Corrosion Science 53 (2011) 16211626

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Corrosion behavior of lanthanum-based conversion coating modied with citric

acid on hot dip galvanized steel in aerated 1 M NaCl solution
Gang Kong , Lingyan Liu, Jintang Lu, Chunshan Che, Zheng Zhong
School of Materials Science and Engineering, South China University of Technology, Guangzhou 510640, China

a r t i c l e

i n f o

Article history:
Received 5 November 2009
Accepted 20 January 2011
Available online 31 January 2011
Keywords:
A. Rare earth elements
B. EIS
B. Polarization
B. Erosion
C. Passive lms

a b s t r a c t
Rare earth conversion coating is one of the most promising substitutes to the toxic chromate coating. The
corrosion resistance of lanthanum conversion coating modied with citric acid on hot dip galvanized
(HDG) steel was investigated in aerated 1 M NaCl solution by means of electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization measurements. Equivalent circuits were subsequently
developed from the measurements to elucidate the corrosion behavior of the coating. The surface morphology of the lanthanum conversion coating was observed by scanning electron microscopy (SEM),
and the chemical composition of the coating was characterized by energy dispersive spectroscopy
(EDS). The results showed that the corrosion process of the modied lanthanum conversion coating consisted of three stages. The overall corrosion resistance of the coating was excellent.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Chromate passivation has been used extensively in surface
treatment of metals until recent decades when the toxicity of
Cr(VI) came under scrutiny in increasing number of countries. It
is now generally accepted that the chromate solution, although
an excellent corrosion inhibitor, will eventually be phased out from
manufacturing processes [1].
The research on the chromate replacements for zinc and zinc
coating started in the late 1980s [210]. Molybdate, silane, silicate,
rare earth (RE) and phosphate salts have been studied extensively
as alternative to toxic chromate for the corrosion protection of zinc
and zinc coating.
In the past decade, most research activities focused on single cerium conversion coating [1116] and Ce compound coatings obtained
using single or multiple step processes [1721]. Those works reported the use of cerium salts for corrosion inhibition on different
metals and alloys and proposed a cathodic mechanism to explain
the formation of the RE oxide lm. Recently, more attentions were
paid to the modication of RE conversion coating with organic and
inorganic inhibitors due to the simplicity and low cost of such modications. One of the successful inhibitors used for the modication
was citric acid, which is renewable and non-toxic. Tizpar and Ghasemi [22] reported that citric acid acted as a good corrosion inhibitor
for the lead alloy exposed to sulfuric acid. Tselesh [23] employed
potentiodynamic technique in his study on the anodic behavior of
tin in citrate buffer solution and revealed that the passive layer
Corresponding author.
E-mail address: [email protected] (G. Kong).
0010-938X/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2011.01.038

formed on the anode surface effectively inhibited the further corrosion of tin. There were, however, few studies on the application of
citric acid on the corrosion protection of galvanized steels.
Extending a prior project on the subject [24], the present study
aims to characterize an environmentally acceptable La conversion
coating modied by citric acid on Hot dip galvanized (HDG) steel.
The corrosion behavior of the coating in aerated 1 M NaCl solution
was investigated by means of electrochemical impedance spectroscopy (EIS) and by the polarization measurements. Equivalent circuit was also developed to study the corrosion process of the
conversion coating.

2. Experimental
Cold-rolled mild carbon steel coupons in the size of 40 mm 
30 mm  2 mm were successively degreased, pickled, uxed and
dried before they were dipped in a molten zinc bath at 450 C for
1 min. Coupons were withdrawn from the zinc bath at a controlled
slow speed and were then immediately quenched in water. The
galvanized steel coupons thus obtained were used as the substrate
of the samples. The thickness of the galvanized coating measured
by STH-1 thickness gauge was about 50 lm. Consistent with a
procedure developed in a prior study [12], the passivation solution
was prepared by mixing analytical reagent grade of 20 g/L
La(NO3)36H2O, 10 mL/L 30% (v/v) H2O2 and 10 g/L citric acid
(H3Cit) in aqueous solutions. The HDG samples were immersed
into the passivation solution for 30 min at 70 C and then dried
in air. A modied La-based conversion coating was thereby
obtained on the surface of the hot dip galvanized samples.

1622

G. Kong et al. / Corrosion Science 53 (2011) 16211626

The electrochemical experiments were performed by using a

CHI640B electrochemical measurement workstation. The electrochemical cell consisted of a working electrode with an exposure
area of 1 cm2, a saturated calomel electrode (SCE) as the reference
electrode, and a platinum foil with an exposure area of 10 cm2 as
the counter electrode. Measurements were taken after the sample
was immersed in an aerated 1 M NaCl solution at ambient temperature for 20 min and the corrosion potential had been stabilized.
The perturbation and scan rate for potentiodynamic linear polarization (LPR) was 10 mV and 1 mV/s, respectively. The electrochemical parameters such as the corrosion potential Ecorr, The
corrosion currents icorr and the polarization resistance Rp were obtained by tting the Tafel slope of the polarization branches using
the software package of CHI640B workstation. The measuring
frequency for electrochemical impendence spectra (EIS) ranged
from 105 down to 102 Hz. The acquired measurement data were
analyzed using the software package of CHI640B. The equivalent
circuits of the EIS data were obtained from Zview.
The surface morphology of the coatings was examined by scanning electron microscopy (SEM, PHILIPS; Model: XL-30-FEG). The
chemical composition was analyzed by energy dispersive spectroscopy (EDS, OXFORD, Model: Inca300).
3. Results and discussion
3.1. Electrochemical behavior

by over one order of magnitude from 13.21 to 0.33 lA/cm2 after

the treatment. The decreases in the current densities were attributable to the coverage of the active area on the electrodes by the La
coating. Correspondent to the decrease in corrosion current, the
polarization resistance Rp increased signicantly from 0.49 to
26.84 kX cm2.
The electrochemical impedance was measured under stable
open circuit potential which was obtained after immersing the untreated and treated samples in 1 M NaCl solution for 20 min. The
results are shown in Fig. 2. The impedance was about 600 X cm2
for the untreated samples and about 24 kX cm2 for the La treated
sample. The increase in impedance is consistent with the increase
in polarization resistance discussed earlier. Compared to single RE
salt conversion coating investigated in a prior study [20], the modied La conversion coating possessed superior anti-corrosion
properties.
3.2. Surface morphology of the La coating
The micro appearance of the La conversion coating is shown in
Fig. 3. The surface morphology displayed certain degree of nonuniformity. In addition, a large number of cracks existed in the
coating, forming a network that is reminiscent to a drought riverbed (Fig. 3a). Flaking of the coating can also be observed. It can be
seen under higher magnications (Fig. 3b) that some occules
were attached to the coating surface.

The electrochemical behavior of the HDG samples with or without La conversion coating in an aerated 1 M NaCl solution was
evaluated by potentiodynamic polarization and the results are
shown in Fig. 1. Compared to the untreated HDG sample, the presence of a La coating on the HDG sample led to signicant changes
in both anodic and cathodic responses. The tting results obtained
from potentiodynamic polarization curves in Fig. 1 were listed in
Table 1. As seen from Table 1, the corrosion potential Ecorr, as measured from Fig. 1, was 1.062 V(SCE) for the untreated HDG sample and 1.116 V(SCE) for the sample with a La conversion coating.
The decrease of Ecorr may be attributed to the effects of the La coating on the cathodic reaction. The corrosion currents icorr decreased

3.3. Corrosion process

Fig. 1. Potentiodynamic polarization curves for the hot dip galvanized steel without
and with La coating.

Fig. 2. EIS spectra obtained on the samples without and with La coating after
20 min of immersion in 1 M NaCl solution.

In order to investigate the corrosion process of the conversion

coating, EIS measurements were carried out on the La coating samples at open circuit potential in 1 M NaCl solution for various
times. Results shown in Fig. 4 reveal that the impedance values decreased with the immersion time in the NaCl solution. For the
immersion time up to 18 h, three capacitive time constants can
be easily distinguished in the Bode plots (Fig. 4a and b): one in
the high frequency zone, another in the high to medium frequency
range and the third in the low frequency zone. When the immersion time increased to over 18 h, diffusion impedance appeared.
The two time constants in the higher frequency zones suggest that

Table 1
Electrochemical parameter obtained from potentiodynamic polarization curves in Fig. 1.
Sample

ba (V vs SCE)

bc (V vs SCE)

Ecorr (V vs SCE)

Rp (kX cm2)

icorr (lA cm2)

HDG
La coating

0.018
0.047

0.084
0.036

1.062
1.116

0.49
26.84

13.21
0.33

ba, bc are Tafel slopes for the anodic and cathodic reactions, the accuracy of the Tafel slopes is 1  108. icorr = (ba bc)/[2.3(ba + bc)Rp].

G. Kong et al. / Corrosion Science 53 (2011) 16211626

1623

Fig. 3. SEM micrograph of the La coating, (a) low magnication, (b) high magnication.

Fig. 4. EIS results obtained for HDG sample with La coating immersed in 1 M NaCl solution during 20 min144 h, (a and b) Bode plots, (c and d) Nquist plots.

the conversion coating might consist of two layers, whereas the

third time constant in the low frequency zone may be attributable
to the charge transfer reactions taken place on the substrate.
Three relaxation stages of the La conversion coating can be seen
in the impedance plots from Fig. 4. After 18 h of immersion in 1 M
NaCl solution, Warburg impedances appeared in the medium
frequency range (36 and 72 h of immersion, in Fig. 4d). At this
stage, the charge transfer resistance of the system decreased to
about one-third of its initial value. When the immersion time
increased to 144 h, the Warburg impedances appeared in the low
frequency range. It can therefore be postulated that the corrosion
process could be divided into three stages: the outer layer of
the coating on the surface of the sample was destroyed and the
electrolyte solution were transported through the inner layer at
the initial stage of immersion (<18 h); at the second stage
(18144 h), the main reaction was the damage of inner layer; at
the last stage (>144 h), the corrosion of the substrate played a
dominant role.
Equivalent circuits were also employed to explain the corrosion
process. Results presented in Fig. 4 were tted by using three
equivalent circuits, corresponding to the aforementioned three
stages. The equivalent circuit for the initial stage is presented in

Fig. 5a. It contains three time constants. The time constant in the
high frequency domain is associated with the presence of the outer
layer of the coating, the second constant in the intermediate frequency accounts for the presence of the inner layer, while the third
constant in the low frequency corresponds to charge transfer reactions occurred on the substrate.
In the equivalent circuit of Fig. 5, the resistive elements Rs, Rf1,
Rf2 and Rct represent the resistances of electrolyte, outer layer, inner layer and charge transfer, respectively. The capacitive elements
Cf and CPEf describe the properties of the protective layer, while Cdl
corresponds to the double layer capacity. The errors involved in the
tting procedures were under 10%. Constant phase element (CPE)
is used most often to describe the frequency dependency of
non-ideal capacitive characteristics [25,26]. The impedance of
CPE is given by:

Z CPE

1
jxn
Q

where j is the imaginary unit, x is the angular frequency, Q corresponds to the pseudo capacitance of the coating and n is the depression angle (in degrees) that evaluates the semicircle deformation.
The factor n, dened as a CPE power, is an adjustable parameter that

1624

G. Kong et al. / Corrosion Science 53 (2011) 16211626

Fig. 5. Equivalent circuit for the La coating at the rst stage of corrosion process in 1 M NaCl solution (a) and experimental and tted impedance spectra of the sample with
20 min immersion (bd).

Table 2
Fitted parameters for La coating with different immersion times using the equivalent circuit in Fig. 5a.
Dipping time

Rs (X cm2)

Rf1 (X cm2)

Cf (lF cm2)

Rf2 (kX cm2)

Y0 (CPEf)  106 (X1 cm2 sn)

n (CPEf)

Rct (kX cm2)

Cdl (mF cm2)

20 min
1h
6h
18 h

1.87
1.95
2.60
1.07

29.77
99.98
57.10
39.08

3.72
1.56
2.21
2.57

6.79
10.27
4.65
2.85

14.81
36.14
38.82
96.37

0.86
0.61
0.65
0.60

14.84
16.79
7.35
3.32

0.18
0.17
0.36
0.46

Fig. 6. Equivalent circuit for the optimal La coating at the second stage of corrosion process in 1 M NaCl solution (a) and experimental and tted impedance spectra of the
sample with 36 h immersion (bd).

1625

G. Kong et al. / Corrosion Science 53 (2011) 16211626

Table 3
Fitted parameters for La coating with different immersion times using the equivalent circuit in Figs. 6a and 7a.
Dipping time

Rs (X cm2)

Rf (X cm2)

Rct (kX cm2)

C (lF cm2)

Y0(CPEdl)  103 (X1 cm2 sn)

n (CPEdl)

W-R (X cm2)

W-T (lF cm2)

W-P

36 h
72 h
144 h

1.92
2.27
1.75

2.26
2.89
14.70

6.86
5.82
0.56

2.31
1.32
3.11

1.01
4.99
0.28

0.58
0.63
0.60

1210
735
1886

0.10
0.05
129.30

0.41
0.36
0.63

Fig. 7. Equivalent circuit for the optimal La coating at the third stage of corrosion process in 1 M NaCl solution (a) and experimental and tted impedance spectra of the
sample with 144 h immersion (bd).

always lies between 0.5 and 1. When n equals unity, the CPE describes an ideal capacitor. For 0.5 < n < 1, the CPE describes a distribution of dielectric relaxation times in frequency space. When n
equals 0.5, the CPE represents a Warburg impedance with diffusional characteristics. Fig. 5bd displayed the tting results in direct
comparison with the experimental measurements shown earlier in
Fig. 4 (20 min immersion time). The excellent agreement validates
the model and the underlying corrosion mechanism.
The detailed parameters used in the tting are listed in Table 2.
It is apparent that with increasing immersion time, the coating
resistances Rf1, Rf2 and Rct initially increased and then decreased,
whereas the opposite trend was observed for the capacitances Cf
and Cdl. The changes in these physical properties reect the
changes taken place at the electrolyte/electrode interface during
the immersion. In the initial stage of immersion (1 h), the gradual
coverage of the corrosion product on electrode surfaces led to increases in Rf1, Rf2 and Rct values. As the immersion time increase,
ions in the electrolyte such as the chlorine ions inltrated the La
salt conversion coating through the micro-holes and other defects,
and created more active points and defects in the process. The electrolyte inltration facilitated the charge transfer process and thereby reduced the coating resistance. As a result, Rf1, Rf2 and Rct values
decreased with increasing immersion time.
As the electrolyte permeated through the inner layer of the conversion coating and eventually reached the substrate/lm interface, the exponents of CPE, n, gradually decreased to a value

close to 0.5, indicating the presence of a diffusion eld. In the

corrosion process, the anodic reaction was the dissolution of zinc
whereas cathodic reaction was predominantly the reduction of
oxygen. The current density associated to the hydrogen discharge
at the cathode is negligible compared to the current density related
to oxygen reduction. Thus, the Warburg impedance could be attributed to oxygen transport from the bulk solution to the zinc surface
[27]. The equivalent circuit for this stage of the corrosion process is
presented in Fig. 6a, where Rs is the solution resistance, Rf is the
coating resistance, Rct is the charge transfer resistance, Cf is the
coating capacitance, CPEdl describes the non-ideal capacitive
behavior of the substrate/lm interface, W is the Warburg impedance associated to the oxygen diffusion process. Fig. 6bd display
the results of the tting to one set of the experimental measurements (36 h immersion time) presented in Fig. 4. Excellent agreement was achieved, validating the adequacy of the model. The
parameters used in the tting are listed in Table 3.
During the third stage of corrosion, macro defects appeared in
the coating, the transport process of chemical species proceeded
without barriers. The diffusion eld in the inner layer of the coating disappeared, while a new diffusion eld appeared at the
substrate/solution interface. At this stage, the coating offered no
protection against corrosion, and the severe corrosion of the substrate commenced.
The impedance spectra at the third stage (Fig. 7) contain valuable information concerning the kinetics of the corrosion process

1626

G. Kong et al. / Corrosion Science 53 (2011) 16211626

occurring at the electrode surface. The equivalent circuit was again

developed and the parameters for tting are listed in Table 3. In the
range of high and intermediate frequencies, Nyquist plots exhibit a
complex capacitive loop, which can be related to the active dissolution of zinc. While at low frequencies the impedance behavior reects the transport process of reacting chemical species from
solution to the electrode surface.
4. Conclusion
The La conversion coating modied with citric acid on hot dip
galvanized steel has an excellent corrosion performance in aerated
1 M NaCl solution. Electrochemical tests results show that the
impedance value of modied La conversion coating is around
24 kX cm2, compared to only about 600 X cm2 for bare HDG samples. Rp of the coating tested by the potentiodynamic polarization
curve is 26.84 kX cm2, approximately 50 times as bare HDG sample. The La salt conversion coatings inhibit both the anodic and
cathodic responses of the system and hence prevent the HDG substrate from corrosion. The corrosion process of the coating can be
divided into three stages. At the initial stage of immersion
(<18 h), the outer layer of the coating on the surface of the sample
was destroyed and the electrolyte solution were transport through
the inner layer. At the second stage (18144 h), the main reaction
was the damage of inner layer. At the nal stage (>144 h), the corrosion of the substrate played a dominant role.
Acknowledgements
The investigation is supported by the Fundamental Research
Funds for the Central Universities (2009ZM 0072) and the ILZRO
project (ILZRO/IZA/2009023). The authors are also grateful to
Dr. Daniel Liu from Teck PTC for the modication of English and
several interesting discussions.

References
[1] B. Ramezanzadeh, M.M. Attar, M. Farzam, Surf. Coat. Technol. 205 (2010)
874884.
[2] Abdel Salam Hamdy, D.P. Butt, A.A. Ismail, Electrochim. Acta 52 (2007)
33103316.
[3] Lihui Yang, Junqing Li, Xiang Yu, Milin Zhang, Xiaomei Huang, Appl. Surf. Sci.
255 (2008) 23382341.
[4] Daimon K. Heller, William G. Fahrenholtz, Matthew J. OKeefe, Corros. Sci. 52
(2010) 360368.
[5] Yingwei Song, Dayong Shana, Rongshi Chen, Fan Zhang, En-Hou Han, Corros.
Sci. 51 (2009) 6269.
[6] M.F. Montemor, A.M. Simes, M.G.S. Ferreira, M.J. Carmezim, Appl. Surf. Sci.
254 (2008) 18061814.
[7] Gang Kong, Jintang Lu, Shuanghong Zhang, Chunshan Che, Haijiang Wu, Surf.
Coat. Technol. 205 (2010) 545550.
[8] B.R.W. Hinton, L. Wilson, Corros. Sci. 29 (1989) 967975.
[9] L. Roman, M. Blidariu, C. Cristescu, Trans. Inst. Met. Finish. 75 (1997) 171174.
[10] K. Aramaki, Corros. Sci. 46 (2004) 15651579.
[11] Y. Hamlaoui, F. Pedraza, C. Remazeilles, S. Cohendoz, C. Rbr, L. Tifouti, J.
Creus, Mater. Chem. Phys. 113 (2009) 650657.
[12] C. Motte, N. Maury, M.G. Olivier, J.P. Petitjean, J.F. Willem, Surf. Film Technol.
200 (2005) 23662375.
[13] M.A. Arenas, J.J. de Damborenea, Surf. Film Technol. 187 (2004) 320325.
[14] M.F. Montemor, A.M. Simoes, M.G.S. Ferreira, Prog. Org. Films 44 (2002)
111120.
[15] Ajit Kumar Mishra, R. Balasubramaniam, Corros. Sci. 49 (2007) 1027
1044.
[16] Jin Hu, Shawei Tang, Zaiyang Zhang, Corros. Sci. 50 (2008) 31853192.
[17] Luis M. Palomino, Patrcia H. Suegama, Idalina V. Aoki, M.F. Montemor,
Herclio G. De Melo, Corros. Sci. 51 (2009) 12381250.
[18] Luis M. Palomino, Patrcia H. Suegama, Idalina V. Aoki, M. Fatima Montemor,
Herclio G. De Melo, Corros. Sci. 50 (2008) 12581266.
[19] K. Aramaki, Corros. Sci. 44 (2002) 26212634.
[20] M.G.S. Ferreira, R.G. Duarte, M.F. Montemor, A.M.P. Simes, Electrochim. Acta
49 (2004) 29272935.
[21] K. Aramaki, Corros. Sci. 49 (2007) 19631980.
[22] A. Tizpar, Z. Ghasemi, Appl. Surf. Sci. 252 (2006) 86308634.
[23] A.S. Tselesh, Thin Solid Films 516 (2008) 62536260.
[24] Gang Kong, Lingyan Liu, Jintang Lu, Chunshan Che, Zheng Zhong, J. Rare Earth
28 (2010) 461465.
[25] G. Quartarone, M. Battilana, L. Bonaldo, T. Tortato, Corros. Sci. 50 (2008)
34673474.
[26] Y.X. Qiao, Y.G. Zheng, W. Ke, P.C. Okafor, Corros. Sci. 51 (2009) 979986.
[27] H. Ma, S. Chen, B. Yin, S. Zhao, X. Liu, Corros. Sci. 45 (2003) 867882.