Вы находитесь на странице: 1из 10

IPENZ Transactions, Vol. 24, No.

1/EMCh, 1997

Corrosion behaviour of Zincalume and


galvanised coatings in a sulphur-containing
atmosphere
W Gao,1 BE, ME, DPhil, MIPENZ, MACA, MMRS, MASMI, MTSM
D Gifford,2 CEng, MIM
Z-Y Liu,3 BE, ME, PhD
Atmospheric exposure has been performed in a high sulphur-containing
environment in the Rotorua area, New Zealand. Zincalume (55%Al-Zn) coatings
on mild steel sheets have been found to have a service life more than 5 times longer
than the standard galvanised (Zn) coatings. It is believed that the aluminium-rich
dendrites in the coatings play the main role in resisting corrosion. The corrosion
mechanisms have been studied in detail based on the microstructure and
microanalysis evidences obtained with optical and electron microscopy, X-ray
diffraction phase identification, and electron microanalysis (EDX). The durability
of the coatings was predicted by measuring the remaining volume fractions of the
aluminium-rich phase with the linear intercept method.
Keywords: corrosion – galvanised coatings – Zn/Al coatings – sulphur-containing
atmosphere
1
Associate Professor in Materials Science and Engineering, Department of
Chemical and Materials Engineering, University of Auckland, Private Bag 92-019,
Auckland, 2Marketing Manager, BHP New Zealand Steel, Private Bag 92-121,
Auckland, 3Research Fellow, Department of Chemical and Materials Engineering,
University of Auckland
This paper, received in revised form on 6 October 1997, was originally presented at
the 1996 IPENZ Conference.

1. Introduction
Geothermal fluids are used in a variety of applications. In New Zealand, for instance, geothermal
energy has been used for electricity generation, domestic and industrial heating, and to supply
processing steam for the pulp and paper industry. The geothermal atmosphere in Rotorua, New Zealand
has a complicated composition containing H2S, CO2, NH3, H2, N2, and CH4. The proportions of these
gases also reflect the stability and solubility of certain minerals. The H2S-containing atmosphere is
corrosive to metals (1). Therefore, it is important to provide corrosion resistance to the materials used in
such environments by material selection or applying coatings with high corrosion resistance.
Steel coils coated with an aluminium-zinc alloy, known as Zincalume, are produced on a
continuous hot dip metal coating line (2). The typical coating actually contains about 55% aluminium,
43.5% zinc, and 1.5% silicon. The silicon is introduced primarily to assist in the formation of a thin
layer of intermetallic compound between the coating and base metal, providing coatings with good
adhesion. The alloy bath temperature is around 600°C, and the coating is solidified very rapidly to
obtain a fine-grain microstructure with enhanced corrosion resistance. Compared to the pure zinc
coating (so-called galvanised coating) or pure aluminium coating, the Zincalume alloy coatings have
shown an excellent corrosion resistance in marine, industrial and rural environments (3). However, there
is very little information about the corrosion performance of the Zincalume coatings in geothermal
atmospheres containing high concentrations of H2S.
In the present work, outdoor exposure testing (4) has been performed with Zincalume coated and
galvanised mild steel samples in Rotorua, New Zealand. An exposure site with a high natural H2S
concentration was chosen in order to obtain rapid results. The corrosion performances of the two types
of coating were compared. The corrosion mechanism of Zincalume coatings has been studied with X-
ray diffraction phase analysis, optical microscopy, scanning electron microscopy (SEM) and energy
dispersive X-ray microcompositional analysis (EDX).

51
2. Experiments
2.1 Outdoor exposure tests
Up to 22 months exposure tests have been carried out with unpainted Zincalume coated and galvanised
mild steel sheets at a site in Rotorua, New Zealand. This is a natural geothermal environment with a
high concentration of H2S between 10 and 30 ppm. The dimensions and shape of the tested coupons are
schematically shown in Fig.1. The sample rack was constructed 400 mm above the ground, 150 mm
from a wall, at 45° to the ground. Duplicate samples were exposed on the same rack. The testing site
was visited on a regular basis particularly in the initial period of exposure, and photographs of the
samples were taken, and observations for all samples were recorded. Rust was seen developed on the
top surface of the coupons. The fasteners did not have significant effect on the corrosion performance
of the coupons. The evaluation periods were as follows:
after 10 days, 27 days, 41 days, 57 days, 71 days, 91 days, 6.5 months, 10 months, 12 months, 15
months, 20 months, and 22 months. The outdoor exposure experiments are still ongoing.

FIGURE 1: Dimension and shape of the tested coupon.


2.2 X-ray diffraction phase identification
A Phillips PW 1729 X-ray generator was used in conjunction with a PW 1710 diffractometer to identify
the phases of coatings and corrosion products. A segment was cut from each of the specimens for
analysis. Co Kα radiation was used with a scanning rate of 0.05°/s and 2θ ranging from 10° to 110°.
2.3 Scanning electron microscopy (SEM/EDX)
The cross-section morphologies of the coatings on Zincalume coated and galvanised specimens before
and after exposure were studied. A Phillips 505 Scanning Electron Microscope with a Princeton
Gamma Tech Energy Dispersive X-ray Analyser was used to carry out the analyses. The coatings were
scanned by using a 20 kV accelerating voltage, and photographs were taken from the specimens exposed
for different periods of time. Energy dispersive X-ray (EDX) analysis was carried out across the
corrosion products and coating layers to measure the elemental compositions in each region.
2.4 Measurements of the volume fraction of Al-rich dendrites
The measurement of the volume fractions of the Al-rich phases were carried out by using linear
intercept method. The initial volume fraction of the Al-rich phases was measured as 80% (11). The
following equations are established to calculate the corrected values:
LC /
VfA = ⋅β
L (1) and (2)
h
β= 1
ho

52
where VfA is the volume fraction of the Al-rich phases, L is the total length measured with the linear
intercept method, LC’ is the total length of the Al-rich phases measured with the linear intercept method,
h1 is the coating thickness after corrosion and h0 is the thickness of the original coating.

3. Results and discussion


3.1 Corrosion performance of Zincalume and galvanised steels
Corrosion started almost immediately after the Zincalume specimens were exposed in the corrosion
environment, but the corrosion rate was relatively slow. After exposure for 10 days, light tarnish was
visible on the Zincalume coating surface, indicating that a thin layer of corrosion products has already
formed. After 91 days’ exposure, the coating surface showed a patina type with a grey/brown colour.
After 12 months, the panel looked similar to the specimens without exposure, apart from the visual
dulling/patina effect. After 22 months’ exposure, the panel appearance showed little difference from
that after 12 months.
In contrast with the Zincalume coatings, extensive white corrosion product (WCP) was observed on
both the Z275 (coated with 275g Zn/m2 on both sides) and Z450 (coated with 450g Zn/m2 on both sides)
galvanised steel panels after 10 days’ exposure. After 71 days, the WCP became very heavy, with the
scribe and spangle just visible. After 91 days, scribe marks had almost disappeared, black spots were
visible on tension bends where the galvanised coating was thought to have flaked off. After 6.5 months,
severe red corrosion products (RCP) formed on the entire surface of both Z275 and Z450 panels, with
heavily rusted sections flaking off.
The typical optical micrographs of Zincalume (after 12 months and 22 months exposures) and
galvanised (after 6.5 months' exposure) coatings are shown in Figs 2(a), (b), and (c), respectively. It can
be clearly seen that the galvanised Zn coating has corroded away completely in less than 6.5 months,
forming brown corrosion products on the steel surface, while the Zincalume coating has remained
intact, covered with a thin layer of corrosion products.

FIGURE 2: Optical micrographs of the cross-section microstructure : (a) galvanised sheet after 6.5
months exposure, (b) Zincalume coating after 12 months exposure, and (c) Zincalume coating after
22 months exposure.
3.2 X-ray diffraction phase analysis
The X-ray diffraction spectra of an original Zincalume coating is shown in Fig. 3(a). The diffraction
spectra from the corrosion products formed on the surfaces of the Zincalume and galvanised coatings
are shown in Fig. 3(b) and Fig. 3(c), respectively.

53
FIGURE 3: X-ray diffraction spectra (a) a Zincalume coating before corrosion, (b) a Zincalume
coating after one year corrosion, and (c) a galvanised coating after 6.5 months exposure.
The Zincalume coatings were found to contain mainly metallic Al and Zn, with a small amount of
intermetallic compound Al0.71Zn0.29, which was identified by XRD. During the hot dipping process,
cooling rate was high (9-26K/s). Zn is generally in the form of a solid-solution in an Al matrix (12), as
the strong diffractions were from metallic Al and Zn. Intermetallic Al0.71Zn0.29 is a metastable phase.
This type of compound forms on the top surface of the coating during the decomposition of quenched
Al-29at.%Zn alloy, and may play certain roles in corrosion processes.
The X-ray diffraction patterns from the Zincalume coatings after one year’s exposure showed that
the main corrosion products were γ-Al2S3, Zn1.7Al20.2S32, and Al2S3. Because X-ray penetrated the thin
layer of corrosion products, we also observed the diffractions from metallic Al, Zn, and a small amount
of silicon aluminium (Al-Si) from the coating. The X-ray diffraction spectra from the corrosion products
on galvanised steel after 6.5 months (Fig. 3(c)) showed that the corrosion products consist mainly of
FeS, with a small amount of Fe2O3. No Zn compounds can be identified as they dissolved into water
and were washed away.
3.3 Cross-section morphology
Fig. 4 is a typical optical micrograph showing the morphology of a taper cross-section of a Zincalume
coating.

FIGURE 4: Optical micrograph showing the morphology of a taper cross-section of a Zincalume


coating, taper angle ~9°. The intermetallic compound layer is seen between the coating and substrate.

54
The coating consists mainly of Al-rich dendrites and zinc-rich networks between the Al-rich dendrites.
There is also a layer of intermetallic compound between the coating and the steel substrate pointed by
two arrows. In the Al-dendrite areas, EDX showed a large amount of Al with a relatively small amount
of Zn. The Zn-rich areas also contain Zn, Al and Si. The interface intermetallic layer consists of Al, Fe,
Zn and Si. A quaternary Fe-Al-Zn-Si compound with an approximate composition of 52% Al, 38% Fe,
6% Zn, and 4 wt.% Si has been reported for this intermetallic compound (11). The EDX spectra are given
in Figs 5(a), (b) and (c).

FIGURE 5: EDX spectra of a Zincalume coating from (a) Zn-rich area, (b) Al-rich area, and (c)
interface intermetallic compound.
Figs 6(a) and (b) are typical SEM cross-section micrographs showing the morphologies of the
Zincalume coatings after one year’s and two years’ exposure. There is a layer of corrosion products
formed on the coating surface, as indicated by an arrow. EDX spectra of these corrosion products are
given in Figs 7(a) and (b). After one year’s exposure, the morphologies of the Al-rich dendrites and the
interface layer have changed very little. Only the Zn-rich regions have become darker than before. The
EDX spectrum from a typical dark area is shown in Fig. 8(a). Compared with the composition
measured from the coating before exposure, the Zn content remains almost unchanged, but the S content
is increased significantly. This can be explained as: when metallic Zn has been corroded preferentially
away from the Zn-rich areas, the corrosion product zinc sulphate may have been retained in these
regions (5). EDX analysis also indicated that the Al content in the Zn-rich regions has not been
significantly reduced. EDX analyses from the intermetallic layer and Al-rich area are given in Figs 8(b)
and 8(c), respectively. Compared with the original coatings, the intermetallic layer and Al-rich areas
showed very little differences in elemental compositions. This indicates that after one year’s exposure,
the Al-rich dendrites and intermetallic layer had not been attacked, and were providing a good protective
ability.

55
FIGURE 6:SEM micrographs showing the morphology of a cross-section of Zincalume coatings after
(a) one year, and (b) two years exposure.

FIGURE 7: EDX spectra of the corrosion products on Zincalume coatings after (a) one year and (b)
two years exposure.

FIGURE 8: EDX spectra from (a) a Zn-rich area, (b) the intermetallic compound layer, and (c) an Al-
rich area in a Zincalume coating after one year’s exposure.
Table 1 is a summary of the EDX composition analysis and X-ray diffraction phase identification. After
two years’ corrosion, the fraction of the dark region is greater than that after one year’s corrosion. This
indicates that not only Zn-rich areas but also some Al-dendrites have been corroded. Aluminium
sulphite has formed in these areas.

56
TABLE 1: Summary of the EDX and X-ray diffraction analysis for Zincalume coatings.
Before exposure After one year’s exposure
Coating EDX analysis: Al, Zn EDX analysis: Al, Zn, and S.
surface
X-ray diffraction: Al0.71Zn0.29, metallic Al X-ray diffraction: Zn1.7Al20.2S32,
and Zn
γ-Al2S3, and Al2S3
Al-rich matrix: Al-rich matrix:
EDX analysis: Al, Zn EDX analysis: Al, Zn
Coating X-ray diffraction: metallic Al and Zn X-ray diffraction: metallic Al and Zn
layer Zn-rich network: Zn-rich network:
EDX analysis: Al, Zn, Si EDX analysis: Al, S, Zn, and Si.
X-ray diffraction: metallic Al and Zn X-ray diffraction: metallic Al, Zn, and (Al-Si)
Interface EDX analysis: Al, Fe, Zn, and Si EDX analysis: Al, Fe, Zn, and Si

A typical EDX spectrum from the centre of a dark area is shown in Fig. 9(a). It can be seen that S
content has been increased significantly compared with the composition shown in Fig. 8(a). The EDX
spectrum from the an Al-rich dendrite is given in Fig. 9(b). Again, the chemical compositions of the
remaining Al-rich dendrites after one year’s or two years’ corrosion showed very little change compared
with those in the original coatings (Fig. 5(b)).

FIGURE 9: EDX spectra from (a) a Zn-rich area, and (b) an Al-rich area in Zincalume coating after
two years’ exposure
Figs 10(a), (b) and (c) show the SEM micrographs of the morphologies of Zn-rich regions before
corrosion, after one year, and after two years corrosion, respectively. By comparing Fig. 10(a) with Fig.
10(b), it can be seen that after one year’s exposure, some of the Zn-rich regions became darker because
Zn in Zincalume coatings has been attacked preferentially. As Zn has been removed from the Zn-rich
areas, the space has been taken by sulphur-containing corrosion products, as shown by EDX analysis in
Fig. 8(a). It was reported that these types of corrosion products may be locked into the inter-dendritic

57
areas and therefore provide certain protection (2). After two years’ exposure, both Zn and Al phases have
been attacked, with corrosion products formed in the former Zn-rich regions.

FIGURE 10: SEM micrographs showing the morphologies of Zn-rich phase in Zincalume coatings,
(a) before exposure, (b) after one year’s exposure, and (c) after two years’ exposure

3.5 Corrosion mechanisms of Zincalume coatings in a H2S-containing


environment
On the Zincalume coating surface, a metastable intermetallic phase Al0.71Zn0.29 is formed due to the
decomposition of the Al-Zn alloy. In a H2S-containing geothermal environment, this intermetallic phase
reacts with H2S to form a stable phase Zn1.7Al20.2S32, which was identified by XRD.
Zincalume coatings consist of two phases: Al-rich alloy dendrites and Zn-rich alloy phase. Al has a
much more negative standard electrode potential (E0 = -1.663V) than Zn (E0 = -0.763V) (8). However,
experimental evidence described above indicates that Zn is removed first from the Zn-rich alloy phase.
This can be explained as: Al has a strong oxidation tendency in an oxygen-containing atmosphere, and
forms a layer of alumina:
2Al + 3O2 = 2Al2O3 (3)

Al2O3 is known to be compact and stable. A thin film of Al2O3 has a good protective ability to
prevent the Al-rich alloy phase from further corrosion. The Zn-rich phase, however, does not have the
ability to form this type of protective film. Therefore, corrosion took place first in the Zn-rich phase.
The partial anodic reaction can be written as:
Zn = Zn2+ + 2e- (4)

When H2S dissolves into water, the following reaction takes place:
H2S = H+ + HS- (5)

and the partial cathodic reactions are:


2H+ + 2e- = H2 (6)

58
and 2HS- + 2e- = H2 + 2S2- (7)

The corrosion products will be ZnS:


Zn2+ + S2- = ZnS (8)

The overall reaction can be written as:


Zn + H2S = H2 + ZnS (9)

It is reported (6) that ZnS has certain solubility in water (6.9×102 kg/m3). This is why very little ZnS
can be detected on the specimen surface with X-ray diffraction.
In sulphur-containing environments, Al2O3 does not have perfect protective ability. When water is
present, H2S may decompose and Al2O3 can react with S2- to form Al2S3 and lose its protective ability:
Al2O3 + 3S2- + 6H+ = Al2S3 + 3H2O (10)
Without the protective film, Al will also react with H2S:
2Al + 3H2S = 3H2 + Al2S3 (11)

and 2Al + 3S2- + 6H+ = Al2S3 + 3H2 (12)


The corrosion products are Al2S3. Al2S3 is an unstable phase. It may follow the reverse reaction to
equation (10) in moist air, and decompose into Al2O3:
Al2S3 + 3H2O = Al2O3 + 3H2S (13)
The above reactions have been suggested based on the general observations. More studies are
needed to identify the detailed steps.
The microstructural evidence shown in Fig. 2 indicated clearly that corrosion was progressing from
the outside surface inwards. This also suggests that the corrosive agents diffuse through the areas
between the Al-rich dendrites. Pores exist in these areas because of the dissolution of the corrosion
products. It is reported that H2S can diffuse with a flux of 3×10-5 mol/(m2 s) through pores (1).

4. Durability of Zincalume coating in a geothermal environment


As discussed above, when the specimens were exposed to the sulphur-containing atmosphere, two
changes have been observed from the microstructures: firstly, intermetallic Al0.71Zn0.29, Al-rich and Zn-
rich phases on the coating surface corroded. The total thickness of the coating is therefore reduced.
Secondly, corrosive agents penetrated the coatings, removing metallic elements from the Zn-rich phases
and then, the nearby Al-rich areas. The protective ability is therefore mainly provided by the Al-rich
phase. Based on this mechanism, the overall corrosion resistance should be a function of the quantity of
Al-rich dendrites. It is reasonable to estimate the life-time of the coatings by measuring the remaining
volume fraction of the Al-rich phases, so the remaining volume fraction versus time is plotted in Fig. 11.

FIGURE 11: A plot of the remaining volume fraction of the Al-rich phase in Zincalume coating versus
exposure time

59
The relation of VA to t appears approximately linear, implying a constant reaction rate of Al2O3 in the
H2S-containing atmosphere. The slope is an indication of the corrosion rate. By using this rate, we
predict the life-time of the Zincalume coating in the present sulphur-containing geothermal atmosphere
is 2.5 years. Note the role of the intermetallic layer has not been taken into account in this prediction.
The actual life-time of the Zincalume coatings (before the steel sheet starts corrosion) is likely to be
longer than the above estimation.

5. Conclusions
1. Zincalume coatings on mild steel sheets have a corrosion resistance more than 5 times greater than
a Z450 galvanised coating by service-life in a high sulphur-containing geothermal environment.
2. After hot dipping, an intermetallic compound, Al0.71Zn0.29, is formed on the coating surface.
3. After one year’s exposure to the geothermal atmosphere, the corrosion products formed on the
Zincalume coating surface are Zn1.7Al20.2S32, γ-Al2S3, and Al2S3. Zn1.7Al20.2S32 is believed to form
from corrosion of Al0.71Zn0.29. The typical microstructure of the Zincalume coating consists of two
zones. The upper zone consists of Al-rich dendrites and porous Zn-rich networks. The lower zone
mainly consists of compact Al-rich dendrites.
4. After two years’ corrosion, the microstructure of the coatings consists of Al-rich dendrites and
corrosion product networks. Some of the Al-rich dendrites near the intermetallic layer have also
been corroded, but the intermetallic layer between the coating and substrate is intact.
5. The remaining volume fractions of the Al-rich phases can be used to predict the service-life of the
Zincalume coatings on mild steel sheets. The service-life of the Zincalume coating in this severe
geothermal atmosphere is predicted to be ≥ 2.5 years before visible corrosion of steel occurs. This
method can also be used to estimate the protection time of a Al-Zn alloy coating under other
corrosive environments.

6. Acknowledgement
This work is partially supported by BHP New Zealand Steel. The authors would also like to thank Mr.
S Strover for technical assistance.

7. References
1. M. E. Inman, 1994. Corrosion of carbon steel in geothermal systems, PhD thesis, University of Auckland, New
Zealand.
2. H. E. Townsend, L. Allegra, R. L. Dutton and S. A. Kriner, 1986. Hot-dip coated sheet steels - A review.
Materials performance, 25 (8): 36-46.
3. H. E. Townsend and A. R. Borzillo, 1987. Twenty-year atmospheric corrosion tests of hot dip coated sheet
steel. Materials performance, 26 (7): 37-41.
4. B. P. Cook, 1994. Interm report on ColorsteelR, ColorbondR, Unpainted Galvanised and Unpainted ZincalumeR
panels exposed in the Rotorua Geothermal Zone, Sept. 1994, BHP New Zealand Steel Ltd.
5. G. J. Harvey and P. N. Richards, 1984. Zinc-based coatings for corrosion protection of steel sheet and strip,
Metals forum, 6 (4): 235-247.
6. ASM International, 1987. Metals Handbook, Corrosion . 9th ed., Vol. 13, ASM International. pp435.
7. H. J. Cleary, 1994. The microstructure and corrosion resistance of 55%Al-Zn coatings on sheet steel,
Microstructure Science, Vol. 12-Corrosion, Microstructure and Metallography, ASM and Inter. Metall. Soc.,
Ed. by D. O. Northwood et al, pp103.
8. M G Fontana, 1986. Corrosion Engineering, McGraw-Hill Book Company, New York. pp42.

60