Journal of Materials Science & Technology
Journal of Materials Science & Technology
Journal of Materials Science & Technology
Research Article
a r t i c l e i n f o a b s t r a c t
Article history: The microstructure and corrosion behavior of Al-Zn-Mg alloy (namely 7×××) after natural aging treat-
Received 14 May 2019 ment (NAT) and artificial aging treatment (AAT) in aqueous NaCl solutions containing different aggressive
Received in revised form 31 July 2019 ions have been investigated in current work. Results of microstructure characterization demonstrate that
Accepted 11 September 2019
the aging treatment has a great influence on the grain size and precipitates. The grain size is relatively
Available online 17 November 2019
sizeable and no evident precipitates are observed in alloy after NAT comparable with that after AAT.
The electrochemical corrosion behavior of alloy was studied by polarization curve and electrochemical
Keywords:
impedance spectroscopy (EIS). The corrosion potential (Ecorr ) of the aluminum alloy is more negative in
Al-Zn-Mg alloy
Aging treatment
3.5 wt.% NaCl containing 0.052 wt.% NaHSO3 solution than that in 3.5 wt.% NaCl solutions with or without
Microstructure 0.907 wt.% NaHCO3 . Charge transfer resistance (Rct ) results reveal that alloy after AAT has an enhance-
Corrosion behavior ment of corrosion resistance compare with that after NAT. With the immersion time increasing, mostly
Aggressive ion pitting spreads over the surface of the alloy only in NaCl solution, whereas exfoliation corrosion mainly
occurs in NaCl solutions containing NaHSO3 or NaHCO3 .
© 2019 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science &
Technology.
https://doi.org/10.1016/j.jmst.2019.09.030
1005-0302/© 2019 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & Technology.
86 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Liu et al. [12]. It reported that the appropriate double-aging (DA) electrode, i.e., Ag(s)|AgCl(s)|Cl- (aq, saturated KCl), and platinum
and RRA could better promote the corrosion resistance than T6 sheet with a large area was chosen as the auxiliary electrode.
when alloy was immersed in 3.5 wt.% NaCl solution for an acceler- Three electrolyte solutions containing 3.5 wt.% NaCl solution (S1),
ating test. In the case of T6 temper, the oxide film broke down, and 3.5 wt.% NaCl containing 0.052 wt.% NaHSO3 (S2), 3.5 wt.% NaCl con-
continuous distribution of bare MgZn2 phase and coarse Al7 Cu2 Fe taining 0.907 wt.% NaHCO3 (S3) were chosen to simulate different
could facilitate the development of pit cavity. Wang et al. [8] atmospheric environments by changing anions [23–25]. Besides,
reported that active MgZn2 particles, secondly Al2 MgCu and Mg2 Si it should be noted that the real service environments for alu-
were main factors for the serious corrosion of 7A60 aluminum alloy, minum alloy is quite complicated, which generally includes various
while Al7 Cu2 Fe particles have little impact on pitting corrosion, and kinds of aggressive ions. To use 3.5 wt.% NaCl solution to sim-
the corrosion process of 7A60 alloy in 3.5 wt.% NaCl electrolyte solu- ulate the marine atmosphere environment has been agreed by
tion consists of two distinguished stages based on electrochemical many researchers, and numerous works for simulating the effect of
impedance spectroscopy (EIS) and electrochemical noise (EN) tech- marine atmosphere environment on metal corrosion have been car-
niques results. Meanwhile, Chen et al. [13] studied the influence ried out under such conditions as cited literatures above [23–25]. In
of heat treatment on various properties of 7085 aluminum alloy. current work, the three solutions do not refer to three completely
Studied results demonstrated that high temperature and subse- identical service environments for Al-Zn-Mg aluminum alloy dur-
quent low temperature aging (HLA) played a role of decelerating ing its application. Our idea is to extract three typical aggressive
the enhancement of corrosion resistance and strength from alloy- ions with its suitable concentration based on literatures before
ing, whilst RRA as well as dual-retrogression and re-aging (DRRA) experiment, and investigate the single/ combined effect of these
improve the exfoliation corrosion resistance, and these behaviors aggressive ions on the corrosion behavior of Al-Zn-Mg aluminum
were rationalized by precipitates prevailed over matrix and grain alloy. In addition, all solutions were prepared with deionized water
boundaries. Tang et al. [14] studied the continuous and discontin- and were open under a natural condition.
uous localized corrosion of 2××× series aluminum alloy. Results
showed that discontinuous localized corrosion event was more 2.2. Methods
common due to the Al2 Cu and Al-Cu-Fe-Mn-(Si) particles. In addi-
tion, the grain size is another significant factor for the corrosion of 2.2.1. Microstructure investigations
alloys [15–19]. Tian et al. [20,21] revealed that grain size has, which The microstructure of specimens after NAT and AAT treat-
showed that the frequency of pitting current transients increased, ments were performed by electron back scatter diffraction (EBSD)
and peak value of current transients decreased simultaneously with technique (NordlysNano, Oxford instruments, Inc., Britain) with
increasing ratios of the fine grain, while the number and volume a TESCAN MAIA3 high resolution field emission scanning elec-
fraction of stable pits on alloy obviously decreased. Moreover, Ral- tron microscope (SEM), operated at the voltage of 20 kV and the
ston and Davies et al. [17,22] studied the relationship between grain step size of around 4 m. The results of EBSD were deeply pro-
size and corrosion rate of metals. Results depicted that the tendency cessed by Channel5 software for further analysis. Additionally, in
of corrosion rate decreased with the diminution of grain size in neu- order to observe dispersed precipitates both in grain boundary and
tral solution. What’s more, fine grain structures were beneficial to matrix, test samples were thinned to 75 m in thickness accurately,
create more reactive surfaces. stamped into 3 mm in diameter and further thinned and polished
In this work, the impact of natural aging treatment (NAT) by the mixed solution HNO3 :CH3 OH = 1:3 on a twin-jet electropol-
and artificial aging treatment (AAT) on the microstructure and ishing device (RL-I, Ruiling Innovation, Inc., China) at a constant
corrosion behavior of Al-Zn-Mg alloy under different simulate current mode of 80 mA at around -30 ◦ . Well prepared samples
atmospheric environments was investigated, which was usually were characterized by utilizing FEI Tecnai G2 F20 S-TWIN transmis-
used to manufacture parts of heavy equipment. To understand the sion electron microscope (TEM) carried out at 200 kV. Moreover, in
effect, of varied typical anions, microstructure characterization and order to describe the microstructures after the immersion at 12, 72
various electrochemical tests were implemented, and the relation- and 120 h under three different solutions and characterize compo-
ship between structure and evolution of corrosion resistance was sitions of corrosion products, Hitachi SU8010 SEM in the secondary
figured out. electron mode coupled with EDS was used and operated at 15 keV.
Fig. 1. Typical microstructural images of Al-Zn-Mg aluminum alloy with two different treatments: (a) orientation map with NAT, (b) orientation map with AAT, (c) color
code acted on characterize crystallographic orientations on standard stereographic projection (different color represents diverse orientation projections: red is [0 0 1]; blue
is [1 1 1] and green is [1 0 1]) and distribution diagram of various grain diameter in Al-Zn-Mg aluminum alloy with NAT (d) and AAT (e), respectively.
by immersing the specimens into three different electrolytes with in Fig. 1(d) (especially in the range of 11–35 m). However, as can
Gamry Reference 600 electrochemical workstation in the applied be seen in Fig. 1(b), the difference in size of the grains of Al-Zn-Mg
potential range from −250 mV to +1200 mV vs. OCP with a scan rate aluminum alloy after AAT is around 5−40 m in Fig. 1(e) (mostly
of 1 mV s−1 after OCP was stable. Corresponding corrosion potential in the range of 5–12 m), mainly comprising of anisometric coarse
(Ecorr ) and corrosion current density (icorr ) could be evaluated by grains and a mass of fine sub-grains.
data obtained from polarization curves. TEM was employed to characterize the microstructure, and the
corresponding results are shown in Fig. 2. Many dislocations can
2.2.4. Bulk pH measurements be seen in intragranular areas of samples after NAT in Fig. 2(a)
To illustrate changes in bulk pH values from initial stage to and (b). Furthermore, no apparent precipitates are observed in
120 h immersion. Variations in pH as a function of time for all the grain boundaries while a few speckle-like particles disperse well
investigated solutions were measured by using pH detector (Rex in grains. Moreover, the samples after AAT are varied in com-
Analytical Instruments, Inc., China). A commercial pH sensor was parison to NAT. A high-population density of round-shaped and
calibrated in two universal buffer solutions firstly, which pH val- elongated IMPs is posed in grains and sub-grains, and relatively
ues were 6.86 and 9.18, respectively. Afterwards, we measured sizeable IMPs distribute along grain boundaries in various shapes.
three electrolyte solutions every 12 h for up to 120 h at ambient Besides, the distinct precipitate free zones (PZF) are additionally
temperature of 25 ◦ C in air. seen along grain boundaries (Fig. 2(b) and (d)). Previous studies
[26,27] have reported the precipitation sequence of Al-Zn-Mg alloy
3. Results i.e., main ˛-supersaturated solid solution, Guinier-Preston zones
(GP zones) and then to form ’ phase. The great majority of ’
3.1. Microstructure and composition phase is a semi-coherent metastable MgZn2 phase, and eventu-
ally the ’ phase transforms into equilibrium phase, i.e. MgZn2
Fig. 1(a) and (b) display the microstructures of Al-Zn-Mg alu- phase. GP zones and ’ phase usually precipitate in intragranular
minum alloy with NAT and AAT, respectively. The boundaries of regions, while the phase generally precipitate in grain bound-
each grain, sub-grain and recrystallized- grain are legible, while aries. In addition, the structure of ’ and phases are always easy
different kinds of grains are coded with diverse colors in terms of to distinguish from ˛-matrix phase, and the both are hexagonal
its characteristic crystal orientation on account of [0 0 1] inverse crystallographic microstructure. In order to identify the elements
pole figure as standard presented in Fig. 1(c). It can be observed and composites of those precipitates, the TEM/EDS measurements
that two different aging treatments after solution treatments lead were performed. The STEM-HAADF (scanning transmission elec-
to a significant difference in grain size. Microstructures of samples tron microscope-high-angle annular dark field) images of samples
after NAT consists of anisometric coarse grains, which are along after NAT and AAT and the corresponding EDS analysis results are
the extruding direction, and a few recrystallized-grains well dis- displayed in Fig. 3. Due to in the difference of imaging patterns,
tribute in the metal matrix. Moreover, the range of grain diameter phase contrast of STEM-HAADF images reverses to TEM images.
is around 10−120 m based on grain diameter distribution diagram The brighter contrast in STEM-HAADF images is relative to heav-
88 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Fig. 2. TEM images of Al-Zn-Mg alloy with (a, b) NAT and (c, d) AAT conditions.
ier atomic compositions while the darker contrast signifies minor under S2 solution, which contributes to lose its metallic luster in
ones [28–30]. Line scan mode was used for analyzing particles comparison to that under S1 and S3 solutions. The samples with
elements, with the maximum resolution is around 0.102 nm. The AAT show a similar bubble evolution behavior as the sample with
EDS results of TEM indicate that the main compositions are Al, Mg NAT as shown in Fig. 4, hence which are not exhibited here. Also,
and Zn, which can establish a consistent relation with EDS results Figs. 5 and 6 show the SEM images of the corroded surface mor-
mentioned above. However, three obvious peaks are observed as phologies of samples with NAT and AAT immersed in S1, S2 and S3
scanning through three brighter contrast precipitates for samples solutions over 12, 72 and 120 h with the insets of higher magni-
after NAT in Fig. 3(a), which is likely to attribute to Al3 Zr dispersoid fication of local micro-surface. Two typical corrosion forms can be
combined with results of selected area electron diffraction (SAED) seen, that is pitting corrosion and exfoliation corrosion occur under
[31]. Meanwhile, the results of the samples after AAT in Fig. 3(b) the three different solutions. Sparse pits having a small number of
show that the content of the brighter contrast precipitates is dif- branches are found on the surface of the alloy both with two aging
ferent from the darker ones. The brighter contrast phases are rich conditions at initial immersion stage in S1 solution, moreover, loose
in Mg and Zn, namely phase or its analogues. In addition, the and fine granular white particles are stacked over pits from a close
darker contrast precipitates are of element comparable to that in look. After 120 h corrosion (Figs. 5(g) and 6 (g)), the pits spread
matrix, though the Al content decreases in intragranular regions, over surface and their scale become large, and the size of the visi-
which should be relative to more complex composites [32–34]. ble pits is comparable, ranging from a few to a dozen microns, while
the granular white particles gather together and grow up to form
3.2. Corrosion behavior evolution dense flocculent corrosion products for sample with AAT in Fig. 6(g).
More serious corrosion phenomena are found at the NAT condi-
Immersion tests of Al-Zn-Mg alloy after NAT and AAT were per- tion where corrosion products are changed to network plied on the
formed in S1, S2 and S3 solutions at ambient temperature of 25 ◦ C, surface, the corresponding EDS analysis results of these corrosion
respectively. Fig. 4 presents the typical optical surface images of products are presented below. The five main elements like Al, Zn,
the aluminum alloy after NAT during tests. Numerous bubbles are Mg, Si and O are detected from corrosion production during 120 h
produced and absorbed on the surface or in the vicinity of the alu- under S1 solution, indicating that the corrosion products mainly
minum alloy in S1 and S3 solutions, while few bubbles are observed consist of oxide or hydroxide of Al, Zn, Mg and Si. Besides, chip-like
in S2 solution, indicating corrosion reaction mechanism is probably corrosion pits are apparently seen in S2 solution at the initial stage.
different in solutions containing different cathodic anions. More- With the immersion time increasing, chip-like corrosion products
over, an evident film covering on the surface of the alloy is observed are changed to relatively honeycomb-like in Figs. 5(h) and 6 (h), and
P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98 89
Fig. 3. STEM-HAADF images of the IMPs of samples after NAT (a) and AAT (b) with corresponding EDS results of precipitates.
it can be noted that a sizeable exfoliation region appears obviously, may contain aluminum sulphates except for substances mentioned
as a massive block of corrosion products is exfoliated from the sur- above in S2 solution. In addition, more severe corrosion occurs in
face (Fig. 5(h)). The results of EDS reveal that the corrosion products S3 solution in the case of samples after NAT compared with that
90 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Fig. 4. Typical surface image of Al-Zn-Mg alloy after NAT immersed in (a) S1 solution, (b) S2 solution and (c) S3 solution at ambient temperature of 25 ◦ C during experiments.
after AAT, and some relatively loose tile-like exfoliated corrosion 3.3.2. EIS measurements
products are scattered on the surface (12 h, in Fig. 5(c)). At the last EIS measurements of alloy with NAT and AAT were performed
period of corrosion test (120 h, in Fig. 5(i)), a large chunks of cor- in S1, S2 and S3 solutions at ambient temperature of 25 ◦ C, respec-
rosion products appears and covers on the surface, which seems tively, and tested at 4, 8, 12, 24, 48, 72, 96 and 120 h to estimate
like a block of lifted land after earthquake. Furthermore, the sur- the corrosion behavior as a function of immersion time. Represen-
face of alloy looks like an arid field at a high magnification after the tative EIS experimental results and corresponding fitting curves
immersion time of 120 h. However, no evident tile-like exfoliated using ZSimpWin analysis software are shown in Figs. 9 and 10. The
corrosion products are found for Al-Zn-Mg alloy with AAT, suggest- existence of unsteady state at initial immersion time contributes
ing that the alloy with AAT has a better corrosion resistance than to data points disperse especially in low-frequency regions. How-
that of sample with NAT in S3 solution. ever, with increasing the immersion time up to 24 h, low-frequency
Considering the influence of pH change in experiment, since the data points become regular gradually. Besides, the Nyquist plots
bubble evolution was observed in S1 and S3 solutions during the show the similar characters, i.e., capacitive impedance arcs appear
immersion test, the bulk pH measurements were performed on the at high and middle frequency and inductive impedance arcs appear
three solutions at ambient temperature of 25 ◦ C in air. The results at low frequency range. Nyquist plots of samples with NAT under
are shown in Fig. 7. The bulk pH value of S1 solution (red line with S1 solution present a slant angle at low frequency ranges, implying
circle) keeps neutral at first, and then increases slightly with the diffusion process controls entire corrosion process rather than elec-
immersion time, and this may be attributed to the dissolution of trochemical charge transfer controls in later period of immersion.
carbon dioxide from air in solution. What’s more, the pH value mea- Moreover, the radius of Nyquist loop decreases over the immersion
sured in S2 solution (green line with diamond) decreases mildly at time in S2 solution while it fluctuates in comparison to that in S1
initial and then remains invariant, indicating hydrogen sulfite ion and S3 solution, reflecting that Al-Zn-Mg alloy has the worse cor-
hydrolyzes to form hydrogen ion in solution, which is very oppo- rosion resistance in S2 and more complicated factors in S1 and S3
site to results gained from S3 solution (blue line with square), i.e., solution. This encounter is consistent with results obtained from
bicarbonate radical ion hydrolyzes to alkalify solution. The order the polarization curve measurements. In addition, it is notable that
of bulk pH is as follows: S3> S1 > S2, and no significant variation the radius of capacitive impedance arcs in high and middle fre-
of bulk pH value can be found, indicating the bulk pH of solution quency ranges is large and capacitive impedance arcs digress from
remains relatively stable even obvious bubble evolution presences. the feature of semicircle, so it is rational to believe that capacitive
impedance arcs in high and middle frequency ranges are overlaid
by more than one, i.e., the number of time constant is in excess of
one at this ranges (Figs. 9(a), (c), (e) and 10 (a), (c), (e)). According to
3.3. Corrosion electrochemical behavior the previous literatures [35–37], the dielectric behavior of passive
film on Al-Zn-Mg alloy usually exhibits the capacitive impedance
3.3.1. Polarization curve results arc of high frequency ranges, and low-frequency arc generally cor-
Potential dynamic polarization curves of the alloy under three responds to the process of metal dissolution, such as metastable
electrolyte solutions at ambient temperature of 25 ◦ C with a scan or stable pitting corrosion. Furthermore, absorption of bubbles on
rate of 1 mV s−1 are shown in Fig. 8. The similarly shape of polariza- the surface and stratification caused by accumulation of corro-
tion curves manifests the similar pattern of corrosion reactions goes sion products are normally related to the emergence of inductive
through. The cathodic current densities head towards a plateau impedance arc in low frequency ranges [38–41]. On top of that, it
in three different solution for both samples with different aging can be seen from experimental results that the radius of inductive
treatments, indicating that electrode reactions are controlled by impedance arcs at low frequency regions presents same character
cathodic reactions at a potential negative far from Ecorr . Moreover, as capacitive impedance arcs, and the evident overlaid inductive
the anodic current densities increase rapidly at a potential above impedance arcs appear, indicating that more than one time con-
Ecorr , which could be explained that the compact oxide film on the stant occur at low frequency ranges as well. Moreover, angle phase
surface of aluminum alloy is inclined to broken, and many alu- plots signify information that ideal capacitance deviates to the cor-
minum atoms are separated from the matrix lattice to the solution responding electrochemical interface due to the maximum values
by removing electrons. However, the trend slows down gradu- of phase angle are at most 90◦ .
ally with the potential increasing, which signifies the reactions are The corresponding EEC models are provided in Fig. 11 using
under diffusion control. In addition, no evident passive region can ZSimpWin analysis software in order to further analyze the EIS
be seen from the anodic curves branch. It can be observed explicitly behavior of aluminum alloy immersed in three electrolytes for
that the values of Ecorr of samples in NAT are more negative than different immersion time quantitatively. The definition of EEC ele-
that in AAT in all solutions. Another apparent result is the values ment is presented as follows: Rs is the solution resistance, CPEdl
sequence of Ecorr for samples with both aging conditions are listed is used to represent the non-ideality of double-layer capacitance,
as follows: S2 < S1 < S3.
P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98 91
Fig. 5. SEM images of the microstructure of Al-Zn-Mg alloy after NAT immersed in (a, d, g) S1 solution, (b, e, h) S2 solution and (c, f, i) S3 solution at ambient temperature of
25 ◦ C for (a, b, c) 12, (d, e, f) 72 and (g, h, i) 120 h with corresponding EDS analysis of framed region, (A) S1/120 h, (B) S2/120 h, (C) S3/120 h. The insets are higher magnification
of micro-surface.
92 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Fig. 6. SEM images of the microstructure of Al-Zn-Mg alloy after AAT immersed in (a, d, g) S1 solution, (b, e, h) S2 solution and (c, f, i) S3 solution at ambient temperature of
25 ◦ C for (a, b, c) 12, (d, e, f) 72 and (g, h, i) 120 h immersion with corresponding EDS analysis of framed region, (A) S1/120 h, (B) S2/120 h, (C) S3/120 h. The insets are higher
magnification of micro-surface.
P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98 93
4. Discussion
Fig. 8. Typical polarization curves of Al-Zn-Mg alloy after (a) NAT and (b) AAT immersed in S1 solution (green line), S2 solution (red line) and S3 solution (blue line) at
ambient temperature of 25 ◦ C with a scan rate of 1 mV s−1 .
94 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Fig. 9. Representative Nyquist and Phase angle plots of the Al-Zn-Mg aluminum alloy with NAT under (a, b) S1 solution, (c, d) S2 solution and (e, f) S3 solution at ambient
temperature of 25 ◦ C for 4, 8, 12, 24, 48, 72, 96, 120 h, respectively. The colorful scattered symbols are experimental data and the corresponding fitting results are represented
by black solid lines.
ture alternating with brightness and darkness, whereas mostly it MgZn2 phase decreases with content increment of Cu, and a con-
displays decrease in the content of Al compared with the matrix current decrease in Zn and Mg make the potential of MgZn2 phase
both in brighter and darker contrast precipitates, but a concurrent towards to noble. Moreover, the increasing content of Zn and Cu and
increase in the content of Zn and Mg is observed in brighter con- decreasing content of Mg would also shift MgZn2 phase towards
trast precipitates, and no evident increment of Zn and Mg is found less anodic. In short, the difference of the content of Cu, Zn and Mg
in darker contrast precipitates. Admittedly, the content of these in MgZn2 phase is the difference of potential between MgZn2 phase
two elements to Al is lower than brighter contrast precipitates. and surrounding aluminum matrix, it determines whether they can
Therefore, a conclusion can be drawn combined with phenom- act as electrochemical units or not, whereas the effect of MgZn2
ena mentioned above with results of TEM/EDS in Fig. 3, referring phase would be overwhelmed and even reversely protect matrix
to the composition of precipitates along grain boundaries is dif- from attacking by these factors. Additionally, sparser and strip-like
ference. Furthermore, electrochemical properties of MgZn2 phase precipitates enhance restrain corrosion propagation by reducing
is correlated with alloying element, electrochemical activity of hydrogen concentration and further eliminate the negative impact
P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98 95
Fig. 10. Representative Nyquist and Phase angle plots of the Al-Zn-Mg aluminum alloy with AAT under (a, b) S1 solution, (c, d) S2 solution and (e, f) S3 solution at ambient
temperature of 25 ◦ C for 4, 8, 12, 24, 48, 72, 96, 120 h, respectively. The colorful scattered symbols are experimental data and the corresponding fitting results are represented
by black solid lines.
on metal atom bonding [47]. The higher volume fraction of pre- above. As a result, in present work the finding is proposed, that the
cipitates both in grain boundaries and intragranular regions may corrosion resistance of Al-Zn-Mg alloy after AAT is better than that
contributes to weak the effect of galvanic coupling between grains after NAT. In addition, the electrochemical behavior and corrosion
boundaries and alloy matrix. Otherwise, more dislocation lines and resistance not merely depend on potential of precipitates, but result
loops spreading over the intragranular regions can be noted in alloy from a synergistic effect of the grain size, precipitate composition
with NAT, as the presence thereof in the microstructure may have and distribution as well as other factors.
positive influence on the corrosion rate due to their disorder fea-
ture, which makes them as favorable places for corrosion attack, 4.2. Effect of cathodic ions on electrochemical behavior
thus they may have serious corrosion. Also, high density of dislo-
cation may inevitably accelerate increment of pits in depth with It is concluded from the experimental data that the Ecorr of alloy
spreading pits for coarse grain and hinder repassivation for alloy with NAT is more negative than that with AAT, moreover, the evo-
at the later stage. Moreover, no obvious dispersoid posed in grain lution of fitting values of the Rct with the immersion time in three
boundaries strengthens the effect of galvanic coupling mentioned solutions is shown similarly. Compared with Al-Zn-Mg alloy with
96 P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98
Fig. 11. Equivalent circuit models used to fit EIS data under OCP condition at ambient temperature of 25 ◦ C. Rs is electrolyte solution resistance, CPEdl is double-layer
capacitance at high frequency, Rf is film resistance, CPEhole is associated with hole capacitance, Rhole is associated with hole resistance, CPEct is charge transfer capacitance at
low frequency, Rct is charge transfer resistance, Ws is Warburg impedance, L is the inductance for fitting dispersive inductive tails and R is corresponding resistance at low
frequency.
NAT, the Rct of that in AAT presents one order larger in value. Gener- solutions, whereas rare bubble is found in S2 solution during tests
ally, the reciprocal of Rct is regarded as an important parameter for as shown in Fig. 4. What’ more, the surface of alloy lost its metallic
characterizing the corrosion rate. Therefore, Rct is always taken to luster under S2 solution as if it is covered by a passive film com-
illuminate the evolution of corrosion process and corrosion resis- pared with S1 and S3 solutions in Fig. 4, so it is rational to believe
tance. In this work, Fig. 12 depicts the values of Rct collected from that these bubbles promote lower stability of aluminum hydrox-
the fitting results for S1, S2 and S3 solutions of different immersion ides or oxides on the surface, which is consistent with variation
time. With respect to two aging conditions of alloy, the value of of the Rct value in different solutions. With the immersion time
Rct tends to decrease with ongoing immersion time in S2 solution. increasing, the value of Rct in S2 solution shifts to small, indicating
Meanwhile, the value of Rct goes up and down as a function of time less protection of the corrosion product layer. However, the data
in S1 and S3 solutions. In view of complicated EEC models make obtained from EIS measurements merely reflect average informa-
the components of EEC favorable for interacting with each other tion on entire testing interface, that the alloy after AAT presents
during fitting experimental results. Moreover, in the case of the superior corrosion resistance to that after NAT on a whole.
analysis results of SEM/EDS, component element of corrosion prod- In summary, the better corrosion behavior of Al-Zn-Mg alloy is
ucts on the surface in S2 solution contains sulfur, implying sulfur gained with AAT than that with NAT. This good electrochemical
involved in the reactions, which are varied from corrosion products character is may attributed to more refinement and homogeniza-
formed in S1 and S3 solutions. On the grounds of literature review tion of grain giving rise to the stronger protective layer of corrosion
[48], in the case of the existence of defect, these areas are prone products, which effectively restrains attack from aggressive ions
to be attacked in the presence of chloride ion, thereby, this causes diffused from bulk solution to electrode interface. Although many
high reactivity of localized oxide film to break, and fresh Al matrix previous works have reported that the precipitation of ’ or phase
is exposed to the electrolyte solution. In neutral aerated solution, in grain boundaries would increase susceptibility of alloy to severe
most anodic reaction presents oxidation of matrix from Al to Al3+ , environment, the result of these work reflects that the corrosion
while dominant cathodic reaction contains reduction of dissolved behavior of alloy is associated with various factors, which would
oxygen to form OH− , and then Al3+ combines with OH− to produce be overwhelmed if others act as domination. Moreover, the corro-
aluminum hydroxides or oxides layer. Nevertheless, Al favors to sion resistance is higher during 4 h especially in S2 solution than
react with bisulfite ion (HSO3− ) to produce aluminum sulphates that in S1 and S3 solution, and these phenomena are reverse after
easily caused by the sulfite and its analogues in solution, which a long period of immersion. Additionally, in order to simulate nat-
is mainly composed of Al2 (SO4 )3 ·16H2 O and Al3 (SO4 )2 (OH)5 ·9H2 O ural condition, dissolved oxygen did not deaerate before tests. As
[49]. Compared with aluminum hydroxides or oxides, aluminum a result, bubbles absorbing on the surface have a great effect on
sulphates are more stable, due to the latter characterized by amor- EIS data. However, based on the analysis of SEM and EIS, it can be
phous forms. However, aluminum hydroxides or oxides are prone found that Al-Zn-Mg alloy is inclined to induce localized pitting in
to further transform. Besides, double hydrolysis reaction is carried S1 solution, and serious exfoliation corrosion occurs in S2 and S3
out in S3 solution, and numerous of bubbles are posed in S1 and S3 solution especially with NAT.
P. Liu et al. / Journal of Materials Science & Technology 64 (2021) 85–98 97
Acknowledgements
References
[33] S.P. Knight, K. Pohl, N.J.H. Holroyd, N. Birbilis, P.A. Rometsch, B.C. Muddle, R. [42] S.S. Wang, I.W. Huang, L. Yang, J.T. Jiang, J.F. Chen, S.L. Dai, G.S. Frankel, L.
Goswami, S.P. Lynch, Corros. Sci. 98 (2015) 50–62. Zhen, J. Electrochem. Soc. 162 (2015) 150–160.
[34] N. Birbilis, R.G. Buchheit, J. Electrochem. Soc. 152 (2005) 140–151. [43] S. Li, H. Dong, P. Li, S. Chen, Corros. Sci. 131 (2018) 278–289.
[35] A. Pardo, M.C. Merino, A.E. Coy, F. Viejo, R. Arrabal, S. Feliú, Electrochim. Acta [44] G. Srikant, Corros. Sci. 62 (2012) 90–94.
53 (2008) 7890–7902. [45] T.C. Tsai, T.H. Chuang, Mater. Sci. Eng. A 225 (1997) 135–144.
[36] N. Pebere, C. Riera, F. Dabosi, Electrochim. Acta 35 (1990) 555–561. [46] S. Lee, H.S. White, J. Electrochem. Soc. 151 (2004) 479–483.
[37] G. Song, A. Atrens, D.S. John, X. Wu, J. Nairn, Corros. Sci. 39 (1997) 1981–2004. [47] B.L. Ou, J.G. Yang, M.Y. Wei, Metall. Mater. Trans. A 38 (2007)
[38] P. Deepa, R. Padmalatha,Arab. J. Chem. 10 (2017), S2234–S2244. 1760–1773.
[39] L. Guo, J. Wang, W. Hu, D. Zhou,Surf. Rev. Lett. 26 (2019), 1850203. [48] M. Cao, L. Liu, Z. Yu, L. Fan, Y. Li, F. Wang, J. Mater. Sci. Technol. 35 (2019)
[40] C. Cao, Corros. Sci. 38 (1996) 2073–2082. 651–659.
[41] S.M. Lee, S.I. Pyun, J. Appl. Electrochem. 22 (1992) 151–155. [49] S. Oesch, M. Faller, Corros. Sci. 39 (1997) 150.