Influence of the Dissolution of Al- and Zn-Based Galvanic Anodes on the Composition of Calcareous Deposits

Abstract

The possible incorporation of Al and Zn issuing from galvanic anodes in the calcareous deposit forming on carbon steel surfaces subjected to cathodic protection was studied via three methodological approaches. The calcareous deposits were analyzed by X-ray diffraction for phase composition and X-ray fluorescence spectroscopy for chemical composition. First, a calcareous deposit formed on the steel pile of a seaport installation, sampled far (2 m) from the closest galvanic anode, was found to incorporate a small amount of the pollutants present in the seawater (Zn, Ti, Cu). An in situ experiment performed at another seaport focused on the calcareous deposit formed on steel surfaces close to the anode. In this case, a small amount of Zn directly issuing from the anode was incorporated in the deposit. This amount remained low as it corresponded to Zn(II) species adsorbed on the surface of aragonite crystals. Finally, laboratory experiments were performed with Zn(II) and/or Al(III) chlorides (10⁻³ mol L⁻¹ each) added to seawater. With both Zn(II) and Al(III), a Zn(II)-Al(III) hydroxychloride precipitated in the bulk seawater. With only Al(III), and under a higher cathodic current density, Al(III) could be incorporated in a deposit mainly composed of brucite, but only in small amount.

1. Introduction

Cathodic protection (CP), often associated with an anticorrosion organic coating, is widely used to protect steel structures immersed in marine environments. Because of the high conductivity of seawater, about 5 S m⁻¹ at low depth [1], CP can easily be applied using galvanic anodes, a concept designed and tested as early as 1824 [2]. Basically, the galvanic coupling between a metal or alloy less noble than carbon steel drastically decreases the corrosion rate of steel but, simultaneously, increases that of the less noble metal. Galvanic anodes are then often called “sacrificial” anodes as they are corroding instead of the steel structure, i.e., ship hull, off-shore platform, marine renewable energy device, or seaport installation, to be protected.

The composition of recommended galvanic anodes is given in the relevant ISO standard [3]. The alloys mainly used for these anodes are Zn-based alloys and Al-based alloys. The widely used Al-Zn-In galvanic anodes, mainly composed of Al, contain 3 wt.% to 6 wt.% Zn and about 0.02 wt.% In. Zinc is included in the OSPAR environmental monitoring program for riverine inputs and direct discharges [4]. In contrast, to our best knowledge, there are no guidelines and no regulatory considerations for Al concentration in seawater.

However, the increasing number of galvanic anodes, for both seaport structures and offshore wind farms, has raised questions about the environmental impact of galvanic anodes and driven various studies since about 2010 [5,6,7,8,9,10,11,12,13,14,15,16,17,18]. The obtained results highlight the influence of the solubility of Zn and Al corrosion products, which governs their concentration in the soluble fraction of seawater, their association with suspended particles or their incorporation in the seabed sediments. For instance, the higher solubility of Zn(II) hydroxide with respect to Al(III) hydroxide explains why Zn has a higher impact on bivalves such as oysters, even if it is a minor component in Al-based galvanic anodes [14,18]. Experiments conducted in situ, in the perimeter of seaports using galvanic anodes for the CP of their installations, did not reveal any significant impact of such anodes. A study was for instance devoted to the possible impact of Al-Zn-In galvanic anodes on the health status of the black scallop Mimachlamys varia in two seaports (the commercial port and the marina of La Rochelle, Atlantic coast, France) [15]. The conclusion was that port activities, as well as meteorological conditions, significantly influenced the biomarker results and masked the potential effects of the anodes.

Some studies were focused on the marine sedimentary compartment where particles of solid corrosion products coming from galvanic anodes could accumulate [10,11,17]. In the seaport of Calais (English Channel, France), a minor Al enrichment accompanied by a significant Zn enrichment could be evidenced in sediments located near an Al-Zn-In anode, but only under stagnant conditions in a somehow confined area of the seaport [11]. Finally, it was also proposed that Al(III) species could precipitate within the calcareous deposit [10].

Indeed, the steel surfaces protected by CP are rapidly covered with a mineral layer, the so-called calcareous deposit (e.g., [19,20,21,22,23]). This phenomenon is due to the increase in pH at the steel/seawater interface, an alkalinization induced by CP that accelerates the cathodic reaction rate. Both possible cathodic reactions, namely dissolved O₂ reduction and H₂O reduction, produce OH⁻ ions as follows:

O₂ + 2H₂O + 4e⁻ → 4OH⁻

(1)

and:

2H₂O + 2e⁻ → H₂ + 2OH⁻

(2)

The resulting increase in pH changes the inorganic carbonic equilibrium at the steel/seawater interface:

OH⁻ + HCO₃⁻ → H₂O + CO₃²⁻

(3)

Finally, this process leads to the precipitation of calcium carbonate, mainly under the form of aragonite [22,23], on the steel surface:

Ca²⁺ + CO₃²⁻ → CaCO_3(s)

(4)

Brucite Mg(OH)₂ is also generally observed, though it remains a minor component when the potential of the protected steel remains in the range of potentials associated with the used of Zn or Al-based galvanic anodes [22,23,24,25]. In this case, the Mg²⁺ ions present in seawater react directly with the OH⁻ ions produced at the steel surface:

Mg²⁺ + 2OH⁻ → Mg(OH)_2(s)

(5)

The proportion of Mg(OH)₂, however, increases with a decrease in temperature [21] or a decrease in applied cathodic potential [23,24,26]. Though the studies dealing with the chemical composition of calcareous deposits have mainly focused on the Mg/Ca ratio (e.g., [27]), that is the brucite-to-aragonite ratio, some results indicate that other metallic elements present in seawater could be trapped in the calcareous deposit, such as Na [19,21,28], Si [19,28], Sr [19], K and Al [21]. In addition, a few studies focused on the possibility of trapping toxic metals, via the cathodic polarization of a steel surface, in the resulting calcareous deposit [29,30]. The feasibility of such a process was first assessed by adding Ni²⁺ ions to seawater [29], then Pb²⁺ [30].

From an environmental point of view, the impact of a metallic element is linked to its bioavailability, i.e., the capacity of this element to be transferred into living organisms. Thus, the nature of the solid phases where the elements are trapped is a key point that must be addressed. The current study was then designed to determine if the calcareous deposit could itself incorporate a significant amount of the Zn(II) and Al(III) species produced by galvanic anodes. Three methodological approaches were considered. First, samples of calcareous deposits formed on seaport installations (La Rochelle marina) were scraped from protected steel surface and characterized. Secondly, laboratory experiments were performed with steel electrodes cathodically polarized in seawater after the addition of Al and/or Zn salts. Finally, specific cathodic protection experiments were carried out in situ (L’Estaque harbor, Marseille, Mediterranean Sea, France) so as to characterize the calcareous deposit forming on the steel surface close to a Zn-based anode. In each case, the calcareous deposit was characterized using X-ray diffraction (XRD) and X-ray fluorescence spectroscopy (XRF).

2. Materials and Methods

2.1. Materials

2.1.1. Samples from Seaport Installation

La Rochelle marina mainly consists of piled floating pontoons. One of these piles, located in the inner part of the port, was considered as representative of a confined zone where the accumulation of metal species released from galvanic anodes could have been favored. The calcareous deposit present on the steel surface was scraped from a 5 cm² area located in the lower part of the tidal zone, at a distance of ~2 m from the closest galvanic anode. CP of the steel piles is ensured by Al-Zn-In anodes (alloy A1 as defined in [3], containing 93.6 wt.% min. Al, 2–6 wt.% Zn and 0.01–0.03 wt.% In). Finally, one liter of the seaport water was sampled for analysis by XRF.

2.1.2. Laboratory Experiments

A carbon steel grade commonly used for the manufacture of seaport sheet piles was chosen for laboratory experiments. Its nominal composition is (in wt.%) C ≤ 0.27, Si ≤ 0.6, Mn ≤ 1.7, Al = 0.02, S ≤ 0.055, P ≤ 0.055 and Fe (bal.). Large (10 cm × 9 cm) rectangular plates, 1 cm thick, were used to prepare the working electrodes. A copper wire was welded on the rear of the plate to ensure connection with the potentiostat. The plate was then embedded in epoxy resin so that the weld could not be in contact with seawater and only the front was exposed to seawater (active area = 90 cm²). This surface was abraded with SiC paper (grade 320) to obtain a roughness similar to that obtained with industrial processes such as sandblasting. Two sets of experiments were performed, called E1-AlZn and I2-Al. In both cases, 10 L of natural seawater were collected in the harbor of Angoulins in Charente-Maritime (Atlantic coast, France) to be used for a single experiment. The pH of each seawater sample was checked and measured at 8.0 ± 0.1 at room temperature. For E1-AlZn, AlCl₃∙6H₂O and ZnCl₂ (99% and 98% min. purity, respectively, purchased from Sigma-Aldrich) were added to seawater so that a concentration of 10⁻³ mol L⁻¹ was reached for each salt. For I2-Al, only AlCl₃∙6H₂O was added, with the same concentration. This concentration was chosen to ensure a sufficient amount of Al and Zn in the system during the experiments, so that the possible incorporation of Al and Zn into the calcareous deposit would not lead to a significant decrease in Al and Zn concentrations in the system. This also implies that the relative abundances of Al and Zn were much higher than those of any other metals present in the natural seawater of Angoulins (including the possibly already present Al and Zn). Similarly, the role of the impurities present in the chemicals was considered to be negligible, in particular with respect to seaport experiments where other metal pollutants may have, in the bulk seawater, a concentration similar to that of Zn or Al.

2.1.3. In Situ Experiments in a Seaport

The experimental device used for in situ experiments in L’Estaque seaport is displayed in Figure 1.

A 2 m long and 3 cm in diameter carbon steel rod was used as the protected structure (same steel as for laboratory experiments, see Section 2.1.2). A zinc-based anode (alloy Z3 as defined in [3], containing 99.1 wt.% min. Zn) was fixed at the center of the rod, and the rod itself was covered with an anticorrosion wax-tape (Temcoat 3000^TM primer, paraffin-based Wax-tape 2 wrap and MCO 110^TM outer wrap, Trenton Corporation, Ann Arbor, MI, USA). Two windows were cut in the coating to exposed two small zones (~5 cm² each) of bare metal to the marine environment. The steel rod was set vertically so that one area (zone 1) was located above the anode and the other area (zone 2) below the anode in seawater. The experiment was conducted for 6 months. The potential of the steel rod was controlled at the end of the experiment and was measured at E_CP = −1010 ± 20 mV/_Ag-AgCl−3M.

The calcareous deposits formed in zones 1 and 2 were entirely scraped and carried to the lab for analysis. One liter of the seaport water was also sampled for analysis by XRF.

2.2. Electrochemistry

For laboratory experiments, cathodic polarization was carried out with a VSP potentiostat (BioLogic, Seyssinet-Pariset, France) using an Ag-AgCl-3M reference electrode (E = +0.210 V vs. SHE at 25 °C) and a titanium grid as counter electrode. All experiments were performed at room temperature, without stirring the electrolyte (stagnant conditions). The steel electrode of experiment E1-AlZn was set horizontally, its active area upward, at the bottom of the 15 L glass aquarium used in any case. In contrast, the steel electrode was set vertically for experiment I2-Al.

For experiment E1-AlZn, CP was applied for 53 h at a constant potential mimicking the use of Zn or Al-Zn-In galvanic anodes, i.e., E_CP = −950 mV/_Ag-AgCl-3M [3]. The formation of the calcareous deposit was monitored via the recording of the current density vs. time curve, j(t), and the experiment was actually stopped when j was completely stable, i.e., j(t) reached a plateau.

For experiment I2-Al, CP was applied at a constant strongly cathodic current density j_CP = −100 µA cm⁻², so as to obtain a calcareous deposit mainly composed of brucite [23,26]. The formation of the calcareous deposit was monitored via the recording of the potential vs. time curve, E(t). The potential did not stabilize as it oscillated with a time period of 24 h. This phenomenon was attributed to the differences in temperature between day (max. 23 °C) and night (min. 16 °C). The experiment was stopped after 76 h, i.e., when the two last oscillations of E were similar.

2.3. Characterization of the Calcareous Deposits

All the calcareous deposits were characterized by XRD and XRF. The whole layer was scraped from the metal and ground to powder for analysis.

For the calcareous deposits sampled in La Rochelle Marina and those obtained during laboratory experiments, the XRD analysis was performed with an ARL-INEL EQUINOX 6000 diffractometer (Thermo Fisher Scientific, Waltham, MA, USA) equipped with the CPS-590 curved detector. This detector allows the simultaneous detection of the diffracted photons on a 2θ range of 90°. The acquisition of each XRD pattern was made with a constant angle of incidence (7°) during 35 min using Co-Kα radiation (λ = 0.17903 nm). For the calcareous deposits obtained at the end of the experiments carried out in L’Estaque seaport, the XRD analysis was performed with a D8 Advance diffractometer (Bruker AXS GmbH, Karlsruhe, Germany), using Cu-Kα radiation (λ = 0.15418 nm) in θ–2θ mode (Bragg–Brentano geometry), from 2θ = 6.2° to 2θ = 65.2°. The step size was set to 0.013° and the overall accumulation time was 30 min. The various detected phases were identified using the ICDD (International Center for Diffraction Data) PDF-4 database (ICDD, Newtown Square, PA, USA).

XRF analysis was carried out with an Epsilon 3^X spectrometer (Malvern Panalytical, Malvern, UK) equipped with an Ag X-ray source and a SDD HR detector. The samples (a few grams of grounded powder or 10 mL of liquid) were placed in P1 plastic cells under which was fixed a 6.0 μm polypropylene thin-film (Chemplex Industries Inc., Palm City, FL, USA). The elemental composition was determined using the Omnian extension of the Epsilon 3X software. Elements lighter than Ar were analyzed under helium atmosphere (0.8 bar), elements heavier than Cl were analyzed under air. Note that with this method, elements lighter than Na, such as C and O, cannot be measured. In addition, seawater was acidified with 2% nitric acid in order to maintain all elements as dissolved species. The given XRF data are average values obtained from two similar samples. The accuracy on the proportions of the chemical elements proved to depend on both the sample and the element. However, in any case, the error interval of the relative abundance (R.A.) was, at worst, ±5%.

2.4. Chemical Modelling

Chemical modelling was carried out to study the equilibrium conditions of brucite Mg(OH)₂ and amorphous Al(OH)₃ in seawater at various pH values. This modelling was achieved with the PHREEQC Interactive software [31] (version 3.5, 2019) using the PHREEQC Minteq V4 database, derived from MINTEQ A2 version 4 [32,33]. A simplified seawater composition was used for computations and only the main seawater elements were considered. The concentrations, based on the ASTM D1141 standard [34], were as follows (in mol kg⁻¹): [Cl⁻] = 0.546, [Na⁺] = 0.468, [Mg²⁺] = 0.053, [SO₄²⁻] = 0.02, [Ca²⁺] = 0.010, [K⁺] = 0.010 and [HCO₃⁻] = 0.0023. The temperature was set at 25 °C.

3. Results

3.1. Analysis of the Calcareous Deposits Sampled from Seaport Installation

The XRD pattern of the calcareous deposit sampled from the surface of a carbon steel pile in La Rochelle marina is displayed in Figure 2.

Three phases were identified. The main one was aragonite, in agreement with previous studies (e.g., [19,20,21,22,23]). Two other minor phases were seen, namely brucite and magnetite. Brucite is the usual other compound associated with aragonite in the calcareous deposit. Moreover, the formation of brucite is favored in the tidal zone [35], where the deposit analyzed here was sampled. The third compound identified was magnetite (Fe₃O₄), which is a common corrosion product of iron and steel in seawater [36]. Because cathodic protection strongly decreases the corrosion rate, the resulting corrosion product layer remains very thin, i.e., much thinner than the overlying calcareous deposit [35,37]. In the tidal zone, the formation of magnetite is favored due to the cyclic immersion [35]. Consistently, magnetite is the main component of the corrosion product layer forming on the metal surface under the calcareous deposit [35]. Finally, some calcite (another form of CaCO₃) was also detected. It may come from the small shells, mainly barnacles, that were present on the steel pile surface.

The sample considered in the present study was then fully representative of a calcareous deposit formed on a carbon steel surface in the tidal zone.

The chemical composition of the calcareous deposit was determined using XRF. The results are listed in

Table 1. Composition of calcareous deposit formed on steel piles in tidal zone of La Rochelle marina, including only metallic elements with R.A. > 0.01 wt.%, Na and K excluded. Sum of proportions for these eight elements (wt.%) was set at 100%. Information about Si and Al was added to facilitate comparison with data given in other tables. R.A. ratios are compared to those obtained via analysis of water of the seaport.

Element	R.A. (wt.%)	R.A. Ratios	Calcareous Deposit	Seawater
Ca	74.79	Ca/Sr	37.8	50
Mg	11.16	Mg/Ca	0.15	3.14
Si	0	-	-	-
Fe	11.82	Fe/Mn	83	-
Sr	1.98	Sr/Zn	86	0.32
Al	0	Al/Zn	0	0.56
Ti	0.064	-	-	-
Zn	0.023	Zn/Cu	1.1	0.93
Mn	0.142	-	-	-
Cu	0.021	-	-	-

. In agreement with XRD, the main identified element was Ca, which came from aragonite.

Other important elements were Fe, present in magnetite, and Mg, present in brucite. Five other elements (Sr, Mn, Ti, Zn and Cu) were detected in low amount. Strontium is present in seawater and is chemically similar to Ca. It is more likely present in aragonite, with Sr atoms substituted for Ca atoms in the aragonite crystal structure. Actually, marine aragonites always contain Sr with an average proportion of 1 mol % SrCO₃ [38]. Mn more likely originates from steel as it is the second most abundant (in wt.%) element in the metal. Ti, Zn and Cu may come from the seawater of the marina, as pollutants that can have various origins.

Various relative abundance (R.A.) ratios are given in Table 1. They will be discussed later, and used for comparison with the results of the in situ experiment carried out in L’Estaque seaport. The Mg:Ca ratio relates, for instance, to the Mg(OH)₂:CaCO₃ ratio in the calcareous deposit. The Zn:Cu ratio may indicate if Zn present in the deposit comes specifically from a neighboring galvanic anode or from the bulk seaport water. Actually, the Zn:Cu ratio was close to 1 in both calcareous deposit and surrounding seawater, and thus reflects, in this case, the composition of the environment.

3.2. Laboratory Experiments

The electrochemical monitoring of both experiments E1-AlZn and I2-Al is displayed in Figure 3. For experiment E1-AlZn, the addition of Al and Zn chlorides to seawater resulted in the precipitation of a white solid compound and part of the particles settled down on the horizontal surface of the steel electrode. However, the cathodic current was initially high (in absolute value), i.e., the settled particles did not hinder the cathodic reaction (Figure 3a). Then, |j| decreased rapidly over time as soon as CP was applied. This phenomenon is commonly attributed to the formation of the calcareous deposit, which decreases drastically the steel surface available for O₂ (and H₂O at lower cathodic potentials) reduction (e.g., [20,22,23]). The curve reached a plateau after about 10 h, and |j| remained between 4.5 and 7.5 µA cm⁻² afterwards.

For the I2-Al experiment, the addition of Al(III) chloride also led to the precipitation of a white solid phase but the electrode surface was vertical and could not be covered with these particles. The application of CP induced a rapid decrease in potential, but the subsequent evolution of the E(t) curve was perturbed by oscillations attributed to the changes in temperature linked to the day/night cycling (time period of 24 h). However, it can be seen that the decrease of the potential, associated with the formation of the calcareous deposit, was not significant after 50 h, the two last minimal values, reached at 51 h and 75 h being both equal to −1.29 V/_Ag-AgCl-3M. Such a low E value should lead to a calcareous deposit mainly composed of brucite [23,26].

At the end of the experiment, the whole surface of the electrode was covered with a white homogeneous layer (Figure 4).

The layers formed on the steel electrodes during both E1-AlZn and I2-Al experiments were scraped from the metal surface for analysis and the corresponding XRD patterns are shown in Figure 5. In the case of experiment E1-AlZn, the surface of the coupon was not rinsed so that the precipitate that settled on this surface was not removed and could be identified. As a consequence, the main diffraction peaks observed on the corresponding XRD pattern (Figure 5a) are those of halite, i.e., NaCl, which crystallized when water evaporated. Other moderately intense sharp peaks were those of aragonite, i.e., the main component of the calcareous deposit at the applied potential of E_CP = −950 mV/_Ag-AgCl-3M and at a temperature about 20–25 °C [21,23]. Finally, a third compound was seen. It corresponded to the solid phase that precipitated when Al and Zn chlorides were added to seawater. The XRD pattern of this compound was very close to that of a Zn(II)-Al(III) layered double hydroxychloride with chemical formula Zn₂Al(OH)₆Cl∙1.8H₂O, according to ICDD file 00-057-0045. A slight shift of the observed diffraction peaks toward smaller angles however indicated a slightly higher c parameter of the hexagonal cell, i.e., c = 23.66 ± 0.04 Å instead of c = 23.22 Å for Zn₂Al(OH)₆Cl∙1.8H₂O. Note that pyroaurite, the Mg(II)-Fe(III) layered double hydroxycarbonate, was reported as a minor component of the calcareous deposit [35]. Pyroaurite has a crystal structure similar to that of Zn₂Al(OH)₆Cl∙1.8H₂O, with c = 23.41 Å (ICDD file 01-086-0181), and thus a similar XRD pattern. Consequently, its presence here cannot be discarded.

For experiment I2-Al, the electrode, set vertically, was not covered with any fluffy precipitate (Figure 4) and its surface was gently rinsed. However, the drying of the deposit did induce the formation of halite (NaCl), as revealed by the corresponding XRD pattern displayed in Figure 5b. The main observed diffraction peaks are however those of brucite Mg(OH)₂, in agreement with the low cathodic potential reached by the electrode (Figure 3). It corresponds to conditions favorable for brucite formation [23], and was actually ensured by the considered cathodic current density j_CP = −100 µA cm⁻². Finally, aragonite was also formed as a minor component of the generated calcareous deposit.

The results of the chemical analysis carried out with XRF are listed in

Table 2. Composition of layers formed on steel during laboratory experiments. Sum of proportions for the five considered elements (wt.%) was set at 100% in each case to facilitate comparison with data given in other tables.

Elements	Experiment E1-AlZn (ECP = −950 mV/AgAgCl-3M)	Experiment I2-Al (jCP = −100 µA cm−2)
Ca	29.04	21.58
Mg	24.19	76.64
Fe	16.13	0.23
Al	10.32	1.55
Zn	20.32	-

. First, these results confirmed that the calcareous deposit contained mainly Ca, i.e., aragonite, for experiment E1-AlZn, and mainly Mg, i.e., brucite, for experiment I2-Al. However, for experiment E1-AlZn, Ca was associated with significant proportions (relative abundance > 10 wt.%) of Mg, Zn, Al and Fe. Zn and Al are combined in the Zn(II)-Al(III) layered double hydroxychloride identified via XRD, which was observed to precipitate when Zn and Al chlorides were added to seawater. Conversely, XRD did not reveal any Mg or Fe compound. As noted previously, Mg and Fe can lead to pyroaurite [35], which has an XRD pattern similar to that of Zn₂Al(OH)₆Cl∙1.8H₂O. It must then be forwarded that both pyroaurite and Zn(II)-Al(III) hydroxychloride did form.

For experiment I2-Al, following Mg was Ca, corresponding to aragonite in agreement with XRD analysis. A very low proportion of Fe was also detected, which may correspond to a very small amount of corrosion products not detected by XRD. Finally, a small proportion of Al was also detected by XRF. This result demonstrates that, if present in the surrounding seawater, some Al(III) species can be incorporated in the calcareous deposit formed at low cathodic potentials, i.e., when the formation of brucite is favored with respect to that of aragonite. The Mg:Ca:Al mass ratio of the calcareous deposit is 49(±2.5):14(±1):1(±0.05).

3.3. In Situ Experiment in L’Estaque Seaport

The XRD patterns of the calcareous deposits scraped from zones 1 and 2 of the coated carbon steel subjected to CP are displayed in Figure 6. Both patterns are similar, which shows that the position of zones 1 and 2 with respect to the galvanic anode had no influence on the phase composition of the calcareous deposit. Calcite (CaCO₃) and magnesian calcite (Ca_1-xMg_xCO₃) were predominating in both cases. These compounds may come from seashells. Quartz and clay (illite) were also detected and came from marine sediments. The third main compound was aragonite, in agreement with the measured applied potential value, i.e., E_CP = −1010 ± 20 mV_/Ag-AgCl-3M. Finally, brucite was present as a minor component of the calcareous deposit.

The chemical analysis carried out with XRF (

Table 3. Composition of layers formed on both zones 1 and 2 of coated steel rod used for L’Estaque seaport experiment (see Figure 1). Sum of relative abundances (R.A.) for the 11 considered elements was set at 100% in each case to facilitate comparison with data given in previous tables. R.A. ratios are compared to those obtained via analysis of water of the seaport.

Elements	R.A. (wt.%), Zone 1	R.A. (wt.%), Zone 2	R.A. Ratios	Zone 1	Zone 2	Seawater
Ca	84.03	88.88	Ca/Sr	63.2	64.9	67.7
Mg	4.29	3.99	Mg/Ca	0.05	0.045	3.14
Si	5.39	3.40	Si/Al	4.4	4.5	-
Fe	3.52	1.44	Fe/Mn	130	120	-
Sr	1.33	1.37	Sr/Zn	24	19	0.5
Al	1.23	0.76	Al/Zn	22	11	0 1
Ti	0.121	0.064	-	-	-	-
Zn	0.056	0.071	Zn/Cu	3.5	5.5	0.99
Mn	0.027	0.012	-	-	-	-
Cu	0.016	0.013	-	-	-	-

¹ Al could not be detected.

) proved consistent with XRD results. First, as for the XRD patterns, the compositions of both calcareous deposits were very similar. Taking into account the determined accuracy of the method (± 5%, see Section 2.3), it can be observed that the variations for the main constituents of the calcareous deposits, i.e., Ca and Mg, fall within the error interval. The uncertainty of ±5% would lead to R.A. = 86 (± 4) % for Ca and R.A. = 4.1 (± 0.2) % for Mg.

Ca predominated (84 to 88 wt.%), in agreement with the observed predominance of CaCO₃ phases by XRD. Sr, detected at about 1.35% in each case, was more likely present in CaCO₃ phases as discussed in Section 3.1. Actually, the Ca/Sr R.A. ratio is the same for both zones 1 and 2, and also equal to that measured in seawater. This indicates that Sr, chemically similar to Ca, behaves similarly in the formation processes of the various CaCO₃ compounds. The Ca/Sr R.A. ratio of the calcareous deposit formed in La Rochelle marina had a smaller value of 37.8, as well as the surrounding seawater (Sr/Zn = 50, Table 1).

Mg, present in brucite and magnesian calcite, represented in both cases about 4 wt.% of the considered elements. Fe was also seen and corresponded to corrosion products of steel, present in a too low amount to be detected by XRD. Mn was detected too, and its relative abundance followed that of Fe, as illustrated by the Fe/Mn R.A. ratio that remained approximately constant. Mn, as Fe, came from the steel substrate.

Si was detected in significant amount too (5.4 and 3.4 wt.%), and could be associated with quartz (SiO₂) and illite (clay, i.e., aluminosilicate). Both compounds were detected by XRD. Al was present in the calcareous deposit, and its relative abundance followed that of Si, the Si/Al R.A. ratio remaining at about 4.5 in both cases. Consequently, Al was mainly present as aluminosilicate here, and mostly came from marine clay sediments.

As for the calcareous deposit from La Rochelle marina installation (Table 1), Zn, Ti and Cu were present inside the calcareous deposit. Zn and Cu could be detected in the sample of seawater collected in the seaport, far from any galvanic anode. The Zn/Cu R.A. ratio in seawater was measured at 0.99 in L’Estaque seaport and 0.93 in La Rochelle Marina (Table 1). A similar value was found for the calcareous deposit of La Rochelle marina (Zn/Cu = 1.1), so that in this case both Zn and Cu may relate to the overall composition of the surrounding seawater. However, in both zones 1 and 2 of the steel rod immersed in L’Estaque seaport, where the calcareous deposit formed on steel areas close to the galvanic anode, the Zn/Cu R.A. ratio was significantly higher, measured at 3.5 in zone 1 and 5.5 in zone 2. It can be reasonably forwarded that this additional amount of Zn came from the neighboring galvanic anode.

The Sr/Zn R.A. ratio, which reflects the Ca/Zn R.A. ratio as the Ca/Sr R.A. ratio is constant, was much higher in the calcareous deposit (about 20) than in seawater (about 0.5). This shows that if Zn can be incorporated in the calcareous deposit, it is incorporated only marginally. Similar results were obtained in La Rochelle marina (Table 1): Sr/Zn = 86 in calcareous deposit and Sr/Zn = 0.32 in seawater.

It can finally be noted that the R.A. of Al is not linked to the R.A. of Zn, as the Al/Zn ratio of zone 1 was twice as much as that of zone 2. This shows that Zn and Al did not come from the same source, which confirms that Al present in the calcareous deposit did not come from the galvanic anode but indeed from marine clay sediments (as indicated by the unchanging Si/Al R.A. ratio).

4. Discussion

For this first study of the possible incorporation of Al and Zn issuing from galvanic anodes in the calcareous deposits, various methodological approaches were considered. First, a calcareous deposit was scraped from the surface of a steel pile in a seaport installation. It was sampled at a 2 m distance from the closest galvanic anode, in the lower part of the tidal zone. XRD only revealed the compounds typical of the calcareous deposit, i.e., aragonite and brucite (e.g., [19,20,21,22,23]), a compound associated with the residual corrosion process taking place even under CP [35,37], i.e., magnetite, and a compound coming from the marine environment itself, i.e., calcite from seashells. Al could not be detected while Zn and Cu were both detected. The Zn/Cu R.A. ratio was however typical of that found in the seawater of the port, i.e., close to 1. Because Cu is only present as traces in alloys used for sacrificial anodes [3], the absence of Zn enrichment in the calcareous deposit, with respect to Cu, indicates that the neighboring galvanic anodes did not influence directly the composition of the deposit. Galvanic anodes should have, however, contributed to the presence of Zn in the seaport water.

The galvanic anodes may influence the composition of the calcareous deposits forming on the protected steel structure, but only at a short distance. The in situ experiment carried out in L’Estaque seaport was designed specifically (Figure 1) to address this question. As for the calcareous deposit sampled from La Rochelle Marina installation, XRD analysis only revealed the components of the calcareous deposit and compounds coming from the marine environment, i.e., calcite and magnesian calcite from seashells, quartz (sand) and illite (clay). Al was present in the calcareous deposit but was found associated with Si and was thus mainly coming from clay. In contrast, a Zn-enrichment of the deposit could be evidenced, via the comparison of the Zn/Cu ratio typical of the deposit (3.5 to 5.5) with that, much smaller, typical of the seawater (0.99). This confirms that, at the immediate vicinity of a Zn-based anode, dissolved Zn(II) species issuing from the anode can be incorporated in the calcareous deposit. However, the fraction of Zn(II) species incorporated in CaCO₃ was kept low, as illustrated by the Sr/Zn ratio that increased from 0.5 in seawater to 19–24 in the deposit.

This last result can be explained by the crystal structure a divalent cation gives to its corresponding carbonate. SrCO₃ (strontianite) is an aragonite-like carbonate and Sr²⁺ cations can easily be incorporated into the aragonite crystal structure [38]. In contrast, cations such as Zn²⁺, Fe²⁺ and Mg²⁺ form calcite-like carbonates and can be incorporated in the crystal structure of calcite [39], as it is the case for Mg in so-called magnesian calcite. Such cations are known to inhibit the growth of calcite and favor the formation of aragonite [40,41,42]. The formation of CaCO₃ begins with the formation of aragonite nuclei following Ostwald’s rule and the transformation of aragonite to calcite is subsequently prevented by the adsorption of the cations on the aragonite crystals [40,41,42]. This adsorption phenomenon may hinder both dissolution of aragonite and growth of calcite (Zn²⁺) or the growth of calcite only (Mg²⁺ and Fe²⁺) [42]. Consequently, when aragonite forms because of CP, Zn²⁺ ions cannot be incorporated into the CaCO₃ crystals and remain adsorbed on the surface of those crystals, which limits the amount of Zn likely to be trapped in the calcareous deposit.

As illustrated by experiment E1-AlZn, Zn(II) species could precipitate at the vicinity of the steel surface as another phase. In the experimental conditions considered here, Zn(II) co-precipitated with Al(III) dissolved species as a layered double hydroxide, i.e., a compound with a crystal structure built on Zn(II)-Al(III) hydroxide layers alternating with so-called interlayers composed of water molecules and anions. The identified compound was Zn₂Al(OH)₆Cl∙1.8H₂O, i.e., the anions in the interlayers were Cl^- ions and this phase was then a double layered hydroxychloride. In experiment E1-AlZn, the particles of Zn₂Al(OH)₆Cl∙1.8H₂O were formed in bulk seawater and settled on the horizontal steel electrode. Such situation is really specific and could only be observed in stagnant conditions and confined areas where such particles could accumulate on, or close to, a protected steel surface. The possible changes over time of the Zn(II)-Al(III) layered double hydroxychloride (dissolution, transformation…) could, however, subsequently favor the incorporation of Zn and/or Al in the calcareous deposit, for instance, as described above, via the adsorption of Zn²⁺ cations on aragonite crystals.

Brucite is another component of the calcareous deposit, though a minor one at the cathodic potentials usually obtained with galvanic anodes (e.g., [19,20,21,22,23]). According to literature data, brucite then forms a thin film under a thicker layer of aragonite [23,43,44]. Some metal cations may be mainly incorporated in the brucite film, as it was observed for Ni²⁺ [30], while others may be trapped in the aragonite layer, such as Pb²⁺ [30]. Actually, Ni trapping was favored at higher current densities, i.e., at lower cathodic potentials, than Pb trapping, because Ni was trapped in the calcareous deposit as Ni(OH)₂ while Pb was trapped in its carbonate form [30]. This showed that, depending on the considered metal, the metallic cations can co-precipitate with brucite in their hydroxide form or with aragonite in their carbonate form. Experiment I2-Al was then carried out with Al³⁺ cations in CP conditions adequate for brucite formation, with a vertical steel electrode to avoid the settlement of the particles formed after Al(III) species precipitated, more likely as Al(OH)₃. The obtained deposit was indeed mainly composed of brucite and contained 1.5 wt.% Al, with 76.5 wt.% Mg and 21.5% Ca. XRD did not reveal any specific Al(III)-compound so that Al(III) may have precipitated inside the brucite layer as amorphous Al(OH)₃.

A chemical modelling was then carried out for a better understanding of the results given by experiment I2-Al. For this modelling, the only phases allowed to precipitate were aragonite (that formed for each considered pH value), brucite and amorphous Al(OH)₃. A synthesis of the obtained results is given in

Table 4. Chemical modelling: dissolved Al(III) concentration (mol kg⁻¹) for increasing pH values, corresponding to the equilibrium conditions with amorphous Al(OH)₃, and list of precipitated phases.

pH	7.85	8.20	8.60	8.87	9.31	9.36
Dissolved Al(III) conc.	1.38 × 10−4	3.13 × 10−4	7.73 × 10−4	1 × 10−3	1 × 10−3	1 × 10−3
Precipitated solid phase	Amorphous Al(OH)3	Amorphous Al(OH)3	Amorphous Al(OH)3	None	None	Brucite

. The considered pH values ranged from 7.85 to 9.36. The lower value is the pH of the solution measured after addition of AlCl₃∙6H₂O (1 × 10⁻³ mol L⁻¹). The highest value corresponds to the critical pH required for brucite precipitation, computed here at 9.36. In a recent work [26], this critical pH value was experimentally measured at 9.61, i.e., the precipitation of brucite on the steel surface may require a slight supersaturation of the solution.

A very clear trend can be observed. When the pH is increased, the dissolved Al(III) concentration increases, from 1.38 × 10⁻⁴ mol kg⁻¹ at pH = 7.85 to 3.13 × 10⁻⁴ mol kg⁻¹ at pH = 8.60. For pH ≥ 8.87, amorphous Al(OH)₃ does not precipitate and Al(III) is only present as dissolved species, in this case mainly (>99%) Al(OH)₄⁻. This means that when the conditions required for brucite formation are established at the steel/seawater interface, they do not allow for the formation of amorphous Al(OH)₃.

This result explains why the proportion of Al(III) trapped in the calcareous deposit of experiment I2-Al remained low. The Al atoms may have been incorporated with Mg(II) in the brucite crystal structure itself, adsorbed on brucite crystals, present as Al(OH)₃ that precipitated farther from the steel surface where the pH was smaller, or may correspond to Al(OH)₃ particles formed in suspension and later incorporated in the deposit.

5. Conclusions

Laboratory experiment E1-AlZn favored the formation and settlement of particles of Zn₂Al(OH)₆Cl∙1.8H₂O on the steel surface and must be considered as a very specific situation, not likely to occur in real industrial conditions.

In other considered experimental conditions, only small amounts (if any) of Al and Zn were incorporated into the calcareous deposit, even if this incorporation was favored by the experimental design, i.e., with the addition of 10⁻³ mol L⁻¹ AlCl₃∙6H₂O in I2-Al laboratory experiment, and with a deposition restricted to two zones close to the galvanic anode in L‘Estaque seaport experiment.

Consequently, the proportion of Al and Zn incorporated into the calcareous deposit must be considered negligible compared with the overall amount of Al(III) and Zn(II) produced by the galvanic anode dissolution. Therefore, the calcareous deposit does not play a significant role (if any) on the fate and on the bioavailability of Al and Zn in the environment.