the presence of an
aluminium-rich bonding phase in the coating structure.
Any reaction products caused by the phosphate sealing
treatment could not be found by XRD for the 8Y AP or
25C AP coatings. This may be a consequence of the
low concentration of the bonding phases or their amor-
phous microstructure. However, phosphate-based sealing
affected the phase structure of the 22M coating to some
degree. If only the surface roughness of the 22M OPA
coating was smoothed by grinding before the XRD
analysis, clear zirconium phosphate (ZrP2O7) peaks
were identified. This seemed to be the most probable
crystalline reaction product between the 22M coating
and orthophosphoric acid.
416 S. Ahmaniemi et al. / Surface and Coatings Technology 151 152 (2002) 412417
Fig. 5. EDS elemental mapping of the Al, O and Zr from the fracture surface of the aluminium phosphate-sealed 8Y2O3 ZrO2 coating.
3.3. Detonation gun-sprayed dense top coatings
Detonation gun-sprayed D8Y, DCr2O3 and DZrSiO4
coatings showed considerably lower porosity compared
to the APS-sprayed 8Y coating. The dense and crack-
free structure of the detonation gun-sprayed coatings
indicated sufficient melting of the ceramic powders.
Hardness values for the D8Y, DCr2O3 and DZrSiO4
coatings were 1010, 1180 and 800 HV0.3, respectively.
Optical micrographs of the detonation gun-sprayed coat-
ings are presented in Fig. 6.
8Y2O3 ZrO2 feedstock powder (A727.054) contained
a minor fraction of m-ZrO2. XRD studies showed that
the phase structure of the D8Y coating was t9-ZrO2.
This was also an indication of the proper melting of the
8Y2O3 ZrO2 powder. Cr2O3 powder and the DCr2O3
coating mostly consisted of the eskolaite-type hexagonal
a-Cr2O3. Some minor amounts of Cr, CrO and Cr3O4
were also detected in the sprayed coating. These minor
phases are probably a consequence of the decomposition
of the hexagonal a-Cr2O3 structure during spraying. The
DZrSiO4 coating mostly consisted of the t9- and m-
ZrO2 phases. This indicates that the ZrSiO4 structure
has decomposed during spraying and a certain amount
of SiO2 has formed a solid solution with ZrO2 and
stabilised the t9-ZrO2 structure. The remainder of the
SiO2 has obviously formed an amorphous structure
during rapid cooling in the spray process.
4. Conclusion
In this paper, three separate sealing procedures for
thick thermal barrier coatings were described, namely:
(i) surface melting by laser-glazing; (ii) impregnation
of a phosphate-based sealant; and (iii) spraying of a
dense top coating with a detonation gun. The main
results of this study are listed below.
The thickness of the melted layer in laser-glazed
TTBCs was 50150 mm. The microhardness of this
layer was 11001250 HV0.3. The phase structure of
the as-sprayed 8Y coating changed on laser glazing
from t9-, c- and m-ZrO2 to t9-ZrO2. The lattice
parameters of the t9-ZrO2 phase in the laser-glazed
coatings changed due to the reduced Y2O3 content in
the coating. In the laser-glazing process, the phase
structure of the magnesia-stabilised zirconia changed
from c-ZrO2 to a mixture of rhombohedral
Mg2Zr5O12 and c-ZrO2.
With phosphate-based sealing treatments, the impreg-
nation depth of the sealant was 300400 mm and the
microhardness was approximately 950 HV0.3. Phos-
 S. Ahmaniemi et al. / Surface and Coatings Technology 151 152 (2002) 412417
Fig. 6. Optical micrographs of the D-gun-sprayed dense top layers on 8Y2O3 ZrO2 coating: (a) 8Y2O3 ZrO2; (b) Cr2O3; and (c) ZrSiO4.
phate phases were not detectable by XRD in the 8Y
AP coating, but elemental mapping of Al by EDS
analysis showed some traces of the bonding phos-
phate phase. Unidentified diffraction peaks were
detected in XRD studies of the 22M OPA coating.
These peaks are probably attributable to the phos-
phate-based bonding phase.
Detonation-gun spraying could be used to produce
denser ceramic top layers onto conventionally sprayed
TTBCs. Detonation gun-sprayed top layers were 70
200 mm thick, depending on the coating material and
the number of spray passes. The microhardness of
the detonation gun-sprayed coatings varied between
800 and 1180 HV0.3. Minor decomposition of the
feedstock powder, eskolaite-type hexagonal a-Cr2O3,
was detected by XRD after the spray process. In the
case of the ZrSiO4, the phase structure of the feed-
stock powder was totally decomposed during the
spray process.
Acknowledgements
Results in this paper have been collected from studies
carried out during the years 20002001. The work was
financed by Finnish industry, TEKES and The Academy
of Finland. The authors are grateful to financial sup-
porters and to Ms Sari Iltanen and Mr Mikko Kylmalahti
for assistance in experimental work.
417
References
w1x D.N. Assanis, Int. J. Mater. Prod. Technol. 4 (1989) 232.
w2x I. Kvernes, High-Temperature Materials for Powder Engineer-
ing, Kluwer Academic Publishers, Liege, Belgium, 1990, pp.
843864.
w3x T.M. Yonushonis, J. Therm. Spray Technol. 6 (1997) 50.
w4x M.B. Beardsley, J. Therm. Spray Technol. 6 (1997) 181.
w5x A. Ohmori, Z. Zhou, K. Inoue, et al., in: A. Ohmori (Ed.),
Thermal Spraying: Current Status and Future Trends, High
Temperature Society of Japan, Osaka University, Osaka, Japan,
1995, pp. 549554.
w6x H.L. Tsai, P.C. Tsai, Surf. Coat. Technol. 71 (1995) 53.
w7x H. Kuribayashi, K. Suganuma, Y. Miyamoto, M. Koizumi,
Am. Ceram. Soc. Bull. 65 (1986) 1306.
w8x G. John, T. Troczynski, in: C.C. Berndt (Ed.), Thermal Spray:
Practical Solutions for Engineering Problems, ASM Interna-
tional, Materials Park, OH, USA, 1996, pp. 483488.
w9x K. Moriya, W. Zhao, A. Ohmori, in: A. Ohmori (Ed.), Thermal
Spraying: Current Status and Future Trends, High Temperature
Society of Japan, Osaka University, Osaka, Japan, 1995, pp.
10171021.
w10x R. Rajendran, V.S. Raja, R. Sivakumar, R.S. Srinivasa, Surf.
Coat. Technol. 73 (1995) 198.
w11x T. Troczynski, Q. Yang, G. John, J. Therm. Spray Technol. 8
(1999) 229.
w12x R. Taylor, J.R. Brandon, P. Morrell, Surf. Coat. Technol. 50
(1992) 141.
w13x S. Ahmaniemi, P. Vuoristo, T. Mantyla,
in: C.C. Berndt (Ed.),
Thermal Spray: Surface Engineering via Applied Research,
ASM International, Materials Park, OH, USA, 2000, pp. 1081
1086.