Mechanism of Corrosion of Cast Aluminum-Silicon Alloys in Seawater. Part 1: Characterization and Field Testing of Bare Alloys in the Adriatic Sea

Aluminum alloys are used in transportation due to their low weight (density 2.71 g/dm³) and suitable mechanical properties. For marine applications, both wrought and cast alloys are being used. Among wrought alloys, those of series 5xxx (Al-Mg, i.e., AA5083, 5086, 5754) and 6xxx (Al-Mg-Si, e.g., AA6082, 6061, 6063) are suitable for their ease of assembly by welding and excellent corrosion resistance.^1-3  These alloys are used for marine structural applications such as hull structures, superstructures, decks, etc. Nonheat treatable alloys of series 5xxx belong to strain hardening alloys manufactured by a sequence of hot, then cold, forming operations (rolling for sheets) combined with intermediate and final annealing.⁴  By adding Mg (and Mn), cold work and solid solution strengthening contribute to alloys’ strength, high corrosion resistance, and toughness. Alloys of series 6xxx containing Mg and Si are age-hardening alloys that obtain their mechanical properties through solution heat treatment at high temperatures, followed by quenching and natural or artificial aging. During the latter phase, the alloying elements form fine, dispersed precipitates that increase the metal’s hardness. The formation of Mg₂Si provides these alloys with heat-treatability.¹  Alloys of series 6xxx, complementary to the 5xxx series, are easily machinable into various shapes and forms by rolling, extrusion, and forging.^1,4 

Due to their structural integrity, enhanced mechanical properties, and other properties, wrought alloys account for around 63% of total sales of aluminum alloys.⁵  However, the use of cast aluminum alloys has grown in recent years mainly because of their good castability and high specific strength. Compared with wrought Al alloys, casting alloys have clear economic advantages due to their shorter processing cycle and assembly costs.⁶  Therefore, market demands are high, and the need for aluminum alloys with good material properties is increasing. The 40000 (Al-Si) and 50000 (Al-Mg) series according to EN AC⁽¹⁾ are used in marine applications.^1,7  Alloys of series 40000 have excellent potential for casting components of complex shapes, very good weldability, and mechanical properties. They are used for ship superstructures, structural components, interior fitments, assemblies with mechanical functions, etc. The addition of silicon to aluminum reduces melting temperature and improves fluidity.¹ 

According to the Al-Si phase diagram, at room temperature, the solubility of Si in Al is very low.⁸  The eutectic temperature is 577°C, and the eutectic composition is 12.7% silicon. Depending on the Si content, these alloys are regarded as hypo-eutectic (5% to 11%), eutectic (12.6 wt%), and hyper-eutectic (13 wt% to 23 wt%). The hypoeutectic Al-Si casting alloys exhibit suitable castability, weldability, low thermal expansion coefficient, good corrosion resistance, and machinability.^1,9  The most significant disadvantage of Al-Si alloys is the lack of heat-treatment capability.¹⁰  For that reason, Mg is added, which forms the secondary-phase precipitates Mg₂Si and Mg-rich intermetallics.^1,10-11  Cu is often added to increase the strength by forming the intermetallic phase Al₂Cu.¹  Copper significantly decreases the solidus and eutectic temperatures and, therefore, enlarges the solidification interval of an alloy.

During solidification, the microstructure evolves in two stages: primary dendrite Al-phase formation (α-Al-rich matrix) and the subsequent eutectic transformation (eutectic Si particles in α-matrix). The volume fraction of Al-Si eutectic in commonly used hypoeutectic Al-Si alloys, such as Al-Si7-Mg0.3, Al-Si9-Cu3, and Al-Si6-Cu4, can be more than 50%.⁹  Primary alloying elements are not completely soluble in the α-Al matrix and sometimes combine with impurities and matrix to form intermediate (intermetallic) phases,¹¹  among which are Fe-based, i.e., Al(Fe,Mn)Si,^12-13  two most important being α-phase (“Chinese scripts”) and β-phase (needle or plates), and ε-Mg₂Si. In addition, Cu-containing intermetallic particles (IMPs) such as θ-Al₂Cu and Q-Al₅Cu₂Mg₈Si also form.^14-16  All of these IMPs and eutectic-Si are noble to the α-Al matrix except Mg₂Si, which is less noble.

Aluminum alloys used in marine applications are subject to different types of corrosion, the most important being pitting, trans- and intergranular corrosion, crevice, and exfoliation corrosion.^1,4  Alloys containing more than 3 wt% Mg become susceptible to intergranular corrosion (IGC) when exposed to temperatures as low as 50°C over long periods.¹⁷  IGC can lead to mass loss (through grain fall-out) and intergranular stress corrosion cracking in the presence of stress.¹⁸  Intercrystalline corrosion is caused by the difference in electrochemical potential between the actual grain and the grain boundary zone where intermetallic compounds, such as the β-Al₃Mg₂ phase for magnesium alloys, can precipitate.^17-18 

There are several reports on the electrochemical and corrosion behavior of cast Al-Si alloys. Hypoeutectic aluminum-silicon alloys Al-Si5 and Al-Si9 were studied in 0.5 M NaCl; the increased Si content resulted in a dendritic refinement and a more extensive distribution of the eutectic region with a deleterious effect on the corrosion resistance.¹⁹  The addition of eutectic modifier (commercial flux containing mainly sodium fluoride), which promotes the improvement of mechanical properties, harmed the corrosion properties of Al-Si9 that was ascribed to an increase in boundaries between the Al-rich phase and Si particles growth during the melt.²⁰  Si particles disseminated through Al-rich phase should make the alloy more susceptible to corrosion since Si is nobler than Al.²⁰  However, the effects of silicon on the corrosion resistance of Al-Si-based casting alloys are minimal because of low corrosion current density resulting from the fact that the silicon particles are highly polarized.¹  The protectivity of the Al-12%Si alloy in borate buffer (pH 8.4) containing 0.01 M NaCl was partly attributed to its ability to retard the adsorption of chloride ion, as concluded based on double-layer capacitance measurements.²¹  Namely, the difference between the potential of zero charge and corrosion potential was negative (−61 mV), reflecting a smaller susceptibility to pitting. It was concluded that silicon oxide helps blocking entry sites and restricts the transport of chloride ions through the passive film.²¹  The effect of the casting procedure of hypoeutectic Al-Si alloy was studied in different chloride concentrations.²²  No significant difference in casting procedure on the corrosion behavior was noticed. The mechanism of corrosion was shown to proceed in three steps: (i) initiation of microgalvanic corrosion at Al/Fe-rich intermetallic particle interface, (ii) progression of corrosion inside eutectic regions, and (iii) precipitation of corrosion products—the gel (dry mud)-like layer in the center of corrosion process of cathodically active sites and fine powdery corrosion products at the surrounding sites of lower pH.²² 

The corrosion resistance of a series of hypoeutectic alloys was investigated in 0.6 M NaCl.²³  All alloys exhibited good corrosion resistance; those with higher Cu and Si content had more positive corrosion potential. Al-Si7-Mg0.3 exhibited the lowest corrosion current density and more negative corrosion potential than Al-Si9-Cu3 alloy.²³  Similar behavior of hypo- and hypereutectic Al-Si alloys in 0.5 M NaCl was noticed, but the latter exhibited a lower corrosion current density.²⁴  Corrosion behavior of cast Al-Si alloys in NaCl solution was improved by adding rare earth elements.²⁵ 

Several papers reported that the oxide film formed in chloride-containing solution is affected by the presence of alloying elements. It seems, however, that the detailed examination of the composition, structure, and electrochemical properties of hypoeutectic cast Al-Si alloys is still missing. Further, the characterization of these alloys during long-term immersion in seawater will bring some novel data on corrosion mechanisms during field exposure. In this study, we focus on two cast alloys: EN AC⁽¹⁾ 42100 (Al-Si7-Mg0.3) and EN AC 46000 (Al-Si9-Cu3(Fe)). Our previous studies dealt with these alloys in terms of their protection using zirconium conversion coatings.^26-28  For Al-Si7-Mg0.3, we studied its corrosion behavior in artificial seawater in the presence of sodium sulfide.²⁹  The surface-analytical analysis of metals exposed to artificial seawater and bacteria was explored in our previous studies on stainless steel.^30-31  This study investigated the microstructure, composition, and electrochemical corrosion characteristics of bare and coated Al-Si9-Cu3 and Al-Si7-Mg0.3 alloys. The microstructure and composition were investigated using focused ion beam scanning electron microscopy (FIB-SEM) with energy dispersive x-ray spectroscopy (EDS), x-ray photoelectron spectroscopy (XPS), and glow discharge optical emission spectroscopy (GDOES). In Part 1 of the series, we investigated the bare alloys before and after immersion in seawater. In Part 2 of the series, we investigated the coated Al-Si9-Cu3 and Al-Si7-Mg0.3 alloys before and after immersion in seawater.³²  Two types of sol-gel coatings synthesized in our laboratory were applied as coatings. Field immersion testing was performed in the Adriatic Sea. The antifouling performance was qualitatively determined using sonication to remove the biofouling attached to the surface.

Two types of aluminum alloys were manufactured by Talum d.d. Kidričevo, Slovenia, and used as substrates (Table 1): (1) Al-Si9-Cu3 alloy (380.0/EN AC⁽¹⁾ 46000/EN AC Al-Si9-Cu3 (Fe)/UNS⁽²⁾ A0380) and (2) Al-Si7-Mg0.3 alloy (A356.0/EN AC 42100/EN AC Al-Si7-Mg0.3/UNS A13560).

Samples were cut in the form of plates with dimensions 80 mm × 39 mm × 4 mm (Al-Si9-Cu3) and 80 mm × 34 mm × 4 mm (Al-Si7-Mg0.3). On each shorter side of the rectangle, a hole with a diameter of 3 mm was made to fix the sample on the rack using a plastic string. Samples were then water-ground on both sides using SiC emery paper of 320-grit (LabPole-5^†, Struers, Ballerup, Denmark). Twenty-four bare samples were prepared (12 of each alloy type, Figure 1).

For SEM/EDS analysis, the nonimmersed samples were water-ground successively using SiC-paper up to 4000 grit.

Out of each set of 12 bare samples, three were used as benchmarks for the characterization before the immersion in the Adriatic Sea (Figure 1). Electrochemical measurements, XPS, GDOES, and SEM analyses were performed on these samples. Nine bare samples of each alloy type were immersed in the Adriatic Sea, with three replicates for each immersion period (1, 3, and 8 months).

The immersion of the samples was performed in collaboration with the Marine Biology Station (MBS) of the National Institute of Biology in Piran, Slovenia. The MBS infrastructural center operates a research vessel (12 m), a boat (5 m), and a diving base with SCUBA facilities. Oceanographic buoy Vida provides real-time meteorological (air T, humidity, wind) and physical and chemical oceanographic data (seawater T, salinity, currents, waves, dissolved oxygen).

Figure 2(c) presents the image of the Vida buoy and its geographical position approximately 2 nautical miles away from Piran at the Adriatic coast in Slovenia (coordinates 45° 32.925′ N, 13° 33.042′ E) (Figures 2[d] and [e]). The framework carrying the racks with samples was lowered from the boat and positioned at a depth of 20 m (video given in Supplemental Video SV1). The frame with racks was secured at the sea bottom and tied to the buoy (Figure 2[f]). Therefore, the samples were exposed to the natural sea movements at this depth but firmly attached to the frame.

The test duration was 8 months, from July 25, 2017 to March 13, 2018. Samples were taken out on August 25, 2017 (after 1 month), October 26, 2017 (after 3 months), and March 13, 2018 (after 8 months).

Data on sea temperature at the sea bottom and the surface, salinity at the surface, and concentration of oxygen at the sea bottom were provided by MBS based on the data collected at the Vida buoy (Figure S1). From July 2017 to March 2018 (32 weeks), the surface temperature of seawater gradually changed from summer to winter conditions decreasing from ca. 26°C to 27°C to 8°C. The bottom temperature reached the maximum of ca. 23°C in October 2017 and then decreased in the winter, when bottom and surface temperatures were equal. Salinity slightly varied depending on the season and increased from summer to winter from 35 to 38 per mille. These variations are related to the freshening by rivers.³³  The oxygen concentration at the sea bottom varied between 320 μmol/L and 280 μmol/L, reaching a minimum in October 2017.

Once samples were delivered to the laboratory (Figures S2[d] through [f]), they were processed further according to the scheme presented in Figure 1. Two vials filled with Ringer solution were sonicated in the same solution (vortexed 30 s, sonicated at 40 kHz for 10 min). The purpose of sonication was to remove the biofilm (biofouling) and microorganisms attached to the samples’ surface during immersion. Samples were photographed using a high-resolution digital camera before and after sonication to document biofilm removal. After sonication, samples were taken out of sonicates, air-dried, and stored in new polyethene vials for further measurements, i.e., electrochemical measurements, XPS, GDOES, and SEM/EDS analyses. Electrochemical measurements were performed on sonicated samples immersed for 1, 3, and 8 months, XPS and GDOES analyses after 1 month, and SEM analysis after immersion for 1 and 8 months.

The biofilm-coated samples placed in a vial filled with seawater were processed further in the laboratory to prepare the surface for SEM analysis. Samples covered with biofilm and attached microorganisms require chemical fixation to preserve and stabilize their structure.³⁴  The fixation process was performed by immersion for 5 min in a 25 vol% solution of glutaraldehyde (25 vol%, Sigma). This step was followed by dehydration of the fixed biological matter using an organic solvent, i.e., ethanol (EtOH), which replaces water in the cells and preserves their structure. Immersion in progressively increasing EtOH concentration was applied: 5 min in 70% EtOH, 5 min in 95% EtOH, and 5 min in absolute EtOH (Carlo Erba, Milan, Italy). Samples were then air-dried and stored in new polyethene vials until the SEM analyses. SEM analyses were performed on biofilm-coated samples immersed for 1 and 8 months.

A scanning electron microscope (SEM, JEOL JSM 5800^†) and a field-emission scanning electron microscope (FE-SEM, JEOL JSM 7600F^†, Tokyo, Japan) were used to analyze the morphology of the following samples: (i) nonimmersed samples, (ii) samples covered with biofilm, and (iii) sonicated samples after removal of the biofilm. Before SEM analysis, a thin carbon layer was deposited on the samples’ surface. The imaging was performed using a secondary electron detector (SE mode) at 10 kV (JSM 5800) and 0.5 kV (JSM 7600F).

Morphology and composition of nonimmersed and immersed, sonicated samples were analyzed by a FIB-SEM (FEI Helios NanoLab 650^† Dual-beam). SEM imaging was performed using CBS (circular backscattered detector) and ICE (ion charged detector) modes at 15 kV beam voltage. The cross sections of selected regions on the samples were obtained after the deposition of a thin layer of platinum on the surface (the first layer, 0.2 μm thick, was deposited using an electron beam at 2 kV and 0.4 nA, the second layer, 1 μm thick, was deposited using Ga FIB beam at 30 kV and 0.24 nA), followed by cutting the coating using Ga FIB beam at 30 kV and 9.4 nA. The surface was polished with a Ga beam at 30 kV and 0.4 nA in the last step. Imaging along the cross section was performed by SEM using SE mode at 15 kV. The top surface and cross-section compositions were determined by EDS using an Oxford^† Instruments AZtec system with an X-max SDD 50 mm² (silicon drift detector). The EDS analyses were performed at 15 kV, 5 kV, or 3 kV in a point analysis. The data were normalized to atomic percentages (at%). The amount of carbon was excluded from the quantitative analysis. EDS mapping was recorded at 15 kV.

The CASINO^† program (version v.2.51, University of Sherbrooke, Canada), a single scattering Monte Carlo simulation of electron trajectory in solid, specifically designed for low energy beam interaction was used to generate many recorded signals (x-rays and backscattered electrons) in a SEM.

The surface characterization of elemental composition and chemical state of the elements was performed by XPS using Thermo Electron ESCALAB 250^† spectrometer, with a monochromated Al Kα x-ray source (hν = 1,486.6 eV). The base pressure in the analytical chamber was maintained at 10⁻⁹ mbar. The spectrometer was calibrated using Au 4f_7/2 at 84.1 eV. The take-off angle of analyzed photoelectrons was 90°. The analyzed area was a 500 µm diameter disk. Survey spectra were recorded with a pass energy of 100 eV at a step size of 1 eV and high-resolution spectra with a pass energy of 20 eV at 0.1 eV. Curve fitting of the spectra was performed with the Thermo Electron^† software Avantage using iterative Shirley-type background subtraction. All spectra were aligned with the C1s peak for carbon involved in C−C and C−H bonds, located at 285 eV. For some samples, a strong charging effect was observed. The position of all XPS peaks was aligned with the position of the reference C1s peak. The values of the photoionization cross sections (σ_X) at 1,486.6 eV were taken from Scofield,³⁵  the inelastic mean free paths (⁠⁠) were calculated by the TPP2M formula.³⁶  The compositions presented in the tables are deduced from high-energy resolution spectra.

The additional elemental in­depth profiling was performed using GDOES (GD-PROFILER2^†, HORIBA Scientific), identifying the elemental composition rapidly through the coating. Operating conditions were: anode diameter 4 mm, argon pressure 850 Pa, and applied radiofrequency (RF) power 35 W.

Electrochemical measurements were performed using a three-electrode flat cell (KO235 Flat Cell Kit^†, Ametek, Berwyn, PA, USA) with a volume of 250 mL. Specimen embedded in a Teflon holder leaving an area of 1 cm² exposed to solution served as the working electrode. A silver/silver chloride (Ag/AgCl, 0.205 V_SHE [standard hydrogen electrode]) served as a reference electrode, and a platinum mesh as a counterelectrode. Potentials in the text refer to the Ag/AgCl scale. Electrochemical measurements were performed at room temperature in artificial seawater (Burkholder formulation B, pH 7.4). The composition of artificial seawater was given in Table 2.³⁷  The chemicals were supplied by Sigma-Aldrich (MgSO₄ × 7H₂O, NaHCO₃, KCl), Acros Organics (MgCl₃ × 6H₂O, CaCl₂ × 6H₂O, KBr, SrCl₂ × 6H₂O), Fisher Chemicals (NaCl), and Riedel-de-Haën (H₃BO₃). Measurements were conducted using an Autolab potentiostat/galvanostat Model 302N^† (Utrecht, The Netherlands).

Electrochemical measurements were performed on nonimmersed samples and samples immersed in the Adriatic Sea for 1, 3, and 8 months. On each sample, measurements were performed on three sites after biofilm removal by sonication. After 8 months of immersion, it was difficult to remove the biofouling by sonication and, consequently, to find several spots at the surface to perform more repetitions. Therefore, the measurements are taken with caution and are presented with different symbols than those after 1 and 3 months of immersion.

Each sample was subject to four successive measurements from least destructive to more destructive methods, including open-circuit stabilization, linear polarization resistance (LPR) measurements, electrochemical impedance spectra, and potentiodynamic polarization curves. Before switching to the following electrochemical technique, the system was allowed to rest for 5 s. (i) The sample was first allowed to rest at open-circuit potential for approximately 1 h to reach a stable, quasi steady-state open-circuit potential (E_oc) at the end of the rest period. (ii) LPR curves were then recorded in the range from −10 mV to 10 mV vs. E_oc using a 0.1 mV/s potential scan rate. Polarization resistance (R_p) was determined as the slope of the fitted potential (E) vs. current density (j) curve using Nova 2.1^† software. (iii) Electrochemical impedance spectra (EIS) were recorded at E_oc in the frequency (f) range from 100 kHz to 0.01 Hz using an AC potential amplitude of 10 mV (rms). (iv) Potentiodynamic polarization curve was recorded in the potential region starting from −250 mV vs. E_oc in the anodic direction until the current density reached 0.001 A/cm². Corrosion current density (j_corr) and corrosion potential (E_corr) were determined using Nova 2.1 software using extrapolation of the linear portion of polarization curves, i.e., Tafel extrapolation, based on the mixed potential theory.³⁸  j_corr was determined as the intersection of the anodic and cathodic Tafel slopes at E_corr. When it was impossible to assure a sufficient range of linearity of Tafel curves (up to 120 mV), especially in the anodic range, j_corr was determined as the intersection of the linear portion of the cathodic curve and the straight line passing through E_corr.

Quantitative electrochemical parameters are presented in tables as mean values±standard deviations. In graphs, representative measurements (at least two similar repetitions) were chosen to be presented.

EDS mapping of one of the IMP sites is presented in Figure 4. Several typical particles are recognized: Al-Fe-Mn longitudinal particle, Al-Cu-Si particle located along with the Fe, Mn-containing particle, patterned Al-Cu-Si-Mg bright area, and Si-rich dark grains.

The composition of Al-Si7-Mg0.3 is given in Table 1. Compared to Al-Si9-Cu3 alloys, Al-Si7-Mg0.3 contains less Si, Mg, Fe, Mn, and Zn, and significantly less Cu; overall, it contains a smaller amount of alloying elements and thus more Al.

EDS mapping of one of the IMP is presented in Figure 6. Several typical particles are recognized: Al-Si-Fe-Mn particle, so-called Chinese scripts, Al-Si-Mg(Fe,Mn) particles, and Si-rich gray grains.

Copper acts like biocide when added to paints for marine vessels.⁴¹  We did not observe the effect that Cu present in the alloy would mitigate the amount of biofoul, probably because it acted as a cathode in IMP. A more focused study would be required to address the effect of Cu in the alloy on the amount of biofouling.

These layers also differ in composition, as identified by point EDS analysis (Table 6). The layer formed on the α-Al matrix (spectrum 3, Figure 13[c]) was composed mainly of Al-oxide; it contains some Mg (ca. 1 at%) and Si (ca. 2 at%). In contrast, the Al-oxide layer formed in the eutectic region was rich in silicon (spectrum 1, Figure 13[c]), with incorporated Si-rich grains (spectrum 2). The latter contained a double concentration of Si (ca. 38 at%) compared to the surrounding layer (ca. 17 at%). The concentration of Mg remained around 1 at%. Similar ratios were obtained when the EDS analysis was made using 5 kV beam voltage, indicating that the Si signal originates from the surface layer and the alloy bulk (depth analysis was 3.7 μm at 15 kV and 0.9 μm at 5 kV, as explained in Figure S7). Therefore, the layer formed on the α-Al matrix was mainly Al-oxide with minor amounts of Mg and Si, whereas the layer formed in the eutectic region was Al-oxide mixed with Si. Compared to the oxide formed during immersion on the Al-Si9-Cu3 alloy, this layer contained only a small amount of Mg, around 1 at%.

The composition of the oxide at the surface of the nonimmersed and immersed Al-Si9-Cu3 sample is presented in Table 7. The composition was calculated with and without oxygen content to look closer at the cationic fraction (ratio of metal species in the oxide layer). High-resolution spectra of the main elements Al, O, Si, and Mg are presented in Figure 15; survey and high-resolution spectra of minor elements are presented in Figures S8 and S9, respectively.

The surface of the nonimmersed sample is covered by a naturally formed oxide layer containing mainly Al oxide (Al/Si = 10.5 and Al/Mg = 26.7). Two peaks appeared in the Al2p spectrum, at 75.7 eV and 72.8 eV reflecting the presence of Al(III) oxide/hydroxide on top of Al metal (Figure 15). An oxide layer thickness of 2.8 nm is calculated from the ratio of Al metal and Al(III) peaks. The O1s peak is centered at 533.1 eV. Si is the most intense among alloying elements, showing Si2p peaks at 99.4 eV and 103.6 eV, respectively, aligned with the presence of Si metal and Si(IV) oxide.⁴²  Mg2p peak is located at 51.8 eV, which may be related to the presence of Mg(II) in MgO, Mg(OH)₂, and/or Mg-Al-Si-oxides.⁴²  Cu2p and Zn2p peaks are much less intense (Figure S9). In the Cu2p core level, the peak at 933.8 eV can be attributed to Cu metal in IMPs or Cu(I).⁴³  The Zn peak at 1,023.7 eV can be attributed to metallic or oxidized Zn.⁴³ 

After immersion, a single peak at 75.7 eV was present in the Al2p spectrum; similarly, a single peak at 103.6 eV was present in Si2p spectrum (Figure 15). The peaks related to Al and Si metals were no longer visible due to the coverage of the surface with the oxide layer formed upon immersion in seawater. The Mg2p peak was strongly increased, reflecting Mg oxide or mixed Mg-Al oxide formation. The related ratios are Al/Si = 2.4 and Al/Mg = 2, showing a strong reduction compared to the nonimmersed sample (Table 7). Zn and Cu are either dissolved or buried below the surface layer, so these elements were not detected. Nitrogen and chlorine peaks appeared (Figure S9). Nitrogen N1s peak at 400 eV is presumably related to the organic matter of biofouling remaining after sonication. C1s peak is broader than a nonimmersed sample, indicating additional carbon species, presumably related to residual biological matter.

To summarize, after 1 month of immersion, the Al signal is much less intense, but Si and Mg are enriched at the surface compared to the spectra before immersion (Table 7, Figure 15). Seven-fold enrichment of the oxide layer in Mg and almost three-fold enrichment in Si was observed. The oxygen signal remained high. The intensity of Cl2p is very small, indicating that the incorporation of chloride from seawater in the solid surface layer remaining after sonication is negligible. Therefore, XPS results confirm the SEM/EDS results (Figures 10 and 11[a]) concerning the enrichment of the oxide layer in Si and especially Mg.

After immersion in seawater, the GDOES profile is strongly modified. Particularly, O is present in the early stages of the profile (0 s to 40 s). Combined with the decrease or absence of characteristic major elements of the alloy (Al, Cu, Si), it can be concluded that a thin oxide is present at the surface of the alloy. During this time, the Al signal presents a plateau (not visible for Cu and Si), suggesting the formation of mainly aluminum oxide. The stabilization of all signals (after a sputtering time of 60 s) at values similar to those obtained before immersion indicates that the bulk alloy has been reached.

Figure 16(b) shows recorded signals of Ca, Mg, Na, O, S, and Cl during the GDOES profiling of the immersed alloy. Two distinct domains can be observed during this chemical profiling with the presence of an oxide layer in the first seconds of sputtering (0 s to 25 s) followed by an interfacial domain enriched in Ca, Na, and S. It is interesting to note the presence of chloride, probably at low levels, in the oxide film, suggesting that the interfacial layer blocks the penetration the chlorides. The evolution of the Mg signal shows enrichment of Mg in the oxide layer, confirming the results of EDS and XPS analyses. Magnesium follows the same trend as chlorides, but it is difficult to define whether magnesium comes from bulk or seawater.

The surface of the nonimmersed sample is covered by a naturally formed oxide layer containing mainly Al oxide (Al/Si = 14 and Al/Mg = 21.8). Components associated with Al metal, Al(III) oxide, Si metal, Si(IV) oxide, and Mg(II) oxide are observed at the same binding energies as on the Al-Si9-Cu3 sample (Figure 15). The thickness of 3.4 nm of the oxide layer is calculated from Al metal and Al(III) peaks.

After immersion, the Al signal is much less intense and exhibits only a single peak related to Al(III) oxide. Similarly, a single Si(IV) oxide is present. Compared to the layer formed on Al-Si9-Cu3, the most considerable change occurs for Mg. Namely, Mg was highly enriched in the layer on Al-Si9-Cu3, but only two-fold enrichment was observed for Al-Si7-Mg0.3 (Table 7, Figure 17). The ratio of Al/Si after immersion is 5.8, and that of Al/Mg is 9.5.

After 130 s, signals of the alloy elements (Al, Si, and Mg) slowly increased but never reached the bulk levels. This observation may suggest the presence of a strongly roughened interface that the corrosion process in seawater may generate.

As for the Al-Si-Cu alloy, Figure 18(b) shows the presence of two distinct domains, the first corresponding to the oxide and the second to an interfacial zone enriched in Ca and Na. Chlorides seem to be present mainly in the oxide layer, unlike magnesium, which is slightly present in the oxide and shows no peak with maximum as in the case of Al-Si9-Cu3. The measured levels of Ca and Na are lower than for Al-Si9-Cu3 alloy (around 9 V for Ca and 0.4 V for Na for Al-Si7-Mg0.3 and 60 V for Ca and 7 V for Na for Al-Si9-Cu3), indicating that this layer is much less enriched. It seems possible to correlate the enrichment of this layer with the thickness of the oxide formed in the seawater.

The curves recorded in artificial seawater for samples immersed in the Adriatic Sea for 1, 3, and 8 months are presented in Figure 19(a). The cathodic current density and j_corr increased with E_corr shifting somewhat positively (Table 8). In the anodic range, however, the increase in current density significantly slowed down, forming a pseudo-passive range, where the current density is not entirely independent of potential, but the increase is considerably diminished. At ca. −0.45 V, the passivity breakdown took place. The curve recorded after 3 and 8 months of immersion were similar to that after 1 month.

The electrochemical response of Al-Si9-Cu3 samples immersed in the Adriatic Sea indicates that the immersion does not significantly deteriorate the surface but rather contributes to its passivation: the electrochemical parameters remained similar to that of nonimmersed alloy, but the narrow passive range was established up to −0.45 V. Therefore, the mixed Al-Mg-Si oxide identified by SEM/EDS, XPS, and GDOES (Figures 10, 15, and 16) is protective. After 8 months of immersion, the j_corr values increased from 2.97 µA/cm² to 9.98 μA/cm² (Table 8). However, compared to the nonimmersed sample, the increase of the anodic branch is significantly slower, confirming that the surface layer formed during immersion is still protective.

The polarization curve recorded for bare Al-Si7-Mg0.3 in artificial seawater is presented in Figure 19(b). This alloy exhibits more negative E_corr, smaller j_corr, and higher R_p than the Al-Si9-Cu3 alloy (Table 8). In the course of immersion in seawater, these two alloys differ significantly. The passive layer formed on Al-Si7-Mg0.3 developed a broad passive region over 3 months of immersion, resulting in a 15-fold decrease in j_corr and a 26-fold increase in R_p. Moreover, a broad passive range of over 700 mV was established. The curve recorded after 8 months of immersion showed the deterioration of electrochemical passivity compared to after 3 months but still much better than that of nonimmersed alloy. The passive range was broad, over 600 mV, indicating that the surface oxide layer still assured a high level of protection of the underlying substrate.

EIS data in Figures 20(c) and (d) corroborated potentiodynamic data. The impedance plot before immersion is similar to that of Al-Si9-Cu3, but with immersion, the impedance increased, showing a broad range of frequencies where the layer with capacitive character is formed. After 3 months of immersion, impedance magnitude at the lowest frequency was higher for Al-Si7-Mg0.3 than for Al-Si9-Cu3. After 8 months, the impedance decreased for the latter as well.

Two cast aluminum-silicon alloys were investigated for marine applications by field immersion testing in the Adriatic Sea. Before and after immersion, the samples were investigated using surface analytical techniques (FIB-SEM/EDS, XPS, and GDOES) to characterize microstructure and composition. Further, electrochemical measurements in artificial seawater were conducted to account for long-term corrosion properties. Biofouling was noted during immersion.

• Al-Si9-Cu3 alloy comprises α-Al-rich matrix, eutectic Si grains, and various Al-Si-Fe(Mn,Cu,Mg) IMPs. Al-Si7-Mg0.3 alloy comprises α-Al-rich matrix, larger eutectic areas, and a smaller number of intermetallics than Al-Si9-Cu3.

• During immersion in the Adriatic Sea, the biofouling was progressively formed, first as micro- and then as macrobiofoul. After 1 month of immersion, biofoul could be almost completely removed by sonication (simulating removal due to vessel in motion). At more prolonged immersion, sonication could not entirely remove the macro biofoul, e.g., diatoms.

• After 1 month of immersion, a micrometer thick oxide layer was formed on the Al-Si9-Cu3 alloy. EDS, XPS, and GDOES analyses confirmed that this layer was rich in Mg, probably incorporated from seawater. The protective oxide layer ensured that the electrochemical parameters did not deteriorate during immersion. A passive range of 110 mV to 140 mV was established.

• On Al-Si7-Mg0.3 alloy, two distinct regions were formed during immersion: Al-oxide was formed on the matrix with small amounts of Mg and Si; on the eutectic region, an Al-oxide rich in Si grains developed. This part contained doubled concentration of Si compared to the underlying alloy. The oxide layer formed on Al-Si7-Mg0.3 alloy during immersion in the Adriatic Sea showed better passivation behavior than Al-Si9-Cu3, which improved up to 3 months of immersion, showing the extent of the passive region of over 700 mV.

• Cast aluminum-silicon alloys are suitable for marine applications, especially alloys containing magnesium (EN AC 42100).

This work is a part of the M-ERA.NET project entitled “Design of corrosion resistant coatings targeted for versatile applications” (acronym COR_ID). The financial support of the project by MESS (Ministry of Education, Science, and Sport of the Republic of Slovenia) and ANR (The French National Research Agency) is acknowledged. This work is also a part of the bilateral Proteus program between Slovenia and France entitled “DCOIN: Disentangling COrrosion and its INhibition” and INCOR: INterfaces relevant for CORrosion and its inhibition”, financed by the Slovenian Research Agency (SRA) and the ANR (Grant No. BI-FR/21-22-008). The financial support from the SRA (research core funding No. P2-0393 and P1-0134) is acknowledged. Région Ile-de-France is acknowledged for partial funding of the XPS equipment. The authors acknowledge the Centre of Excellence in Nanoscience and Nanotechnology—Nanocenter (CENN), Ljubljana, Slovenia, to access the scientific equipment (FIB-SEM/EDS). The authors acknowledge D. Zimerl, G. Šekularac, D. Hamulić, and U. Tiringer for valuable technical help and Prof. A. Cör and Dr. K. Šuster for valuable discussions. The authors also thank co-workers of the Marine Biology Station of the National Institute of Biology in Piran, Slovenia: Dr. A. Ramšak for the expert discussion, Dr. V. Malačič for providing data from the Vida buoy and diver T. Makovec for handling the samples and recording the video material in the Supplemental Material. The company Talum d. d. Tovarna aluminija Kidričevo, Slovenija, is acknowledged for production and providing the cast Al-Si alloys.