- Email: [email protected]
Characterization and corrosion resistance of duplex electroless Ni-P composite coatings on magnesium alloy
Surface & Coatings Technology 232 (2013) 432–439
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Characterization and corrosion resistance of duplex electroless Ni-P composite coatings on magnesium alloy Elsa Georgiza ⁎, Jelica Novakovic, Panayota Vassiliou School of Chemical Engineering, National Technical University of Athens, Iroon Polytechniou Str, GR- 157 80, Athens, Greece
a r t i c l e
i n f o
Article history: Received 25 January 2013 Accepted in revised form 30 May 2013 Available online 10 June 2013 Keywords: Magnesium AZ31 Electroless nickel Composite coatings Heat treatment
a b s t r a c t Three different types of electroless nickel (EN)-phosphorus plating were applied on AZ31 wrought magnesium alloy. The first type was plain mid-phosphorus coating, the second one was duplex, with the first layer having a mid-phosphorus content and the upper one consisting of high-phosphorus whereas the third type of coating was also duplex with a first mid-phosphorus layer and a second high-phosphorus one with incorporated ceramic TiO2 and ZrO2 microparticles. Characterization of deposits by means of scanning electron microscopy and X-ray diffraction analysis proves the production of adherent, defect free coatings with differences in crystallinity depending on P content. Surface roughness was maintained at acceptable levels while a great increase in microhardness was observed that was further enhanced by ceramic microparticles incorporation. Electrochemical testing, that was performed by Tafel polarization in 3.5% NaCl solution, show that all the coatings effectively protect the Mg alloy substrate. Additionally, the composite coatings exceptionally enhance the protective capability of the system after heat treatment at 200 °C. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In various fields of industry, where light weight is a prerequisite, such as automobile, aerospace, aeronautics, telecommunications, electronics, magnesium and its alloys are predominant materials. They combine excellent physical and mechanical properties, such as electrical and thermal conductivity, high specific strength, excellent anti-shock resistance, vibration absorption and good electromagnetic shielding effectiveness. All of the above, as well as the fact that magnesium alloys are easy to cut, form, of low cost and recycle are more than enough to be characterized as green engineering materials and of great importance for the 21st century. However, these significant advantages are overshadowed by one major drawback that becomes an obstacle for developing new applications and improving the existing ones. Their high chemical and electrochemical reactivity make them prone to oxidation and corrosion in humid atmosphere, fresh water, seawater and most of organic and inorganic acids and their salts [1]. This highly reactive nature is enhanced by alloying constituents which introduce electrochemical heterogeneity, hence microgalvanic corrosion [2]. Therefore, appropriate methods that improve the corrosion resistance are highly recommended. Electroless nickel-phosphorus (EN) plating is considered one of the most effective surface treatment techniques due to its perfect comprehensive properties, such as good corrosion and wear resistance, deposit uniformity over complex geometries and solderability [3]. However, magnesium alloys are challenging ⁎ Corresponding author. Tel.: +30 2107723063. E-mail address: [email protected] (E. Georgiza). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.05.047
materials as far as plating is concerned, mainly because of MgO that can be formed on their surface, which causes the deterioration of the coatings adhesion. Furthermore, the inhomogeneous microstructure consisting of the primary phase and the second phase can be the reason of non-uniform coatings. Additionally, since magnesium is one of the most electrochemically active metals, EN coatings are cathodic to the alloy substrate so they can only provide a physical barrier against corrosion attack. The considerable potential difference between the substrate and the coating will result in pitting corrosion when cracks and pores are present, a fact that highlights the importance of defect-free coatings [4,5]. The ability to co-deposit fine particulate matter within an electroless metal matrix has led to a new generation of composite coatings. Successful co-deposition is dependent on various factors including particle catalytic inertness, particles charge, electroless bath composition, bath reactivity, compatibility of the particles with the metallic matrix, plating rate and particle size distribution [6]. Composite coatings have been a source of disagreement among many researchers. According to one group, the corrosion resistance of electroless Ni-P coatings is believed to be significantly less than that of the electroless Ni-P composite coatings [7]. Dehghanian et al. reported that the electroless deposited Ni-P-nano-TiN coating on low carbon steel displayed very weak corrosion properties [8]. Zang et al. have found that co-deposition of either SiC or PTFE particles slightly decreased corrosion resistance of the coatings on mild steel in 3% NaCl aqueous solution but had insignificant effects on their corrosion resistance in 1 N H2SO4 [9]. On the other hand, electrochemical, salt spray and immersion tests have proved that the corrosion resistance of Ni-P-ZrO2 composite coating was superior to
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
that of Ni-P coating on AZ91D magnesium alloy in the study of Han et al. [10]. Furthermore, Ni-P-CeO2, Ni-P-TiO2 and Ni-P-Si3N4 composite coatings offer better corrosion protection on medium carbon steel, according to Balaraju et al. [11]. In the present study, electroless Ni-P composite duplex coatings with the upper coating containing crystalline TiO2 and ZrO2 microparticles on AZ31 wrought magnesium alloy were prepared. Their physical properties, surface roughness and microhardness, as well as their corrosion behavior were tested and compared to those of duplex and plain coatings. 2. Experimental The substrate for this work was AZ31 wrought magnesium alloy, whose composition is presented in Table 1. Rectangular coupons (30 × 20 × 4 mm) were gradually ground down from 1000 to 2000 grit using silicon carbide papers to achieve similar surface roughness, and they are then subjected to the pretreatment process (Table 2). Alkaline cleaning removes oils and greases from the coupons surface whereas acid pickling creates an oxide-free substrate. To minimize air contact and thus atmospheric oxidation, coupons are rinsed thoroughly with deionized water and as quickly as possible after each step of the pretreatment process. Ethanol rinses were also applied before water rinses to assure that each specimen would be wetted equally by the solutions, i.e. ethanol was employed as a wetting agent. The weight loss during pretreatment was negligible. For the present study two different EN plating baths were prepared (Tables 3–4). Their common compounds are nickel sulfate and sodium hypophosphite which serve as the source of nickel cations and metal reducing agent respectively. Thiourea in the first bath and Pb2+ in the second one serve as stabilizers in order to prevent bath decomposition, whereas the complexing agents are sodium acetate for the first EN plating process and propionic and lactic acids for the second one. After pretreatment, the specimens were immersed in the first EN plating bath for 1 h at 90 °C. They were then removed, dried and weighed before their subsequent 1-h immersion in the second EN plating bath, that contained microparticles of TiO2 or ZrO2, and was kept under stirring by a rotating stirring rod with a rate of 250 rev/min. The grain size of TiO2 was 0.3–0.5 μm and that of ZrO2 powder was up to 5 μm. Afterwards the coupons were once more dried and weighed. For comparison, plain Ni-P deposits by immersing coupons in the 1st EN plating bath for 2 h and duplex Ni-P deposits by immersing coupons in the 1st EN plating bath for 1 h and then in the 2nd EN plating bath without microparticles were also prepared. The total thickness of the deposits was estimated at approximately 50 μm for each specimen. The microhardness of the substrate and of the coatings was calculated as the average of five measurements taken on each side of the coupons with a load of 150 g for 10 s using a Leitz Wetzlar tester with a Vickers diamond indenter. The surface morphology of the specimens was examined by means of Siemens X-ray diffractometer 5000 (XRD) equipped with a Cu Kα X-ray source and FEI QUANTA 2000 scanning electron microscopy (SEM) along with EDS analysis in order to determine the chemical composition of the deposits. The latter was also used to study cross-sections of the specimens for possible defects of coatings as well as for an estimation of the adhesion of the coatings. Surface roughness test was also conducted with a Talysurf instrument and the average of five measurements on each side of the coupons was recorded.
Table 1 Composition of AZ31 wrought magnesium alloy. Alloy
Mg
Al
Zn
Mn
Si
Cu
Ni
Fe
AZ31
Balance
3.0
1.0
0.44
b0.1
b0.01
b0.003
b0.005
433
Table 2 Pretreatment of coupons to be electroless nickel plated. Pretreatment Stage
Solution
Parameters
Alkaline cleaning Acid pickling
NaOH 45 g/L HNO3 (65% w/w) 100 mL/L CrO3 125 g/L HF (50% w/w) 280 mL/L
10 min, 65 °C 1 min
Fluoride activation
10 min
Tafel electrochemical tests were carried out in a three-electrode cell, having a saturated calomel electrode (SCE) as reference, a platinum sheet as counter electrode and the tested specimen as a working electrode, exposed in a 3.5% NaCl aqueous solution at room temperature in the range of − 0.25 to 0.25 V vs. open circuit potential and at a constant scan rate of 1 mV/s. The instrumentation was a CMS 100 Gamry potentiostat, computer controlled, with commercial software for the obtained data process. Thermal treatment of the system was also tried to test cracking and the overall behavior of the coatings in high temperatures. The heat treatment was carried out in a furnace with constant airflow, at 200 °C for 2 h, and then the coupons were left to cool at room temperature. This level of temperature was selected, as it is the average operating temperature of many helicopters engine. The heat treated coupons were then studied for any changes in microhardness and their surface roughness as well as for the effect of temperature in the crystallinity of deposits and their overall corrosion behavior by XRD and Tafel analysis respectively. Cross-sections of heat treated specimens were examined for cracks by means of the aforementioned scanning electron microscope.
3. Results and discussion 3.1. Characterization of the as-plated specimens The surface roughness of the specimen to be plated should be considered before the beginning of any other plating procedure. The initial roughness plays an important role in the thickness and adhesion of the deposits as well as in their wear properties. In a polished substrate with a very low surface roughness, a thinner, less adherent coating will develop in contrast with a rougher substrate that favors an enhanced interlocking force. On the other hand, on a rough substrate, coatings of high friction coefficient will be produced [12]. Therefore, the coupons of this study were grinded down to 2000 grit, so as their average surface roughness (Ra) reached 0.1 μm (Fig. 1). The pretreatment process is the key to successful EN plating on magnesium alloys. Their high chemical reactivity with air will result in the formation of an oxide film on their surface that has a detrimental effect on the coatings adherence and uniformity. Therefore, air contact of specimens must be minimized during this process. The acidic solution corrodes the surface, making it rougher but more
Table 3 Composition of the first, medium-phosphorus EN plating bath. 1st EN plating bath NiSO.46H2O NaC2H3O2 NH4HF2 HF (50% w/w) NaH2PO2 CH4N2S NaOH(aq) for adjusting pH T = 90 °C, pH = 6.2, t = 60 min
15 g/L 13 g/L 8.5 g/L 10 mL/L 14 g/L 1 mg/L
434
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Table 4 Composition of the second, high-phosphorus EN plating bath. 2nd EN plating bath NiSO.46H2O C3H6O2 C3H6O3 NaH2PO2 Pb2+ TiO2 or ZrO2 NaOH(aq) for adjusting pH T = 90 °C, pH = 4.7, t = 60 min, mechanical agitation: 250 rev/min
21 g/L 3 ml/L 23 mL/L 24 g/L 1 mg/L 1 g/L
chemically active, providing surface pits to act as sites for mechanical interlocking of the EN coating (Figs. 1–3) [13]. The last step of the pretreatment, fluoride activation, is an indispensible one and of great importance due to the microstructure of magnesium alloys. This microstructure is characterized by two different phases: the α matrix phase and the secondary β phase, which in Mg-Al series consists of intermetallic compounds Mg17Al12 with body-centered cubic crystal structure. Each phase has different electrochemical properties resulting in electrochemical heterogeneity that affects and deteriorates the EN deposits. Thus, an equipotentialized surface, homogeneous and suitable for subsequent procedures, is created by the MgF2 thin film that is formed on specimens surface during fluoride activation [1]. This film makes the specimen surface even rougher, as can be observed in Fig. 1, where the increasing value of Ra during pretreatment process is demonstrated. The presence of hydrofluoric acid and ammonium bifluoride in the first EN plating bath is justified by the fact that the above-mentioned MgF2 film must be stabilized while the coupon is immersed in the bath. These two compounds act as a buffer solution as well. However, the concentration of fluorides in the bath must not exceed a certain limit for the plating to take place. According to some researchers, the deposition of electroless nickel in the bath takes place by replacement of MgF2 film [14], whereas others have proved the existence of a F-rich interface between the deposit and substrate [15]. EDS analysis of coupons has detected the presence of fluoride and therefore it can be deduced that a MgF2 film acts as an intermediate layer between the substrate and the coating. It has been reported, that during the initial stages of EN deposition, some Ni nucleates beneath this film [15], preferentially on β phase. Possible explanation is that the initial stage of deposition has been influenced by a galvanic coupling
Fig. 1. Progress of ssurface roughness of as-deposited coupons throughout pretreatment and plating process.
between β phase and adjacent eutectic α phase. The electrons produced by the anodic dissolution of magnesium from the α phase are consumed by the cathodic deposition of electroless nickel on β phase [14,16]. SEM images (Fig. 2) show the development of a typical nodular structure for all deposits. Neither pores nor cracks were observed on the coupons surface. The apparent difference in appearance may be attributed to the different phosphorus content, confirmed by EDS analysis. Fig. 2b shows a more compact grain structure of a duplex deposit whose upper layer consists of 13% w/w phosphorus obtained from a bath with pH = 4.7, whereas “cauliflower” grains appear in a plain deposit with 6.5% w/w phosphorus obtained from a bath with pH = 6.2 (Fig. 2a). It is known that pH is one of the major factors that affect the content of phosphorus in the deposits. The latter decreases at higher pH values as the reducing ability of sodium hypophosphite increases with increasing pH [17,18] and the structure of the deposit changes from amorphous to microcrystalline [2]. Fig. 2c and d show the incorporation of TiΟ2 και ZrO2 microparticles, that appear as tiny gray dots. Their percentage in coating has been estimated at approximately 10% and 14% w/w respectively by EDS analysis. In these figures, some agglomeration of ceramic microparticles is observed, that can be ascribed to the inefficient mechanical stirring of plating bath during EN plating process. The light weight of specimens did not permit a more intense stirring, resulting in an accumulation of ceramic powder on the coupons surface. These composite deposits were produced from baths, whose oxide load (1 g/L) has been chosen based on the fact that at high loads the rate of deposition decreases and some particles may be physically adsorbed on the catalytic surface. In this way, they suppress available active sites for the deposition process, as it has been pointed out in previous works [19]. Lower bath concentrations (0.5–2.0 g/l) result in better particles dispersion and stronger particles–matrix adhesion [20]. Additionally, it has been reported that there is a critical concentration of particles in an EN bath, that depends on bath composition, operating conditions, size and nature of particles. Beyond this critical concentration, there is a possibility of grouping or agglomeration of these second phase particles due to the decrease in the mean distance between them [21]. It is essential that the electroless composite plating will not take place right after the pretreatment process and a middle plain EN coating is applied. Otherwise, some of the ceramic particles could be absorbed in the pores of the activating film preventing the origination of nickel particles from these sites resulting to non-adherent and porous deposits [4]. The above-mentioned agglomerated particles seem to slightly affect the roughness of the deposits. Nevertheless, it is still maintained in sufficient values, similar enough to the respective ones of non-composite deposited samples, as it can be seen in Fig. 1. Roughness is sometimes an undesirable property in actual usage, especially if it exceeds a certain limit, as it may cause unacceptably high levels of frictional heating, causing coating damage and failure. On the other hand, it is sometimes beneficial, as it allows surfaces to trap lubricants and prevents them from welding together [22]. As far as its progress during pretreatment and plating process is concerned, the surface pits that are created during pretreatment are “filled” with EN coating (Fig. 3) resulting in a slightly smoother surface in most cases on one hand (Fig. 1), and in adequately adherent coatings on the other (Fig. 3). Factors that can affect surface roughness are particle size and particle content in the deposit. For instance, it has been reported that there is a direct correlation between roughness and particle size in the case of diamond microparticles [6]. Alirezaei et al. have also found a direct correlation between roughness and Al2O3 content in composite coatings on a steel substrate [23]. On the other hand, nano-sized particles can be swept away from electrode surface due to agitation during EN plating leading to lower incorporation or they can agglomerate more compared with microparticles, resulting to poor distribution in the EN matrix [24]. To prevent particle
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
435
Fig. 2. SEM images of as-plated coupons, a) plain Ni-P coating, b) duplex Ni-P coating, c) duplex Ni-P-TiO2 coating. An area of the surface is zoomed in in the right hand corner, where the arrows mark the gray dots that are ceramic particles, d) duplex Ni-P-ZrO2 coating. An area of the surface is zoomed in in the right hand corner, where the arrows mark the gray dots that are ceramic particles.
localization in the coating which can lead to areas with lower concentration of particulate matter exhibiting less favorable mechanical and physical properties surfactants are often introduced to act as dispersing agents [25,26]. It is known that high phosphorus deposits are completely amorphous whereas medium and low phosphorus deposits consist of a mixture of amorphous and nanocrystalline nickel phase or simply crystalline nickel with small amounts of amorphous phase. In X-ray diffractograms (Fig. 4) for medium phosphorus plain Ni-P deposits, a sharp intense peak appears at 2θ = 45°, certifying the above mentioned nanocrystalline nickel phase. It has to be pointed out that slightly different X-ray diffraction peaks can be taken from different areas of a same coupon, some of them broader indicating a semi-amorphous structure, some of them very sharp, indicating the dominance of the nanocrystalline Ni. Crystallinity, texture and perhaps composition inhomogeneity can happen in the electroless Ni-P deposits at the as-deposited condition especially
for low and medium phosphorus EN deposits [27]. The structure of the deposits changes from nanocrystalline to amorphous with increasing phosphorus content. Therefore, the peak is extremely different for high phosphorus duplex deposits, as it appears broadened and with lower intensity, indicating the amorphous Ni-P phase. For the composite coatings, no other peaks than the above mentioned and the ones for ZrO2 and TiO2 are apparent. As far as microhardness is concerned, a dramatic increase is observed for all the deposits that reaches 1000% for the composite coatings (Fig. 5). The slight difference of 80 HV between the plain and duplex-coated coupons can be attributed to the existence of nanocrystalline structure for the first ones. Although this difference is within the limits of standard deviation, a similar trend has been reported by Apachitei et al. [28]. Since hardness is related to the ease with which plastic deformation can be made to occur, by reducing the mobility of dislocations, the mechanical strength may be
436
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Fig. 3. SEM images of cross-sections of as-plated coupons, a) plain Ni-P coating, b) duplex Ni-P coating, c) duplex Ni-P-TiO2 coating, d) duplex Ni-P-ZrO2 coating. Ceramic microparticles appear as tiny grey dots.
enhanced; that is, greater mechanical forces will be required to initiate plastic deformation. A fine-grained material, like the nanocrystalline, plain Ni-P deposits, is harder and stronger than one that is coarse
grained (high-P deposits), since the former has a greater total grain boundary area to impede dislocation motion [29]. The incorporation of TiO2 and ZrO2 microparticles in the duplex deposits further increases
Fig. 4. X-ray diffractograms of as-plated coupons.
Fig. 5. Microhardness of as-plated coupons.
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
437
their microhardness from 700 HV to 850 HV as a result of the strong particles/matrix adhesion. When a dislocation propagates and meets precipitated particles, that are coherent to the Ni-P matrix, it has to take a path around them, whereas in the case of incoherent particles the dislocation cuts through them [18]. This is in accordance with the reported increase in hardness of electroless Ni-P composite coatings with the incorporation of ceramic (hard) particles [20,30]. 3.2. Characterization of the heat-treated specimens After 2 h of heat treatment at 200 °C, no discoloration, cracks, blisters, or peeling of the coating was observed. X-ray diffractograms show no apparent change of structure for all specimens. Crystallization and formation of nickel-phosphide particles do not take place so the creation of small active/passive corrosion cells by these particles, that contributes to the destruction of the deposits, is avoided [7]. The amorphous character of the duplex coatings and the nanocrystalline-amorphous character for the plain ones is sustained, so no diffusion paths due to distinctive grain boundaries are formed. Furthermore, the ceramic microparticles that are embedded in the Ni-P matrix have no effect in structure or crystallization process, as expected [31]. The lack of crystallization during heat treatment at 200 °C keeps the microhardness of the coatings at the same level as the as-deposited coupons. A slight increase for each coating is observed, but as it is in the limits of standard deviation, it can be considered negligible. 3.3. Corrosion resistance The corrosion resistance of the as-deposited and heat treated coatings was investigated by electrochemical polarization curves in a 3.5% w/w NaCl aqueous solution. In Fig. 6a notable shift to more positive potentials in regard to magnesium alloy can be observed for plain coated and duplex-coated coupons. These coatings have a difference of 0.2 V in their corrosion potentials and a difference of 0.081 mm/yr in their corrosion rates, with the duplex deposits appearing nobler and with lower corrosion current density (Table 5). This can be attributed to the different phosphorus contents of the two deposits, the plain ones having a 6.5% w/w phosphorus and the duplex ones being high phosphorus deposits. It has been reported that the electroless Ni-P coatings with various phosphorus content have different electrochemical and mechanical properties, i.e. the corrosion resistance to attack in neutral and acidic environments is increased with phosphorus content in the deposit, whereas the reverse is true in alkaline corrosive environments [7]. Since all the specimens in the present study have approximately the same thickness, it can be deduced that the phosphorus content is a major factor that affects the electrochemical properties of the coatings, which in accordance to several researches [7,32]. The complete amorphous character of high phosphorus deposits provides a better corrosion behavior due to the exclusion of diffusion paths among grain boundaries, which are usually encountered in polycrystalline deposits. The corrosion characteristics of the composite deposits appear similar to the duplex ones and the incorporation of ceramic particles does not seem to have a significant effect. Heat treatment for 2 h at 200 °C transfers the corrosion potential of plain-coated specimens to more negative areas and intensely increases their corrosion current density and consequently their corrosion rate. Micro-cracks and micro-voids that form during heat treatment, due to the difference in the coefficients of thermal expansion between the coating and the substrate that results in development of internal stresses, expose the substrate to attack (Fig. 7). On the other hand, the corrosion potential of duplex non-composite coatings after heat treatment remains unchangeable and the only change in their characteristics lays in the corrosion current density and corrosion rate that are increased by 25 μA/cm2 and 0.25 mm/yr respectively. Composite Ni-P coatings appear unaffected after heat
Fig. 6. Comparative polarization curves of a) as-plated coupons and substrate, b) coupons after heat treatment for 2 h at 200 °C.
treatment (Fig. 6b). The incorporation of ceramic TiO2 and ZrO2 microparticles restrains the anodic dissolution reaction of the deposit by means of reducing its active surface as well as enhances the barrier effect of the coating on the substrate [4,11,19]. This fact is highlighted
Table 5 Corrosion characteristics of magnesium alloy, as-plated and heat treated coatings. TAFEL RESULTS As - plated
Magnesium alloy Plain Ni-P Duplex Ni-P Duplex Ni-P-TiO2 Duplex Ni-P-ZrO2
Ecorr (V)
icorr (μA/cm2)
−1.5 −0.6 −0.4 −0.3 −0.4
350 12.4 4.3 2.4 2.2
Ecorr (V)
icorr (μA/cm2)
Rp (Ohm.cm2)
CorrRate (mm/year)
−1.2 −0.3 −0.4 −0.4
203.5 29.3 5.7 3.8
175 10620 3552 5316
2.034 0.293 0.058 0.039
Rp (Ohm.cm2) 37 3817 8900 8415 9306
CorrRate (mm/year) 7.8 0.124 0.043 0.024 0.022
Heat Treated, 200 °C
Plain Ni-P Duplex Ni-P Duplex Ni-P-TiO2 Duplex Ni-P-ZrO2
438
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Fig. 7. Optical microscope images (500×) of heat treated coupons a) plain Ni-P, the crack reaches the coupon surface b) duplex Ni-P-TiO2, the crack is hindered by the composite upper layer c) duplex Ni-P-ZrO2, the crack is hindered by the composite upper layer.
by Fig. 7, where the cross-sections of three different heat treated coatings are shown. In Fig. 7a of a plain plated sample, the crack is developed all the way from substrate to surface providing a diffusion path and making the Mg alloy susceptible to corrosion. On the other hand, Fig. 7b and c show that the cracks developed after heat treatment are hindered by the composite layer of Ni-P-ceramic particles. Composite coatings have been characterized as more compact than plain ones. During co-deposition some particles are adsorbed at nodules boundaries, forming, in this way, new nodules at these sites. The composite coating becomes thicker in accordance with this kind of growth model and the tortuous nodule boundaries are not the weak sites of the coating any more, especially when compared with a plain coating [10]. As a result, the protection characteristics of composite coatings are impressively maintained at very high levels, compared to the plain and duplex non-composite specimens. 4. Conclusions Electroless nickel plating was performed on AZ31 wrought magnesium alloy, with different types of obtained deposits: plain
mid-phosphorus deposit and duplex deposit with the first layer having a mid-phosphorus content and the upper one consisting of high-phosphorus. The third type was duplex deposit with a first mid-phosphorus layer and a second high-phosphorus one with incorporated ceramic microparticles. The plain coatings were found to consist of a mixture of nanocrystalline and amorphous phase, with the plated specimens exhibiting crystallinity heterogeneity throughout their surface. The duplex coatings were found to be completely amorphous, a fact that does not affect the microhardness and surface roughness of these two different coatings. Composite deposits offer a slight increase in microhardness, which is either way increased by approximately 1000% in regard to the substrate. No changes in crystallinity, microhardness and surface roughness can be observed after heat treatment at 200 °C for 2 h, as expected. All the coatings provide excellent corrosion resistance, with the most amorphous ones having the lowest corrosion rates due to the exclusion of diffusion paths among grain boundaries. The beneficial effect of the incorporation of ceramic TiO2 and ZrO2 microparticles
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
can be observed after heat treatment. The composite layer acts as a stronger barrier than the non-composite one against cracks progression that form due to the internal stresses that develop during heat treatment, thus providing an exceptional protection for magnesium alloy in an aggressive medium. References [1] L.P. Wu, J.J. Zhao, Y.P. Xie, Z.D. Yang, Trans. Nonferrous Met. Soc. 20 (2010) 630–637. [2] A.K. Sharma, M.R. Suresh, H. Bhojraj, H. Narayanamurthy, R.P. Sahu, Met. Finish. 96 (1998) 10–16. [3] T.N. Qin, L.Q. Ma, Y. Yao, C. Ni, X.Y. Zhao, Y. Ding, Trans. Nonferrous Met. Soc. 21 (2011) 2790–2797. [4] Y.W. Song, D.Y. Shan, E.H. Han, Electrochim. Acta 53 (2008) 2135–2143. [5] X.M. Chen, G.Y. Li, J.S. Lian, Trans. Nonferrous Met. Soc. 18 (2008) 323–328. [6] N. Feldstein, in: G.O. Mallory, J.B. Hajdu (Eds.), Electroless plating: fundamentals and applications, American Electroplaters and Surface Finishers Society, 1990, pp. 269–288. [7] A.S.M. Handbook, Surface Engineering, vol. 5, ASM, Material Park, OH, 1991. [8] I.R. Mafi, Ch. Dehghanian, Appl. Surf. Sci. 258 (2011) 1876–1880. [9] Y.S. Huang, X.T. Zeng, X.F. Hu, F.M. Liu, Electrochim. Acta 49 (2004) 4313–4319. [10] Y.W. Song, D.Y. Shan, E.H. Han, Mater. Corros. 58 (2007) 506–510. [11] J.N. Balaraju, T.S.N. Sankara Narayanan, S.K. Seshadri, J. Solid State Electrochem. 5 (2001) 334–338. [12] Z. Liu, W. Gao, Surf. Coat. Technol. 200 (2006) 5087–5093. [13] C. Gu, J. Lian, G. Li, L. Niu, Z. Jiang, J. Alloy. Compd. 391 (2005) 104–109. [14] R. Ambat, W. Zhou, Surf. Coat. Technol. 179 (2004) 124–134.
439
[15] Z. Liu, W. Gao, Surf. Coat. Technol. 200 (2006) 3553–3560. [16] Y.H. Xiang, W.B. Hu, X.K. Liu, C.Z. Zhao, W.J. Ding, T. I. Met. Finish. 79 (2001) 30–32. [17] M. Schlesinge, in: M. Schlesinger, M. Paunovic (Eds.), Modern Electroplating, John Wiley & Sons, Inc., New Jersey, 2010, pp. 447–458. [18] M. Momenzadeh, S. Sanjabi, Mater. Corros. 63 (2012) 614–619. [19] J. Novakovic, P. Vassiliou, Kl Samara, Th. Argyropoulos, Surf. Coat. Technol. 201 (2006) 895–901. [20] J. Novakovic, P. Vassiliou, Electrochim. Acta 54 (2009) 2499–2503. [21] J.N. Balaraju, K.S. Rajam, J. Alloys Compd. 459 (2008) 311–319. [22] J. Sudagar, J.S. Lian, Q. Jiang, Z.H. Jiang, G.Y. Li, R. Elansezhian, Prog. Org. Coat. 74 (2012) 788–793. [23] Sh. Alirezaei, S.M. Monirvaghefi, M. Salehi, A. Saatch, Surf. Coat. Technol. 184 (2004) 170–175. [24] J.N. Balaraju, Kalavati, K.S. Rajam, Surf. Coat. Technol. 200 (2006) 3933–3941. [25] N. Nwosu, A. Davidson, C. Hindle, M. Barker, Ind. Eng. Chem. Res. 51 (2012) 5635–5644. [26] I.R. Mafi, C. Dehghanian, Appl. Surf. Sci. 257 (2011) 8653–8658. [27] W. Sha, X. Wu, K.G. Keong, Electroless copper and nickel–phosphorus plating, processing, characterization and modeling, first ed. Woodhead Publishing Limited, New Delhi, 2011. [28] I. Apachitei, J. Duszczyk, Surf. Coat. Technol. 132 (2000) 89–98. [29] W.D. Callister Jr., Materials Science and Engineering: An Introduction, seventh ed. John Wiley & Sons, New York, 2007. [30] J.N. Balaraju, T.S.N. Sankara Narayanan, S.K. Seshadri, J. Appl. Electrochem. 33 (2003) 807–816. [31] L. Yu, W. Huang, X. Zhao, J. Alloys Compd. 509 (2011) 4154–4159. [32] W.X. Zhang, Z.H. Jiang, G.Y. Li, Q. Jiang, J.S. Lian, Appl. Surf. Sci. 254 (2008) 4949–4955.
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Characterization and corrosion resistance of duplex electroless Ni-P composite coatings on magnesium alloy Elsa Georgiza ⁎, Jelica Novakovic, Panayota Vassiliou School of Chemical Engineering, National Technical University of Athens, Iroon Polytechniou Str, GR- 157 80, Athens, Greece
a r t i c l e
i n f o
Article history: Received 25 January 2013 Accepted in revised form 30 May 2013 Available online 10 June 2013 Keywords: Magnesium AZ31 Electroless nickel Composite coatings Heat treatment
a b s t r a c t Three different types of electroless nickel (EN)-phosphorus plating were applied on AZ31 wrought magnesium alloy. The first type was plain mid-phosphorus coating, the second one was duplex, with the first layer having a mid-phosphorus content and the upper one consisting of high-phosphorus whereas the third type of coating was also duplex with a first mid-phosphorus layer and a second high-phosphorus one with incorporated ceramic TiO2 and ZrO2 microparticles. Characterization of deposits by means of scanning electron microscopy and X-ray diffraction analysis proves the production of adherent, defect free coatings with differences in crystallinity depending on P content. Surface roughness was maintained at acceptable levels while a great increase in microhardness was observed that was further enhanced by ceramic microparticles incorporation. Electrochemical testing, that was performed by Tafel polarization in 3.5% NaCl solution, show that all the coatings effectively protect the Mg alloy substrate. Additionally, the composite coatings exceptionally enhance the protective capability of the system after heat treatment at 200 °C. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In various fields of industry, where light weight is a prerequisite, such as automobile, aerospace, aeronautics, telecommunications, electronics, magnesium and its alloys are predominant materials. They combine excellent physical and mechanical properties, such as electrical and thermal conductivity, high specific strength, excellent anti-shock resistance, vibration absorption and good electromagnetic shielding effectiveness. All of the above, as well as the fact that magnesium alloys are easy to cut, form, of low cost and recycle are more than enough to be characterized as green engineering materials and of great importance for the 21st century. However, these significant advantages are overshadowed by one major drawback that becomes an obstacle for developing new applications and improving the existing ones. Their high chemical and electrochemical reactivity make them prone to oxidation and corrosion in humid atmosphere, fresh water, seawater and most of organic and inorganic acids and their salts [1]. This highly reactive nature is enhanced by alloying constituents which introduce electrochemical heterogeneity, hence microgalvanic corrosion [2]. Therefore, appropriate methods that improve the corrosion resistance are highly recommended. Electroless nickel-phosphorus (EN) plating is considered one of the most effective surface treatment techniques due to its perfect comprehensive properties, such as good corrosion and wear resistance, deposit uniformity over complex geometries and solderability [3]. However, magnesium alloys are challenging ⁎ Corresponding author. Tel.: +30 2107723063. E-mail address: [email protected] (E. Georgiza). 0257-8972/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.surfcoat.2013.05.047
materials as far as plating is concerned, mainly because of MgO that can be formed on their surface, which causes the deterioration of the coatings adhesion. Furthermore, the inhomogeneous microstructure consisting of the primary phase and the second phase can be the reason of non-uniform coatings. Additionally, since magnesium is one of the most electrochemically active metals, EN coatings are cathodic to the alloy substrate so they can only provide a physical barrier against corrosion attack. The considerable potential difference between the substrate and the coating will result in pitting corrosion when cracks and pores are present, a fact that highlights the importance of defect-free coatings [4,5]. The ability to co-deposit fine particulate matter within an electroless metal matrix has led to a new generation of composite coatings. Successful co-deposition is dependent on various factors including particle catalytic inertness, particles charge, electroless bath composition, bath reactivity, compatibility of the particles with the metallic matrix, plating rate and particle size distribution [6]. Composite coatings have been a source of disagreement among many researchers. According to one group, the corrosion resistance of electroless Ni-P coatings is believed to be significantly less than that of the electroless Ni-P composite coatings [7]. Dehghanian et al. reported that the electroless deposited Ni-P-nano-TiN coating on low carbon steel displayed very weak corrosion properties [8]. Zang et al. have found that co-deposition of either SiC or PTFE particles slightly decreased corrosion resistance of the coatings on mild steel in 3% NaCl aqueous solution but had insignificant effects on their corrosion resistance in 1 N H2SO4 [9]. On the other hand, electrochemical, salt spray and immersion tests have proved that the corrosion resistance of Ni-P-ZrO2 composite coating was superior to
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
that of Ni-P coating on AZ91D magnesium alloy in the study of Han et al. [10]. Furthermore, Ni-P-CeO2, Ni-P-TiO2 and Ni-P-Si3N4 composite coatings offer better corrosion protection on medium carbon steel, according to Balaraju et al. [11]. In the present study, electroless Ni-P composite duplex coatings with the upper coating containing crystalline TiO2 and ZrO2 microparticles on AZ31 wrought magnesium alloy were prepared. Their physical properties, surface roughness and microhardness, as well as their corrosion behavior were tested and compared to those of duplex and plain coatings. 2. Experimental The substrate for this work was AZ31 wrought magnesium alloy, whose composition is presented in Table 1. Rectangular coupons (30 × 20 × 4 mm) were gradually ground down from 1000 to 2000 grit using silicon carbide papers to achieve similar surface roughness, and they are then subjected to the pretreatment process (Table 2). Alkaline cleaning removes oils and greases from the coupons surface whereas acid pickling creates an oxide-free substrate. To minimize air contact and thus atmospheric oxidation, coupons are rinsed thoroughly with deionized water and as quickly as possible after each step of the pretreatment process. Ethanol rinses were also applied before water rinses to assure that each specimen would be wetted equally by the solutions, i.e. ethanol was employed as a wetting agent. The weight loss during pretreatment was negligible. For the present study two different EN plating baths were prepared (Tables 3–4). Their common compounds are nickel sulfate and sodium hypophosphite which serve as the source of nickel cations and metal reducing agent respectively. Thiourea in the first bath and Pb2+ in the second one serve as stabilizers in order to prevent bath decomposition, whereas the complexing agents are sodium acetate for the first EN plating process and propionic and lactic acids for the second one. After pretreatment, the specimens were immersed in the first EN plating bath for 1 h at 90 °C. They were then removed, dried and weighed before their subsequent 1-h immersion in the second EN plating bath, that contained microparticles of TiO2 or ZrO2, and was kept under stirring by a rotating stirring rod with a rate of 250 rev/min. The grain size of TiO2 was 0.3–0.5 μm and that of ZrO2 powder was up to 5 μm. Afterwards the coupons were once more dried and weighed. For comparison, plain Ni-P deposits by immersing coupons in the 1st EN plating bath for 2 h and duplex Ni-P deposits by immersing coupons in the 1st EN plating bath for 1 h and then in the 2nd EN plating bath without microparticles were also prepared. The total thickness of the deposits was estimated at approximately 50 μm for each specimen. The microhardness of the substrate and of the coatings was calculated as the average of five measurements taken on each side of the coupons with a load of 150 g for 10 s using a Leitz Wetzlar tester with a Vickers diamond indenter. The surface morphology of the specimens was examined by means of Siemens X-ray diffractometer 5000 (XRD) equipped with a Cu Kα X-ray source and FEI QUANTA 2000 scanning electron microscopy (SEM) along with EDS analysis in order to determine the chemical composition of the deposits. The latter was also used to study cross-sections of the specimens for possible defects of coatings as well as for an estimation of the adhesion of the coatings. Surface roughness test was also conducted with a Talysurf instrument and the average of five measurements on each side of the coupons was recorded.
Table 1 Composition of AZ31 wrought magnesium alloy. Alloy
Mg
Al
Zn
Mn
Si
Cu
Ni
Fe
AZ31
Balance
3.0
1.0
0.44
b0.1
b0.01
b0.003
b0.005
433
Table 2 Pretreatment of coupons to be electroless nickel plated. Pretreatment Stage
Solution
Parameters
Alkaline cleaning Acid pickling
NaOH 45 g/L HNO3 (65% w/w) 100 mL/L CrO3 125 g/L HF (50% w/w) 280 mL/L
10 min, 65 °C 1 min
Fluoride activation
10 min
Tafel electrochemical tests were carried out in a three-electrode cell, having a saturated calomel electrode (SCE) as reference, a platinum sheet as counter electrode and the tested specimen as a working electrode, exposed in a 3.5% NaCl aqueous solution at room temperature in the range of − 0.25 to 0.25 V vs. open circuit potential and at a constant scan rate of 1 mV/s. The instrumentation was a CMS 100 Gamry potentiostat, computer controlled, with commercial software for the obtained data process. Thermal treatment of the system was also tried to test cracking and the overall behavior of the coatings in high temperatures. The heat treatment was carried out in a furnace with constant airflow, at 200 °C for 2 h, and then the coupons were left to cool at room temperature. This level of temperature was selected, as it is the average operating temperature of many helicopters engine. The heat treated coupons were then studied for any changes in microhardness and their surface roughness as well as for the effect of temperature in the crystallinity of deposits and their overall corrosion behavior by XRD and Tafel analysis respectively. Cross-sections of heat treated specimens were examined for cracks by means of the aforementioned scanning electron microscope.
3. Results and discussion 3.1. Characterization of the as-plated specimens The surface roughness of the specimen to be plated should be considered before the beginning of any other plating procedure. The initial roughness plays an important role in the thickness and adhesion of the deposits as well as in their wear properties. In a polished substrate with a very low surface roughness, a thinner, less adherent coating will develop in contrast with a rougher substrate that favors an enhanced interlocking force. On the other hand, on a rough substrate, coatings of high friction coefficient will be produced [12]. Therefore, the coupons of this study were grinded down to 2000 grit, so as their average surface roughness (Ra) reached 0.1 μm (Fig. 1). The pretreatment process is the key to successful EN plating on magnesium alloys. Their high chemical reactivity with air will result in the formation of an oxide film on their surface that has a detrimental effect on the coatings adherence and uniformity. Therefore, air contact of specimens must be minimized during this process. The acidic solution corrodes the surface, making it rougher but more
Table 3 Composition of the first, medium-phosphorus EN plating bath. 1st EN plating bath NiSO.46H2O NaC2H3O2 NH4HF2 HF (50% w/w) NaH2PO2 CH4N2S NaOH(aq) for adjusting pH T = 90 °C, pH = 6.2, t = 60 min
15 g/L 13 g/L 8.5 g/L 10 mL/L 14 g/L 1 mg/L
434
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Table 4 Composition of the second, high-phosphorus EN plating bath. 2nd EN plating bath NiSO.46H2O C3H6O2 C3H6O3 NaH2PO2 Pb2+ TiO2 or ZrO2 NaOH(aq) for adjusting pH T = 90 °C, pH = 4.7, t = 60 min, mechanical agitation: 250 rev/min
21 g/L 3 ml/L 23 mL/L 24 g/L 1 mg/L 1 g/L
chemically active, providing surface pits to act as sites for mechanical interlocking of the EN coating (Figs. 1–3) [13]. The last step of the pretreatment, fluoride activation, is an indispensible one and of great importance due to the microstructure of magnesium alloys. This microstructure is characterized by two different phases: the α matrix phase and the secondary β phase, which in Mg-Al series consists of intermetallic compounds Mg17Al12 with body-centered cubic crystal structure. Each phase has different electrochemical properties resulting in electrochemical heterogeneity that affects and deteriorates the EN deposits. Thus, an equipotentialized surface, homogeneous and suitable for subsequent procedures, is created by the MgF2 thin film that is formed on specimens surface during fluoride activation [1]. This film makes the specimen surface even rougher, as can be observed in Fig. 1, where the increasing value of Ra during pretreatment process is demonstrated. The presence of hydrofluoric acid and ammonium bifluoride in the first EN plating bath is justified by the fact that the above-mentioned MgF2 film must be stabilized while the coupon is immersed in the bath. These two compounds act as a buffer solution as well. However, the concentration of fluorides in the bath must not exceed a certain limit for the plating to take place. According to some researchers, the deposition of electroless nickel in the bath takes place by replacement of MgF2 film [14], whereas others have proved the existence of a F-rich interface between the deposit and substrate [15]. EDS analysis of coupons has detected the presence of fluoride and therefore it can be deduced that a MgF2 film acts as an intermediate layer between the substrate and the coating. It has been reported, that during the initial stages of EN deposition, some Ni nucleates beneath this film [15], preferentially on β phase. Possible explanation is that the initial stage of deposition has been influenced by a galvanic coupling
Fig. 1. Progress of ssurface roughness of as-deposited coupons throughout pretreatment and plating process.
between β phase and adjacent eutectic α phase. The electrons produced by the anodic dissolution of magnesium from the α phase are consumed by the cathodic deposition of electroless nickel on β phase [14,16]. SEM images (Fig. 2) show the development of a typical nodular structure for all deposits. Neither pores nor cracks were observed on the coupons surface. The apparent difference in appearance may be attributed to the different phosphorus content, confirmed by EDS analysis. Fig. 2b shows a more compact grain structure of a duplex deposit whose upper layer consists of 13% w/w phosphorus obtained from a bath with pH = 4.7, whereas “cauliflower” grains appear in a plain deposit with 6.5% w/w phosphorus obtained from a bath with pH = 6.2 (Fig. 2a). It is known that pH is one of the major factors that affect the content of phosphorus in the deposits. The latter decreases at higher pH values as the reducing ability of sodium hypophosphite increases with increasing pH [17,18] and the structure of the deposit changes from amorphous to microcrystalline [2]. Fig. 2c and d show the incorporation of TiΟ2 και ZrO2 microparticles, that appear as tiny gray dots. Their percentage in coating has been estimated at approximately 10% and 14% w/w respectively by EDS analysis. In these figures, some agglomeration of ceramic microparticles is observed, that can be ascribed to the inefficient mechanical stirring of plating bath during EN plating process. The light weight of specimens did not permit a more intense stirring, resulting in an accumulation of ceramic powder on the coupons surface. These composite deposits were produced from baths, whose oxide load (1 g/L) has been chosen based on the fact that at high loads the rate of deposition decreases and some particles may be physically adsorbed on the catalytic surface. In this way, they suppress available active sites for the deposition process, as it has been pointed out in previous works [19]. Lower bath concentrations (0.5–2.0 g/l) result in better particles dispersion and stronger particles–matrix adhesion [20]. Additionally, it has been reported that there is a critical concentration of particles in an EN bath, that depends on bath composition, operating conditions, size and nature of particles. Beyond this critical concentration, there is a possibility of grouping or agglomeration of these second phase particles due to the decrease in the mean distance between them [21]. It is essential that the electroless composite plating will not take place right after the pretreatment process and a middle plain EN coating is applied. Otherwise, some of the ceramic particles could be absorbed in the pores of the activating film preventing the origination of nickel particles from these sites resulting to non-adherent and porous deposits [4]. The above-mentioned agglomerated particles seem to slightly affect the roughness of the deposits. Nevertheless, it is still maintained in sufficient values, similar enough to the respective ones of non-composite deposited samples, as it can be seen in Fig. 1. Roughness is sometimes an undesirable property in actual usage, especially if it exceeds a certain limit, as it may cause unacceptably high levels of frictional heating, causing coating damage and failure. On the other hand, it is sometimes beneficial, as it allows surfaces to trap lubricants and prevents them from welding together [22]. As far as its progress during pretreatment and plating process is concerned, the surface pits that are created during pretreatment are “filled” with EN coating (Fig. 3) resulting in a slightly smoother surface in most cases on one hand (Fig. 1), and in adequately adherent coatings on the other (Fig. 3). Factors that can affect surface roughness are particle size and particle content in the deposit. For instance, it has been reported that there is a direct correlation between roughness and particle size in the case of diamond microparticles [6]. Alirezaei et al. have also found a direct correlation between roughness and Al2O3 content in composite coatings on a steel substrate [23]. On the other hand, nano-sized particles can be swept away from electrode surface due to agitation during EN plating leading to lower incorporation or they can agglomerate more compared with microparticles, resulting to poor distribution in the EN matrix [24]. To prevent particle
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
435
Fig. 2. SEM images of as-plated coupons, a) plain Ni-P coating, b) duplex Ni-P coating, c) duplex Ni-P-TiO2 coating. An area of the surface is zoomed in in the right hand corner, where the arrows mark the gray dots that are ceramic particles, d) duplex Ni-P-ZrO2 coating. An area of the surface is zoomed in in the right hand corner, where the arrows mark the gray dots that are ceramic particles.
localization in the coating which can lead to areas with lower concentration of particulate matter exhibiting less favorable mechanical and physical properties surfactants are often introduced to act as dispersing agents [25,26]. It is known that high phosphorus deposits are completely amorphous whereas medium and low phosphorus deposits consist of a mixture of amorphous and nanocrystalline nickel phase or simply crystalline nickel with small amounts of amorphous phase. In X-ray diffractograms (Fig. 4) for medium phosphorus plain Ni-P deposits, a sharp intense peak appears at 2θ = 45°, certifying the above mentioned nanocrystalline nickel phase. It has to be pointed out that slightly different X-ray diffraction peaks can be taken from different areas of a same coupon, some of them broader indicating a semi-amorphous structure, some of them very sharp, indicating the dominance of the nanocrystalline Ni. Crystallinity, texture and perhaps composition inhomogeneity can happen in the electroless Ni-P deposits at the as-deposited condition especially
for low and medium phosphorus EN deposits [27]. The structure of the deposits changes from nanocrystalline to amorphous with increasing phosphorus content. Therefore, the peak is extremely different for high phosphorus duplex deposits, as it appears broadened and with lower intensity, indicating the amorphous Ni-P phase. For the composite coatings, no other peaks than the above mentioned and the ones for ZrO2 and TiO2 are apparent. As far as microhardness is concerned, a dramatic increase is observed for all the deposits that reaches 1000% for the composite coatings (Fig. 5). The slight difference of 80 HV between the plain and duplex-coated coupons can be attributed to the existence of nanocrystalline structure for the first ones. Although this difference is within the limits of standard deviation, a similar trend has been reported by Apachitei et al. [28]. Since hardness is related to the ease with which plastic deformation can be made to occur, by reducing the mobility of dislocations, the mechanical strength may be
436
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Fig. 3. SEM images of cross-sections of as-plated coupons, a) plain Ni-P coating, b) duplex Ni-P coating, c) duplex Ni-P-TiO2 coating, d) duplex Ni-P-ZrO2 coating. Ceramic microparticles appear as tiny grey dots.
enhanced; that is, greater mechanical forces will be required to initiate plastic deformation. A fine-grained material, like the nanocrystalline, plain Ni-P deposits, is harder and stronger than one that is coarse
grained (high-P deposits), since the former has a greater total grain boundary area to impede dislocation motion [29]. The incorporation of TiO2 and ZrO2 microparticles in the duplex deposits further increases
Fig. 4. X-ray diffractograms of as-plated coupons.
Fig. 5. Microhardness of as-plated coupons.
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
437
their microhardness from 700 HV to 850 HV as a result of the strong particles/matrix adhesion. When a dislocation propagates and meets precipitated particles, that are coherent to the Ni-P matrix, it has to take a path around them, whereas in the case of incoherent particles the dislocation cuts through them [18]. This is in accordance with the reported increase in hardness of electroless Ni-P composite coatings with the incorporation of ceramic (hard) particles [20,30]. 3.2. Characterization of the heat-treated specimens After 2 h of heat treatment at 200 °C, no discoloration, cracks, blisters, or peeling of the coating was observed. X-ray diffractograms show no apparent change of structure for all specimens. Crystallization and formation of nickel-phosphide particles do not take place so the creation of small active/passive corrosion cells by these particles, that contributes to the destruction of the deposits, is avoided [7]. The amorphous character of the duplex coatings and the nanocrystalline-amorphous character for the plain ones is sustained, so no diffusion paths due to distinctive grain boundaries are formed. Furthermore, the ceramic microparticles that are embedded in the Ni-P matrix have no effect in structure or crystallization process, as expected [31]. The lack of crystallization during heat treatment at 200 °C keeps the microhardness of the coatings at the same level as the as-deposited coupons. A slight increase for each coating is observed, but as it is in the limits of standard deviation, it can be considered negligible. 3.3. Corrosion resistance The corrosion resistance of the as-deposited and heat treated coatings was investigated by electrochemical polarization curves in a 3.5% w/w NaCl aqueous solution. In Fig. 6a notable shift to more positive potentials in regard to magnesium alloy can be observed for plain coated and duplex-coated coupons. These coatings have a difference of 0.2 V in their corrosion potentials and a difference of 0.081 mm/yr in their corrosion rates, with the duplex deposits appearing nobler and with lower corrosion current density (Table 5). This can be attributed to the different phosphorus contents of the two deposits, the plain ones having a 6.5% w/w phosphorus and the duplex ones being high phosphorus deposits. It has been reported that the electroless Ni-P coatings with various phosphorus content have different electrochemical and mechanical properties, i.e. the corrosion resistance to attack in neutral and acidic environments is increased with phosphorus content in the deposit, whereas the reverse is true in alkaline corrosive environments [7]. Since all the specimens in the present study have approximately the same thickness, it can be deduced that the phosphorus content is a major factor that affects the electrochemical properties of the coatings, which in accordance to several researches [7,32]. The complete amorphous character of high phosphorus deposits provides a better corrosion behavior due to the exclusion of diffusion paths among grain boundaries, which are usually encountered in polycrystalline deposits. The corrosion characteristics of the composite deposits appear similar to the duplex ones and the incorporation of ceramic particles does not seem to have a significant effect. Heat treatment for 2 h at 200 °C transfers the corrosion potential of plain-coated specimens to more negative areas and intensely increases their corrosion current density and consequently their corrosion rate. Micro-cracks and micro-voids that form during heat treatment, due to the difference in the coefficients of thermal expansion between the coating and the substrate that results in development of internal stresses, expose the substrate to attack (Fig. 7). On the other hand, the corrosion potential of duplex non-composite coatings after heat treatment remains unchangeable and the only change in their characteristics lays in the corrosion current density and corrosion rate that are increased by 25 μA/cm2 and 0.25 mm/yr respectively. Composite Ni-P coatings appear unaffected after heat
Fig. 6. Comparative polarization curves of a) as-plated coupons and substrate, b) coupons after heat treatment for 2 h at 200 °C.
treatment (Fig. 6b). The incorporation of ceramic TiO2 and ZrO2 microparticles restrains the anodic dissolution reaction of the deposit by means of reducing its active surface as well as enhances the barrier effect of the coating on the substrate [4,11,19]. This fact is highlighted
Table 5 Corrosion characteristics of magnesium alloy, as-plated and heat treated coatings. TAFEL RESULTS As - plated
Magnesium alloy Plain Ni-P Duplex Ni-P Duplex Ni-P-TiO2 Duplex Ni-P-ZrO2
Ecorr (V)
icorr (μA/cm2)
−1.5 −0.6 −0.4 −0.3 −0.4
350 12.4 4.3 2.4 2.2
Ecorr (V)
icorr (μA/cm2)
Rp (Ohm.cm2)
CorrRate (mm/year)
−1.2 −0.3 −0.4 −0.4
203.5 29.3 5.7 3.8
175 10620 3552 5316
2.034 0.293 0.058 0.039
Rp (Ohm.cm2) 37 3817 8900 8415 9306
CorrRate (mm/year) 7.8 0.124 0.043 0.024 0.022
Heat Treated, 200 °C
Plain Ni-P Duplex Ni-P Duplex Ni-P-TiO2 Duplex Ni-P-ZrO2
438
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
Fig. 7. Optical microscope images (500×) of heat treated coupons a) plain Ni-P, the crack reaches the coupon surface b) duplex Ni-P-TiO2, the crack is hindered by the composite upper layer c) duplex Ni-P-ZrO2, the crack is hindered by the composite upper layer.
by Fig. 7, where the cross-sections of three different heat treated coatings are shown. In Fig. 7a of a plain plated sample, the crack is developed all the way from substrate to surface providing a diffusion path and making the Mg alloy susceptible to corrosion. On the other hand, Fig. 7b and c show that the cracks developed after heat treatment are hindered by the composite layer of Ni-P-ceramic particles. Composite coatings have been characterized as more compact than plain ones. During co-deposition some particles are adsorbed at nodules boundaries, forming, in this way, new nodules at these sites. The composite coating becomes thicker in accordance with this kind of growth model and the tortuous nodule boundaries are not the weak sites of the coating any more, especially when compared with a plain coating [10]. As a result, the protection characteristics of composite coatings are impressively maintained at very high levels, compared to the plain and duplex non-composite specimens. 4. Conclusions Electroless nickel plating was performed on AZ31 wrought magnesium alloy, with different types of obtained deposits: plain
mid-phosphorus deposit and duplex deposit with the first layer having a mid-phosphorus content and the upper one consisting of high-phosphorus. The third type was duplex deposit with a first mid-phosphorus layer and a second high-phosphorus one with incorporated ceramic microparticles. The plain coatings were found to consist of a mixture of nanocrystalline and amorphous phase, with the plated specimens exhibiting crystallinity heterogeneity throughout their surface. The duplex coatings were found to be completely amorphous, a fact that does not affect the microhardness and surface roughness of these two different coatings. Composite deposits offer a slight increase in microhardness, which is either way increased by approximately 1000% in regard to the substrate. No changes in crystallinity, microhardness and surface roughness can be observed after heat treatment at 200 °C for 2 h, as expected. All the coatings provide excellent corrosion resistance, with the most amorphous ones having the lowest corrosion rates due to the exclusion of diffusion paths among grain boundaries. The beneficial effect of the incorporation of ceramic TiO2 and ZrO2 microparticles
E. Georgiza et al. / Surface & Coatings Technology 232 (2013) 432–439
can be observed after heat treatment. The composite layer acts as a stronger barrier than the non-composite one against cracks progression that form due to the internal stresses that develop during heat treatment, thus providing an exceptional protection for magnesium alloy in an aggressive medium. References [1] L.P. Wu, J.J. Zhao, Y.P. Xie, Z.D. Yang, Trans. Nonferrous Met. Soc. 20 (2010) 630–637. [2] A.K. Sharma, M.R. Suresh, H. Bhojraj, H. Narayanamurthy, R.P. Sahu, Met. Finish. 96 (1998) 10–16. [3] T.N. Qin, L.Q. Ma, Y. Yao, C. Ni, X.Y. Zhao, Y. Ding, Trans. Nonferrous Met. Soc. 21 (2011) 2790–2797. [4] Y.W. Song, D.Y. Shan, E.H. Han, Electrochim. Acta 53 (2008) 2135–2143. [5] X.M. Chen, G.Y. Li, J.S. Lian, Trans. Nonferrous Met. Soc. 18 (2008) 323–328. [6] N. Feldstein, in: G.O. Mallory, J.B. Hajdu (Eds.), Electroless plating: fundamentals and applications, American Electroplaters and Surface Finishers Society, 1990, pp. 269–288. [7] A.S.M. Handbook, Surface Engineering, vol. 5, ASM, Material Park, OH, 1991. [8] I.R. Mafi, Ch. Dehghanian, Appl. Surf. Sci. 258 (2011) 1876–1880. [9] Y.S. Huang, X.T. Zeng, X.F. Hu, F.M. Liu, Electrochim. Acta 49 (2004) 4313–4319. [10] Y.W. Song, D.Y. Shan, E.H. Han, Mater. Corros. 58 (2007) 506–510. [11] J.N. Balaraju, T.S.N. Sankara Narayanan, S.K. Seshadri, J. Solid State Electrochem. 5 (2001) 334–338. [12] Z. Liu, W. Gao, Surf. Coat. Technol. 200 (2006) 5087–5093. [13] C. Gu, J. Lian, G. Li, L. Niu, Z. Jiang, J. Alloy. Compd. 391 (2005) 104–109. [14] R. Ambat, W. Zhou, Surf. Coat. Technol. 179 (2004) 124–134.
439
[15] Z. Liu, W. Gao, Surf. Coat. Technol. 200 (2006) 3553–3560. [16] Y.H. Xiang, W.B. Hu, X.K. Liu, C.Z. Zhao, W.J. Ding, T. I. Met. Finish. 79 (2001) 30–32. [17] M. Schlesinge, in: M. Schlesinger, M. Paunovic (Eds.), Modern Electroplating, John Wiley & Sons, Inc., New Jersey, 2010, pp. 447–458. [18] M. Momenzadeh, S. Sanjabi, Mater. Corros. 63 (2012) 614–619. [19] J. Novakovic, P. Vassiliou, Kl Samara, Th. Argyropoulos, Surf. Coat. Technol. 201 (2006) 895–901. [20] J. Novakovic, P. Vassiliou, Electrochim. Acta 54 (2009) 2499–2503. [21] J.N. Balaraju, K.S. Rajam, J. Alloys Compd. 459 (2008) 311–319. [22] J. Sudagar, J.S. Lian, Q. Jiang, Z.H. Jiang, G.Y. Li, R. Elansezhian, Prog. Org. Coat. 74 (2012) 788–793. [23] Sh. Alirezaei, S.M. Monirvaghefi, M. Salehi, A. Saatch, Surf. Coat. Technol. 184 (2004) 170–175. [24] J.N. Balaraju, Kalavati, K.S. Rajam, Surf. Coat. Technol. 200 (2006) 3933–3941. [25] N. Nwosu, A. Davidson, C. Hindle, M. Barker, Ind. Eng. Chem. Res. 51 (2012) 5635–5644. [26] I.R. Mafi, C. Dehghanian, Appl. Surf. Sci. 257 (2011) 8653–8658. [27] W. Sha, X. Wu, K.G. Keong, Electroless copper and nickel–phosphorus plating, processing, characterization and modeling, first ed. Woodhead Publishing Limited, New Delhi, 2011. [28] I. Apachitei, J. Duszczyk, Surf. Coat. Technol. 132 (2000) 89–98. [29] W.D. Callister Jr., Materials Science and Engineering: An Introduction, seventh ed. John Wiley & Sons, New York, 2007. [30] J.N. Balaraju, T.S.N. Sankara Narayanan, S.K. Seshadri, J. Appl. Electrochem. 33 (2003) 807–816. [31] L. Yu, W. Huang, X. Zhao, J. Alloys Compd. 509 (2011) 4154–4159. [32] W.X. Zhang, Z.H. Jiang, G.Y. Li, Q. Jiang, J.S. Lian, Appl. Surf. Sci. 254 (2008) 4949–4955.