On the Corrosion Properties of Aluminum 2024 Laser‐Textured Surfaces with Superhydrophilic and Superhydrophobic Wettability States

In this work, the mechanism of the corrosion behavior of laser‐treated aluminum is studied. Two different laser techniques are used to fabricate the samples, direct laser interference patterning (DLIP) and direct laser writing (DLW), using nanosecond laser sources. The DLIP treatment uses a two‐beam optical configuration producing line‐like periodic structures. The DLW technique is employed to produce non‐periodic structures on the Al‐surface with the same cumulated fluences as in DLIP. The surface topography is analyzed by confocal microscopy, and the formation of oxide layers is investigated by scanning electron microscopy of cross‐sections produced using a focused ion beam. Wetting measurements performed on the laser‐treated samples exhibit a contrasting behavior, leading to either superhydrophobic or superhydrophilic states. In the case of the DLIP treatment, the static water contact angle is increased from 81° up to 158°, while for DLW, it decreases to 3°. Electrochemical tests demonstrate a decreased corrosion rate after laser treatment. Additionally, findings indicate no correlation between wettability and corrosion reduction. Therefore, the improvement in corrosion resistance is mainly attributed to the oxide layer formed by laser treatment. Although similar corrosion rates are achieved for both treatments, surfaces produced with DLIP can be beneficial when additional surface properties are required.


Introduction
In recent years, several studies have been performed regarding the development of methods for modifying the wettability DOI: 10.1002/admi.202300607 of metallic surfaces based on bio-inspired surfaces. [1]Superhydrophobic surfaces are known to repel water, forming a contact angle between the water droplet and the surface greater than 150°.In contrast, superhydrophilic surfaces allow water droplets to spread on surfaces with contact angles lower than 20°. [2]The aforementioned wettability characteristics depend on the chemistry and roughness of the surface, which define the surface energy.3][4] Surfaces with a superhydrophobic characteristic have shown to be relevant in numerous applications requiring for instance antibacterial, [5] self-cleaning, [6] and high corrosion resistance properties. [7]Thus, several physical and chemical methods have been developed for the fabrication of superhydrophobic surfaces.[10][11] For example, Huang et al. used a chemical etching method to produce superhydrophobic surfaces in aluminum (reaching water contact angles, WCA, of 156°) that simultaneously showed a decrease in the corrosion current densities (0.023 μA cm −2 ) and improved corrosion resistance. [12]Wang et al. also fabricated superhydrophobic surfaces (WCA of 158°) on Al samples that exhibited an improvement in the corrosion behavior but applied the sol-gel method. [13] significant drawback of chemical-based methods lies in their reliance on environmentally toxic reagents [10,14] that can lead also to high production costs. [15,16]aser-based microfabrication methods have been developed to create superhydrophobic surfaces on metals.These methods involve producing diverse textures that facilitate the formation of topographical elements, giving rise to air pockets that effectively prevent water from wetting the surface, as discussed earlier. [6,17][20] For instance, Xin et al. used a Nd:YVO4 laser with a pulse duration of 13 ns to produce pillar-like structures on Al-surfaces by applying after the laser process a solution of fluoride agent (FAS-17, C 16 F 17 H 19 O 3 Si).These surfaces reached WCA up to 163°and corrosion inhibition efficiency up to 95.7%. [20]In another work, Yang et al. fabricated superhydrophobic aluminum also combining a chemical agent and nanosecond laser source (with a pulse duration of 50 ns).As a result, the treated aluminum showed a decrease in the corrosion current as well as a positive shift in the corrosion potential. [19]he study of the corrosion resistance of superhydrophobic laser-treated surfaces without utilizing chemical treatments has recently become of great interest, to avoid the environmental impact due to the toxic reagents. [7,21]Ahuir-Torres et al. used a picosecond laser to fabricate different patterns in Al2024 generating three diverse patterns: dimples, crossed grooves, and concentric rings.Their findings indicated that there was no improvement in terms of corrosion resistance even though they attained the condition of hydrophobicity. [22]Also, Lara et al. used direct laser writing (DLW) with a laser source emitting pulses with a pulse duration of 30 ns, producing micropillars on aluminum sheets.The corrosion current density of the laser-treated samples was reduced by 95% for superhydrophobic surfaces.This study attributed the enhanced corrosion resistance to the superhydrophobicity state [7] Indeed, Pariona et al. also proposed that the topography of the laser-treated samples, the homogeneity of the pattern, and the phases of the grew aluminum oxide, are the main factors for the improvement of the corrosion resistance (14 times). [23]][26] All these studies agree that the superhydrophobic wetting state plays a key role in the enhancement of corrosion resistance.However, the correlation between these two properties has not been sufficiently evaluated in the literature.To the best of the author's knowledge, there are no published studies on the correlation between the wettability state and corrosion resistance of laser-treated surfaces exhibiting both superhydrophobic and superhydrophilic characteristics, without utilizing chemical treatments and employing similar irradiation conditions.
This work focuses on the fabrication of superhydrophobic and superhydrophilic surfaces on aluminum 2024 using laser-based microfabrication technologies, with the aim of determining the relationships between their wetting behaviors and corrosion resistance.In order to generate topographies with very different wetting conditions two laser-based techniques were used, namely DLW and Direct Laser Interference Patterning (DLIP).In both cases, ns pulses from infrared (1064 nm) laser sources were utilized.In the DLIP process, line-like structures were fabricated using a two-beam configuration with a spatial period of 6.0 μm.Scanning Electron Microscopy (SEM) and Confocal Microscopy methods were used to characterize the surface topography.In addition, a Focused Ion Beam (FIB) technique was used to perform cross-sectional analyses of the laser-treated surfaces.Finally, the wetting behavior was determined by static contact angle (SCA) measurements, and the corrosion resistance of the samples was studied using polarization curves as well as Electrochemical Impedance Spectroscopy (EIS).

Materials
Aluminum 2024-T351 sheets (Gemmel Metalle, Germany) with a thickness of 1 mm and a total area of 1 cm 2 were used in this study.Aluminum alloys of the 2000-series are preferently used for their high mechanical strength.By nature, they exhibit low corrosion resistance.The effect is originated by the CuAl 2 -precipitation with a higher electrochemical resistance than the bulk material, which poses the risk for intercrystalline corrosion. [22,27]he initial surface roughness (S a ) was 0.35 μm ± 0.01 μm.Prior to laser treatment, the samples were cleaned with ethanol, and after the laser process, the textured samples were stored under atmospheric conditions.No further treatments were performed.

Laser Surface Texturing
The aluminum surfaces were treated using both DLIP and DLW methods.A schematic drawing of both configurations is shown in Figure 1.In the case of DLIP (Figure 1a), the structuring process was performed using a nanosecond solid-state laser (Edgewave InnoSlab-IS400-3-GH, Würselen, Germany) with a fundamental wavelength  of 1064 nm and a pulse duration  of 6.3 ns.The frequency of the laser source can be set up to 12 kHz, leading to an average output power of 130 W (at 5 kHz).The main beam is then directed to an xDLIP optical head (SurFunction GmbH, Germany), which produces an interference pattern over the sample surface.This pattern is characterized by a periodic distribution of laser intensity, which, depending on the number of overlapping beams, can generate different shapes, e.g., line-like or crater-like patterns. [28]In this case, the used xDLIP optics overlapped two sub-beams, and thus a line-like geometry was obtained.The distance between the generated periodic microstructures is called spatial period (Λ) and its size depends on the used laser wavelength as well as the intercepting angle between the beams. [28]In this case, the obtained spatial period Λ was 6.0 μm, corresponding to an intercepting angle of 5.1°.
The shape of the overlapped beams is controlled by the xDLIP optics, which utilize cylindrical lenses that lead to an elongated spot geometry that contains the two-beam interference pattern.The use of a cylindrical lens with a focal distance of 450 mm permitted the laser sub-beams to be focused on the material surface with a dimension of 5 mm and 50 μm in the X and Y directions, respectively (D x , D y ).In addition, the xDLIP optics provide a depth of focus of ≈10 mm, which is ≈50 times longer than previous optical setups. [29]By adjusting the laser power between 54 and 68 W at a constant repetition rate of 5 kHz, different cumulated laser fluences from 5.3 to 13.4 J cm −2 were applied.The pulse-to-pulse distance was adjusted to 50 μm for the fluence of 5.3 J cm −2 , 30 μm for 7.3 and 9.1 J cm −2 , and 20 μm for the highest fluences of 10.8, 12.2, and 13.4 J cm −2 .
In the DLW method, a nanosecond pulsed laser source was also implemented.The used experimental setup is shown in Figure 1b.In this case, the laser wavelength  was also 1064 nm and the pulse width (pulse duration) was 4.0 ns.Using an ftheta lens with a focal distance of 254 mm, a laser spot diameter of 70 μm was reached.To compare the results with those of the DLIP experiments, equal cumulated fluences were applied.This was achieved by adjusting the scanning speed, pulse overlap, and laser power.The scanning speeds were varied from 122 to 378 mm s −1 at a constant laser power of 0.9 W using 5, 7, 9, 12, 14, 16, and 17 overscans and therefore resulting in cumulated fluences of 5.3, 7.3, 9.1, 10.8, 12.2, and 13.4 J cm −2 , respectively.

Characterization of the Treated and Untreated Al-Surfaces
The topographic characterization of the treated and untreated Alsurfaces was performed using a confocal microscope (CM, Sensofar S-Neox, Spain) with a magnification objective of 150x.This led to lateral and vertical resolutions of 140 and 1 nm, respectively.The surface roughness of the treated samples was analyzed using the software Sensomap 7.3 (Sensofar, Spain).The presented results represent the average surface roughness of the five samples for each treatment, measured within an area of 160 μm 2 .In addition, Scanning Electron Microscopy (SEM, Ther-moFischer Scientific Quattro S, USA) was employed to obtain high-resolution images of the structured surfaces.
In order to characterize the possible formation of oxide layers as a consequence of the laser treatments, cross-sections of selected samples were performed with a Focused IFIB system, employing Ga Ions at an acceleration voltage of 30 kV.To prevent damage from Ga-ion imaging or milling, the regions of interest were previously coated with a 100-500 nm thick Pt layer deposited by Electron Beam-Induced Deposition (EBID).On top of this layer, a second thicker Pt layer with a thickness between 1 and 2 μm was deposited by Ion Beam Induced Deposition (IBID), improving the damage resistance and allowing the preparation of homogeneous cross sections.To characterize the reference sample, Scanning Transmission Electron Microscopy images were acquired with an electron acceleration of 30 kV.Then, images of the cross sections of the laser-treated samples were obtained using SEM at an accelerating voltage of 5.0 kV and a tilting angle of 52°.The oxide layer thickness was measured in at least three positions directly from the SEM images by correcting the vertical scale using the viewing angle.

Wettability Analysis
Wetting behavior was assessed by measuring the SCA.Measurements were performed five times per treatment using a drop shape analyzer (Krüss, DSA 100, Germany) with 4 μL of deionized water droplets under ambient conditions (22 °C and 16% relative humidity).The SCA values were calculated using the Young-Laplace fit.[32] The error was described by the standard deviation of the measured values.

Corrosion Resistance Measurements
Corrosion experiments were carried out in a three-electrode cell connected to a potentiostat (1010T, Gamry, USA).A Saturated Calomel Electrode (SCE) was used as a reference and a platinum sheet as a counter electrode.All tests were performed at room temperature (23 °C) in a 3.5%w/w NaCl solution.Potentiodynamic polarization curves were conducted from −0.25 to +0.25 V voltages versus the Open Circuit Potential (OCP) at a scan rate of 0.16 mV s −1 .The corrosion current density (i corr ), as well as the corrosion potential (E corr ) were calculated by Tafel extrapolation to the anodic and cathodic parts of the curves. [33]The potential values were expressed relative to the SCE.The results are presented as the average of the three measurements per treatment.
After applying the pitting corrosion potential, the samples were stored in a 3.5%w/w NaCl solution to study the distribution of the localized corrosion on the surface.After 48 h, samples were dried under ambient conditions.The resulting surface was characterized using SEM.Furthermore, to address the barrier properties of the oxide layer and for a deeper understanding of the electrochemical reactions occurring at the oxide-metal interface, EIS measurements were conducted in another potentiostat (Vortex, IVIUM, Netherlands).The EIS measurements scans were carried out over a frequency range from 100 kHz to 0.01 Hz.An amplitude of 10 mV was selected under potentiostatic conditions.All electrochemical results were fitted using Echem Analyst software (Gamry Instruments, USA).

Results and Discussion
For the purpose of studying the main factors affecting the corrosion resistance of the laser-treated samples, two different methods were employed to produce surfaces with dissimilar wetting conditions.In the following subsections, the differently produced textured Al-2024 samples are presented, including the characterization of both their surface topography and wettability properties, as well as that of the oxide layer induced by the laser treatment.Finally, corrosion properties were determined and compared.

Laser Structuring of the Aluminum 2024 Surfaces
The samples designed to be superhydrophobic were fabricated using the DLIP method, with two-elongated beams and applying nanosecond pulses.The used processed parameters are based on a previous publication of the authors. [34]The resulting topographies of the laser-treated aluminum surfaces obtained by means of DLIP as well as the reference are shown in Figure 2a-c,f-h.
For the laser treatments, fluences between 5.3 and 13.4 J cm −2 were applied.The DLIP-treated samples show a well-defined line-like pattern geometry with a spatial period Λ of 6.0 μm (as expected).For instance, the confocal image in Figure 2h shows the average height of line-like produced topography of 3.5 μm (irradiated with a cumulated fluence of 13.4 J cm −2 ).This pattern arises from the local interaction at the maximum positions of the periodic intensity distribution with the material surface.In particular, at these zones, the material is molten and partially ablated (evaporated) forming the valleys of the structure which is driven by the recoil pressure as explained elsewhere. [13,14]n the case of DLW treatment, significantly different surface structures were produced, as shown in Figure 2d,e,i,j.As can be seen, the topography of the samples with equal laser fluences depicts a more random surface architecture (see SEM images in Figure 2d,e).
Since the DLW laser process parameters were chosen to apply the same cumulated fluence as in the DLIP samples, the pulse-topulse distance was adjusted to a range from 13 to 4 μm, which is significantly shorter than the beam diameter (70 μm).Thus, no defined texture geometry is obtained.The laser-structured surface exhibited several randomized topographical elements consisting of resolidified melt attachments.These elements have for instance lateral dimensions (average) of 0.54 μm and heights up to 3.1 μm as shown in Figure 2j for an accumulated fluence of 13.4 J cm −2 .
Next to the SEM images (for both the DLIP and DLW methods), the confocal analysis reveals in general an increase in surface roughness with increasing laser fluence, as shown in Figure 2f-j.To further examine the effect of the applied laser fluence on the resulting surfaces, the roughness values determined by the arithmetic mean height, S a , were measured, and are summarized in Figure 3.
The results indicate that for the DLW technique, the roughness increased with higher cumulated laser fluences.For this method, the maximum surface roughness achieved was 0.64 ± 0.02 μm at the highest used cumulated fluence (13.4 J cm −2 ).This increase is due to the higher recoil pressure that is induced in the interaction zone causing a higher accumulation of molten material that is been ejected. [34,37,38]n the case of DLIP, the arithmetic mean height roughness (S a ) increases with the laser fluence until it reaches a value of  ≈1.0 ± 0.1 μm at 10.8 J cm −2 .At higher fluences, the S a values decreased because a larger amount of recast material was produced, which can accumulate within the periodic DLIP fringes.[41][42] In summary, it can be seen that for both laser treatments, very different surface topographies were produced.Using the DLIP approach, higher surface roughness values were obtained for the aluminum surface.
Following the surface characterization of the Al-2024 lasertreated samples, FIB cuts were performed on the selected samples and compared with the reference.The obtained images are shown in Figure 4.For the reference sample depicted in Figure 4a,b, an oxide layer was observed over the raw aluminum surface (note that the protective platinum layer was deposited in the cut areas to prevent damage during the FIB cut, as described in the experimental section).This oxide layer is formed due to the natural tendency of aluminum to react with ambient oxygen and consists mainly of Al 2 O 3 .Its thickness generally varies between 2 and 20 nm. [43]In our case, the oxide layer had a thickness of 10 nm ± 2 nm, which is in agreement with the literature (see Figure 4b).
The cross-section of the laser-treated samples by DLIP presents a thicker oxide layer with a thickness of 120 nm ± 20 nm, and extends uniformly over the entire treated surface, as shown in Figure 4c,d.In addition to the greater thickness, the oxide layer exhibited a slightly porous character.The increase in the thickness of the oxide layer responds to the thermal effects produced by laser irradiation. [44,45]he FIB cut samples treated by DLW depict a very different morphology compared to the DLIP-treated surfaces.In this case, the layer is characterized by a thick and highly porous morphology, as illustrated in Figure 4e,f.Although this layer extended over the entire irradiated surface, it had present a constant thickness owing to the pores.The average thickness of the oxide layer was 1.2 μm ± 0.5 μm, which is ≈10 times thicker than the oxide layer produced in the DLIP treatment.

Wetting Properties
The wetting behavior of the aluminum samples was evaluated by determining the static water contact angle (SWCA for 55 days.In Figure 5a, the evolution of the SWCA with time is reported for the laser-treated samples fabricated using the DLIP and DLW techniques, compared to the reference. For the DLIP treatment, the samples first became hydrophilic.[48] This theory has been based on the findings of unsaturated elements in the formation of the oxide layer of Al 2 O 3 that behave as Lewis acid and base pairs, as demonstrated by Hass et al. [49] After 15 days, the samples reached the SWCA of the reference surface (81°± 2°), after which they became hydrophobic or superhydrophobic.After 55 days, all the tested samples achieved an SWCA in the range of 128.1°± 3.8°-158.4°±4.0°.The last value corresponded to the sample treated with the highest fluence (13.4 J cm −2 ).Moreover, the transition from the hydrophilic to the hydrophobic/superhydrophobic state has been already reported by several authors and explained by the increment in the number of non-polar sites on the surface. [47,50]ndeed, studies on aluminum superhydrophobic surfaces found that, after the laser-structured samples are exposed to atmospheric conditions, the adsorption of non-polar organic compounds occurs. [34,38,46,47]This was confirmed for instance by Long et al. using X-ray photoelectron spectroscopy (XPS). [47]n contrast, the SWCA values of the DLW samples decreased below 20°, denoting a superhydrophilic character.The DLW samples preserved their superhydrophilic property during the 55 days without significant variations, presenting final SWCA values between 2.5°± 1.8°and 7.0°± 4.0°.Several studies have reported the fabrication of superhydrophilic laser-treated aluminum surfaces using thermal [45,51,52] or chemical [21,53] treatments.Only a limited number of studies have reported the generation of superhydrophilic laser-structured aluminum surfaces without additional treatments. [45,52,54]For example, Guan et al. used a nanosecond pulsed laser also with a wavelength of 1064 nm, to fabricate superhydrophilic samples.However, after 10 days, the surfaces stored in ambient air experienced an increase in their SWCA nature, with the contact angle reaching 31°. [54]In another study, Zhao et al. reported that a grid pattern also created by laser treatment with nanosecond pulses preserved its superhydrophilicity after 20 days.Both contributions demonstrated the presence of polar compounds in the surface composition and associated the superhydrophilic character of the chemical composition and the increasing roughness produced by the laser treatment. [52,54]or better visualization of the obtained results, the final SWCA values (after 55 days) were plotted as a function of the used cumulated fluence for each laser treatment, as shown in Figure 5b.As mentioned before, the SWCA angles achieved by the DLIP samples were situated in the hydrophobic and superhydrophobic range, whereas for the DLW-treated samples, the SWCA was superhydrophilic.Because the transition of the wettability states has been attributed to both chemical and topographical changes induced by laser treatment, [5,25,26] and in the case of the experiments performed with DLIP and DLW, the same energy and storage conditions were used.We can assume that the different wettability states were induced mainly by changes in the surface topography and morphology of the oxide layer.

Corrosion Behavior
The corrosion behavior of the aluminum-2024 samples was characterized using potentiodynamic polarization curves and Electrochemical Impedance SpectrEIS techniques.The polarization curves of the laser-treated samples obtained using the DLIP and DLW techniques are compared to the reference sample, as shown in Figure 6.The values of the parameters calculated from the potentiodynamic curves are presented in Table 1.
The curves corresponding to the DLIP laser-treated samples are shown in Figure 6a.Compared with the untreated reference sample, the curves show lower current densities.For example, the DLIP sample treated with a laser-cumulated fluence of 12.2 J cm −2 (see also Table 1) shows a decrease in the current densities from 1220.5 nA cm −2 (reference) to 35.4 nA cm −2 .The results showed a trend of decreasing corrosion current densities with increasing laser fluence.It should be noted that the samples fabricated at a laser fluence of 13.4 J cm −2 exhibited similar corrosion resistance to those fabricated at 12.2 J cm −2 .Furthermore, from the figure, it can also be observed that when higher laser fluences are used, the corrosion potential shifts toward more positive potentials.The nobler corrosion potentials reached were −656 and −660 mV, for the laser fluences of 12.2 and 13.4 J cm −2 , respectively.The reference surface was −725 mV.
In the case of the aluminum samples treated by the DLW method, the recorded curves show lower current densities than the reference.Interestingly, the samples fabricated at lower laser fluences exhibited the lowest current densities.For instance, the lowest current density was 27.9 nA cm −2 , obtained by the sample fabricated at 5.3 J cm −2 .In general, the DLW treatment showed a clear tendency to protect the surface from localized corrosion, as represented by the passive region of the curves.However, only the curves corresponding to the two lowest fluencies, 5.3 and 7.3 cm −2 , presented a visible nobler corrosion potential than the reference, scaling up to −622 and −698 mV, respectively.
Overall, for both laser treatments, it can be seen that the polarization curves are located at current densities lower than the reference.This indicates that due to the laser treatment, the corrosion resistance of the Al-surface was improved.
After the potentiodynamic evaluation, a study on the localized corrosion mechanism of the laser-treated samples was performed.After applying the pitting corrosion potential, the samples were stored in the same 3.5% NaCl solution to induce localized corrosion on the surface.SEM images of selected DLIP and DLW laser-treated samples are presented in Figure 7.As can be observed in Figure 7a,b, the pits that originated in the DLIP samples show a rounded form and are not oriented toward the DLIP fringes.This situation is more favorable compared to a selective direction of propagation since this can represent a disadvantage for corrosion resistance in the long term. [56]For instance, Trdan et al. reported a pitting propagation in the direction of the laser textures on stainless steel surfaces. [56]dditionally, a uniform texture was visible within the cavities of the pits formed on the DLIP surface.For aluminum, pitting is usually initiated by the deterioration of the oxide layer and depends on the electrical properties of the layer, as well as the presence of defects in its composition.A less defective oxide layer is more resistant to pitting. [57,58]Thus, in the case of the DLIP textures, the morphology observed could be related to the thicker and relatively compact oxide layer generated by the laser treatment, as already discussed (see Figure 4c,d).
In the DLW-structured samples, it can be observed that they exhibited crystallographic pits, as shown in Figure 7c,d.These type of pits are commonly seen in homogeneous aluminum alloys and is characterized by rounded walls in the crystallographic planes due to the presence of copper in their chemical composition. [59]he propagation of these pits then follows the orientation of the substrate grains below the oxide. [60]From the comparison of the pits formed on the surfaces treated by both techniques, it can be concluded that the pits formed on DLW-treated surfaces are shallower than those found on the DLIP surfaces, which could be  related to the thickness of the protective oxide layer.In summary, this study demonstrates that the orientation of laser-fabricated structures is not a point of propagation of localized corrosion when these surfaces are exposed to saline solutions and specific electrochemical conditions that induce the formation of this type of corrosion.Furthermore, to examine the influence of thelaser parameters on the corrosion resistance, the efficiency of the laser treatment to inhibit the corrosion of the Al samples was calculated as the inhibition efficiency (), which is commonly used in corrosion inhibitors.This parameter () can be calculated using the current densities determined from the potentiodynamic curves for each treatment, using Equation 1: [19,61] n% = i 0 corr − i where i 0 corr and i j corr correspond to the corrosion current densities of the reference sample and a specific laser-structured surface, respectively.The inhibition efficiency was analyzed as a function of the cumulated laser fluence, as shown in Figure 8a.
The results showed that the inhibition efficiency of the DLIP treatment increased with the applied fluence from 57% to 97%.The increment in the inhibition efficiency with the laser fluence can be attributed to an increase in the thickness of the oxide layer with higher fluences as observed in other research studies. [17,18,62]n contrast to the trend described for the DLIP behavior, the DLW treatment proved to be more efficient when lower cumulative fluences were used, with a maximum efficiency of 98% at the lowest fluence (5.3 J cm −2 ).A possible explanation for this phenomenon could be related to the fact that the observed porous structure with increasing laser fluence trends to crack and the electrolyte penetrates into the oxide layer thus reducing the corrosion efficiency. [63,64]or both laser treatments, the lowest calculated efficiencies for preventing corrosion were above 55% (55% improvement compared to the flat reference).The values reported here represent a positive advance for the use of laser treatments (with pulsed ns laser sources) in corrosion prevention since they are comparable to efficiencies reached with other anticorrosion  treatments, such as the application of coatings or the use of corrosion inhibitors. [65,66]o study the relationship between corrosion resistance and wettability state, the corrosion rates for each laser-structured sample were determined.The corrosion rate was calculated from the Faraday's law, described by Equation ( 2): [67] where i corr is the corrosion current density of the sample, M is the molecular weight of aluminum, N is the number of electrons involved in the oxidation process (for aluminum is 3), F is the constant of Faraday, and  the density of aluminum.
The dependency between the SWCA and the corrosion rate is shown in Figure 8b.Since a lower corrosion density is related to a lower corrosion rate, lower corrosion rates are expected for high laser fluences for samples fabricated with the DLIP technique.Regarding the untreated reference Al-2024 surface, the calculated corrosion rate was 12.6 ± 0.5 μm yr −1 .As for the samples treated with laser fluences in the range between 9.1 and 13.4 J cm −2 , the measured corrosion rates varied between 1.9 ± 0.2 μm yr −1 and 0.40 ± 0.02 μm yr −1 .These values represent an improvement of 97% compared to the reference.
For the DLW treatment, the lowest corrosion rates were obtained for lower applied laser fluences.For instance, the laser treatment performed with 5.3 J cm −2 led to a corrosion rate of only 0.32 ± 0.02 μm yr −1 , being this value similar to the best result for the DLIP treatment.For the other applied fluences (from 7.3 to 13.4 J cm −2 ), the corrosion rate varied from 0.96 ± 0.03 to 4.0 ± 0.1 μm yr −1 .
Interestingly, the data depicted in Figure 8b show that whether the Al-surfaces are superhydrophobic or superhydrophilic, they exhibit a significantly lower corrosion rate compared to the untreated reference.In other words, both the superhydrophobic and superhydrophilic surfaces performed well in terms of corrosion rates.In other words, both superhydrophobic and superhydrophilic surfaces performed well in terms of corrosion rates.It can be finally concluded that the calculated corrosion rates did not correlate with the observed SWCA values.
To further investigate the reasons for the observed improvements in the corrosion rate owing to both DLIP and DLW laser treatments, and in particular, considering the formed oxide layers (with different morphologies), Electrochemical ImpedaEIS measurements were also performed.The Bode plots of the samples that provided the lowest corrosion rates, which are 13.4 J cm −2 for DLIP and 5.3 J cm −2 for DLW, are presented in Figure 9.
The black symbols represent the variation of the impedance module (|Z|) as a function of frequency, whereas the white symbols depict the phase angle variation with the frequency.The results show that both DLIP and DLW laser treatments provided the samples with larger impedances as well as a prompter increase in the phase angle.At 0.1 Hz the reference sample denotes a |Z| value of only 4.44 kOhm cm 2 meanwhile in the DLIP and DLW samples such values increased to 59.69 and 110.80 kOhm cm 2 , respectively.The observed increment in the impedance at low frequencies indicates an increasing charge transfer resistance at the interface between the oxide layer and the substrate. [68,69]The  phase maxima of the laser-treated samples shifted to lower capacities indicating a higher layer thickness.
The effect of laser treatment on the resistance of the aluminum surface can also be observed in detail in the Nyquist plot presented in Figure 10.The investigated specimens corresponded to the lowest and highest laser fluences, 5.3 and 13.4 J cm −2 for each treatment.The Nyquist plot presents the real component of the impedance (Z′) versus the imaginary component (Z″).In the case of the DLIP treatment (Figure 10a), there is a large capacitive loop at the high-frequency regions (HF) and a small capacitive loop at the low frequency.The second capacitive loop depicts a diffusion behavior characterized by the tendency of the loop to a line.This phenomenon is explained by the diffusion of ions to the protective layer. [70,71]The diameter of the capacitive loop increased in the DLIP-treated samples.The samples treated at the higher fluence, 13.4 J cm −2 presented a larger diameter than those fabricated at 5.3 J cm −2 .An increase in the diameter of the capacitive loop indicates a higher resistance against corrosion reactions. [20]he EIS results of the DLW-treated samples also agree with the conclusions drawn from the potentiodynamic polarization curves.The samples treated at 5.3 J cm −2 (with the highest impedances) presented a significantly higher loop compared to the surfaces treated at the highest fluence (13.4 J cm −2 ) denoting a large increase in the resistance of the oxide layer.
With the information obtained from the EIS measurements, it is possible to reconstruct the mechanism of ion migration occurring in the interface between the electrolyte and the surface, by fitting the data to an equivalent electric circuit (EEC).The fitting of the EEC provides quantitative information on the EIS analysis by explaining the electrochemical process between the interface of the electrolyte and the surface of the metal in terms of electrical circuit elements. [72]Based on the experimental results, the equivalent circuits shown in Figure 11 were proposed for both the DLIP-and DLW-treated samples as well as the reference.
The parameters used to model the data in the three different proposed circuits include the resistance of the electrolyte solution (R s ), resistance of the oxide layer (R o ), and resistance to charge transfer at the interface between the film and the aluminum substrate, (R ct ).In the analysis of the impedance data, a constant phase element is preferably used instead of an ideal capacitor, since this element allows to define the interfacial heterogeneities that may be present in the oxide layer by using the power index number (n) (for an ideal capacitor, n = 1). [68]Thus, in the fitted circuits, the capacitance of the oxide layer in terms of the constant phase element (C o ) and the double layer capacitance, (C dl ), are included in the circuits with their respective n values.In the case of the DLIP-treated samples, a Warburg resistance (W) had to be included, being the last connected with the R ct , indicating the presence of a diffusion process at the interface between the oxide layer and the aluminum substrate. [72]he DLW samples exhibited an equivalent circuit similar to the reference sample.In this case, the large pores present in the oxide layer were considered non-conductive.Overall, similar EECs were fitted to porous oxide layers, in order to explain the phenomenon taking place in the interface of the electrolytemetal surface when a porous layer or film is on the aluminum surface. [63,64]The polarization resistance represents the total resistance of the surface to corrosion reactions owing to the presence of the oxide layer.The polarization resistance (R p ) was calculated by adding the resistances extracted from the fitting of the EIS data to the equivalent circuits, excluding the resistance of the circuit due to the inner resistance of the electrolyte solution, as indicated by Equation (3): [73] R The values of each component of the circuit for all tested samples are listed in Table 2.
The results show an increment in the resistance to charge transfer for the laser-treated samples by two times for the DLIP samples (41 vs 17.8 kOmh) and 10 times for the DLW samples (179.0 vs 17.8 kOmh).Moreover, the double layer capacitance decreased from 4.06 × 10 −5 to 2.31 × 10 −6 S cm −2 s n for the DLIP samples and 1.26 × 10 −6 S cm −2 s n to the DLW.The R p calculated for the reference sample was 26.4 kOhm cm 2 .The calculated R p for DLIP-treated samples at 5.3 and 13.4 J cm −2 was 39.1 and 72.4 kOhm cm 2 .In the case of the DLW samples, the polarization resistance was calculated in 31.6 and 179 kOhm cm 2 , for the laser fluences of 5.3 and 13.4 J cm −2 .Trdan et al. also reported an improvement in corrosion resistance due to laser treatment.By using laser peening, the polarization resistance of the same aluminum alloy was improved to 30.2 kOhm cm 2 . [27]The achieved polarization resistances indicate that the laser-treated samples present a higher resistance to the corrosion reactions that can occur due to exposure to salty water.
Although the corrosion rates obtained for the surfaces treated with both DLIP and DLW methods were very similar (and ≈36 times lower than the untreated surface), the surfaces produced with DLW presented higher resistance to corrosion.Nevertheless, the superhydrophobic character of the textures produced by DLIP provides the opportunity to have additional functionalities, such as enhanced ice-repellency or self-cleaning properties, which occur in wet environments and thus require lower corrosion rates, which have to be further investigated in the future.

Conclusion
In this work, the corrosion resistance behavior of laserfunctionalized surfaces was investigated.To examine the influence of the laser parameters on the corrosion properties, two different types of topographies were fabricated using the DLIP and DLW methods, with varying process parameters.The DLIP approach was to create line-like patterns with a spatial period of 6.0 μm, thereby achieving a surface roughness S a, of 1.02 μm.For the DLW technique, the Al-2024 surfaces were scanned using the same cumulated laser fluences as for the DLIP treatment, and random surface topographies with roughness values up to 0.64 μm could be generated.
The FIB cross-section revealed a thicker oxide layer for both treatments than the untreated surface.For the DLIP samples, the thickness increased from 10 to 120 nm, whereas for the DLW process, it reached 1.2 μm.In the latter case, this layer showed a remarked porosity.Wettability measurements showed that the DLIP samples could reach either the hydrophobic or superhydrophobic state, while all Al-surfaces processed with DLW became superhydrophilic.
The evaluation of the corrosion behavior of the samples provided different results.First, for both methods, which means also for superhydrophilic and hydrophobic/superhydrophobic conditions the corrosion rate could be decreased from 12.6 to ≈0.3 μm yr −1 .In addition, the inhibition efficiency was improved up to 98% for the laser-treated samples.Thus, the corrosion resistance of the laser-treated samples can be mainly attributed to the thicker oxide layers that were produced, which was confirmed by the increment of the impedances shown in the EIS measurements.
Finally, although the corrosion rates obtained for both treated surfaces were very similar, the surfaces produced with DLIP could be beneficial when additional surface properties are required, such as ice repellency or self-cleaning properties.
Given the established fabrication throughputs achieved in DLIP, such as the range of 0.1-1.0m 2 min −1 utilizing a 250 W laser source, and through conducting a comprehensive cost-productivity evaluation as detailed by Zwahr, [74] the resultant processing costs approximate to 0.2-0.5 € m −2 , subsequent to the amortization of initial equipment investments.Consequently, the technology herein represents a promising approach with a substantial productivity potential for industry integration.

Figure 1 .
Figure 1.Schematic drawing of the used laser configurations to treat the Al-2024 samples using a) DLIP (obtained 6.0 μm line-like periodic patterns) and b) DLW (with a 254 mm f-theta lens, reaching a spot size of 70 μm).

Figure 2 .
Figure 2. SEM images of a) reference and samples treated with DLIP at a laser fluence of b) 5.3 and c) 13.4 J cm −2 as well as using DLW with cumulated fluences of d) 5.3 and e) 13.4 J cm −2 .f-j) Confocal images of the corresponding SEM images.

Figure 3 .
Figure 3. Surface roughness (arithmetic mean height, S a ) of the lasertreated samples using both DLIP and DLW techniques.

Figure 4 .
Figure 4. STEM images of the FIB cut of the a,b) reference sample compared to the SEM images of FIB cuts of the c,d) DLIP and e,f) DLW treated samples at 9.1 J cm −2 a cumulated laser fluence.Images (c-f) were obtained with the sample tilted at 52°.A protective platinum layer is deposited during the preparation of the sample for the FIB cuts, as described in the experimental section.

Figure 5 .
Figure 5. a) Wettability evolution of the DLIP samples over time and b) final sSWCA achieved after 55 days for each cumulated laser fluence.

Figure 6 .
Figure 6.Polarization curves of the laser-treated samples compared to the reference using a) DLIP and b) DLW techniques in 3.5% NaCl solution at room temperature (23 °C).

Figure 7 .
Figure 7. SEM images of pitted laser-treated surfaces in 3.5% w/w NaCl for the a,b) DLIP and c,d) DLW treatments.The cumulated laser fluence was 9.1 J cm −2 .

Figure 8 .
Figure 8. a) Inhibition efficiency for the DLIP and DLW laser treatments as a function of the cumulated laser fluence and b) correlation between the WCA and the corrosion rate (in μm per yr) compared to the reference untreated sample.

Figure 9 .
Figure 9. a) Bode plots of the EIS measurements in 3.5%w/w NaCl at room temperature (23 °C) for the reference as well as the DLIP and DLW samples treated at 13.4 and 5.3 J cm −2 , respectively.

Figure 10 .
Figure 10.Nyquist plots of the EIS measurements in 3.5%w/w NaCl at room temperature (23 °C) for a) DLIP and b) DLW laser-treated samples at 5.3 and 13.4 J cm −2 compared to the reference.

Figure 11 .
Figure 11.Schematic of the equivalent electric circuits (EEC) obtained from fitting the EIS data of both DLIP and DLW textured Al-2024 surfaces as well as the reference sample.

Table 1 .
Electrochemical parameters determined from the potentiodynamic polarization curves of the laser-treated samples compared with the reference in 3.5% NaCl at room temperature (23 °C).SCE: Saturated Calomel Electrode; E corr : corrosion potential; I corr : corrosion current density.