Mechanical, tribological and anti-corrosive properties of polyaniline/graphene coated Mg-9Li-7Al-1Sn and Mg-9Li-5Al-3Sn-1Zn alloys

1. Introduction

Ultralight-weight (1.35-1.65 g/cm³), high specific strength, and excellent formability of Mg-Li based alloys offers them extensive application in aerospace, weapons, automotive and electronics industries [1,2]. The duplex Mg-Li alloys with Li percentage between 5-11 wt% show super plasticity behaviour [2,3]. Mg-Li alloys, i.e. Mg-9Li-7Al-1Sn (LAT971) and Mg-9Li-5Al-3Sn-1Zn (LATZ9531) showed superior mechanical- (hardness $\widetilde{1}$.5 GPa and modulus of elasticity $\widetilde{6}$0 GPa) and wear-properties (coefficient of friction, COF of 0.50-0.60) as compared to that of conventional Mg-3Al-1Zn (AZ31) alloy (elastic modulus/hardness of 36.8 and COF of 0.69) [4,5]. However, these alloys cannot be widely used due to their poor corrosion resistance (corrosion current density, I_corr = 0.82 mA/cm² and 0.34 mA/cm² for LAT971 and LATZ9531 alloy, respectively) in 0.6MNaCl solution as compared to that of conventional AZ31 alloy (0.054 mA/cm²) [6,7]. The presence of Li enhances the corrosion rate of the alloys due to its higher chemical activity (E =-3.04 V vs. SHE) than that of Mg (E_o of ̶ 2.36 vs. SHE). Primarily Song et al. reported that Mg-Li based alloys shows the higher corrosion rate than that of pure Mg by forming a poorly bonded inhomogeneous oxide and hydroxide corrosion product of Mg and Li on the alloy's surface [3]. However, recently, Li et al. observed that the Mg-Li alloys with lower Li contents (4 wt% and 7 wt%) show a higher corrosion rate that of higher Li content (14 wt%) [8]. It was proposed that the formation of more compact layer Li₂CO₃ on Mg-14Li provides corrosion protection when compared with that of lower Li contents alloys [9].

Researchers have used two approaches for improving the corrosion resistance of Mg based alloys: (i) first by alloying addition, severe plastic deformation and subsequent heat treatment that promotes the formation of stable precipitates [10], and (ii) second by utilising various surface modification techniques to improve the oxidation resistance of the Mg-based alloys including conversion film [11,12], anodizing [12,13], plasma electrolytic oxidation [14], hydrothermal [15], micro arc oxidation [12,16], thermal spray [17], chemical vapour deposition [18], electrodeposition [19], and depositing electroless- [12] and organic-coatings [20,21]. Recently, there has been a keen interest to enhance the coating efficiency on Mg alloys for economical- and ecological-reasons. In this regards, the polymeric coatings are one of the best substitute to the metal and ceramic based coatings [22]. There are few reports available on the protection of Mg-based alloys from the corrosion by using polymer based organic coatings [1,23,24].

In recent years, conducting polymers such as polyaniline (PANI), polypyrrole (PPy) and their derivatives has been gained attention due to their applications in sensors, batteries, electronic devices and the corrosion prevention of metallic components [25,26]. Among all available conducting polymers, PANI is one of the most promising polymers for the corrosion protection due to its redox ability [27], optical activity and environmental stability etc. [28]. Furthermore, PANI can be synthesized very easily by the oxidation polymerization via chemical or electrochemical route [28]. PANI based coatings can be directly applied on the substrate by electrochemical deposition or in the combination with the commercial organic coatings (epoxy, polyester, acrylic, polyurethane) in form of pigment [28]. It is mentioned that the pure PANI show a poor adhesion with the substrates, thus it needs to mix with the conventional resins for better adhesion [29]. It is reported [30] that addition of secondary component in form of nanoparticle increases its functionality and enhances the mechanical, electrical and structural performance [30,31]. Nanoparticles can be metallic (Ag, Au, Pt, Cu), metal oxide (TiO₂, SiO₂, Fe₃O₄) and carbon compound such as carbon nanotube (CNT), graphite and graphene (Gr) etc. [30]. These nanoparticles interact with the polymer via weak columbic and vender wall forces rather than forming any coordination bond. However, such interaction improves the electronic interactions and charge transfers.

PANI based composite coatings have been used for the corrosion protection of steel [22,27], Al alloys [32,33] and Mg-based alloys [23,24,[34], [35], [36]]. Sathiyanarayanan et al. [36] applied paint coating containing PANI pigment in the protection of ZM21 Mg alloy and the composite coating was capable to protect the underline substrate even after the 75 days exposure in NaCl solution. Sathiyanarayanan et al. [34] also used PANI - TiO₂ composite pigment in the acrylic resin based paint for the corrosion protection coating on ZM21 Mg alloy and observed that the composite coating was more protective than that of pure PANI coating [34]. However, only a few scientific reports are available on the corrosion protection of Mg-Li based alloys via protective PANI coating. Shao et al. [1] used epoxy coating containing PANI pigment to improve the corrosion resistance of Mg-5Li alloy with the charge transfer resistance of the alloys enhanced up to $\widetilde{1}$0⁶ Ω cm² as compared to that of uncoated one ($\widetilde{1}$0² Ω cm²). Chen et al. [37] used PANI/SiO₂ composite coating to protect Mg-11.5Li-1.5Al-0.7(La, Pr, Ce) alloy and observed that PANI based coating showed an enhanced corrosion protection with lower I_corr value (4.01 × 10^-10 A/cm²) than that of pure epoxy coating (2.19 × 10^-7 A/cm²) and uncoated alloy (1.383 × 10^-4 A/cm²) [37]. Therefore, the composite coatings were found to be more promising for corrosion protection of the epoxy-based coatings.

Recently, study on the Gr/PANI composite coatings [22,27,30,38] has gained much attention due to its good mechanical integrity [39], electrorheological property [40], and barrier performance [41]. It is reported that the lower density and higher aspect ratio of the conducting Gr sheet make it an advanced gas barrier polymeric coating than that of the other non-conducting additive [38]. Gr, helps not only in the electrical conduction but also offers a large surface area used for the dispersion of the PANI nanoparticles [42]. Firstly, Chang et al. [38] used the PANI/Gr coating to improve the corrosion resistance of steel and found it to be an excellent barrier to corrosive environment (O₂ and H₂O) as compared to that of pure PANI and clay (insulator)/PANI coating. Later on, researchers have used the PANI/Gr coating to enhance the corrosion resistance of various alloys [22,27,38]. But, there is no report on the corrosion protection of Mg or Mg-Li based alloys via Gr/ PANI composite coating.

Targeted Mg-Li alloys, i.e. LAT971 and LATZ9531, possess potential application in the field of aerospace and automotive, where superior performance is required along with the higher corrosion resistance. Thus, the PANI/Gr based coatings on alloys surface may provide superior mechanical and tribological properties due to the high strength and modulus of Gr particle and higher corrosion resistance (via redox ability of PANI particle and barrier ability of Gr particle). In the present work, PANI /Gr composite pigments were synthesised by using in-situ oxidative polymerization. PANI/Gr pigment containing epoxy coatings were applied on the LAT971 and LATZ9531 alloys. Morphology of the coating was analysed using scanning electron microscopy (SEM). Mechanical and tribological properties of the coated alloys were evaluated using nanoindentation and micro scratching, respectively. The corrosion performance of the coated alloys has been examined in 3.5 wt% NaCl electrolyte via performing potentiodynamic polarization test and using electrochemical impedance spectroscopy.

2. Materials and method

2.1. Synthesis of PANI/Gr pigment and preparation of epoxy coating containing PANI/Gr pigment

PANI and Gr nanosheet (referred to as Gr) composite pigment was synthesized via in-situ polymerization. The mechanism of PANI/Gr composite pigment formation is schematically shown in Fig. 1. First, 0.4 mg Gr (particle size: <15 μm, thickness: $\widetilde{1}$5 nm and surface area: 50-80 m²/g procured from XG Science, Inc.US), 1 ml aniline monomers (molecular weight: 93.13 g/mol, procured from LOBA Chemie, India) and 95% pure ethyl alcohol (10 ml) was added into the HCl solution (1 M). As-prepared mixture was sonicated for 1 h and then cooled from room temperature ($\widetilde{3}$3 °C) to 0 °C with the use of ice water bath. Afterwards, the polymerization is initiated by the addition of the solution (aqueous ammonium persulphate, APS of 2.5 g dissolved in 40 ml of HCl) drop by drop into the mixture kept in stirred ice water bath. Lastly, as-prepared PANI/Gr composite pigment was washed using distilled water and ethanol repeatedly and then the cleaned pigments were dried in an oven at 50 °C for 12 h.

Pure PANI pigments were also prepared, using the above-mentioned process without addition of Gr particles for comparison. Dispersion of PANI and Gr in synthesized pigments were analyzed using transmission electron microscopy (TEM, Model: FEIUT 20, accelerating voltage: 200 kV).

The coatings were prepared on the as-rolled LAT971 and LATZ9531 alloys of dimension 15 mm × 15 mm × 2 mm. Prior to the coating, the alloys were polished and cleaned ultrasonically in acetone. As synthesized PANI/Gr powders were crushed using a mortar and then used as pigment (10%), epoxy resin, polyamide used as binder and a curing agent, respectively in the ratio of 10:1. The coating was cured at 60 °C for 12 h in air environment. Surface morphology of coated alloys was analyzed using scanning electron microscopy (SEM, Model: Zeiss EVO50). The alloys coated with pure epoxy were referred to as LAT971_Epoxy and LATZ9531_Epoxy for LAT971 and LATZ9531 alloys, respectively. Whereas, alloys coated with PANI/Gr pigment containing epoxy were referred to as LAT971_PANI/Gr and LATZ9531_PANI/Gr for LAT971 and LATZ9531 alloys, respectively.

2.2. Phase analysis

The constituent phases of PANI/Gr composite pigments were analyzed by X-ray diffraction (XRD, Model: Bruker's D8 Focus, Panalytical Empyrean) using Cu-Kα radiation (wavelength, λ: 1.5406 Å). The XRD was performed using scanning rate of 0.12° per second and a step size of 0.026° at 30 kV and 15 mA. The formation of PANI pigment and the chemical grafting of Gr in PANI matrix is confirmed using Fourier transform infrared (FT-IR, Model: Bruker, vortex 70) spectroscopy. The FT-IR spectra were obtained in the range of 2000-500 cm^-1 using KBr pellets.

Micro- Raman spectroscopy (Renishaw Reflex Raman Microscope) was used to characterize the Gr powder and synthesized PANI and PANI/Gr composite pigments. Raman analysis was done using monochromatic radiation via an Ar ion laser source having the wavelength of 514 nm.

2.3. Thermogravimetric analysis

The thermal degradation of synthesized pigments was studied using thermogravimetric analysis (TGA, Model: STA-8000, Perkin Elmer, USA). The sample was placed in an Al₂O₃ crucible and analyzed under N₂ environment in the temperature range from 25 °C to 700 °C at the heating rate of 5 °C min^-1.

2.4. Nanoindentation test

Mechanical properties of coated alloys were obtained using nanoindentation test (Model: Hysitron^® TI 750-D Ubi-1) enabled with a diamond Berkovich indenter (tip radius = $\widetilde{1}$50 nm; elastic modulus (E_i) = 1141 GPa; and Poisson's ratio (ν_i) = 0.07 [17]). Prior to testing, instrument was calibrated by using a standard sample of fused silica (elastic modulus = 69.6 GPa and hardness = 9.25 GPa). For the repeatability, around 20 indents were made on each coated alloy with a maximum load of 500 μN at the loading rate of 100 μN/s and holding of 2 s at the peak load. Oliver and Pharr method was utilized to calculate the elastic modulus and hardness of the coating [43].

2.5. Micro-scratch test

The scratch resistance of the neat epoxy and PANI/Gr reinforced epoxy coated, alloys was examined by using micro-scratch tester (MHT, Model: CSM Instrument, Switzerland). A Rockwell diamond indenter of 100 μm tip radius was used to scratch the surfaces. A progressive load from 30 to 3000 mN was applied, with the speed of 0.5 mm/min on 1 mm long track. During the progressive load the simultaneous friction force and depth of the penetration were recorded. For each coated alloy, three scratch tests were performed to measure the average critical load for delamination. The plowing hardness H_p, which is the measure of parallel deformation along the scratch direction, was calculated using following equation [44]:

H_p=$\frac{8μP}{wd}$ (1)

The scratch hardness, H_s is the normal deformation along the scratch direction was calculated using following equation [44]:

H_s=$\frac{8P}{πw^2}$ (2)

where, μ is COF, P is the applied normal load, w is the scratch width and d is the penetration depth recorded during the scratch test. The scratch width, w was also calculated theoretically by using the following relation [44]:

w_th=$2\sqrt{2Rd-d^2}$ (3)

The wear volume (W_v) was also calculated by using following equation [45]:

W_v=L[$R^2arcos(1-\frac{d}{R})-(R-h)\sqrt{2Rh-d^2}$] (4)

where, L is scratch length, R is indenter radius and d is penetration depth. Further Archard, equation was utilised to estimate the specific wear rate (W_sp) according to subsequent relation [46]:

W_sp=$\frac{V}{PL}$ (5)

2.6. Electrochemical testing

The electrochemical cell used in this experiment was a flat bottom vessel of approximately 270 ml in total volume. The coated alloys of size 1 cm² were exposed to the electrolyte. The counter electrode (CE) consisted of platinum (Pt) mesh 3 cm × 3 cm in size, welded to the Pt wire inserted through another orifice on the top of the cell, and positioned on the side of the cell. A PARSTAT 2263 potentiostat, (Model: Princeton Applied Research, USA) was used to control potential in the present study. Electrochemical studies were performed under sea water (3.5% NaCl) condition to examine the corrosion behaviour of coated alloys. The open circuit potential (OCP) of the coated alloys was measured via potential (V) verses time (h) plot. After stabilizing the OCP of coated alloys for 1 h, electrochemical impedance spectroscopy (EIS) analysis was performed from 100 kHz down to 10 mHz frequency range at OCP with the sinusoidal perturbation amplitude of 10 mV. The potentiodynamic polarization (PP) test was performed in the potential range of ± 250 mV from the OCP at a scan rate of 0.1660 mV/s. Tafel extrapolation method was utilised to calculate corrosion current density (I_corr) and corrosion potential (E_corr)

The instantaneous corrosion rate, P_i (mm/y) was measured by using I_corr, according to the following relation [47,48]:

P_i=22.85×I_corr (6)

Corrosion protection efficiency, η was calculated from the I_corr value for pure epoxy and epoxy containing PANI/Gr pigments using the following equation [49]:

η=(1-$\frac{I_{corr}^c}{I_{corr}^o}$)×100 (7)

where, $I_{corr}^o$ and $I_{corr}^c$ are the corrosion current density of the neat epoxy coating and PANI/Gr containing epoxy coating, respectively.

3. Results and discussion

3.1. Microstructural analysis of pigments

During polymerization, PANI particles form agglomerates in worm-like structure of150 nm in length and $\widetilde{5}$0 nm in width (Fig. 2(a)). When the monomers of aniline and Gr are added into the HCl solution, the aniline monomers bond to the surfaces of Gr sheets due to the electrostatic attraction [50]. The Gr sheet and aniline act as an electron acceptor and electron donor, respectively and form a weak charge transfer compound [50]. Fig. 2(b) shows the TEM image of Gr sheet, which provides nucleation sites for the formation of PANI particle. The Gr sheets coated with homogeneous PANI particles is shown in Fig. 2(c).

3.2. Phase analysis of pigments

Fig. 3 shows the FT-IR spectra of pure PANI, Gr and PANI/Gr composite pigments. For pure PANI particles, the bands at 1560, 1487, 1301 and 1137 cm^-1 correspond to the quinoid ring with C=C stretching, benzenoid ring with C=C stretching, benzenoid unit with C-N stretching and quinoid unit with C-N stretching, respectively [42]. The band 1566 cm^-1 in the spectra of Gr is corresponds to the aromatic C=C stretching vibrations [42]. Conversely, the FT-IR spectrum of PANI/Gr composite powders shows the same vibrational bands of PANI, suggesting that the PANI is positively covered on the Gr surface.

Fig. 3(b) shows the Raman spectra of the pure PANI, pure Gr and PANI/Gr composite powders. The Raman spectra of pure PANI show the band at 1172, 1331, 1503 and 1600 cm^-1 corresponding to C–H bending of the quinoid ring, C-N stretching of the bipolaron structure, N–H deformation vibration associated with the semiquinoid structure and C–C stretching of the benzene ring [[51], [52], [53]]. For pure Gr, well documented D, G and 2D band were observed corresponding to 1345, 1580 and 2688 cm^-1, respectively. In case of PANI/Gr, the G (1597 cm^-1) and D (1344 cm^-1) band of graphene were broadened and merged together along with the reduced 2D band intensity. The broadening and merging of the peak may occur due to the interaction or coupling between Gr and PANI particle due to strong long-range π-π and electrostatic interactions [51]. The I_D/I_G ratio was observed to decrease (from 1.24 in Gr) to 1.16 for the PANI/Gr composite, suggesting the exfoliation of pristine. Graphene exfoliation [54] may occur during the in-situ polymerization process which involves continuous stirring of the aniline monomer and Gr particle solution for 24 h. The shifting of the 2D in the PANI/Gr to the higher wavenumber (2708 cm^-1) when compared to that of pure Gr (2688 cm^-1) also suggests an increase in the graphene layer thickness or exfoliation of Gr structure.

The XRD patterns of the pure PANI pigment, Gr and PANI/Gr composite pigment are shown in Fig. 4. For Gr powder, the diffraction at 2θ of 26.5° and 43.34° are ascribed to the Gr structure, which correspond to (002) and (100) diffraction planes, respectively. The intense peak at 2θ = 25.2° corresponds to (200) plane of emeraldine salt form of PANI [50,55]. Crystalline peak similar to pure PANI was observed for PANI/Gr composite pigment, which reveals that there was no additional phase formation in the composite pigments. No distinct peak of Gr was observed in XRD pattern of PANI/Gr pigment, which reveals that Gr particles may have been completely covered by PANI particles.

3.3. Thermal gravimetric analysis (TGA) of pigments

Fig. 5 shows the thermal behaviour of Gr, synthesized PANI pigment, and PANI/Gr composite pigment. The TGA plot of Gr shows a $\widetilde{2}$0% of weight loss at a maximum temperature of 550 °C (Fig. 5(a)). The increase in the weight loss of Gr with increase in the temperature corresponds to pyrolysis of carbon [56]. PANI and PANI/Gr pigments show two distinct stages of weight loss. For pure PANI, the first stage is observed with 2%-15% weight loss ranging from 60 °C to 300 °C, due to the vaporization of moisture and liquid residue in the powder. The major amount of weight loss ($\widetilde{8}$0%) was observed at above 310 °C, which is mainly ascribed to the degradation of PANI particles [57]. A small blip (of decrease in temperature) may be noticed for PANI at $\widetilde{3}$50 °C, with no change in the weight. Herein, the change in the weight of PANI is plotted against the actual sample temperature (rather than the programmed temperature). The spike at 350 °C shows a decrease in the sample temperature, which may be attributed to the formation of some stable by product (see compared 'heat-flow' pattern, Fig. 5(b)) after the degradation of polymeric phase. This blip may not be observed, if we will plot the weight change of PANI against the programmed temperature. However, it can be noticed that the degradation temperature for PANI/Gr composites has increased, when compared to that of neat PANI, which shows that the degradation mechanisms (i.e. two distinct weight loss stages of pure PANI) may not have changed even after the addition of Gr particles. However, an increase in the thermal degradation temperature was observed for PANI/Gr pigment due the higher thermal stability of Gr particles. Thus, the strong interaction between the PANI and Gr filler results in an enhanced thermal stability of PANI/Gr composite pigments than that of pure PANI.

3.4. Surface morphology of coating

The SEM micrographs of neat epoxy and PANI/Gr pigment containing epoxy-coated alloys are given in Fig. 6. Neat epoxy coated LAT971 (Fig. 6(a)) and LATZ9531 (Fig. 6(b)) shows micro-pores, which generated due to the air gap in the coating. After the addition of PANI/Gr pigments in the epoxy matrix, the coating shows the dense and smooth surfaces for both LAT971 (Fig. 6(c)) and LATZ9531 (Fig. 6(d)) alloys as the uniform distribution of pigments vanishes air gaps in the coating. Gupta et al. have also observed a similar type of air bubbles in the neat epoxy coating on Al alloys substrate [32]. It is observed that the air bubbles in the epoxy coating disappeared as the concentration of the PANI-lignosulfonate pigment increased from 1 wt% to 20 wt% in the epoxy matrix [32].

Air pockets in the epoxy coating are more often formed by a process called outgassing, which occurs during curing process and the entrapped air creates these bubbles (Fig. 6(a)-(b)). The air pockets in the layer of epoxy form because of the high viscosity of the epoxy entrapping the outgoing gases formed during the curing process. These air pockets could have been eliminated by curing in vacuum. Whereas, in case of PANI/Gr composite coating, PANI and Gr particles have prevented the entrapment of outgassing during the curing process that has resulted in dense coating (Fig. 6(c)-(d)) even when cured under atmospheric pressure.

3.5. Mechanical behaviour of coatings

The characterization of mechanical properties is very essential to explore the effectiveness of PANI/Gr pigment reinforcement in addition to epoxy matrix. Fig. 7 shows the typical load vs. indentation depth curves obtained for all coated Mg-Li alloys. By adding PANI/Gr pigment in epoxy matrix, load verses displacement curves shifted towards left and, thus, the maximum penetration depth was reduced due to increased hardness as compared to that of neat epoxy coated alloys. The elastic modulus and hardness values of pure epoxy coated and PANI/Gr containing epoxycoated alloys is summarised in Table 1, which shows that the PANI/Gr pigment containing coating exhibited a high elastic modulus (7.05-7.51 GPa) and hardness (0.63-0.68 GPa) when compared to that of neat epoxy coated alloys. For instance, the hardness and elastic modulus of PANI/Gr based coated alloys was improved by $\widetilde{4}$9% and $\widetilde{3}$0% for LAT971 and $\widetilde{4}$8% and $\widetilde{2}$1% for LATZ9531 alloy, respectively, when compared to that of neat epoxy resin coating. The yield strength (σ_y) theoretically calculated using Tabor's law as the one-third of the nano-hardness, is reported in Table 1.

The elastic recovery (r_e) was also calculated using the following expression [58].

r_e=$\frac{h_m-h_f}{h_m}$ (8)

where, h_m is maximum depth and h_f is final depth of indentation.

A high plasticity ratio (E/H$\widetilde{1}$7) and lower elastic recovery ($\widetilde{0}$.62) was observed for the neat epoxy coated alloys when compared to that of PANI/Gr containing coated alloys, where plasticity ratio was found down to $\widetilde{1}$1 with elastic recovery of 0.85. Thus, the indenter interacted with Gr sheets in PANI/Gr containing coating resists the plastic deformation produced by the normal load due to its higher hardness and showed higher elastic recovery.

3.6. Tribological behaviour of coatings

Scratch-induced damage in the polymeric coating is a very complex phenomenon because of its nonlinear viscoelastic behaviour and scratch-induced stress field [59]. For the PANI/Gr containing coatings and neat epoxy coated alloys, the scratch track was not visible at low load (30 mN), due to the limited plastic deformation and viscoelastic recovery [60]. Therefore, a higher load (3 N) was utilized to observe the scratch response and permanent deformation as lateral and frontal pile-up, i.e. coating cracking and delamination. All micrographs are captured after completion of the scratch and the removal of the indentation load to observe the scratch response (Fig. 8).

At the maximum load of 3 N, a wide scratch damage (with periodic chevron micro cracks) was observed due to stick-slip oscillations in case of neat epoxy coatings (Fig. 8(a)-(b)) when compared to that of PANI/Gr containing coatings (Fig. 8(c)-(d)) [59]. A similar type of stick-slip patterns was observed on epoxy based coatings in earlier studies, which resemble the brittle polymeric behaviour [61]. This type of behaviour occurs during the scratch testing at slower traverse speed than that of the crack propagation rate during scratching [61]. On the contrary, for PANI/Gr containing coatings (Fig. 8(c)-(d)), no sign of micro-cracking and delamination was observed even at maximum applied load of 3 N, which is attributed its strong interfacial bonding between with the substrate [60]. Sacre et al. proposed the mechanism for scratch damage in the polymeric coating containing ZnO thin films [62] and observed the micro-cracks in the coating similar to that observed in the present study, (Fig. 8(a)-(b)) and are known as 'in-track' cracks, which usually starts from the starting of scratch and propagate along the scratch direction until the deviation of scratch track at an angle of 45° [62].

The scratch coefficient of friction (COF) parameter for each scratch is presented in Fig. 9, which is the ratio of the lateral load to the normal applied load. The COF is sensitive to the material pile-up which gives the information about the different changes through scratch-induced deformation [61]. It is demonstrated in Fig. 9 that PANI/Gr containing coatings has much lower COF (0.01-0.40) than the neat epoxy coatings (0.23-0.53), which is ascribed to its low surface friction (lubricating offered by Gr sheets) and high strength and modulus values (Table 1).

The average width of the scratch is measured from the SEM images of the scratch to calculate the plowing hardness (H_p) and scratch hardness (H_s). The H_p, H_s, scratch width (w), wear volume (W_v) and specific wear rate (W_sp) were calculated using Eqs. (1), (2), (3), (4), (5) and its values are reported in Table 2. The plowing hardness of the PANI/Gr containing coatings was found to be higher (up to 0.61 GPa) as compared to neat epoxy coatings (up to 0.40 GPa) due to the lubricating nature of the PANI/Gr containing coatings. Additionally, the H_s of the PANI/Gr containing coatings was observed to be superior (up to 0.35 GPa), when compared with that of neat epoxy coated alloys (up to 0.17 GPa), which is attributed to the higher hardness (Table 1) of the PANI/Gr containing coatings, which results in lower scratch width at maximum load of 3 N.

The specific wear rate of PANI/Gr containing coatings was obtained to be less (up to 13%) as compared to neat epoxy coated alloy due to the high hardness and low COF of the coatings, which results in shallow penetration depth thus the lower wear rate. Also, the elastic contact of the coated alloys (a) with the indenter is estimated by using following equation [44]:

a³=$\frac{3PR}{4E}$ (9)

where, P is the applied load, R is the radius of the indenter and E is the elastic modulus. The representative strain (ε_r) was estimated using following relation [44]:

ε_r=$frac{0.2a}{R}$ (10)

The maximum value of initial contact stress acting at the interface between indenter and coating was calculated using the following equation [63]:

$Q_{max}=(\frac{6PE^2}{π^3R^2}^{0.33})$ (11)

Table 2 summarizes the estimated values of contact radius (a), representative strain (ε_r) and subsequently contact stress (Q_max) during the scratching for all the coated alloys. The elastic contact radius was observed to be less (10.6 μm) for PANI/Gr containing coatings as compared to that of neat epoxy coatings (14.2 μm). The representative strain of the PANI/Gr containing coating was much lower (0.021) than that of neat epoxy coatings (0.028), which has resulted in higher contact stress (1.20 GPa) for smaller contact radius, when compared with that of neat epoxy coatings having lower contact stress (0.99 GPa) for larger contact radius (14.2 μm). Hence, the tribological results show that the resistance to scratch is improved significantly after adding PANI/Gr pigments to the neat epoxy coating by avoiding the micro-cracking, plowing damages and decreasing the scratch COF. After the mechanical and tribological performance of the coatings, the electrochemical testing was performed to observe the corrosion resistance offered by the coatings with PANI/Gr pigments in NaCl (3.5 wt%) corrosive medium.

3.7. Electrochemical behaviour of the coatings

The chemical stability or corrosion resistance of coated Mg-Li alloys has been examined in 3.5 wt% NaCl solution by electrochemical methods, such as OCP, potentiodynamic polarization and EIS. The OCP of alloys coated with neat epoxy, and epoxycontaining PANI/Gr pigments coatings were measured for 1 h. The change in OCP for coated alloys at room temperature is provided in Fig. 10.

The pure epoxy coated LAT971 and LATZ9531 alloys shows considerably low (more negative) OCP (-1.56 V and -1.54 V, respectively) to that of epoxy containing PANI/Gr pigments coated LAT971 and LATZ9531 (-1.45 V and -1.45 V, respectively). Thus, the more positive OCP values of PANI/Gr composite coated LAT971 and LATZ9531 alloys suggest that the coating provides an enhanced corrosion resistance [64]. Thus, substantially lower OCP values of PANI/Gr composite coated LAT971 and LATZ9531 alloys confirm that the coatings provide an enhanced protection and higher resistance towards oxidation. Fig. 11(a) illustrates the overall potentiodynamic behaviour of pure epoxy and epoxy containing PANI/Gr pigments coatings in 3.5 wt% NaCl solution. It can be observed that the composite coatings show quite different behaviour than that of the pure epoxy coating.

Table 3 summarizes E_corr and I_corr values extracted via potentiodynamic polarization plots using Tafel extrapolation method, calculated corrosion rate and corrosion protection efficiency. It is observed that the E_corr for epoxy based coatings shows a substantial shift to more positive value (-1.38 V and -1.42 V, for LAT971 and LATZ9531 alloys, respectively) when compared to that of pure epoxy coatings on alloys ($\widetilde{1}$.56 V). This confirms that the PANI/Gr pigment coated alloys possess an enhanced corrosion resistance than that of epoxy coated alloys.

Compared to that of neat epoxy coated alloys, PANI/Gr coated alloys exhibited a drastic drop in I_corr ($\widetilde{2}$-3 orders of magnitude), thus revealing an enhanced corrosion resistance than that of neat epoxy coated alloys substrate. The instantaneous corrosion rate, Pi (μm/y) was determined by using I_corr. It can be observe from the Table 3 that the corrosion rate of the PANI/Gr coated LAT971 and LATZ9531 alloy is 0.40 μm/y and 2.42 μm/y, which is $\widetilde{1}$ and $\widetilde{2}$ orders of magnitude lower than that of the neat epoxy LAT971 (92.77 μm/y) and LATZ9531 (246.78 μm/y) alloys, respectively. The corrosion rate of the PANI/Gr coated alloys was obtained to be $\widetilde{5}$ and $\widetilde{3}$ order of magnitude lower than that of the value reported [7] for the uncoated LAT971 and LATZ9531 alloy, respectively. Fig. 11 shows the digital image of coated sample after corrosion test. The epoxy coated LAT971 and LATZ9531 alloys (Fig. 11(b-c)) was found to swell or delaminate after corrosion test. Whereas, PANI/Gr based coating was intact or remained adhered on the LAT971 and LATZ9531 alloys surface (Fig. 11(d-e)) after the corrosion test.

Electrochemical impedance spectroscopy (EIS) has been used to estimate the corrosion resistance values of the coatings. Fig. 12(a) shows the Nyquist plot obtained for neat epoxy and PANI/Gr pigment containing epoxy coated substrates. Nyquist plots of PANI/Gr coatings shows large, incomplete semicircle diameter as compared to neat epoxy-based coatings (given as the inset image in the Fig. 12(a)), which signifies the lower corrosion rate [65].

The equivalent fitted electrical circuit is shown in Fig. 12(b), where RE and WE are reference electrode and working electrode, respectively. Other elements such as R_s, R_P, R_ct correspondes to solution resistance, porosity resistance and charge transfer resistance, respectively. In the electrical circuit, Q is used for constant phase element (CPE) and is defined as Y = Y_o(jω)ⁿ, where Y_o is admittance, j is imaginary unit, ω is angular frequency and n is power index (n₁ for coating and n₂ for double layer) [66]. In general, CPE is used in place of capacitor, for the better-quality fit [66]. The values of the all fitted elements are provided in Table 4. At the starting of the immersion when the coated surface interacts with the electrolyte then both neat epoxy and PANI/Gr based coatings gives the capacitive response where the n₁ = 1. As the electrochemical reactions proceeds, the electrolyte starts to penetrate towards the metal surface through the coating.

It can be observed from the Table 4, that neat epoxy-based coatings exhibit lower pore resistance (R_p, 2-5 order of magnitude lower) that allows more electrolyte to penetrate through it when compared to that of PANI/Gr based coatings. R_p represents the ionic transfer resistance through the pore presents in the coatings [24]. The presence of micro-pores (Fig. 6(a)-(b)) in the epoxy-based coatings may easily allow the penetration of electrolyte that resulted in the lower R_p values as compared to that of PANI/Gr based coatings. As the electrolyte penetrates in to the coatings, it forms a double layer capacitor (Q_dl). The values of n₁ = 1 stands for pure capacitor, whereas n₂ < 1 associated with the double layer which is related to the heterogeneity of the surface due to the degradation of coating and formation of corrosion products on exposed surface in an aggressive environment [67]. The value of n₂ for PANI/Gr based coating was observed to be slightly higher than that of neat epoxy coating, which signifies the higher capacitive behavior, and indicates a lowered water absorption by the coating in comparison to that of neat epoxy coatings. The higher R_p of the PANI/Gr based coatings (10⁶-10⁷Ω. cm²), resulted in a higher charge transfer resistance (R_ct) than that of neat epoxy coating. The R_ct values of PANI/Gr based coatings were estimated as $\widetilde{2}$-3 orders of magnitude higher than that of the neat epoxy coating. Higher R_ct values of PANI/Gr based coatings suggest an enhanced corrosion protection or oxidation resistance compared to that of neat epoxy coatings. According to aforementioned studies, at low frequencies, the impedance modulus lZl value is attributed to the corrosion protection ability of the coating [24,35,68]. It can be seen in Fig. 12(c), Bode plot for PANI/Gr based coatings showed a higher lZl value ($\widetilde{1}$.0-1.1 × 10⁷ Ω. cm²) at 0.01 Hz frequency than that of neat epoxy coating (lZl $\widetilde{3}$.0-3.7 × 10⁵ Ω. cm²), which indicating that PANI/Gr based coatings exhibited a effective barrier against corrosive ions from electrolyte. Furthermore, at high frequency, the phase angle value is another parameter for estimating the corrosion resistance of coatings. Phase angle (Fig. 12(d)) at high frequency (10⁵ Hz) for PANI/Gr based coatings was observed to be $\widetilde{9}$0°, whereas a lower phase angle of $\widetilde{8}$4° was observed for the neat epoxy coatings. The lower phase angle for neat epoxy coatings suggests the delamination of coatings, while for PANI/Gr based coatings, the higher phase angle indicates a strong interaction or bonding of coatings with the underlying metallic surface. Thus, on the basis of results obtained from the electrochemical analysis, it can be confirmed and concluded that the corrosion protection of epoxy-based coating was enhanced (R_ct$\widetilde{2}$ orders of magnitude) with the addition of PANI/Gr conducting pigments. Barrier protection and corrosion inhibition properties of PANI based coatings are main possible corrosion protection mechanisms [24], which is proposed by several researches.

Generally, the electrolyte penetrates through real microscopic pores and virtual pores present in the surface exposed to the electrolyte. According to microstructural results (Fig. 6), PANI/Gr based coatings had more compact microstructure as compared to neat epoxy coating. Therefore, pore-free PANI/Gr nanocomposite coating (Fig. 6(c)-(d)) with high scratch and plowing hardness (Table 2) provides the strong adhesion of interaction of coating with the underlying substrate i.e. LAT971 and LATZ9531 alloys surface. Consequently, the strong adhesion of coating with the substrate avoids the penetration of electrolyte (via delamination of coating) through it and, thus, resulted in the improved corrosion resistance. Meanwhile, the Gr particles covered with the PANI matrix (Fig. 2(c)) might delay the penetration of electrolyte through the composite coatings by increasing the tortuosity in the diffusion pathways of electrolyte ions (H₂O and O₂ molecules) [35], as shown schematically in Fig. 13.

Apart from the barrier properties, the corrosion protection capability is also attributed to the anodic protection effect of PANI due to redox ability [69]. During the electrochemical test, when the electrolyte penetrates into the space between the coating and the substrate, then it forms a double layer as explained earlier (Section 3.7). The reaction between the PANI/Gr pigment and underlying substrates occurs in the formed double layer. As suggested by the Song et al. [3], when Mg-Li based alloys expose to corrosive medium then Mg and Li establish a micro-galvanic coupling, where Mg site acts as micro cathode and Li as micro anode. Oxidation of Mg and Li occurs at the Li anodic site along with the H₂ evolution reaction at the Mg cathodic site according to the following reactions:

Mg-2e^-→Mg²⁺ (12)

Li-e^-→Li⁺ (13)

2H₂O+2e^-→2OH^-+H₂ (14)

The corrosion product such as hydroxide and oxides of Mg and Li forms in the double layer formed on exposed surface. Oxidation or corrosion of alloys surface promotes the reduction of the polyaniline coating i.e. Emeraldine salt (ES) form of PANI to LeucoEmeraldine base (EB) form of PANI and maintains the passivity state of the coating [23].

Mg²⁺+PANI(ES)+2H₂O→Mg(OH)₂++PANI(EB)-2H⁺ (15)

Li⁺+PANI(ES)+H₂O→LiOH+PANI(EB)+H⁺ (16)

As the reaction proceeds, the reduced EB form of PANI again oxidised to its initial state of ES according to Eq. (17) [32]. Thus, this cyclic reaction occurs throughout the process which maintains the passivity of the exposed surface and protect it from the further corrosion [69].

2O₂+4H₂O+PANI(EB)→PANI(ES)+8OH^- (17)

Thus, the cyclic reactions in PANI based coatings in the double layer maintain the passivity of the exposed surface. Therefore the higher charge transfer resistance R_ct (by $\widetilde{2}$ order of magnitude) was observed for the PANI/Gr containing coating as compared to neat epoxy coating.

4. Conclusions

PANI/Gr composite pigment prepared by in-situ polymerization of aniline monomer and Gr sheets. The TEM analysis of PANI/Gr composite indicated that Gr sheets were homogeneously dispersed in the PANI matrix. The enhanced thermal stability of the PANI/Gr pigments indicates a strong interface between the PANI matrix and Gr sheets as demonstrated by TGA analysis. PANI/Gr containing coatings were displayed enhanced mechanically integrity with higher elastic modulus and hardness (E = $\widetilde{8}$ GPa and H = 0.68 GPa) when compared to that of neat epoxy coating (E = $\widetilde{6}$ GPa and H = 0.35 GPa). Due to the superior mechanical properties, PANI/Gr containing coatings showed lower wear volume W_v = 4.53 × 10^-3 m³) than that of neat epoxy coating (W_v = 5.15 × 10^-3 m³). The PANI/Gr containing coatings were capable to maintain nobler OCP (-1.45 V) in comparison to that of pure epoxy coatings (-1.56 V) in corrosive medium. The PANI/Gr containing coatings showed highly protective nature with the higher charge transfer resistance (R_ct) (>10⁶ Ω cm²) as compared to that of neat epoxy coatings (R_ct $\widetilde{1}$0⁵ Ω cm²), evincing superior barrier properties and anodic protection of PANI based coatings justifying enhanced corrosion resistance of the PANI/Gr coatings. In summary, PANI/Gr based coatings can serve as promising candidates for wear and corrosion resistance in marine environments.

Acknowledgements

The authors acknowledge funding from Space Technology Cell (IIT Kanpur, and Indian Space Research Organisation). Dr. Govind is acknowledged for processing Mg-Li based alloys (at Vikram Sarabhai Space Center, Trivandrum, India) and Prof. Kallol Mondal is acknowledged for extending electrochemical facilities. KB acknowledges Swarnajayanti Fellowship (DST/SJF/ETA-02-2016-17) from Department of Science and Technology.

The authors have declared that no competing interests exist.

---