Effect of Superhydrophobic Surface on Corrosion Resistance of Magnesium-Neodymium Alloy in Artificial Hand Sweat

Abstract

A superhydrophobic surface can endow metals with some intriguing characteristics such as self-cleaning behavior. In this study, a simple solution-immersion method based on the concept of predesigned corrosion is developed to enhance the corrosion resistance of a magnesium-neodymium alloy. The Mg alloy is directly soaked in potassium dihydrogen phosphate solution with the addition of ultrasound, and a layer of rough but dense coating is uniformly formed on the Mg-Nd alloy after the immersion process, which is mainly composed of MgHPO₄∙3H₂O. A superhydrophobic surface with an average wetting angle of 150.5° and a sliding angle of about 4.5° can be obtained on the Mg alloy by further chemical surface modification with perfluorodecyltriethoxysilane. This superhydrophobic surface has an interesting self-cleaning effect as well as good corrosion resistance in artificial hand sweat. In brief, this study provides a feasible way to prepare a superhydrophobic surface on the Mg-Nd alloy and reveals the effect of a superhydrophobic surface on the corrosion behavior of the Mg-Nd alloy, offering new technical insights into the corrosion protection of magnesium alloys.

1. Introduction

Benefiting from the characteristics of low density, high specific strength and natural biodegradation, magnesium alloys have attracted much attention in the aerospace, transportation, building and biomedical fields [1,2,3,4,5,6]. However, their poor corrosion resistance in aqueous solutions always hampers their wider applications, which is mainly due to the high chemical activity of magnesium and the loose natural surface film on most magnesium alloys [7,8,9]. Therefore, it is very crucial to control the corrosion rate of magnesium alloys in engineering applications, and in the past years, surface treatments including electroplating, electroless plating, chemical conversion, thermal spraying, polymer coating, and vapor deposition have been considered among the most effective strategies to enhance their corrosion resistance [10,11,12,13,14,15,16].

Zhang et al. proposed a concept for corrosion mitigation of magnesium alloys in which predesigned surface corrosion is conducted before their service. In their study, an effective method was developed using a uniform self-layered coating composed of an inner compact layer and top Mg-Al layered double hydroxide (LDH) micro sheets, prepared on AZ80 Mg alloys in a hydrothermal environment for corrosion protection [17]. Yao et al. immersed a magnesium-neodymium alloy in potassium dihydrogen phosphate solution at 50 °C for 4 h and obtained a layer of uniform MgHPO₄·3H₂O coating on the Mg alloy, significantly improving the corrosion resistance of the Mg-Nd alloy in Cl^--containing simulated concrete pore solution [18]. Gradually, it is becoming an acceptable strategy of retarding corrosion that corrosion products may be designed in advance to form a protective surface coating on magnesium alloys. Nonetheless, there have been few studies on the development of this strategy.

Superhydrophobic surfaces are often believed to provide metals with high corrosion resistance because of their ability to repel water [19]. For example, Lin et al. prepared a superhydrophobic Ni-Co/Cu composite coating on carbon steel and successfully improved the corrosion resistance in seawater [20]. In this study, we first immersed an Mg-Nd alloy in KH₂PO₄ solution with the addition of ultrasound to spontaneously form a layer of protective coating based on the aforementioned idea of predesigned corrosion. Then, the surface chemical modification was applied to fabricate a super-hydrophobic surface after the immersion process. Finally, the effect of superhydrophobic surface on the corrosion resistance of Mg-Nd alloy was investigated in artificial hand sweat and the associated corrosion mechanism was also discussed.

2. Materials and Methods

As-cast Mg-3.3 wt. % Nd alloys with dimensions of 10 mm × 10 mm × 5 mm were used as the substrates. The fabrication process of a superhydrophobic surface on the Mg-Nd alloy is shown in Figure 1a. Firstly, Mg-Nd alloys were mechanically ground with silicon carbide abrasive paper up to # 2000, ultrasonically cleaned in ethanol for 3 min, and rapidly dried with air. The samples are denoted as MA. Secondly, three samples were immersed into 120 mL of 0.1 mol·L⁻¹ potassium dihydrogen phosphate (KH₂PO₄) solution, whose pH value was equal to 4.6. The solution was heated to 30 ℃ before the immersion process and the beaker containing this solution was placed in an ultrasonic cleaner during the immersion process. In this process, the samples were treated at a frequency of 100 kHz for 60 min to achieve uniform corrosion. The evolution of temperature and pH value during the treatment are shown in Figure 1b. The samples were taken and washed with deionized water and ethanol in turns, followed by rapid drying. The corresponding samples are denoted as MA/UC. Finally, 2 wt. % perfluorodecyltriethoxysilane (C₁₆F₁₇H₁₉O₃Si, PFDS) ethanol solution was prepared with analytical grade reagent PFDS and alcohol. Then, MA/UC samples were soaked in the solution for 2 h. After soaking, the samples were taken out and heated at 140 °C for 1 h in an oven. After drying, the samples were taken out and cooled in the air. Here, the processed samples are named as MA/UC/CM and the MA samples directly modified by 2 wt. % C₁₆F₁₇H₁₉O₃Si ethanol solution are denoted as MA/CM. Figure 1c gives the appearance of all samples, showing that the color of the surface of Mg-Nd alloy changes from bright white to gray-white after the ultrasound-assisted corrosion process.

The surface and cross-section morphologies of the coated samples were observed by scanning electron microscopy (SEM) equipped (Zeiss, Oberkochen, Germany) with energy-dispersive X-ray spectroscopy (EDS) (OXFORD Xplore, Oxford, UK). The phase compositions of the coated and uncoated samples were analyzed by X-ray diffraction (XRD) with Cu Kα radiation. X-ray photoelectron spectroscopy (XPS) with an Al Kα irradiation was conducted to reveal the elemental composition and chemical state of the PFDS-modified surface. The binding energies were referenced to the C 1s peak (284.8 eV).

The static water contact angle and sliding angle of the samples were measured with 4 μL of the ultrapure water droplet on a JC2000D1 instrument (POWEREACH, Shanghai, China). Each static water contact angle was tested at least three different times to calculate the average values. In order to evaluate the self-cleaning ability of the samples, they were placed on a platform with an inclination angle of about 8° and a small amount of grit was uniformly dispersed on their respective surfaces. In the test, blue ink droplets dropped from a height of about 0.01 m above the samples and the state of the grit on the surfaces of different samples was accordingly recorded by a digital camera.

Corrosion behavior was evaluated by electrochemical measurement and an immersion test in simulated hand sweat. Here, the simulated hand sweat was prepared with a ratio of 0.5 wt.% NaCl, 0.5 wt.% lactic acid and 0.1 wt.% urea. Its pH value was adjusted to 5.5 with 1 mol/L NaOH (pH = 13.01) solution [21,22,23]. All the electrochemical measurements were carried out on an electrochemical workstation (Reference 600+, Gamry, Warminster, PA, USA) using a conventional three-electrode electrochemical cell. The cell was composed of a sample as the working electrode, a platinum sheet as the counter electrode, and a saturated calomel electrode as the reference electrode. In the test, each sample with an exposed surface area of 1 × 1 cm² was immersed in 100 mL of simulated hand sweat for 1 h, and then the electrochemical impedance spectrum (EIS) was collected from 100 kHz to 100 MHz. Similarly, the polarization test was performed at a scanning rate of 1 mV·s⁻¹. The corrosion potential (E_corr) and corrosion current density (I_corr) were derived from the polarization curve, and the impedance data were also fitted using the respective electrical circuits. Furthermore, the immersion test in simulated hand sweat was conducted for evaluating the corrosion resistance of samples. Each investigated sample was immersed in 100 mL of simulated hand sweat and taken out for cleaning and drying after immersion for 24 h. Subsequently, the surface morphology of the corroded samples was captured by scanning electron microscope.

3. Results

Uniform corrosion is one of the key steps to fabricating the superhydrophobic surface on the Mg-Nd alloy in this study. Figure 2 shows the XRD patterns of the pristine sample and the sample treated by uniform corrosion in potassium dihydrogen phosphate solution. It is identified by comparison that the substrate is composed of α phase (Mg) and β phase (Mg₁₂Nd) and the newly-formed coating material is MgHPO₄·3H₂O. Figure 3 exhibits the surface and cross-section morphologies of the coated sample after PFDS modification. It is revealed in Figure 3a,b that the surface consisting of hierarchical regular polygons crystals becomes block-like. The cross-section shown in Figure 3c,d proves that a rough but dense coating is uniformly formed on the surface with an estimated thickness of 27.11 ± 4.44 μm. Figure 4 further shows the EDS elemental mappings of the cross-section, which indicates that the coating mainly contains magnesium, phosphorus, and oxygen. On the one hand, it proves that this composition is in accordance with the result of XRD tests. On the other hand, the distribution of neodymium exhibits the position of β phase (Mg₁₂Nd). Generally, most common phosphate conversion coatings have massive cracks due to the dehydration process or the entrapped hydrogen [24,25]. Relatively speaking, the number of cracks in our prepared coating is very small. Furthermore, the crack shown in Figure 4 possibly stems from the process of mechanical polishing for the preparation of SEM samples. Figure 5a shows the XPS survey spectrum of the coated sample after PDFS-modification, which discloses the presence of elements C, F, and Si. Figure 5b–d shows the high-resolution XPS spectra of C 1s, F 1s, and Si 2p, respectively. The F 1s spectrum shows one peak at 688.8 eV, the Si 2p spectrum shows one peak at 102.2 eV, and the C 1s spectrum can be deconvoluted into 4 peaks, namely -CF₃ at 293.6 eV, -CF₂ at 291.3 eV, -C-O at 286 eV, -C-C at 284.8 eV, suggesting that the surface has been successfully modified by PFDS.

Figure 6a shows the macro-wettability of MA, MA/UC, MA/CM, and MA/UC/CM captured by a digital camera. The droplet spreads out on the surface of MA, whereas it is absorbed by the surface of MA/UC. In contrast, the droplet keeps an apparent ball-like shape on MA/CM and MA/UC/CM, respectively. Figure 6b gives the measured contact angles of samples fabricated by different methods. MA, i.e., the Mg-Nd alloy, is hydrophilic with a contact angle of 18.9 ± 2.1° and MA/UC, i.e., the Mg-Nd alloy processed by uniform corrosion, exhibits much stronger hydrophilicity and thus much smaller contact angle, which corresponds to the phenomenon that the blue water droplets are directly absorbed, as shown in Figure 6a. After PFDS-modification, the surfaces of the samples become hydrophobic. The contact angles of MA/CM and MA/UC/CM are 119.5 ± 3.6° and 150.5 ± 0.6°, respectively. In addition, the sliding angles of MA/CM and MA/UC/CM are about 19° and approximately 4.5°, respectively.

Figure 7a shows the results of the self-cleaning tests that were performed according to the schematic diagram shown in Figure 7b. For MA and MA/UC, the blue water droplet directly infiltrates the grits due to the hydrophilicity of the surfaces. Although the surface of MA/CM is hydrophobic with a contact angle close to 120° and a sliding angle of about 19°, the water droplet cannot roll down from the surface covered with grits because the inclination of the platform is only 8° in this study. In contrast, the water droplet which has fallen on MA/UC/CM can immediately blend with grits and rapidly roll down, taking away the grits simultaneously. Therefore, it is concluded that the surface of MA/UC/CM has an exclusive self-cleaning effect compared to others.

Figure 8a presents the polarization curves of all the investigated samples after immersion in artificial hand sweat for 1 h. The average corrosion potential (E_corr) and corrosion current density (I_corr) calculated from polarization curves are summarized in Figure 8b. The corrosion potential of MA is comparable to that of MA/CM, but both are far lower than those of MA/UC and MA/UC/CM, indicating that the MgHPO₄∙3H₂O coating can significantly improve the corrosion potential of the alloy. In addition, MA/UC/CM exhibits the lowest corrosion current density and the corrosion current density of MA/CM is comparable to that of MA/UC. Since smaller I_corr corresponds to better corrosion resistance [26], this indicates that MA/UC/CM has the best corrosion resistance among all the samples.

Figure 9 shows the electrochemical impedance spectroscopy of all the samples after immersion in artificial hand sweat for 1 h. Figure 9a,b give the Bode plots and Nyquist plots, respectively. The impedance at the lowest frequency is generally used to estimate corrosion resistance, and a bigger impedance at the lowest frequency means better corrosion resistance [27,28]. As shown in Figure 9a, MA/UC/CM has the highest impedance and MA has the lowest one at the frequency of 100 mHz in the Bode plots. In addition, the impedance of MA/CM at the frequency of 100 mHz is close to that of MA/UC. The results of the Nyquist plots are in accordance with those of the Bode plots. Three equivalent circuits in Figure 9d are proposed for fitting those spectra based on the previous literature [27,28,29]. R_s(Q_dl R_ct)(Q_diff R_diff) is used to fit the EIS data of MA and MA/CM, R_s(Q_f(R_pore(Q_dl R_ct)(Q_diff R_diff))) is used in an attempt to fit the EIS data of MA/UC, and R_s(Q_f(R_pore(Q_dl (R_ct(Q_diff R_diff))))) is applied to fit the one of MA/UC/CM. In the aforementioned circuits, R_s refers to the solution resistance, Q_dl and R_ct represent the electric double layer capacity and the charge transfer resistance, respectively. Q_diff represents the capacitance pertaining to the diffusion and R_diff denotes the relevant resistance. Q_f refers to the capacitance of the whole film and R_pore represents the total resistance of the pores in the whole film. The fitting data and calculated polarization resistance R_p are listed in

Table 1. Fitted EIS data based on the equivalent circuits.

Sample		MA	MA/CM	MA/UC	MA/UC/CM
Rs (Ω·cm2)		42 ± 6	55 ± 21	37 ± 2	44 ± 1
Qf	Yf (Ω−1∙cm−2∙sn)	/	/	(1.42 ± 0.36) × 10−5	(1.76 ± 0.01) × 10−6
	nf	/	/	0.60 ± 0.02	0.83 ± 0.02
Rpore (Ω·cm2)				7 ± 4	31 ± 3
Qdl	Yd (Ω−1∙cm−2∙sn)	(1.41 ± 0.10) × 10−3	(5.14 ± 0.73) × 10−4	(7.72 ± 2.81) × 10−4	(8.31 ± 0.09) × 10−6
	ndl	0.75 ± 0.18	0.70 ± 0.23	0.45 ± 0.05	0.81 ± 0.01
Rct (Ω·cm2)		781 ± 92	1566 ± 125	555 ± 389	5105 ± 753
Qdiff	Ydiff (Ω−1∙cm−2∙sn)	(1.38 ± 0.09) × 10−5	(5.31 ± 0.77) × 10−6	(2.81 ± 0.18) × 10−5	(2.99 ± 1.81) × 10−3
	ndiff	(1.41 ± 0.10) × 10−3	0.80 ± 0.26	0.98 ± 0.02	0.91 ± 0.01
Rdiff (Ω·cm2)		713 ± 48	1668 ± 216	2204 ± 433	1172 ± 346
Rp (Ω·cm2)		1494 ± 45	3234 ± 322	2766 ± 45	7339 ± 1188

. Generally, the polarization resistance R_p is believed to be used to evaluate the corrosion resistance, i.e., the larger R_p is, the better the corrosion resistance is [26,27,30,31]. Consequently, it is concluded that MA/UC/CM has the best corrosion resistance due to having the highest polarization resistance, which also agrees well with the estimation from the Bode plots.

Figure 10 shows the surface morphology of all the investigated samples after immersion in simulated hand sweat for 24 h. After immersion for 24 h, MA, MA/CM and MA/UC exhibit severe damage on their respective surfaces, especially big corrosion pits occurring on MA and MA/UC. SEM observation reveals that localized corrosion such as pits dominates the damage mode of MA and MA/UC (Figure 10a,c), whereas MA/CM presents web-like localized corrosion with some tiny cracks (Figure 10b). Fortunately, there is no obvious damage on the surface of MA/UC/CM from the macroscopic observation and even under the observation of SEM, the substrate is still completely covered by the coating though there is slight uniform corrosion of the coating in simulated hand sweat. The result further corroborates that the super-hydrophobic Mg-P coating can significantly enhance the corrosion resistance of the magnesium-neodymium alloy.

4. Discussion

Yuan et al. introduced Nd, one of the light rare earth elements, into magnesium alloys for improving their mechanical properties by solid solution strengthening and precipitation strengthening and successfully developed a kind of Mg-Nd-Zn-Zr alloy (denoted as JDBM) mainly consisting of α-Mg and Mg₁₂Nd for biomedical applications [32,33,34]. The corrosion potential of Mg₁₂Nd is slightly more positive than that of α-Mg and the small potential difference between them is believed to slow down galvanic corrosion and thus lead to highly uniform degradation under physiological conditions [33,35,36]. In our study, the microstructure of the investigated Mg-Nd alloys is extremely similar to that of the JDBM alloy, providing the feasibility of uniform corrosion in potassium dihydrogen phosphate (KH₂PO₄) solution for forming a uniform protective coating.

Figure 11a,b show the schematic diagram of the coating formation. The coating formation mechanism from the perspective of chemical reactions is discussed below: firstly, H⁺ and HPO₄²⁻ are ionized from H₂PO₄⁻ in the solution. Then, the Mg matrix rapidly reacts with H⁺ in the solution to generate H₂ due to its high chemical activity, giving rise to the rapid increase of the pH value during the coating process. Subsequently, Mg²⁺, HPO₄⁻ and water are combined and deposited on the surface of the substrate. The associated chemical reactions are shown as follows [18,37,38]:

H₂PO₄⁻ → H⁺ + HPO₄²⁻

(1)

Mg + 2H⁺ → Mg²⁺ + H₂ ↑

(2)

Mg²⁺ + HPO₄²⁻ + 3H₂O → MgHPO₄∙3H₂O

(3)

During the coating process, the magnesium matrix is corroded and dissolved into KH₂PO₄ solution, and the second phases, Mg₁₂Nd, remain undissolved. Here, some second phases may be encapsulated in the corrosion products (MgHPO₄∙3H₂O), which is revealed in the cross-sectional EDS mapping (Figure 3).

Ultrasound is known to play an effective role in coating formation. For example, Lv et al. introduced ultrasonic vibration into the micro-arc oxidation process for fabricating Zn-modified TiO₂ coating. With the help of ultrasonic vibration, a uniform surface layer was successfully produced to enhance corrosion resistance, increase adhesive strength and improve biological performance [39]. In our study, it is speculated that the generation and bursting of H₂ will lead to the non-synchronous deposition of MgHPO₄∙3H₂O on the surface. Thus, the coating becomes rough enough to establish a necessary condition for obtaining the superhydrophobic behavior.

Figure 11c shows a schematic diagram explaining the superhydrophobic behavior of the coating. A suitable rough microstructure has been constructed on the surface of the Mg-Nd alloy. Usually, polysiloxane catching oxhydryl on the surface of materials can form an effective bonding and exposed perfluoroalkyl groups can reduce the surface free energy of materials to achieve the hydrophobic effect [40]. After the surface modification with PFDS, one of the low surface energy materials, the surface has acquired sufficient conditions for superhydrophobic behavior. The water static contact angle on MA/UC/CM is greater than 150° and the sliding angle is less than 10°, thus indicating that water droplets can easily roll off the surface once they contact the sample. Moreover, a superhydrophobic surface has a small fraction of the solid/liquid contact area, inducing a lot of air pockets trapped underneath the liquid [17,19,41,42]. Based on the nature of superhydrophobic behavior, the anti-corrosion mechanism of the coating in artificial hand sweat solution can be inferred (Figure 11d). If the surface is immersed in artificial hand sweat, air pockets will be formed among the rough structure to reduce the actual contact between the sample and the external medium. Consequently, the water-repellent behavior integrated with the barrier effect of the coating can prevent the corrosion of the substrate in artificial hand sweat significantly, based on the findings of this study.

In summary, a facile process has been successfully developed to form an anti-corrosion coating on an Mg-Nd alloy based on the concept of predesigned corrosion. This coating with chemical modification can not only improve the corrosion resistance but also exhibit a good self-cleaning effect. Nonetheless, this strategy needs to be refined to produce more effective techniques in the future for actual applications. For instance, the range of substrate materials should be expanded to other commercial alloys and the sustainability of the coating in artificial hand sweat solution should be further improved.

5. Conclusions

In this study, a superhydrophobic coating was successfully fabricated on a magnesium-neodymium alloy by ultrasonic-assisted immersion process and post chemical surface modification. The immersion process can generate a layer of rough but dense MgHPO₄∙3H₂O coating on the Mg-Nd alloy and post chemical surface modification is responsible for resulting in the super-hydrophobic behavior. The contact angle of the super-hydrophobic coating is 150.5 ± 0.6° and the sliding angle is about 4.5°, which endows the surface of the Mg-Nd alloy with an excellent self-cleaning effect. Both electrochemical measurement and immersion testing show that the Mg-Nd alloy samples with a super-hydrophobic surface have significantly improved corrosion resistance in artificial hand sweat compared to the samples without a super-hydrophobic effect. This study provides an effective strategy for corrosion mitigation of magnesium alloys.