Structure and corrosion resistance of ZrO2 ceramic coatings on AZ91D Mg alloys by plasma electrolytic oxidation

Article abstract of DOI:10.1016/j.jallcom.2011.06.011

The aim of this work is to study the structure and the corrosion resistance of the plasma electrolytic oxidation ZrO₂ ceramic coatings on Mg alloys. The ceramic coatings were prepared on AZ91D Mg alloy in Na ₅P₃O₁₀ and K₂ZrF₆ solution by pulsed single-polar plasma electrolytic oxidation (PEO). The phase composition, morphology and element distribution in the coating were investigated by X-ray diffractometry, scanning electron microscopy and energy distribution spectroscopy, respectively. The results show that the coating thickness and surface roughness were increased with the increase of the reaction time. The ceramic coatings were of double-layer structure with the loose and porous outer layer and the compact inner layer. And the coating was composed of P, Zr, Mg and K, of which P and Zr were the main elements in the coating. P in the coating existed in the form of amorphous state, while Zr crystallized in the form of t-ZrO₂ and a little c-ZrO₂ in the coating. Electrochemical impedance spectra (EIS) and the polarizing curve tests of the coatings were measured through CHI604 electrochemical analyzer in 3.5% NaCl solution to evaluate the corrosion resistance. The polarization resistance obtained from the equivalent circuit of the EIS was consistent with the results of the polarizing curves tests.

Full text of DOI:10.1016/j.jallcom.2011.06.011

Journal of Alloys and Compounds 509 (2011) 8469–8474
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Structure and corrosion resistance of ZrO ceramic coatings on AZ91D Mg alloys
2
by plasma electrolytic oxidation
a,b,∗
a
a
a
a
a
Zhongping Yao
, Yongjun Xu , Yunfu Liu , Dali Wang , Zhaohua Jiang , Fuping Wang
a
School of Chemical Engineering and Technology, Harbin Institute of Technology Harbin 150001 PR China
Postdoctoral Station of Chemistry Engineering and Technology, Harbin Institute of Technology, Harbin 150001, PR China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 May 2011
Received in revised form 26 May 2011
Accepted 3 June 2011
Available online 12 June 2011
The aim of this work is to study the structure and the corrosion resistance of the plasma electrolytic
oxidation ZrO2 ceramic coatings on Mg alloys. The ceramic coatings were prepared on AZ91D Mg alloy in
Na5P3O10 and K2ZrF6 solution by pulsed single-polar plasma electrolytic oxidation (PEO). The phase com-
position, morphology and element distribution in the coating were investigated by X-ray diffractometry,
scanning electron microscopy and energy distribution spectroscopy, respectively. The results show that
the coating thickness and surface roughness were increased with the increase of the reaction time. The
ceramic coatings were of double-layer structure with the loose and porous outer layer and the compact
inner layer. And the coating was composed of P, Zr, Mg and K, of which P and Zr were the main elements
in the coating. P in the coating existed in the form of amorphous state, while Zr crystallized in the form of
t-ZrO2 and a little c-ZrO2 in the coating. Electrochemical impedance spectra (EIS) and the polarizing curve
tests of the coatings were measured through CHI604 electrochemical analyzer in 3.5% NaCl solution to
evaluate the corrosion resistance. The polarization resistance obtained from the equivalent circuit of the
EIS was consistent with the results of the polarizing curves tests.
Keywords:
Ceramic coatings
Structure
Corrosion resistance
Plasma electrolytic oxidation
Mg alloy
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
et al. for the first time reported the preparation of ZrO coatings on
2
Al alloys in K ZrF6 solution in 1996 [17–21]. Except adding dis-
2
Magnesium and its alloys have been applied in many fields
persed ZrO2 or Zr(OH)4 particles in the other electrolytes [22–24]
or adopting Zr(SO ) electrolyte directly [19], fluorozirconate elec-
like aerospace, automobile, computer, etc., due to the excellent
low density, high strength to-weight ratio and mechanical casta-
bility [1–3]. Unfortunately, the high susceptibility to corrosion of
magnesium alloys greatly restricts their further application. There-
fore, some kinds of techniques like surface oxidation, PVD/CVD
coating or ion implanting are studied by many researchers [4–7]
to improve the corrosion resistance of magnesium alloys. Among
these methods, plasma electrolytic oxidation (PEO) is considered as
one of the most promising surface treatment techniques of the Mg
alloys [8–10]. At present, much research on PEO technique is being
focused on the preparation of the ceramic coatings on magnesium
alloys in widely adopted alkaline electrolytes such as phosphate,
silicate, or sodium hydroxide solutions and so on. And meantime
the growth characteristics of the coating on magnesium alloy in
these electrolyte systems were also studied by some researchers
4
2
trolyte is the most often adopted one to prepare ZrO2 ceramic
coating up to now [25–29]. In most cases of PEO process, such addi-
tives as H PO4 and NaH PO4 was used fluorozirconate electrolyte
3
2
system to keep the PEO process carrying out under an acid environ-
ment. In this work the sodium tripolyphosphate (Na5P O10) was
3
used to make the pH of the electrolyte solution nearly equals to
7 because of the pollution of the acid electrolyte to the environ-
ment. Meantime, the phase composition, structure and corrosion
resistance of the ZrO ceramic coatings on an AZ91D Mg alloy were
2
investigated.
2. Experimental details
2
.1. Preparation of ceramic coatings by plasma electrolytic oxidation technique
[
11–16].
Besides, ZrO2 ceramic coatings on Al, Ti or Mg alloys by PEO
Plate samples of AZ91D Mg alloy with reaction dimension of
a
method have been attracted much attention recently since Schukin
15 mm × 20 mm × 0.6 mm were used as working electrode and the electrol-
yser made of stainless steel served as the counter electrode. The electrolyte used
in the experiment was Na5P3O10 with the concentration of 3 g/L and K2ZrF6 with
the concentration of 1 g/L. A home-made high power pulsed single-polar electrical
source with power of 5 kW was used for plasma electrolytic oxidation under the
∗ Corresponding author. Tel.: +86 0451 86413710; fax: +86 0451 86415647.
pulsed current density of 10 A/dm for different time. The duty ratio was 50% and
2
E-mail address: yaozhongping@hit.edu.cn (Z. Yao).
the working frequency was 500 Hz. The reaction temperature was controlled to be
0
925-8388/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2011.06.011

8
470
Z. Yao et al. / Journal of Alloys and Compounds 509 (2011) 8469–8474
Table 1
The relative contents of main elements on the surface of the coating by EDS analysis
(
atom%).
Time/min
Zr
8.9
10.4
37.5
32.5
Mg
Al
P
K
5
0
0
0
14.9
12.9
7.8
2.2
0.1
0
73.0
75.3
51.8
54.3
1.0
1.3
2.9
3.6
1
2
3
9.6
0
3.2. The morphologies and the element distribution of the
coatings
Fig. 2 is the surface images of the coatings for different time.
There were some residual discharging channels and many micro-
holes on the coating surface. Meantime, there were some cracks
and bumps. Increasing the PEO time, the cracks and the bumps
were enlarged and the coating became rough and uneven. The
relative contents of the main elements on the surface of the coat-
ing was analyzed by EDS under the spatial resolution of 200 s,
with the results shown in Table 1, it can be noted that the coat-
ing was composed of P, Zr, Mg and K, of which P and Zr were
the main elements in the coating. At the initial stage of the PEO
reaction, much P from the electrolyte joined the PEO reaction.
But increasing the reaction time, the amount of P in the coating
surface was gradually decreased so that the amount of Zr was
more than that of P for in the coating. The amount of Mg in the
coating surface was decreased with the reaction time in general,
which illustrated that the amount of Mg joining PEO process was
reduced with the reaction time. Meantime, a little Al from the sub-
strate was also checked on the surface of the coating of the short
time.
Fig. 1. Thickness of the coatings prepared for different time.
ꢀ
below 30 C by adjusting the cooling water flow. After the treatment, the coated
samples were flushed with water and dried in air.
2
.2. Analysis of phase composition and structure of the coatings
Phase composition of the coatings was examined with XRD (D/max-rB, RICOH,
Japan) with a Cu K␣ source. Surface images of the produced coatings were stud-
ied with SEM (S-570, Hitachi, Japan). The elemental distribution was investigated
by energy-dispersive spectroscopy (EDS; US PN5502). The coatings’ thickness was
measured, using an eddy current based thickness gauge (CTG-10, Time Company,
China). In this experiment, the average thickness of each of the triplicate samples
was obtained from 10 measurements at different positions.
Fig. 3 is the section morphologies of the coatings for differ-
ent time. There were many micro-sized cracks and pores in the
coating. The coating for 10 min was comparatively compact and
its thickness was generally uniform. Increasing the reaction time,
the coating thickness was increased, but the coating thickness
was much less than the measured results by CTG-10. This is
because the coating surface was so loose and coarse that it was
easily embedded or destroyed by the mounting resin. Besides,
the coating was of double layer structure: the outer layer was
much loose and porous whereas the inner layer was comparatively
compact.
2
.3. EIS of the ceramic coatings
EIS spectra of the coatings were measured in a three-electrode cell (a Pt plate
was used as a counter electrode, a calomel electrode as an auxiliary electrode, and
the coated sample as a working electrode with the reaction area of 1 cm ) through
CHI604 electrochemical analyzer (Shanghai, China) in 3.5% NaCl solution. A sinu-
soidal AC perturbation of 5 mV was applied to the electrode at the open circuit
potential of the coated samples over the frequency range 0.1 Hz–100 kHz.
2
2.4. The corrosion resistance of the coatings
The corrosion resistance of the coatings was evaluated by the polarizing curves
in 3.5% NaCl solution, using the same three-electrode cell as in 2.3. The polarization
curve’s scanning rate was 1 mV/s, with a scanning range from −0.25 V of open circuit
potential to +0.25 V of open circuit potential.
3.3. Phase analysis of the coatings for different time
Fig. 4 is the XRD patterns of the coatings prepared for differ-
ent time. From which it follows that the coating was composed
3
. Results and discussion
of t-ZrO2 and a little c-ZrO . According to the intensities of the
diffraction peaks, the amount of ZrO2 can be approximately com-
pared for the coatings of different time. Increasing the reaction
2
3
.1. The thickness of the coatings
At the initial stage of the PEO process, the anode pulse voltage
time, the amount of ZrO was increased in general. However, when
2
was increased quickly and reached the breakdown voltage of the
oxide film within 2 min. From this moment the reaction time was
recorded. The voltage was increased gradually to the designed reac-
tion time under the fixed current density. Fig. 1 is the thickness of
the coating prepared for different time. Clearly, the coating thick-
ness was increased linearly with the reaction time. Through the line
fitting of the thickness vs. time, with the result shown in formula
the reaction time was 20 min, the intensity of diffraction peaks
corresponding to the ZrO2 was lower than that when reaction
time was 10 min, which seemed that the amount of ZrO2 was
reduced. This conflict may be related to the penetration depth of
X-rays. The extending of the reaction time increased the coating
thickness, and correspondingly decreased the penetration depth of
X-rays, which maybe led to the decrease of the detected amount
(
1):
of ZrO . However, P and K in the coating were not crystallized,
2
Thickness (␮m) = 6.476 + 1.537 × Time (min)
(1)
whether its amount was more or less; and Mg and Al from the
substrate were not in the form of the crystalline in the coating
yet. This may be related to the quick cooling speed by the elec-
trolyte during the PEO process. A large amount of Zr and P from
the electrolyte solution joined the PEO reaction and formed the
coating. Therefore, the coating thickness was mainly determined
The growth rate of the coating was 1.537 ␮m/min or so under
the experimental conditions. Actually, it can be found from Fig. 1
that in the initial 2 min stage of PEO process, the coating with the
thickness of about 6.476 ␮m already generated on the substrate.

Z. Yao et al. / Journal of Alloys and Compounds 509 (2011) 8469–8474
8471
Fig. 2. The surface morphologies of the coatings for different time (a): 5 min; (b): 10 min; (c) 20 min; (d): 30 min.
by the contents of Zr and P. Besides, the diffraction peaks corre-
sponding to the substrate can also exist in the patterns, but their
intensity was generally decreased with the increasing time. But,
there also existed the exceptions, which may be related to the
pores and the cracks induced in the coatings discussed in Section
for the Mg alloy substrate, which was obviously corresponding to
the capacitance of the double layer of the electrode interface (Cdl).
But, there were two differentiated time constants for the coatings of
1 min and 5 min, which should also be corresponding to the capac-
itance elements of the equivalent circuit. Increasing the reaction
time, the coating electrode became porous and rough, the relax-
ation effect of the electrode surface was strengthened, by which
the two time constants may be overlapped and seemed to become
one.
3
.2.
3.4. EIS analysis of the coatings
Fig. 5 is EIS (Bode plot) of PEO ceramic coatings for different time
Based on the above analyses and the porous double-layer struc-
ture of the coatings in Fig. 3, the EIS equivalent circuit for the PEO
coatings on magnesium alloy AZ91D is presented in Fig. 6 with the
fitted results shown in Table 2. Rs was the solution resistance, which
was determined by the conductance of the NaCl solution, therefore,
all the Rs were almost the same. R1 and C1 were corresponding to
in 3.5% NaCl solution, which clearly shows that the impedance val-
ues and phase angles look greatly different as the reaction time was
increased. The impedance values were increased with the decrease
of the frequency to different extents. It can be noted from the phase
angle extremum in panel (b) that there was only one time constant

8
472
Z. Yao et al. / Journal of Alloys and Compounds 509 (2011) 8469–8474
Fig. 3. The section morphologies of the coatings for different time. (a): 10 min; (b):
0 min.
2
Fig. 5. EIS (Bode plot) of PEO ceramic coatings for different reaction time and the
AZ91D substrate.
the outer layer of the PEO coating. R2 and C2 were corresponding
to the inner layer of the PEO coating.
With the increase of the reaction time, the value of C1 was
decreased gradually. Because C1 was in series with C2 and much
less than C2, therefore, the whole capacitance of the coating was
determined by the C . The phase angle extremums in panel (b) were
1
decreased and moved to the low frequency area with the increas-
ing reaction time, which meant that the capacitance interface was
weakened. This just complied with the changing regularity of the
value of C . The value of the C2 in Table 2 changed little, which
1
showed the inner layers of all the coatings had the similar capaci-
tance features. Besides, the values of R and R were decreased first
1
2
and then increased, which was related to the density and the thick-
Fig. 4. XRD patterns of the coatings and the AZ91D substrate.
ness of the coatings. R2 was much larger than R , which showed
1
Table 2
Impedance fitting values of the coatings prepared for different time.
RS (ohm cm2)
R1 (ohm cm2)
C1 (F cm2)
R2 (ohm cm2)
C2 (F cm2)
−7
−7
−7
−8
−8
−6
−6
−6
−6
−6
1
5
1
2
3
min
min
0 min
0 min
0 min
24.10
25.34
28.22
26.80
24.65
260.3
256.1
105.8
401.8
896.1
2.017 × 10
1.793 × 10
1.616 × 10
6.241 × 10
2.947 × 10
8260
7517
4151
5594
5902
2.187 × 10
3.306 × 10
2.733 × 10
4.749 × 10
8.208 × 10

Z. Yao et al. / Journal of Alloys and Compounds 509 (2011) 8469–8474
8473
determined by the density of the coatings, not the coating thick-
ness. And the extending of the PEO time was not liable to improve
the density of the coatings although increasing the coating thick-
ness.
4. Conclusions
Fig. 6. Equivalent circuit used for PEO coatings on AZ91D magnesium alloy in 3.5%
NaCl solution. Rs is the solution resistance, R1 and C1 are corresponding to the outer
layer of the PEO coating, R2 and C2 are corresponding to the inner layer of the PEO
coating.
Ceramic coatings on AZ91D Mg alloy were prepared by plasma
electrolytic oxidation in tripolyphosphate-potassium fluorozir-
conate solution, the research on the composition, structure and
corrosion resistance of coatings allowed the following conclusions
to be drawn:
that the density of the inner layer was much more than that of
the outer layer although the thickness of the inner layer was less
than that of the outer layer. The values of R1 and R2 of the coat-
ings for short PEO time were comparatively large; this was mainly
due to the high density of the formed coatings at the initiate stage.
Increasing the reaction time, the decrease of the values of R1 and
R2 was due to the increase of the porosity and the cracks of the
coatings. However, increasing the reaction time further, the values
of R1 and R2 were increased again, which was mainly attributable
to the increase of the coating thickness. Therefore, the EIS of the
coatings reflected the structure of the coating with the increasing
PEO time.
(
1) The PEO ceramic coatings on AZ91D Mg alloy were composed
of t-ZrO2 and a little c-ZrO . With the increase of the reaction
2
time, the coating thickness was increased linearly and the coat-
ing turned rough. And the coating was of double-layer structure
and the out layer was loose while the inner layer was com-
paratively dense. EDS analysis showed that the coating was
composed of P, Zr, Mg and K, of which P and Zr were the main
elements in the coating.
(
2) EIS and the established equivalent circuit were corresponding
to the double-layer structure of the coatings. Besides, the polar-
ization resistance Rp (the sum of the R and R in the equivalent
1
2
circuit) can reflect the corrosion resistance of the coatings for
different time, which was consistent with the corrosion current
density of the samples obtained through the polarizing curves.
The corrosion resistance of the coatings was much better than
that of the Mg alloy substrate. And the coating prepared for
3
.5. The corrosion resistance of the coatings for different time
Generally, it is considered that the impedance values measured
at the low frequency area in the EIS spectra can reflect the cor-
rosion resistance of the coatings. Considering the established EIS
equivalent circuit in Fig. 6, the polarization resistance Rp of the
coatings should approximately equal to the sum of the R1 and
R2, which was completely corresponding to the impedance val-
ues measured at the low frequency area in the EIS spectra. The
corrosion resistance of all the coatings was much better than that
of the Mg alloy substrate, and the coating prepared for the 1 min
was the best, and then increasing the reaction time, the corrosion
resistance of the coating was decreased first and then increased
again. In order to assess the corrosion resistance of the coating
samples, the polarizing curve method was used to calculate the
corrosion current density of the coating samples for different time
and the substrate in 3.5% NaCl solution with the results shown
in Fig. 7, which also presented the same regularity as the results
of Rp obtained from EIS analyses. According to the above analy-
ses, the general corrosion resistance of the coatings was mainly
1
min had the best corrosion resistance mainly due to its better
density.
Acknowledgement
This work was financially supported by Special Foundation for
New Teachers of Doctor course in Chinese Education Ministry
(grant no. 200802131065) and China Postdoctoral Science Special
Foundation (grant no. 201003426).
References
[
[
[
1] J.E. Gray, B. Luan, J. Alloys Compd. 336 (2002) 88.
2] R. Ambat, N.N. Aung, W. Zhou, Corros. Sci. 42 (2000) 1433.
3] Y. Zhang, C. Yan, F. Wang, H. Lou, C. Cao, Surf. Coat. Technol. 161 (1) (2002)
36.
[
4] Y. Mizutani, S.J. Kim, R. Ichino, M. Okido, Surf. Coat. Technol. 169–170 (2003)
143.
[
[
5] H. Fukuda, Y. Matsumoto, Corros. Sci. 46 (2004) 2135.
6] L. Kouisni, M. Azzi, M. Zertoubi, F. Dalard, S. Maximovitch, Surf. Coat. Technol.
185 (2005) 58.
[
[
7] H. Altun, S. Sen, Surf. Coat. Technol. 197 (2005), 193.1757.
8] Y. Ma, X. Nie, D.O. Northwood, H. Hu, Thin Solid Films 469–470 (2004)
472.
[
9] A.L. Yerokhin, A. Shatrov, V. Samsonov, P. Shashkov, A. Leyland, A. Matthews,
Surf. Coat. Technol. 182 (2004) 78.
[
10] L.O. Snizhko, A.L. Yerokhin, A. Pilkington, N.L. Gurevina, D.O. Misnyankin, A.
Leyland, A. Matthews, Electrochim. Acta 49 (2004) 2085.
[
[
[
[
[
[
[
11] H.F. Guo, M.Z. An, Appl. Surf. Sci. 246 (2005) 229.
12] H.P. Duan, C.W. Yan, F.H. Wang, Electrochim. Acta 52 (2007) 5002.
13] H.Y. Hsiao, W.T. Tsai, J. Mater. Res. 20 (2005) 2763.
14] C. Blawert, V. Heitmann, W. Oietzel, Surf. Coat. Technol. 200 (2005) 68.
15] J.G. Qian, D. Li, F. Zhang, Mater. Sci. Forum 475–479 (2005) 3905.
16] A.V. Timoshenko, Yu.V. Magurova, Surf. Coat. Technol. 199 (2005) 135.
17] V.S. Rudnev, T.P. Yarovaya, K.N. Kilin, I.V. Malyshev, Prot. Met. Phys. Chem. Surf.
4
6 (2010) 456.
18] H.H. Luo, Q.Z. Cai, B.K. Wei, B. Yu, J. He, D.J. Li, J. Alloys Compd. 474 (2009)
51.
19] V.S. Rudnev, K.N. Kilin, P.M. Nedozorov, A.Yu. Ustinov, T.P. Yarovaya, T.A.
Kaidalova, Prot. Met. 43 (2007) 542.
[
[
[
[
5
20] Z.P. Yao, Y.L. Jiang, Z.H. Jiang, F.P. Wang, Z.D. Wu, J. Mater. Process. Technol. 190
(
2007) 117.
Fig. 7. The corrosion current density of the coatings and AZ91D substrate in 3.5%
NaCl solution by polarizing curve method.
21] G.L. Schukin, A.L. Belanovich, V.P. Savenko, L.S. Ivashkevich, V.V. Sviridov, J.
Appl. Chem. 69 (1996) 939 (in Russian).

8
474
Z. Yao et al. / Journal of Alloys and Compounds 509 (2011) 8469–8474
[
22] K.M. Lee, K.R. Shin, S. Namgung, B. Yoo, D.H. Shin, Surf. Coat. Technol. 205 (2011)
[26] W.Y. Mu, Y. Han, Surf. Coat. Technol. 202 (2008) 4278.
[27] Z.P. Yao, H.H. Gao, Z.H. Jiang, F.P. Wang, J. Am. Ceram. Soc. 91 (2008)
555.
[28] L. Wang, L. Chen, Z.C. Yan, H.L. Wang, J.Z. Peng, J. Alloy. Compd. 493 (2010)
445.
3
779.
[
[
23] S.G. Xin, R.G. Zhao, H. Du, L.X. Song, J. Inorg. Mater. 24 (2009) 107–110.
24] E. Matykina, R. Arrabal, P. Skeldon, G.E. Thompson, J. Appl. Electrochem. 38
(
2008) 1375.
25] P. BalaSrinivasan, J. Liang, C. Blawert, W. Dietzel, Appl. Surf. Sci. 256 (2010)
265.
[
[29] V.S. Rudnev, T.P. Yarovaya, K.A. Gaivarotskaya, P.M. Nedozorov, T.A. Kaidalova,
Prot. Met. Phys. Chem. Surf. 46 (2010) 710.
3