The Influence of Water Content in Ethylene Glycol Electrolyte on Magnesium Plasma Electrolytic Fluorinated Coating

Abstract

Plasma electrolytic fluorination (PEF) of AZ31 magnesium alloy was carried out by adding different ratios of water to the ethylene glycol-ammonium fluoride electrolyte. The structural composition of the coatings was characterized using SEM, XRD, and EDS, and the effects of water content on the microstructure and corrosion resistance of the PEF coatings were analyzed. The results showed that the addition of water promoted the ionization of ammonium fluoride and increased the conductivity of the glycol electrolyte, which led to a decrease in the termination voltage. However, the coating thickness was not changed by the addition of water. The O element in water was not enough to compete with the F element in the electrolyte and had a small effect on the PEF coating composition, which was still dominated by MgF₂. The addition of water had an effect on the structure of the coating: with an increase in water content, the number of coating penetration holes decreases, and the continuity is enhanced. The pores on the surface of the coating tended to be levelled off and transitioned to the typical coating structure of PEO (plasma electrolytic oxidation). The addition of water to the glycol electrolyte was conducive to improving the corrosion resistance of the coatings. The corrosion resistance of PEF coatings in neutral NaCl corrosive medium firstly increased and then decreased, and the strongest corrosion resistance was obtained when the ratio of glycol and water is 6:4.

1. Introduction

Plasma electrolytic oxidation (PEO) is a surface modification technology that generates ceramic coatings in situ on metal surfaces by high-voltage discharge [1,2,3,4]. The principle is based on the micro-arc discharge of the metal substrate in the electrolyte, which induces a localized high-temperature and high-pressure environment, prompting the metal to react with the electrolyte components to form a dense oxide layer [5,6]. Compared with traditional anodic oxidation, the PEO process combines high deposition efficiency, excellent coating adhesion, and wear resistance, and is widely used in the field of magnesium alloy corrosion protection [7,8,9]. As the core reaction medium of PEO, the composition of the electrolyte directly affects the discharge behavior and coating composition. The ionic species, concentration, and pH of the electrolyte regulate the plasma discharge intensity and reaction path, which in turn determine the microstructure and properties of the coating [10,11,12,13]. Therefore, optimization of the electrolyte system is a key research direction for enhancing the functionality of PEO coatings.

Conventional PEO electrolytes predominantly exhibit alkaline characteristics, with the three primary systems comprising silicates, phosphates, and aluminates [14,15,16], though their mixtures are more frequently employed. Under aqueous conditions, the anodically formed PEO coating demonstrates chemical reactivity towards reactive oxygen species. Owing to this characteristic, magnesium oxide (MgO) consistently forms the predominant phase in the surface architecture of PEO-treated magnesium alloys, with this compositional stability remaining independent of variations in electrolyte composition [9,17].

However, due to the porous structure and easy hydration characteristics of the MgO coating, its corrosion resistance may not meet the engineering requirements [18,19]. Recent studies have shown that introducing fluoride salts (KF) [20,21] or zirconium salts (K₂ZrF₆) [22,23,24] into the electrolyte can promote the formation of highly stable phases such as magnesium fluoride (MgF₂) or zirconium oxide (ZrO₂) in the coating. Among them, MgF₂, due to its low solubility and dense structure, can significantly inhibit the penetration of Cl⁻ and enhance the durability of the magnesium matrix in corrosive media. Nevertheless, such additives can only take effect at high concentrations or under specific pH conditions, and excessive introduction may cause an imbalance in the electrolyte’s conductivity, resulting in a decrease in coating uniformity. Therefore, exploring new electrolyte systems to optimize the formation mechanism of fluoride coatings remains of great value.

To mitigate aqueous-phase-induced MgO formation, investigators have developed plasma electrolytic fluorination (PEF) methodologies utilizing non-aqueous media—typically organic solvents or ionic melts—combined with ammonium fluoride (NH₄F) [25,26,27]. These systems leverage thermal decomposition of NH₄F to supply fluoride ions, enabling oxygen-free synthesis of monolithic MgF₂ layers with superior anti-corrosion performance [28]. While fluoride-based ionic melts necessitate elevated temperatures (accelerating equipment degradation and energy demands), glycol-based systems enable ambient-temperature processing. Nevertheless, rapid NH₄F decomposition in glycol solvents causes electrolyte instability, coupled with hazardous hydrogen fluoride (HF) emissions during byproduct release. Furthermore, industrial scalability of such approaches is constrained by the elevated operational costs and restrictive parameter tolerances inherent to non-aqueous platforms. Consequently, formulating cost-effective, compositionally stable hybrid electrolytes represents a critical challenge requiring resolution in this research domain.

This investigation introduces regulated aqueous additions into glycol-ammonium fluoride systems to explore their modulation effects on magnesium alloy fluorination via plasma electrolytic processing. By precisely controlling the aqueous-to-glycol ratio, we hypothesize that water content regulates surface reactions through two counteractive mechanisms: controlled aqueous additions potentially enhance NH₄F dissociation kinetics to facilitate fluoride ion liberation, whereas elevated hydration levels may trigger competitive oxidation processes yielding magnesium oxide. Strategic balancing of these antagonistic pathways enables directional growth of MgF₂-dominant surface architectures while circumventing inherent limitations of anhydrous approaches. This methodology establishes a novel paradigm for developing energy-efficient magnesium fluorination technologies with mitigated operational hazards.

2. Materials and Methods

The substrate material was AZ31 magnesium alloy, size 30 mm × 20 mm × 6 mm, with the following composition (in mass fraction): Al 3.12%, Zn 1.04%, Mn 0.44%, Si 0.006%, Fe 0.001%, Cu 0.001%, and the balance of Mg. It was sanded by SiC sandpaper in a graded manner, polished to 2000 mesh, then washed in distilled water, ultrasonicated with ethanol solution, blow-dried, and set aside. The electrolyte consisted of ethylene glycol and ammonium fluoride (NH₄F), and different ratios of ethylene glycol and water were selected, with the mass fraction of ammonium fluoride in the electrolyte being 5 wt.%, and phosphoric acid being added at 5 g/L. There were six groups of ethylene glycol to water ratios, which were 10:0, 8:2, 6:4, 4:6, 2:8, and 0:10. The plasma electrolytic fluorination process was carried out using a 10 kW unipolar pulsed DC power supply. The temperature of the electrolyte was kept below 30 °C using a cooling water system while the electrolyte was stirred with a magnetic stirrer. The current density was maintained at 5 A/dm². The frequency and pulse width were 500 Hz and 400 μs, respectively. Based on the ratio of glycol to water in the electrolyte, the samples were denoted as 10E0H, 8E2H, 6E4H, 4E6H, 2E8H, and 0E10H, in that order.

The morphology and elemental composition of the PEF coatings were observed by a scanning electron microscope (JEOL, Tokyo, Japan) equipped with an energy dispersive X-ray spectrometer (EDS, JEOL, JSM-IT500A, Tokyo, Japan)—the microscopic surface of the coating was observed using the secondary electron mode, while its microscopic cross-section and elemental distribution (after being embedded in resin) were examined using the backscattered electron mode. The coating phase was determined by X-ray diffraction (Rigaku, Tokyo, Japan). Coating thickness was measured using an eddy current thickness gauge (FMP20, Fischer AG, Achern, Germany), and the average value was obtained from ten measurements taken on the coating surface. To determine the insulation resistance of the coatings, the electrical insulation properties of the coatings were measured using a DC withstand voltage/insulation resistance tester, with the range voltage range of the instrument set to 0–2000 V, and the average was obtained from ten measurements. For these tests, EIS was immersed in the corrosion solution for 30 min in the contact area of the magnesium alloy coating before entering. The parameters are as follows: the amplitude is 10 mV and the frequency range is 10,000 to 0.01 Hz. The corrosion protection properties of the coatings were tested using an electrochemical workstation (Metrohm Autolab PGSTAT302 N) by performing electrochemical tests in a 3.5 wt.% NaCl solution at room temperature. The coating contact area was 1 cm², and after 1 h of impregnation, the kinetic potential polarization curves were tested at a scan rate of 10 mV/s; each sample was tested three times to minimize errors, and corrosion current density and corrosion potential were fitted using Origin software (Origin 2018). The emission spectra of the discharge sparks during the PEF process were investigated with an optical emission spectrometer (Ideaooptics, Shanghai, China); background light interference was eliminated during testing, with the probe positioned parallel to and 3 cm away from the sample surface.

3. Results

3.1. The Influence of Water Content on the Voltage–Time of Magnesium Plasma Electrolytic Fluorination Process

The voltage versus time curves during plasma electrolytic fluorination with different water-to-glycol ratios are given in Figure 1, from which it can be seen that, similar to the aqueous PEO, the voltages went through three stages with time: the passivation stage, breakdown discharge stage, and plasma electrolytic fluorination stage. In the passivation stage (stage I), their voltage growth rate decreased with increasing water content due to the different conductivity of the electrolyte—the greater the conductivity, the smaller the resistance and the smaller the voltage. When the time reaches 40 s, the reaction is in the breakdown discharge stage (stage II)—the voltage growth in the PEF process is consistent with the first stage, but at this time, mild, small sparks appeared on the surface of the specimen. Finally, at 152 s, in the plasma electrolytic fluorination stage (stage III), the voltage of all the curves were approximately in the linear growth stage, at which time the coating grew rapidly and the growth rate accelerated. After 10 min of treatment, the termination voltages were 639 V, 575 V, 534 V, 493 V, 455 V, and 377 V as the water content increased, and the final voltages were reduced by the addition of different amounts of water during the treatment process, which had a significant effect on the final voltages.

3.2. Effect of Water Content on the Structure of Magnesium Plasma Electrolytic Fluorination Coatings

The thickness of the coatings formed in electrolytes with different water contents is shown in Figure 2. The coatings in the 10E0H group had an average thickness of 49.5 μm, while those in the 8E2H, 6E4H, 4E6H, 2E8H, and 0E10H groups measured 47.6 μm, 48.3 μm, 49.5 μm, 48.4 μm, and 44.2 μm, respectively. These results indicate that water content had a minimal effect on the thickness of the PEF coatings. However, the coating thickness reached its lowest value (44.2 μm) when the water content was 100% (0E10H group).

Figure 3 shows the surface morphology of the coatings for different groups. All coatings exhibited typical porous structures. As the water content increased, the continuity of the coatings improved, resulting in denser microstructures that approached those of conventional PEO coatings. The 10E0H coatings displayed pores of varying sizes, with larger pores attributed to intense discharge events, accompanied by scattered protruding particles on the surface. In contrast, the 0E10H coatings showed interconnected structures with significantly fewer pores. These pores were predominantly elliptical in shape, uniformly distributed, and formed under milder discharge conditions.

The surface roughness of coatings prepared in different electrolytes is presented in Figure 4. As shown in the figure, the surface roughness of the coatings decreased initially and then increased with increasing water content in the glycol-based electrolyte. Therefore, the water content in the electrolyte should not exceed 20%, which helps to reduce the roughness of the coating surface. The cross-sectional morphology of the coatings (Figure 5) reveals that variations in coating thickness were not significant. However, as the water content increased, the coating structure transitioned from discontinuous large-sized discharge channels to a continuous and dense morphology. This transformation enhanced the coating continuity and reduced the number of penetration holes. At 100% water content, the coating cross-section resembled that of a typical PEO coating.

3.3. The Influence of Water Content on the Composition of Magnesium Plasma Electrolytic Fluorinated Coatings

The EDS analysis of coating elements in electrolytes with varying water content is shown in Figure 6. As the water content in the glycol-based electrolyte increased, the elemental composition of the coatings did not change very much: The F content decreased slightly (from ~48.87 wt.% to ~45.29 wt.%), while the O content showed a minor increase (from ~9.54 wt.% to ~13.19 wt.%). Mg and Al elements originating from the AZ31 magnesium alloy substrate remained consistent, with stable concentrations of approximately 38 wt.% and 1 wt.%, respectively. Specifically, the anhydrous electrolyte-prepared 10E0H coatings contained 48.87 wt.% F and 9.54 wt.% O, whereas the 0E10H coatings (100% water content) exhibited 45.29 wt.% F and 13.19 wt.% O. The maximum variation in F and O content between these two extremes did not exceed 4 wt.%, confirming that Mg and F remained the dominant elements across all coating groups.

To further investigate the effect of water on the plasma electrolytic fluorination (PEF) coatings, EDS elemental distribution analysis was performed for all groups, with the results shown in Figure 7. The cross-sectional elemental maps revealed that Mg and F were uniformly distributed throughout the coatings and remained the dominant elements, regardless of water content. Oxygen, however, showed surface enrichment, likely caused by atmospheric oxidation of the coating. However, as can be seen from Figure 6, the content of elemental oxygen does not account for much of the variation.

The XRD analysis of the coatings formed in electrolytes with different water contents is presented in Figure 8. The diffraction patterns reveal that the coatings were predominantly composed of MgF₂, regardless of water content. However, the intensity of the MgF₂ peaks varied significantly: they increased initially and then decreased with increasing water content. This trend was attributed to differences in crystallization behavior during coating formation. Additionally, weak Mg substrate peaks were observed, resulting from X-ray penetration through the coating to the underlying substrate during analysis. Based on these results, the phase composition of the coatings was confirmed to be primarily MgF₂.

Although the EDS results showed a small amount of elemental oxygen (O) in the coatings prepared at different water contents—especially in the coating prepared in aqueous ammonium fluoride solution (0E10H), even as high as 13.19 wt.%—no oxides or oxygen-containing compounds were detected in the XRD analysis results. The reason for this may be that the oxygen (O) content in the coating itself was low, but it was involved in the formation of various compounds, which are present in the coating in an amorphous form. Therefore, no signals of oxides or oxygenated compounds appeared in the XRD results. It can be seen that at a certain concentration of ammonium fluoride in the electrolyte, the addition of water has little effect on the composition of the coating; there was no significant competition between elemental oxygen (O) from the water and elemental fluorine (F) from the main salt (NH₄F).

3.4. Spectral Analysis of Magnesium Plasma Discharge Process

The OES spectra of the PEF process at different voltages are shown in Figure 9, with spectral curves corresponding to 250, 300, and 350 V for the 10E0H group and 450, 500, and 550 V for the 0E10H group. The spectra were calibrated against the National Institute of Standards and Technology’s Atomic Spectral Database (NIST-ASD) [29]. Analysis revealed that spectral intensity increased with rising voltage, and the characteristic peaks became more pronounced at higher voltages. Post-calibration data confirmed the participation of elements from both the substrate (Mg) and electrolyte (F, N, H, C) in the discharge process. Specifically, the F emission line was initially observed at 292.31 nm, followed by N peaks at 287.08 nm and 365.67 nm. With increasing voltage, additional spectral lines corresponding to Mg, H, and C emerged sequentially, reflecting their progressive involvement in plasma reactions.

3.5. Effect of Water Content on the Performance of Magnesium Plasma Electrolytic Fluorination Coatings

Figure 10a,b show the Bode curves of the PEF coatings in different electrolytes in 3.5 wt.% NaCl solution. The value of Z_f→0 in the Bode plot can reflect the corrosion resistance of the coatings, and the larger the value of Z_f→0, the better the corrosion resistance of the coatings. It can be seen that at low frequencies, the 6E4H group has pitting corrosion and thus has no tendency to move to higher resistance values, while the 4E6H group is relatively stable and has no scattering. Therefore, it can be judged that 4E6H is more stable than 6E4H.

The electrochemical properties and specific structural features of the coatings were analyzed using an equivalent circuit model (Figure 10c). The corresponding fitting results are shown in

Table 1. Fitted electrochemical parameters obtained from Figure 10.

Specimens	CPE1-T	CPE1-P	R1 Ωcm2	CPE2-T	CPE2-P	R2 Ωcm2	L	R3 Ωcm2
10E0H	3.2 × 10−8	0.9	791	4.4 × 10−6	0.5	9444	8778	18765
8E2H	5.75 × 10−8	0.9	1235	1.96 × 10−6	0.6	10592	9157	12044
6E4H	5.44 × 10−7	0.7	9127	3.59 × 10−6	0.7	178380	117420	41705
4E6H	3.73 × 10−7	0.8	13619	4.4 × 10−6	0.7	156010	--	--
2E8H	3.92 × 10−7	0.8	10410	4.4 × 10−6	0.7	59420	--	--
0E10H	3.92 × 10−7	0.8	3242	4.4 × 10−6	0.4	52481	433820	27658

. In the fitted circuit model, R₁ is the outer layer resistance, R₂ is the inner layer resistance, CPE₁-T and CPE₂-T are constant phase angle components, R₃ is the inductive resistance, and L is the inductance. The presence of inductance (L) indicates the occurrence of localized corrosion in the magnesium alloy. From the table, it was found that the corresponding ground resistance values (R₁, R₂) of the specimens first increase and then decrease with the increase in water content. The highest R₂ value (178380 Ω cm²) was observed for the 6E4H coating after immersion in 3.5 wt.% NaCl solution for 1 h, which indicates that the optimum corrosion resistance of the coating is achieved when the ratio of glycol to water content is 6:4.

Figure 11 presents the potentiodynamic polarization curves of PEF coatings immersed in 3.5 wt.% NaCl solution for 1 h, with corresponding fitting data listed in

Table 2. Polarization curve parameters of PEF coatings in different electrolytes.

Specimens	Ecorr (mV vs. Ag/AgCl)	Icorr (A/cm 2)
10E0H	−1246	1.68 × 10−7
8E2H	−1447	8.63 × 10−8
6E4H	−1352	6.41 × 10−8
4E6H	−1447	7.52 × 10−8
2E8H	−1407	3.30 × 10−7
0E10H	−1475	1.00 × 10−6

. The 6E4H coating exhibited a corrosion potential (E_corr) of −1352 mV and a corrosion current density (i_corr) of 6.41 × 10⁻⁸ A/cm². As water content increased, E_corr values shifted negatively, indicating that water addition compromised the thermodynamic stability and corrosion resistance of the PEF coatings. Notably, the 6E4H coating’s significantly lower i_corr confirmed its superior corrosion resistance among all groups, and this phenomenon may be attributed to the lower surface roughness and fewer internal defects in the coating. Collectively, the electrochemical impedance spectroscopy (EIS) findings aligned with the polarization curve results, validating the consistency of the electrochemical analysis.

The breakdown voltages of coatings from different groups are presented in Figure 12. The 10E0H, 8E2H, 6E4H, 4E6H, 2E8H, and 0E10H coatings exhibited breakdown voltages of 985 V, 1088 V, 1272 V, 1372 V, 1408 V, and 1512 V, respectively. The breakdown voltage increased with water content, which is due to the difference in coating densities. Cross-sectional analysis revealed that higher water content enhanced coating continuity, reduced the number of large pores formed during intense discharge events, and improved overall coating densification. These structural changes led to increased breakdown strength.

4. Discussion

The conductivity of the electrolyte plays a critical role in the coating growth process. As shown in Figure 13, the conductivity of the glycol-based electrolyte increased significantly with higher water content. Compared to the pure glycol solution (conductivity: 5.1 mS/cm), the aqueous solution exhibited a conductivity of 45.1 mS/cm—approximately nine-fold higher. All glycol electrolytes contained 5 wt.% ammonium fluoride (NH₄F). The addition of water enhanced NH₄F ionization (NH₄F → NH₄⁺ + F⁻), which improved electrolyte conductivity by facilitating fluoride ion mobility. This enhanced ion transport also altered the voltage–time dynamics during film formation, directly linking conductivity to the PEF process kinetics.

According to the researchers’ analysis, the termination voltage is a key factor affecting the coating thickness, and the higher the termination voltage, the thicker the coating. Contrary to this analysis, it can be seen in Figure 5 that the coating thickness does not vary significantly with increasing termination voltage. It was stated that as the water content in the electrolyte increases, it promotes the degree of ionization of ammonium fluoride (NH₄F), and more fluoride ions move toward the anode, which causes a two-fold effect. On the one hand, the conductivity of the electrolyte increases sequentially as shown in Figure 13, and the resistance of the solution is inversely proportional to the conductivity, implying a decrease in the solution resistance. In the electrolyte-coating combination resistance, the lower the resistance of the electrolyte, the lower its partial voltage; the coating partial pressure is high, which means that most of the energy provided is allocated to the film-forming process, thereby promoting rapid coating growth. On the other hand, the XRD results (Figure 8) show that fluoride ions (F⁻) ionized from ammonium fluoride are the primary film-forming agents for coating formation. At the same mass fraction of ammonium fluoride, the higher the ionization degree, the higher the fluoride ion (F⁻) concentration, which facilitates rapid coating growth. Therefore, as the water content in the electrolyte increases, coatings with minimal thickness variation can be obtained at lower voltages.

In summary, the addition of water promoted the ionization of ammonium fluoride, and thus, the electrolyte was more conductive, resulting in a slower voltage increase without a significant change in the thickness of the coating obtained. In the NH₄F-EG non-aqueous electrolyte, the addition of water on the one hand made the ionization of ammonium fluoride more sufficient, and on the other hand, the hydroxide ions provided by the ionization of water provided the possibility of generating oxides for the coating. However, from the XRD and EDS results, the addition of water did not significantly oxidize the magnesium surface, the oxygen content did not change much, and the coating was dominated by magnesium fluoride. The molar Gibbs free energies of generation of magnesium fluoride (MgF₂) and magnesium oxide (MgO) are listed in

Table 3. Molar Gibbs generation free energies of magnesium fluorides and oxides.

Compound	MgF2	MgO
ΔfGmϴ (KJ/mol)	−1071.1	−569.3

, with parameter values of −1071.1 KJ/mol and −568.9 KJ/mol, respectively, which suggests that magnesium fluoride is more thermodynamically stable than magnesium oxide and is more likely to be generated. Therefore, although the addition of water provided the element oxygen, the competitive ability of the element oxygen was much lower than that of the element fluorine, and the reaction was more likely to produce stable magnesium fluoride than unstable magnesium oxide in this process. Although the addition of water changed the electrolyte properties (conductivity), there was very little change in the elemental fluoride and oxygen in the coating, which was still dominated by magnesium fluoride.

5. Conclusions

• Water had a facilitating effect on the growth of coatings in the PEF process. The addition of water promoted the ionization of ammonium fluoride in the glycol electrolyte and increased the conductivity of the electrolyte, which led to a lower termination voltage. Coatings of comparable thickness could be obtained at lower termination voltages.
• The O element in the water was not enough to compete with the F element in the electrolyte and had little effect on the composition of the PEF coatings, which were still dominated by MgF₂.
• The addition of water had an effect on the structure of the coating. With the increase in water content, the number of holes in the coating decreased and the continuity was enhanced. The pores on the surface of the coating tended to calm down and transitioned to the typical coating structure of PEO.
• The addition of water to the glycol electrolyte improved the corrosion resistance of the coatings. The corrosion resistance of the PEF coatings in neutral NaCl corrosive medium increased and then decreased, and the strongest corrosion resistance was found when the ratio of glycol to water was 4:6.