Corrosion protection performance of nano-TiO₂-containing phosphate coatings obtained by anodic electrochemical treatment

3.1 Electrodeposition and characterization of the phosphate layer

The electrodeposition process of phosphate coatings obtained by an anodic electrochemical treatment as a function of process variables’ current density and temperature was well described by Jegannathan et al. (2005). Therefore, in this work, only the mechanism of coating formation and relevant details are outlined for the sake of brevity. In the electrodeposition of phosphate coatings, LCS was used as the anode, and graphite rod as the cathode. Electrodeposition of the phosphate coating takes place in two stages: nucleation and growth. During this process, the potential of LCS was controlled continuously as a function of time by following the OCP. It was found that any normal phosphating process takes place in two steps: metal dissolution followed by phosphate deposition.

It was observed that impressed current is the key factor for the high rate of metal dissolution. As the first of its kind, potential is shifted toward the anodic region and attains a maximum value. Hydrogen evaluation can be seen at the graphite rods (cathodic sites). The increase in deposition time leads to a shift in the potential of substrate toward the cathodic region. Several literature reports (Banczek et al., 2006; Jegannathan et al., 2006a,b; Yantapure et al., 2015) revealed that the observed anodic shift at the start of deposition is due to the depletion of H⁺ ions in the electrode/electrolyte interface caused by repulsive forces of ferrous ions and the positive charge on the substrate. In this electrodeposition process, initially, hydrogen evaluation was shifted and a rise in interfacial pH is expected near the graphite rod (cathodic region) until it attains maximum value (Figure 1). At this point, rapid growth of phosphate coatings may be expected. Meanwhile, more dissolution of LCS was observed, probably due to the impressed anodic current, thus by generating greater amounts of ferrous ions. However, maximum potential suggests the conversion of primary soluble phosphate into tertiary insoluble phosphate, which makes phosphate coatings insoluble in nature.

Figure 1:

Variation in potential with respect to time during electrodeposition of nano-TiO₂ containing phosphate coating and normal phosphate on low carbon steel.

The potential significantly declines with time after shifting toward the cathodic sites until it attains stability (150 s). At this stage, the phosphate coatings formed were denser. This can be explained by displacement of protons away from the LCS/phosphating solution interfaces, and this reflects a slight increase in interfacial pH and reveals the anodic shift of the potential. This rise in interfacial pH has to assist the deposition of phosphate coating on LCS anode. This agrees well with the study reported by Jegannathan et al. (2005). Conversion of primary soluble phosphate into tertiary insoluble phosphate suggests the attainment of maximum potential.

The average weight of nano-TiO₂-incorporated phosphate layers obtained by electrodeposition and chemical deposition (Bagal et al., 2018) at various times of immersion is shown in Figure 2. The average weight of phosphate coatings obtained by electrochemical means is higher than that of its counterpart, chemically deposited coatings. This could be justified by the presence of the nano-TiO₂ trace as an additive in the phosphate layer. It is evident from Figure 2 that, for a given time, iron dissolution is of high order (approximately 40–55 g/m² per 10 min). However, phosphate coatings showed dissolution of low order (6–9 g/m²) for the same time. It is anticipated that the immersion time is required in order to characterize the coating stabilization. The relationship between coating weight and coating thickness is given elsewhere in the literature (Weng et al., 1996). Most of the phosphate coatings exhibit relation of 1-μm thickness to 1.5–2 g/m².

Figure 2:

Variation in phosphate coating weight at various electrodeposition time.

The reactions occurring at the metallic surface and the mechanism of formation of the phosphate coating are illustrated in Figure 3 and below:

Figure 3:

Mechanism of phosphate coating formation on metal substrate.

The first reaction reveals a metal attack by an acid in the phosphating solution, followed by the other two, which lead to the formation of phosphate crystals and deposition on metallic surface. The Zn₃(PO₄)₂.4H₂O (hopeite) is the main constituent in phosphate coating. The reaction 3 above advances continuously until the nucleation and growth of the coating, thus ensuring the maximum coverage of all micro anodic sites.

The XRD studies reveal an identification of various phases that would occur in the phosphate layer. The phases present in the phosphate layers were mainly identified as a trace of Zn₃(PO₄)₂.4H₂O (hopeite) and Zn₂FE(PO₄)₂.4H₂O (phosphophylite) (Figure 4). The XRD analysis identifies the peaks of hopeite and phosphophylite in the phosphate layer formed on LCS. The peaks of Fe are anticipated to contribute porosity in the phosphate coatings.

Figure 4:

X-ray diffraction (XRD) patterns of the zinc phosphate layer deposited on low carbon steel.

3.2 Morphology characterization of the phosphate layer

The surface morphology of the electrodeposited zinc phosphate coatings obtained without and with nano-TiO₂ at an optimized condition (current density 4 mA/cm², deposition time 60 min and 0.5 wt % nano-TiO₂) was studied to observe the growth of the phosphate coating on LCS. In both cases, formation of a well-crystallized coating was observed. It is worth recalling Machu’s hypothesis (Machu, 1943) of precipitation of metal phosphates; these crystals can be initiated on cathodic sites. The color of the coatings observed was dark gray for the normal electrodeposited phosphate and gray for nano-TiO₂-containing electrodeposited phosphate and has needle-like structures. The significant difference was traced in the surface morphology of these two coatings. Compared to the normal electrodeposited phosphate coating surface, the surface of nano-TiO₂-containing coatings was more homogeneous.

It may be expected that the number of pores was less in the case of the coatings containing nano-TiO₂ due to the sealing effect by nano-TiO₂ particles. It is possible that nano-TiO₂ particles play a crucial role in preventing the subsequent coatings from their inhabitant behavior. The incorporation of nano-TiO₂ into the phosphating bath caused structural refinement of the crystal and also helped to achieve maximum surface coverage (Figure 5). The SEM micrographs confirm that the results obtained by the EIS studies are in line with the previous studies reported by Zhang et al. (2008b).

Figure 5:

Representative images of scanning electron microscopy study for electrodeposited zinc phosphate coatings indicating (a) normal phosphate coatings, (b) incorporated with nano-TiO₂ (optimized condition: current density 4 mA/cm², deposition time 60 min and 0.5 wt % nano-TiO₂) and (c) Zn-rich powder.

The EDX signals of the electrodeposited normal zinc phosphate coatings and those obtained at the optimized condition are depicted in the chemical composition of the phosphate coatings (Figure 6). The presence of nano-TiO₂ was confirmed by above EDX studies. Further EDX analysis reveals a significant increase in Zn/P ratio from 3.58 to 3.80 for the coatings obtained using normal phosphating bath and nano-TiO₂-containing phosphating bath, which, in turn, led to the increase in the content of Zn₃(PO₄)₂·4H₂O.

Figure 6:

Energy dispersive spectroscopy (EDS) analysis of electrodeposited zinc phosphate coatings (a) normal phosphate coatings (b) phosphate coatings containing 0.5 g/L nano TiO₂ (at optimized condition).

3.3 Electrochemical characterization

The polarization curves corresponding to bare, chemically deposited nano-TiO₂ phosphate-coated and electrodeposited nano-TiO₂ phosphate-coated LCS are shown in Figure 7. Polarization methods are faster experimental techniques for the determination of corrosion rates than the conventional weight loss method. The coated samples were immersed in 3.5% NaCl electrolyte to obtain the Tafel plots. Here, in this work, we obtained a reasonably steady Tafel plots (both anodic and cathodic) after the stabilization of OCP. This technique permits measurement of accurate and precise corrosion rates (even corrosion rate <0.2 milli inches per year (mpy)). At the end, the framework calculates the corrosion rates by following formula:

Figure 7:

Representative polarization curves for bare low carbon steel, chemically deposited nano TiO₂ containing phosphate coating and nano TiO₂ containing electrodeposited phosphate coatings on low carbon steel in 3.5% NaCl electrolyte.

where icorr=1Rp×(βaβc)2.303 (βa + βc), E.W. is the equivalent weight and ρ is the equivalent density of the substrate, respectively. The values of polarization resistance (R_p) and anodic and cathodic Tafel constants (β_a and β_c, respectively) are obtained from Tafel extrapolation curves (framework). Potentiodynamic polarization assumes dissolution of coating in its own electrolyte. These corrosion rates were calculated when coatings were in an undamaged state. However, we followed an ASTM G59, G1, G4 and G31 standards for preparing and corrosion testing of samples.

It is evident from the polarization curves that anodic polarization reveals the corrosion protection offered by the coating, while cathodic polarization suggests the evaluation of hydrogen. The results showed that electrodeposited nano-TiO₂ phosphate-coated samples exhibit more corrosion resistance. The experimental findings reveal that the cathodic shift in E_corr and the increased magnitude of I_corr, essentially with increase in current density, make the coating more porous and discontinuous. This was observed during the experiments 8 and 9 (Table 4), as increased current density under the influence of anodic current caused a dissolution of coatings. However, cathodic reactions in polarization curves tend to show a diffusion-controlled mechanism. In comparison with bare and chemically deposited coatings, electrodeposited coatings show a decrease in corrosion current density (−8.445 mA/cm²), and corrosion potential is shifted toward the positive side. The addition of nano-TiO₂ in the phosphate bath is expected to enhance the corrosion protection performance of coatings by forming the more compact phosphate layer. The observed results are in agreement with the OCP and SEM measurements. Figure 7 shows the Tafel polarization curves for bare, chemically deposited nano-TiO₂ phosphate-coated and electrodeposited nano-TiO₂ phosphate-coated LCS. The corrosion rate of electrodeposited phosphate-coated samples were found to be in the range of 0.8 to 2 mpy (Table 5), which was approximately eight times less than that of bare LCS in 3.5% NaCl solution. The corrosion inhibition efficiency for each sample was evaluated using Equation 6:

where η_corr is the corrosion efficiency, ∂corro is the corrosion rate of the substrate and ∂corro is the corrosion rate of the coated substrate.

Corrosion rate is one of the output responses in this work. Corrosion rate obtained from the above potentiodynamic polarization study using the Tafel analysis was investigated for regression analysis through the MINITAB 15 software. Measured and predicted values were compared (Figure 8). The regression model proposed is given below:

Figure 8:

Variation in experimentally measured and predicted values of corrosion rate.

Corrosion rate=0.524+0.132 (current density)+0.0112 (deposition time) – 0.882 (wt % TiO₂)

Generally, it has been observed that phosphate coatings are porous in nature, which is a detrimental factor for corrosion rate. Hence, the presence of nano-TiO₂ particles helps in the reduction of porosity. In this work, coatings were electrodeposited by an anodic electrochemical treatment using current density and deposition time as major process parameters. It was observed from the above results that too high deposition time along with high current density results in unstable coatings, and such coatings are susceptible to damage. ANOVA was performed, which helps in assessing the contribution of each process variable and the chance of process improvement.

The EIS results obtained in the 3.5% NaCl solution are demonstrated in Figure 9 (Nyquist plots) and Figure 10 (Bode plots) and compared with the previous work of Jegannathan et al. (2006a,b). For bare LCS, only a time constant is shown. The Nyquist plot shown below reveals two capacitive arcs, suggesting two time constants: semicircle in the high-frequency region followed by diffusion tail in low-frequency region. The existence of the diffusion tail in the low-frequency region suggests that the phosphate coating formation is a diffusion-controlled process. However, we found that the obtained Nyquist curves are in line with the previous established work of Jegannathan et al. (2006a,b). Coatings with a current density of 4 mA/cm² showed high impedance values. The Bode plot at high frequency shows the high impedance value and interaction between the time constants. Figure 10 manifests the higher impedance values for the phosphate coatings on LCS obtained by the electrodeposition process and are compared with those obtained by Jegannathan et al. (2006a,b). This shows the pores resistance to electrolyte penetration. This Bode plots support the previous results indicating the superior corrosion resistance of the electrodeposited phosphate layer. The higher impedance values of nano-TiO₂ electrodeposited phosphate coatings reveals higher coating resistance. However, we observed that the Bode plots tend to show a single inflation point in the plot of Z_mod vs. frequency.

Figure 9:

Nyquist plots for (a) chemically deposited and electrodeposited nano TiO₂ phosphate coating on low carbon steel and (b) bare low carbon steel in 3.5% NaCl solution.

Figure 10:

Bode plots for bare low carbon steel, chemically deposited and electrodeposited nano TiO₂ phosphate coated low carbon steel and in 3.5% NaCl solution.

From the results, it was observed that the value of Z_mod decreases as the current density increased from 4 to 6 mA/cm² (Table 6). However, coating capacitance showed an increased trend as a function of deposition time. This eventually confirms that the increased current density leads to an increase in porosity in the coatings and that the coating may get damaged easily. Therefore, in the current work, being more porous, the electrodeposited nano-TiO₂ phosphate coating at a current density of 6 mA/cm² showed less corrosion resistance. Rise in interfacial pH can be an essential condition for deposition of phosphate coating on the steel substrate. The magnitude of coating weight is substantially less at high current density.

Further the coating resistance increased as the deposition time was increased from 20 to 60 min at the same current density. Increased deposition time yields increased coating thickness (Sankara Narayanan, 2005; Yantapure et al., 2015). The Gamry framework instrument allows electrochemical behavior to be modeled using constant phase element (CPE) instead of pure capacitance. So, to study the corrosion behavior of obtained electrodeposited nano-TiO₂ phosphate coatings on LCS, an equivalent electrical circuit is drawn. Figure 11 shows equivalent circuit to demonstrate the impedance behavior of all samples. Here, in this study, we assumed that proposed model/circuit accounts for coating-metal/electrolyte interface. Our proposed model closely resembles the one used by (Meszaros et al., 1989; Weng et al., 1996). R₁ is the ohmic resistance of the solution between the working and reference electrodes. This aforementioned resistance is strongly influenced by the solution concentration and sample area. R₂ is the electrolytic resistance through the pores and parallel to coating capacitance (C_c). The charge transfer resistance R₃ is parallel to double layer capacitance (C_dl) at the coating/metal interface (CPE). The purpose of including resistance and capacitance in the model is to encounter the changes taking place at the metal/electrolyte interface. The electrochemical behavior of the phosphate layers on zinc-coated LCS obtained using EIS corresponds with the previous literature (Weng et al., 1996; Wolynec, 2003). The double layer capacitance (C_dl) was calculated by obtaining the best curve fit to EIS curves using the Kaleida 4.0 software. The R² value 0.9674 indicates goodness of fit and less variability of fitted data around the mean. As the first kind, the double layer capacitance (C_dl) showed an increasing trend with increase in current density from 4 to 6 mA/cm².

Figure 11:

Equivalent circuit proposed for the impedance behaviour of nano TiO₂ containing phosphate coated samples. A detailed explanation is given in the text.

It should be noted that corrosion protection performance of the phosphate coating is dependent on the degree of insulation of these coatings. When these phosphate-coated steel are immersed in 3.5% electrolyte, it is liable to form an electric circuit from the cathodic to the anodic site. The corrosion rate of these phosphate-coated samples depends on the resistivity of the circuit. The electrochemical corrosion reaction will occur at the bottom of the pores within the coating due to its insulating nature. It is anticipated that the increase in corrosion products in the occluded zone is mainly due to the diffusion resistance between the inside and the outside of the pores. As a result, the dissolution of the phosphate layer will start with a decrease in pH value of the bath. This suggests that corrosion events occurring in phosphate-coated steel are diffusion-controlled processes.

The regression model for coating resistance was proposed. It was found that results obtained experimentally are in agreement with results predicted by the regression model. Measured and predicted values were compared (Figure 12). The regression model proposed is given below:

Figure 12:

Experimentally measured and predicted values of coating resistance.

Coating resistance=3.42 – 0.119 (current density)+0.141 (deposition time) – 9.78 (wt % TiO₂)

From the ANOVA analysis of the responses coating resistance and corrosion rate, it was found that the deposition time influences the phosphate coating process by 60%, followed by current density 18.8% and wt % TiO₂ 2.48%. It is expected from the above exercise that increased deposition time increases the coating weight by increasing coating thickness. However, a too high deposition time makes the coating to lower down and spill off easily because of attainment of the equilibrium state, which, in turn, causes the coating to be a less stable.

3.4 Comparative study of coatings deposited chemically and electrochemically

In this section, comparison of phosphate coatings deposited by chemical and electrochemical means is discussed. Table 7 shows the potentiodynamic polarization and EIS results for these coatings are comparable. It can be inferred from the table that the electrodeposited zinc phosphate coating using 0.5 g/l of nano-TiO₂ reflects the formation of thick coatings.

The corrosion rate of nano-TiO₂ containing the electrodeposited phosphate-coated sample was found to be 1.38 mpy, while it was 3.5 mpy for the nano-TiO₂-containing phosphate-coated sample obtained by chemical deposition and can be attributed to reduction in the porosity of electrodeposited coatings. The existence of open pores is the major problem in phosphate coatings and, thus, affects the coating/substrate interface. Since the corrosion reaction intends to occur mostly at the coating/substrate interface, the determination of coating porosity becomes extremely important. As the first of its kind, this current campaign investigates the role of nano-TiO₂ particles in coating porosity. Thus, the performance evaluation of these coatings can be carried by electrochemical measurements. The coating porosity can be estimated using Equation 7 (Aromaa et al., 1991; Creus et al., 2000; Vanessa de Freitas et al., 2006).

where R_ps is the polarization resistance of the sample, R_p is the polarization resistance of the coated sample obtained from the EIS study, ΔE_corr is the potential difference between the sample and the coated sample and β_a is the anodic Tafel constant obtained from the potentiodynamic polarization study. As the first of its endeavor, the porosity of nano-TiO₂-containing electrodeposited-coated samples (~0.10%) was found to be lower than that of chemically deposited nano-TiO₂-containing phosphate-coated samples (0.32%). The optimization studies revealed that the deposition time leads to the formation of thick coatings. However, it is worth mentioning here the significant observation made by Jegannathan et al. (2006a,b), that too much deposition time will make the coating unstable and breakdown could happen more easily. The authors truly believe that the obtained results are in agreement with the experimental observation of Shibli and Chacko (2011). The EIS results revealed a low coating capacitance in all the electrodeposited nano-TiO₂ phosphate-coated samples. This will further oppose the penetration of external agencies (i.e. electrolyte or water), which minimizes the dissolution of zinc, thus increasing the corrosion resistance.

3.5 Multiobjective optimization of coating process parameters

The main motto of this work is to achieve maximum coating resistance and minimum corrosion rate for the optimal sets of input controlled process variables. However, coating resistance and corrosion rates are opposite output responses. So, multiobjective optimization helps in this regard. It uses the Taguchi signal-to-noise ratio to predict the rank of the input controlled process parameters, which contributes to the coating process significantly. This signal-to-noise ratio is of importance in this Taguchi method.

As signal should always be maximum and noise should be minimum, provision was made such that the signal-to-noise ratio should always be larger. So, selecting a larger better signal-to-noise ratio (Figure 13), the optimal setting of the process variables can be proposed (Table 8).

Figure 13:

Signal to noise ratio for optimal setting of process variables.

Current density of 4 mA/cm² with deposition time of 60 min and 0.5 wt % of nano-TiO₂ gave the maximum coating resistance and minimum corrosion rate. So, to confirm these results, a confirmation test was taken, and both EIS and potential polarization studies were carried out. It was found that confirmation test results are in agreement with previous results.

In this work, an attempt has been made to propose optimal settings of input process parameters for the evaluation of effective corrosion protection performance of zinc phosphate coatings electrodeposited by anodic electrochemical treatment. Use of the Taguchi method was found to be very significant in performing the experiments and to achieve the maximum coating resistance (10.21 kΩ) and the minimum corrosion rate (1.38 mpy). The optimal settings of input process variables for the aforementioned output response could be proposed at respective levels of current density (4 mA/cm²), deposition time (60 min) and 0.5 wt % of nano-TiO₂. Confirmation test and ANOVA analysis of the same set of input process variables were carried out to verify the accuracy and correctness of regression. The coating resistance and corrosion rate of the optimized sample in the confirmation run was found to be ~9.75 kΩ and 1.28 mpy, respectively, which again confirm that the results were in agreement. Further, the improvement in the signal-to-noise ratio was observed from 3.605 to 4.125 (Table 8), which is approximately 12.60% of the original.