See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/226897301

EIS testing of new aluminium sacrificial anodes

Article in Journal of Applied Electrochemistry · August 2002

DOI: 10.1023/A:1020547508321

CITATIONS READS

16 255

5 authors, including:

Socorro Valdez Baltasar Mena

Universidad Nacional Autónoma de México Universidad Nacional Autónoma de México
45 PUBLICATIONS 106 CITATIONS 65 PUBLICATIONS 613 CITATIONS

SEE PROFILE SEE PROFILE

Juan Genesca
Universidad Nacional Autónoma de México
161 PUBLICATIONS 973 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

recubrimientos inteligentes para diversas aplicaciones View project

Effect of hydrodynamic conditions on the mechanism and kinetics of some corrosion processes View
project

All content following this page was uploaded by Juan Genesca on 13 August 2015.

The user has requested enhancement of the downloaded file.

 Journal of Applied Electrochemistry 32: 897–903, 2002. 897
Ó 2002 Kluwer Academic Publishers. Printed in the Netherlands.

EIS testing of new aluminium sacrificial anodesq

M.A. TALAVERA1, S. VALDEZ2, J.A. JUAREZ-ISLAS2, B. MENA2 and J. GENESCA1,*

1
Dpto. Ingenierı´a Metalúrgica, Facultad Quı´mica, Universidad Nacional Autónoma Me´xico, UNAM,
Ciudad Universitaria, 04510 Me´xico DF
2
Instituto Investigación Materiales, Universidad Nacional Autónoma Me´xico
(*author for correspondence, e-mail: [email protected])

Received 22 June 2001; accepted in revised form 9 April 2002

Key words: aluminium alloys, anode corrosion efficiency, EIS, sacrificial anodes

Abstract

Aluminium alloys In/Hg free, suitable as sacrificial anodes for cathodic protection, have been developed. Short-term
electrochemical tests were performed on these new alloys to obtain their electrochemical efficiency and to reveal any
tendencies to passivation. Samples were galvanostatically held at different current densities, which changed every
24 h according to DNV RP 401. Electrochemical impedance spectroscopy (EIS), polarization curves and electrode
potential against time measurements were employed to study the main features of the processes taking place at the
alloy–solution interface.

1. Introduction Unalloyed Al adopts a relatively noble solution

potential in saline media as a result of its protective
The most frequently used sacrificial metals for cathodic oxide film. The oxide is the cause of rapid polarization
protection systems are alloys of magnesium, zinc and when Al is placed under corrosion load in a cathodic
aluminium. Although aluminium is relatively new as an protection circuit. Numerous alloying combinations
anode material, its undeniable economic advantages have been made with Al to reduce anodic polarization
over magnesium and zinc have pushed these alloys into traits. With few exceptions, the alloying approaches
an offshore leadership role. Performance characteristics which are used to eliminate passivation, thus promoting
of the Al anodes are related to alloy chemistry and the surface activation, have been largely empirical, with the
application environment. Unfortunately, many alloys of result that several acceptable commercial alloys are now
Al, having been promoted as anode materials, have available. Unfortunately, it is not always clear why these
failed upon continued application. Of the 30–35 Al activating materials are successful or how they work.
anodes that have appeared on the market from 1960 up Nonetheless, the success of the Al anode is dependant
to the present more than half have disappeared from upon the alloying of certain metals whose surface role is
commercial use for technical reasons. ultimately to prevent the formation of a continuous,
Aluminium has considerable merit as the basis for a adherent and protective oxide film on the alloy, thus
galvanic anode principally because of its low density, permitting continued galvanic activity of the Al.
large electrochemical equivalent (high current capacity), Aluminium alloys suitable for cathodic protection
availability and reasonable cost. Its electrochemical have been developed in recent years, and the influence of
equivalent is 1.6 and 1.35 times that of zinc and alloying elements such as Zn, titanium (Ti), mercury
magnesium, respectively. The low electrode potentials (Hg), and indium (In) has been studied by several
of such anodes are readily adapted to a variety of saline researchers [1, 2]. Much of the work done refers to the
environments such as seawater, marine muds and addition of Hg, gallium (Ga), tin (Sn), and In to
brackish waters, though their interesting properties are aluminium alloys. Each of these elements has been
often hindered by the presence of a passive oxide film, demonstrated to improve aluminium activation in neu-
which causes several difficulties regarding their use as tral chloride media; however, the good results obtained
sacrificial anodes. in this field clash with the increased sensitivity to
environmental protection. The use of Hg in particular,
which may be dangerous during the manufacture of the
q
This paper was initially presented at the 5th International
anodes, together with In, pollutes marine life, thus
Symposium on Electrochemical Impedance Spectroscopy at Marilleva, giving rise to environmental concern. In a 500 kg Al-
Trento, Italy, June 2001. anode, 50–100 g correspond to In and 175–200 g
898
corresponds to Hg [3, 4]. Research towards the substi- IIA system, in the a-Al matrix, with the aim of achieving
tution of dangerous elements such as In and Hg involves two ends: (i) a good surface activation of the anode and
scientific knowledge of solidification processes, the (ii) corrosion products similar to those found in the salts
kinetics of precipitation of primary binary/ternary of the sea water, thus preventing the pollution of sea life.
intermetallics, the distribution of intermetallics and/or
precipitates in a-Al solid solution after heat treatment,
and the corrosion mechanisms involved during the 2. Experimental details
dissolution of this kind of anode.
Technical knowledge involves the melting/casting of An Al–Zn (5 at %)–Mg (5 at %)–Li (0.1 at %) alloy was
aluminium alloys with ternary additions of elements prepared with commercially available Al, Zn and Mg
belonging to the groups IA, IIA and IIB of the periodic with purities of 99.98%. Li was used as a wire of 3.2 mm
table, and the implementation of suitable electrochem- in diameter and 99.9% of purity with 4.5 mg cm)3 of
ical techniques for the evaluation of anode current sodium (Na). From previous experiments, and to
efficiency. prevent losses of Mg, Zn and Li during the melting
The resulting Al-anode must have the properties, in process of these alloys, the elements were placed in Al
terms of current efficiency, of commercial Al–Zn–In capsules. Initially, the Al was placed in an alumina/
and/or Hg, must be easy to produce, and must not graphite-coated crucible and melted in a resistance
contain In and/or Hg. furnace under an argon atmosphere. Once the Al was
It has been reported [5] that in the Al–IIB–IIA system, melted, the liquid bath was overheated 150 °C and the
the a-Al solid solution and the s-phase with a compo- Al capsules containing Zn and Mg were added. The bath
sition close to Al2 IIB3 IIA3 are present [6]. In the as-cast was stirred with argon for 10 min to obtain a uniform
condition the main microstructure is formed by the a-Al distribution of Zn and Mg. Immediately after this
solid solution and the second phase s [6]. The eutectic operation, the Al capsule containing Li was added to
formed by a fine dispersion of the a þ s is segregated at the liquid bath, which was stirred with a flux of
grain boundaries and, to some extent, the s-phase is argon for another 5 min, after which time the liquid
partially dispersed in the core of the aluminium grains. alloy was poured into a copper mould of dimensions
To promote further fine dispersion of the s-phase in the 8  8  50 cm. The top and bottom of the ingot were cut
matrix, thermal treatment can be carried out on the off for chemical analysis, thus obtaining a nominal
resulting ingots by taking advantage of the fast kinetic composition of Al-Zn (4.8 at %)–Mg (5.1 at %) and Li
reactions occurring in the solid state. The addition of a (0.09 at %).
combination of IIA and IIB elements in Al is practised A representative microstructure observed in the as-
with the aim of assuring that the aluminium oxide does cast ingot, as shown in Figure 1, consisted of a-Al
not develop protective properties against dissolution. dendrites with sizes between 130 to 150 lm. In the
The amount and distribution of intermetallic com- interdendritic regions, the presence of eutectic and black
pounds is relevant on the electrochemical and corrosion spherical particles was observed. The eutectic showed a
behavior of the alloy. white colour with a maximum width of 10 lm, always
With reference to the Al–IIB–IIA ternary phase following the contour of the dendritic arms. This
diagram [7], in the isotherm at 400 °C (720 °F), the a- eutectic, instead of presenting a platelet morphology
Al solid solution is present plus eutectics of IIA2 Al3 and such as has been reported [5], showed the presence of
Al2 IIB3 IIA3, which ensure, via ageing treatments, a fine rows formed by grey spherical particles.
dispersion of these eutectics which help the non-passi- A short-term electrochemical test [9] was performed
vation of the anode. To make the Al–Zn–Mg ternary on the Al–Zn–Mg–Li alloy anode, corresponding ap-
system more efficient with respect to the superficial proximately to Al–5Zn–5Mg–0.1Li. Cubic anodes were
activation of the anode (preventing the formation of machined from the ingot, thus eliminating the structure
superficial aluminium oxide films) additions of IA are formed by direct contact with the mould wall. The
appropriate [8] with the aim of precipitating Al3 IA ðd0 Þ samples were put in a specially designed electrochemical
at grain boundaries and the matrix, by taking advantage cell, leaving only the cross section area of 1 cm2 exposed
of the fact that Zn lowers the solid solubility of IA in the and connected anodically to a d.c. current source, the
a-Al phase and also that Zn can be incorporated in to cathode of which was a sheet of carbon steel. All the
the d0 phase. specimens were polished with 1200 emery paper, then
This research has as its main objective, the substitu- degreased in acetone, washed thoroughly in double
tion of Al–Zn–In and Al–Zn–Hg alloys used as alumin- distilled water and kept in a dessicator.
ium sacrificial anodes for alloys of the Al–IIB–IIA–IA An accelerated electrochemical test was carried out to
type in order to avoid pollution of the marine environ- determine anode efficiency and working potential [9]. All
ment by elements such as Hg and In, without at the potentials were measured relative to a saturated calomel
same time decreasing the current efficiency of the reference electrode (SCE). Samples were held galvano-
resulting anodes. The choice of the alloy system, has statically at different current densities, Figure 2. Each
as its basis the precipitation and distribution of fine level was applied for a 24 h period. The total, duration
intermetallic phases belonging to the ternary Al–IIB– of the test was 96 hours. During the test, the anode
 899

Fig. 1. Microstructure of the Al–Zn–Mg–Li anode in the as-cast condition.

Fig. 2. Experimental set-up short-term electrochemical test (after DNV RP B401) [9].

potential was determined and electrochemical imped- anode specimen and the corresponding electrochemical
ance spectroscopy (EIS) measurements were made at the equivalent.
beginning and end of each current density level period
(every 24 h). This revealed any tendencies to passiva-
tion. The test solution was aerated substitute ocean 3. Results and discussion
water with an initial pH of 8.3 according to ASTM
D1141 [10]. Electrochemical impedance was conducted The trend of the anode potential (open-circuit potential)
by using a Gamry CMS300 Instrument over a frequency against time when the experimental alloys are immersed
range of 10 kHz to 100 mHz (10 points per decade) in a 3% NaCl is reported in Figure 3. The alloy needs
superimposing a 10 mV a.c. signal and interpreted in time to reach a steady potential. This potential,
terms of equivalent circuit with frequency dependent (1020 mV vs SCE) is very close to the electrode
components. The equipment was controlled with a PC potential reported by Barbucci et al. [5] and Foley and
running EIS software, also from Gamry. EIS measure- Trzaskoma [11], who used an aluminium alloy very
ments were used to monitor the corrosion process on the similar to that used in this study in a chloride containing
aluminium anode daily for the 96 test hours. solution.
At the end of the test, the samples were removed, The calculated electrochemical efficiency [9], e, for the
cleaned and weighed to obtain the weight loss. The as-cast Al anode samples was 61.43%, corresponding to
anode efficiency was obtained by calculating the total 1832 Ah kg)1. The closed-circuit potentials measured
charge passed through the system, the weight loss of the were in the range )1.05 V vs SCE at 0 mA cm)2 to
900

Fig. 3. Anode potential of Al–Zn–Mg–Li alloy versus time of different samples tested. Key: (r) sample A, (j) sample B, (m) sample C, ()
sample D, () sample E and (d) sample F.

)0.86 V vs SCE at 4.0 mA cm)2. An active closed- (lower melting point than aluminium). Under polariza-
circuit potential is desirable because a relatively noble tion, this local composition variation will favour the
potential could indicate the presence of passivation. initiation and propagation of macro- and microlocal
Anodes must also possess high faradaic efficiency to events (galvanic corrosion and pitting for example).
prevent frequent anode replacement. The NACE [12] These events are responsible for lowering anode effi-
and DNV [9] tests specify that an Al anode should have ciency by electrochemical or mechanical mass loss [13].
a closed-circuit potential active to )1.0 V vs SCE and The low efficiencies shown by the studied anodes are
a e between 2300 and 2700 Ah kg)1. Then the anode similar to those obtained by Salinas et al. [13] for Al–
analysed met the potential criterion, but appeared to 5%Zn alloy. A possible explanation is that the main
exhibit only moderate efficiency. Theoretically, uniform efficiency loss can be ascribed to the secondary reac-
anode dissolution will give maximum efficiency. Either tions, since they are produced by the relatively high
secondary cathodic reaction on the same interface and/ content of impurities, particularly Fe and Cu, that cause
or mechanical grain loss due to local macro or micro- local cell action.
corrosion cells will reduce the anode efficiency [13]. In Another possible reason for the low efficiency of the
Al–Zn alloys, Zn tends to be rejected to interdendritic anode alloys tested is the effect of the formation of
zones or grain boundaries. This effect is favoured by the corrosion products on the self-corrosion rate of the
cooling rate and the alloying element characteristics anode. The formation of these corrosion products on

Fig. 4. Corrosion products resistance as a function of current density. Estimated values from corresponding EIS diagrams, at the beginning of
applied current density (r), after 24 h (j) and average (m).
 901

Fig. 5. Typical Nyquist diagrams of Al–Zn–Mg–Li alloy obtained at the different current densities applied.
902
the anode surface significantly affects the electrochem- an inductive semicircle, Figure 5, which may be attrib-
ical efficiency of the material, as pointed out by the uted to pitting corrosion. The preferred dissolution
results shown in Figure 4, in which it can be observed morphology is general attack rather than pitting, since
that the electrochemical behaviour (performance) of the pitting attack has been correlated to less than optimal
Al anode alloy is strongly dependent on the formation performance.
of surface corrosion products. Figure 4 summarizes the Galvanostatic polarization was applied to the anode
resistance of the corrosion product film as a function of after the test. Under these conditions, the actual active
applied current density. It is clear that the increase in electrode area had already reached a steady state value.
current density produces a decrease in the resistance The polarization test was performed at the end of the
value of the corrosion products, Rcp, formed on Al short-term electrochemical test to determine the anode
anode. This is probably due to the fact that a heavier behaviour type. The anodic polarization behaviour
and denser corrosion product forms at a higher current showed that the anode efficiency obtained at low current
density, thus promoting the self-corrosion activity [14]. densities would be the same as that obtained at high
This seems to prove the effect of the formation of current densities. These results consolidate the acceler-
corrosion products on the self-corrosion rate of the ated dissolution test performed.
anode. Several authors [15, 16] have shown a relation Formerly an argument against using the short-term
between the electrochemical efficiency of anode materi- electrochemical test stated that in the two-week [9] or
als and the anodic current densities. The anodic current four-day [12] test, only the outer 260 lm is consumed,
density is therefore a test parameter that should be given whereas in the one-year test at least 0.5 cm penetration
due consideration. On specimens without a well-devel- is achieved. Murray et al. [21] have obtained successful
oped corrosion product, the self-corrosion rate will results by showing that the evaluation of both the as-
account for a relatively higher portion of the total cast material simultaneously with the cut surfaces
mass loss at lower anodic current densities. On partly (which are representative of the bulk material) makes
consumed anodes, where corrosion products have set- the short-term test appear representative of the cast
tled, the maintenance and development of the corrosion anode long-term performance.
products will be more efficient at higher current densi-
ties than at lower current densities. Consequently, the
self-corrosion rate due to a more corrosive environment
4. Conclusions
beneath the corrosion product will be correspondingly
higher at high anodic current density [14].
(i) The presence of a secondary phase strongly affects
EIS measurements were performed during the DNV
the electrochemical behaviour of the Al–Zn–Mg–Li
test, Figure 5. The shapes of the diagrams show signif-
alloys studied.
icant differences. At high and medium frequencies two
(ii) A secondary reaction led to significant loss in cur-
depressed capacitive loops are evident, probably due to
rent efficiency and was dependent on the micro-
the oxide and to the aluminium faradaic dissolution. At
structure of the Al anodes.
low frequencies an inductive loop is detected. These
(iii) EIS is a useful tool to study the dissolution behav-
results are in agreement with those obtained by Barbucci
iour of Al-alloy sacrificial anodes when subjected to
et al. [17] and other authors [18–20], although in their
cathodic protection.
research unalloyed aluminium was studied. The capac-
itive behavior at high and medium frequencies might be
due to the oxide passive film and the metal–oxide
interface (the charge transfer processes, Al dissolution Acknowledgement
through the film oxide), whereas for the inductive
behaviour, an oxide–solution interface (oxygen ion This work was supported by a grant from CONACYT
ingress in the oxide film) could be ascribed to the (Mexico), Project NC-204.
faradaic processes.
It is evident from the electrochemical impedance data,
Figure 5, that the current density condition plays a References
critical role on the resistive property of the corrosion
products, suggesting a significant change on the corro- 1. A.R. Despic, J. Appl. Electrochem. 6 (1976) 499.
2. M. Salleh, PhD thesis, UMIST, Manchester, UK (1978).
sion product property formed on the Al anode. Dark
3. J.B. Clark, Trans. Am. Soc. Met. 53 (1961) 295.
grey corrosion products were formed. The surface 4. G.M. Kuznetsov, A.D. Barsukov, Izv. Akad, Nauk, SSSR Met. 4
dissolution morphology exhibited a rough, pitted sur- (1986) 198.
face. The EIS diagrams showed evidence of both 5. A. Barbucci, G. Cerisola, G. Bruzzone and A. Saccone, Electro-
passivation and pitting. The lower electrochemical chim. Acta 42 (1997) 2369.
6. R.J. Kilmer and G.E. Stonere, ‘Light Weight Alloys for Aerospace
activity exhibited by the Al anode samples strongly
Applications II’, TMS (1991), p. 3.
suggests that secondary reactions may be the funda- 7. D.A. Petrov, in G. Petzow and G.E. Effemberg (Eds), ‘Ternary
mental cause of this behavior. In our experiments, the Alloys’, Vol. 8 (VCH Verlagsgesellschaft, Weinheim 1993), p. 57
EIS diagrams obtained at higher current densities show 8. X. Zhand and Y. Wang, Corros. Sci. Prot. Technol. 7(1) (1995) 53.
 903
9. DNV Recommended Practice RP B401 (1993): ‘Cathodic Protec- 15. C.F. Schreiber and R. Murray, Mater. Perform. 27 (1988)
tion Design’, Det Norske Veritas Industry AS, Hovik (1993). 70.
10. ‘Standard Specification for Substitute Ocean Water’, ASTM 16. S.L. Wolfson, Mater. Perform. 33 (1994) 22.
D1141–75. 17. A. Barbucci, G. Cerisola, P.L. Cabot and G. Bruzzone, Mater. Sci.
11. R.T. Foley and P.P. Trzaskoma, in R.P. Frankental and J. Kruger Forum 289–292 (1998) 529.
(Eds), ‘Passivity of Metals’ The Electrochemical Society, NJ 18. J.B. Bessone, D.R. Salinas, C.E. Mayer, M. Ebert and W.J.
(1978), p. 337. Lorenz, Electrochim. Acta 37 (1992) 2283.
12. NACE Standard Test Method ‘Impressed Current Laboratory 19. J.B. Bessone, C.E. Mayer, K. Juttner and W.J. Lorenz, Electro-
Testing of Aluminium Alloy Anodes’, NACE Standard TM0190- chim. Acta 28 (1983) 171.
98. NACE International, Houston, TX (1998). 20. J.H.W. DeWitt and H.J.W. Lenderink, Electrochim. Acta 41 (1996)
13. D.R. Salinas, S.G. Garcia and J.B. Bessone, J. Appl. Electrochem. 1111.
29 (1999) 1063. 21. J.N. Murray, R.A. Hays and K.E. Lucas, ‘Testing Indium
14. B. Spelid, B. Schei and T. Sydberger, ‘Characterization of Activated, Aluminium Alloys Using NACE TM0190-90 and Long
sacrificial anode materials through laboratory testing’, Corro- Term Exposures’, Corrosion’93, NACE, Houston, TX (1993),
sion’96. NACE, Houston (1996), Paper 551. Paper 534.

View publication stats