Influence of laser surface remelting on microstructure and degradation mechanism in simulated body fluid of Zn-0.5Zr alloy

1. Introduction

The Zn alloy has recently been proposed as a novel and promising biodegradable alloy, because its degradation rate is more suitable compared with the Mg and Fe alloys, and its corrosion products are considered to be innocuous for adjacent tissues [[1], [2], [3], [4]]. For demanding application, the ideal biodegradable implants must provide temporary mechanical support and fully degraded after fulfilling its mission [1]. For instance, the biodegradable stent should remain complete mechanical integrity within at least 6 months in scaffolding and artificial remolding, then completely absorbed in 12-24 months [[5], [6], [7]]. In other words, the metals should hold stable early after the implantation and degrade rapidly later. Nevertheless, previous researches indicate that many Zn alloys immersed in a chloride-rich solution for some times are governed by aggressive localized pitting corrosion [[8], [9], [10], [11]]. The pitting corrosion attacks would break down the device geometry and jeopardize the mechanical integrity, and the inherent strength will deteriorate during the degradation process, which may further lead to unexpected early fracture and peeling of the bulk second phases. As a result, the health of the patients will be threatened. Hence, it’s advantageous to regulate the corrosion rate and change the early stage corrosion mode by modifying the microstructure of the Zn alloys, and thus ensuring the mechanical integrity until sufficiently healing of the tissue.

Laser surface remelting (LSR) can refine the grain surface microstructure, homogenize the alloying elements distribution, and also increase the solid solubility without obvious effects on the substrate [12], so it may be one of the most suitable means to modify the surface microstructure and properties to meet the demands for implants. Furthermore, LSR has been performed to modify the corrosion resistance and biocompatibility of a series of biodegradable Mg alloys [[13], [14], [15]]. Besides, laser machining or cutting are usually used in manufacture of biodegradable implant devices, which could alter the surface microstructure and corrosion behavior, so it is also necessary to concern the responses of alloy to laser in the fabrication. However, very few reports regarding LSR of Zn alloy are available, because the Zn alloy usually is subjected to evaporation during LSR due to its low melting point and very high vapor pressure. Therefore, investigations on influence of LSR on microstructure and corrosion mechanism of the Zn alloy are necessary.

For the reasons above, a suitable heat input combined with Ar gas shielding was used in this work to minimize the damage of laser on the Zn alloy. Then the influence of LSR on microstructure and corrosion mechanism of the Zn-0.5 wt%Zr (Zn-Zr) alloy was investigated. The microstructure was characterized in detail, and the relationship between the degradation mechanism in simulated body fluid (SBF) and the transformation of microstructures was discussed. Briefly the objective is to provide research basis for controlling the surface corrosion behavior and eliminate the peeling of the bulky secondary phase, so as to design smarter and more durable implants.

2. Experimental procedure

2.1. Sample preparation

The as-cast Zn-Zr alloy ingot was used as the experimental material in the present work, which was cut into cuboid blocks with a size of 12 mm × 12 mm × 5 mm by wire spark cutting. Prior to laser treatment, the blocks were mechanically ground by emery papers, polished by diamond polishing slurry, cleaned by anhydrous ethanol and dried in warm air at last. The LSR was performed with a Nd:YAG laser with a pulsed wavelength of 1064 nm and a maximum power of 600 W (Chutian industrial JHM-1GY-600D, China). The LSR parameters were chosen based on some early LSR trials and comparison of the surface quality, and the used LSR parameters are as follows: scanning rate 200 mm/min, pulse frequency 10 Hz, pulse duration 3 ms, beam diameter 0.8 mm, overlapping ratio 50% and laser power 100 W. A coaxial Ar shielding gas flow at the rate of 20 L/min was used to suppress the surface oxidation and evaporation of Zn during LSR, and the LSR without Ar gas shielding was also conducted as a comparison.

2.2. Microstructure characterization

After the LSR process, the surface phase constitution of the original sample and LSR-treated sample was identified by X-ray diffraction (XRD, D8 Advance, Germany) with the Cu-K_α radiation from 10° to 100° at the rate of 6°/min. The surface of the samples was observed with the scanning electron microscopy (SEM, FEI Sirion200, America), then the cross section of the samples was ground with emery papers of different grits, followed by disc polishing using diamond polishing slurry with the size of 0.5 μm. To show the grain structure, the surface was swabbed with an etchant of 3% HCl, 3% HNO₃ and 94% deionized water (volume fraction), rinsed with anhydrous ethanol and dried in warm air. The microstructure of the samples was observed by SEM, and the element distribution was detected with an energy dispersive spectrometer (EDS). The microstructure of the LSR-treated sample and the corrosion products were also probed with SEM (Verios G4 UC, America) at higher magnification.

2.3. Corrosion behavior evaluation

All the samples were pickled with the foregoing etchant for 3 s before immersion in SBF (for SBF composition, see ref. [16] and Methods section). To make the Zn alloy degrade a stable state, the SBF was replaced every two days, simulating the metabolism of in human body. After being soaked in SBF for 5, 12 and 20 days, the samples were rinsed with deionized water and anhydrous ethanol for times, then dried in warm air. After immersing for 20 days, the corrosion products were collected in deionized water by ultrasonic and cleaned by deionized water for times. The X-ray photoelectron spectroscopy (XPS) measurements were carried out on a Kratos Axis Ultra DLD (Kratos Analytical Ltd., Manchester, UK) using monochromatic Al K_α radiation to identify the corrosion products and samples before immersion. The data were converted into VAMAS file and imported into the CasaXPS software for curve-fitting. The binding energies were calibrated using the binding energy of contaminant carbon (C 1s = 284.8 eV). Afterwards, the samples were also observed by SEM and probed again after removing the corrosion products with chromate acid (composed of 200 g/L CrO₃ + 10 g/L AgNO₃).

2.4. Electrochemical characterization

The electrochemical impedance spectroscopy (EIS) measurements were performed with an electrochemical workstation (Princeton M273A) after immersion in SBF for 15 min and 20 days. A conventional three-electrode setup was employed with a frequency range of 10 mHz-100 kHz and a perturbation amplitude of ±10 mV, which consists of a reference electrode (saturated calomel electrode, SCE), counter electrode (platinum sheet), and working electrode (specimens). The polished specimens were molded into epoxy resin with only the remelting surface exposed for the test and the original samples were used as comparison. Simulations were performed using the Zview® (II) software.

3. Experiment results

3.1. Microstructure and element distribution

Fig. 1 shows the XRD patterns and expanded views of the spectrums in the vicinity of 31° to 43° of the original and LSR-treated sample. The XRD pattern of the original sample shows strong Zn peaks and weak Zn₂₂Zr peaks, while the pattern of LSR-treated sample presents the broad Zn peaks with indistinguishable Kα₂ lines and some weaker broad peaks associated with Zn₂₂Zr due to the decreased grain size. The Zn peak intensity of LSR-treated sample changes as the grain orientation shifts. The third phase peaks, i.e. the peaks of ZnO, appear and only appear in the sample LSR-treated in air, indicating that the Zn was oxidized during the LSR in air. Furthermore, according to the well-known Scherrer Equation, the full width at half maximum of peaks will rise as the crystallite size reduces. Therefore, the peaks for smaller crystallites become broad and some adjacent peaks in 2θ axis are overlapped. On the other hand, the integral area of every Zn₂₂Zr peak, which represents peak intensity, keeps a basic constant due to the unchanged content, and hence the height of the peaks decreases because of the increased width. Taken together, the peaks of Zn₂₂Zr are weak and broad. The results confirm that samples mainly consist of Zn and Zn₂₂Zr, and the oxidization of Zn during LSR was restrained by Ar gas shielding.

View original graphic| Download| PPT slide

Fig. 1.   XRD patterns of original and LSR-treated samples.

The surface morphologies and element distribution of the original sample are exhibited in Fig. 2. As depicted in Fig. 2(a), the original sample consists of near-equiaxed grains and some randomly dispersed secondary phases. The greater enlargement (Fig. 2(b)) shows the size of the grains and protruding bulk phases is tens of micrometers in various sizes. According to the EDS results and the Zn-Zr phase diagram [17], the irregularly shaped phase with white edge is predicated to be Zn₂₂Zr, and the principal phase is Zn-rich phase with little Zr. A variation in distribution of the Zn and Zr elements is apparent from the elemental maps, as in Fig. 2(c) and (d), the corresponding element distribution diagram demonstrates that the Zr concentrates in the Zn₂₂Zr. The concentration sites are consistent with the Zn₂₂Zr phase in the original sample.

View original graphic| Download| PPT slide

Fig. 2.   Surface microstructures of (a), (b) the original sample and the corresponding element distribution of (c) Zn and (d) Zr in (b).

The surface microstructure and element distribution of the LSR-treated sample are shown in Fig. 3. Fig. 3(a) exhibits the surface morphology of the sample treated in air, without Ar gas shielding. It can be seen that the surface contains lots of oxides and suffers severe ablation, which fits with the XRD result. With the Ar gas shielding, the surface is relatively smooth, with no severe ablation and oxidation (Fig. 3(b)). The laser track composed of overlapping laser spots is clearly distinct with a width of about 400 μm. Fast cooling of the molten metal leads to the formation of near-equiaxed dendrites with a size of about 5-20 μm, as in Fig. 3(c), which presents the microstructure of the remelted surface. A detailed view in Fig. 3(d) shows that the dendrites are highly hierarchical branched patterns with primary-, secondary- and higher-order branches, with an exceedingly small secondary arm dendrite spacing of about 100 nm. According to the EDS results, the white precipitation is still Zn₂₂Zr. The SEM observation results illustrate that the Zn₂₂Zr phase of the upper layer melts during LSR and solidifies into dendritic microstructure within a short time. Also, it can be found that the Zn and Zr elements disperse evenly throughout the remelted surface (see Fig. 3(e) and (f)).

View original graphic| Download| PPT slide

Fig. 3.   Surface microstructures of the LSR-treated sample: (a) in air and (b) with Ar gas shielding; (c), (d) microstructure of the remelting layer and the corresponding distribution of (e) Zn (f) Zr in (c).

3.2. Corrosion behavior

Fig. 4 shows the surface morphologies of the original sample immersed in SBF for different time. After being immersed for 5 days, some white clusters (corrosion products) appear on the surface (Fig. 4(a)) and a higher magnification image (upper right corner) shows that the corrosion products mainly consist of white particles and a dark layer. Fig. 4(b) reveals that the pits initiate at the grain boundaries of the Zn-rich phase. When the immersion progresses to 12 days, the concentration of the corrosion products occurs at some areas (Fig. 4(c)), and many pits pile up together to form bigger and continuous shallow pits under the concentrated corrosion products (Fig. 4(d)). Fig. 4(e) reveals that the white corrosion products deposit compactly in quantity at some certain areas after immersed for 20 days. After removal of the corrosion products, a large hole can be observed in the Zn matrix in addition to lots of separated pits along the grain boundaries, as shown in Fig. 4(f).

View original graphic| Download| PPT slide

Fig. 4.   Surface morphologies of the original samples following immersion in SBF for 5, 12, and 20 days: (a), (c), (e) with corrosion products, (b), (d), (f) after removal of the corrosion products.

Fig. 5 shows the surface morphologies of the LSR-treated Zn-Zr alloy immersed in SBF for different time. At the time of 5 days, it is clear that the corrosion products begin to appear, and the distribution is similar to that of the original sample (Fig. 5(a)). Localized corrosion occurs in some areas and some dendrites are revealed partially, as presented in Fig. 5(b). After an immersion of 12 days, the corrosion products show an aggregation phenomenon, as in Fig. 5(c). Once the corrosion products were removed, a more severe corrosion with larger area is found (see Fig. 5(d)), in which some white dendrites and densely distributed micropores are observed. After 20 days, a lot of corrosion products accumulate and distribute loosely throughout the surface, as shown in Fig. 5(e). Removing the corrosion products, it can be observed that most of the surface of the alloy has been corroded, and only a few small island areas with original surface remains (Fig. 5(f)). The dendrites can be delineated by plenty of white Zn₂₂Zr precipitation, because the Zn₂₂Zr shows higher corrosion resistance and is more likely to be left after a thin layer is eroded.

View original graphic| Download| PPT slide

Fig. 5.   Surface morphologies of the LSR-treated samples following immersion in SBF for 5, 12, and 20 days: (a), (c), (e) with corrosion products; (b), (d), (f) after removal of the corrosion products.

Fig. 4, Fig. 5 display the corrosion process of the original and LSR-treated specimens. As the immersion time goes on, the corrosion products of both the samples increase constantly. For the original sample, the corrosion products concentrate to form a compact layer covering on a site. In contrast, more loose corrosion products distribute throughout the LSR-treated surface, serious localized corrosion does not occur. In general, it is confirmed that the corrosion mode of the Zn-Zr alloy in the SBF converted from pitting/localized corrosion to uniform corrosion by the LSR. From the corrosion morphologies, it is predicted that the corrosion process of the original alloy in SBF can be classified into three stages: the initial stage with independent pits formation, the second stage with pits connection and the final stage with development of localized corrosion. For the LSR-treated specimen, the corrosion also initiates at some small pits, but the corrosion spreads around rather than develops into the depth as that in the original sample.

The XPS spectra of corrosion productions and samples before immersion are presented in Fig. 6. The survey scan of samples before immersion suggests the existence of Zn element (Fig. 6(a)), but it is difficult to split the Zn peak into the Zn-rich phase peak and the Zn₂₂Zr peak, because the binding energies of the photoelectrons from the Zn of the two phase are almost the same. The appearance of O 1s peaks should be contributed to the inescapable oxidation of alloy in air, and the content of Zr is so low that it cannot be detected. Fig. 6(b) illustrates the survey scan of corrosion products, which confirms the existence of Zn and P elements in products, while the C and O elements need to be further confirmed from detail scan. Fig. 6(a) and (b) indicates the XPS result of the LSR-treated sample is almost the same with that of the original sample, because the chemical composition is not altered by LSR. Fig. 6(c) shows a peak associated Zn²⁺ appears at 1022.3 eV. The chemical state differentiation of Zn is difficult to split because of small binding energy (BE) shift in this region, while the composition can be detected by other element spectrums. The peak of C 1s spectrum at the binding BE of 284.8 eV is considered to be adventitious carbon contamination in the environment and the contribution of carbonate species locates at 286.5 eV and 288.6 eV (see Fig. 6(d)) [18], suggesting the formation of ZnCO₃. The single peak at 134.0 eV in P 2p spectrum shown in Fig. 6(e) confirms the existence of PO₄^3-. Fig. 6(f) depicts that the spectrum of O 1s is decomposed by three contributions at BE = 530.0, 531.8 and 533.2 eV, which are related to O^2-, OH^- and average BE for PO₄^3- and CO₃^2- [19,20], respectively. The result simultaneously shows the PO₄^3- and CO₃^2- peak is dominant in O 1s spectrum and the O^2- peak intensity is very weak. No peaks assigned to Zr are detected, which can be attributed to the good corrosion resistance of Zn₂₂Zr. Based on the XPS results and some previous researches [3,7,10,[21], [22], [23], [24]], it can be induced that the corrosion products are ZnCO₃, Zn(OH)₂, Zn₃(PO₄)₂·2H₂O and a little ZnO.

View original graphic| Download| PPT slide

Fig. 6.   XPS spectra: survey scan for (a) samples before immersion, (b) corrosion products of samples after immersion for 20 days and detail scan for corrosion products of original sample (c) Zn 2p, (d) C 1s, (e) P 2p and (f) O 1s.

3.3. Electrochemical impedance spectroscopy

The Nyquist diagrams and Bode plots for the original and LSR-treated alloy recorded at open circuit potential (OCP) after 15 min and 20 days immersion in SBF are presented in Fig. 7. To get further information about the mechanism, the impedance spectra in the complex plane are fitted using suggested equivalent circuits in Fig. 7. In all of the proposed equivalent circuits, R_s is the solution resistance, R_f-Q_f is attributed to the formed corrosion product film properties, R_ct is parallel with the Q_dl corresponds to the charge transfer resistance and double layer capacitance at the sample electrode interface. Also, Q_pit and R_pit in Fig. 7(g) refer to the capacitance and resistance of localized pitting corrosion occurred in original alloy after 20 days immersion in SBF. Moreover, it can be seen that the equivalent circuits shown in Fig. 7(e) and (h) for original alloy after 15 min immersion and LSR-treated alloy after 20 days immersion contain an additional element, W, which is called Warburg impedance for diffusion processes. Its components (W-R, W-T and W-P) have ohmic resistance, capacitance and exponent significance, respectively. A Warburg element occurs when charge carrier diffuses through a material. The electrical equivalent circuits are designed on the basis of the idea of achieving the best fit for the observed experimental EIS data and also physicochemical model of the system. The parameter values obtained by fitting the spectra are presented in Table 1.

View original graphic| Download| PPT slide

Fig. 7.   Nyquist diagram and Bode plots of samples immersed in SBF for (a), (b) 15 min and (c), (d) 20 days; equivalent circuits proposed for original sample immersed in SBF for (e) 15 min and (f) 20 days and LSR-treated samples immersed in SBF (g) for 15 min and (h) for 20 days.

The results show that the proposed equivalent circuit for the LSR-treated sample after 15 min immersion in SBF consists of two time constants, while the Warburg element for the original sample appears in low frequencies. Also comparing data show that the R_ct value for the LSR-treated alloy is about 986 Ω cm², which is higher than that of the original alloy ∼368 Ω cm², while R_f of the original alloy is higher than the LSR-treated one (R_{f, original alloy} = 401 Ω cm² and R_{f, LSR-treated} = 260 Ω cm²). After an immersion of 20 days, the appearance of the third time constant in both Nyquist (three semicircles in Nyquist plot can be seen) and Bode plots of the original alloy may associate with localized severe corrosion in some pits. The LSR-treated sample still has two time constant but an extra Warburg element W is proposed for the dissolution of hydroxide layer.

4. Discussions

4.1. Formation mechanism of microstructure after LSR

The SEM results present that coarse grains transformed into fine dendrites after LSR, which can be attributed to high cooling rate. The granular texture to dendrites transition is subjected to the degree of constitutional undercooling. It is accepted that crystal growth behaviors are affected by the undercooling, and the grain mode changes from planar to cellular, then to columnar dendritic, and finally to equiaxed dendritic as the degree of undercooling continue to increase [25,26]. During the LSR process, the temperature gradient between the melting pool and the substrate is very high when the surface is melted by laser, and the substrate with good thermal conductivity can act as heat sink after the laser pulse is shut, so the undercooling of the liquid metal is high. It is indicated that the morphological instability of the solid-liquid interface can produce dendrites spatially and the inhomogeneous distribution of impurities will form dendritic network [27]. Hence, the dissolved secondary phases in the Zn matrix may become nucleation sites and precipitate along the dendrite arms. Besides, it has been referred that perturbation of the higher cooling rate could reduce the dendrite spacing to a narrower range than the steady state growth condition [28]. The abundant dendrite arms influence each other and the insufficient growing space prevents the neighbor dendrites from growing bigger. Meanwhile, consistent with the homogeneity of the secondary phases, the Zn and Zr elements distribute uniformly.

4.2. General corrosion mechanism of Zn alloy

According to the XPS results and some previous reports, the corrosion of Zn in SBF can be demonstrated by the following chemical reaction [29]:

Anodicreaction:Zn-2e^-→Zn²⁺(1)

Cathodicreaction:O₂+2H₂O+4e^-→4OH^-(2)

The interaction of released Zn ions (Zn²⁺) with hydroxyl ions (OH^-) results in built-up of the hydroxide film and formation of zinc oxide and zinc carbonate [3,24,29]:

Zn²⁺+2OH^-↔Zn(OH)₂(3)

Zn²⁺+CO₃^2-↔ZnCO₃(4)

Zn²⁺+2OH^-↔ZnO+H₂O(5)

The Zn(OH)₂ can transform into more stable ZnO over time [18]. The XPS results confirm the content of ZnO is very low because the PH of the SBF ranges from 7.3 to 7.4, which is not a stable range for ZnO [23,30]. The abundant Cl^- ions from SBF are aggressive to the Zn alloy and will attack the hydroxide film. The film dissolution can be expressed as below [11]:

Zn(OH)₂+2Cl^-→Zn²⁺+2Cl^-+2OH^-(6)

With the PH value of the SBF, the Zn₃(PO₄)₂·2H₂O is the most stable product from the calculated Pourbaix diagram of Zn in SBF [23]. The phosphate ions (PO₄^3-) from SBF reacts with the Zinc ions to form Zn phosphate (Zn₃(PO₄)₂·2H₂O):

2Zn²⁺+3PO₄^3-+2H₂O↔Zn₃(PO₄)₂·2H₂O (7)

Some researches have provided further evidence on the formation of Zn phosphate in vivo or vitro study [11,23,24]. In summary, the ZnCO₃, Zn(OH)₂ and Zn₃(PO₄)₂·2H₂O are main components of the corrosion products, the Zn(OH)₂ will transforms into insoluble phosphate over time.

4.3. Corrosion processes

Fig. 8 exhibits an illustration on the corrosion processes of the original and the LSR-treated samples. For the as-cast Zn-Zr alloy, the grain boundaries represent high energy because of the atomic mismatch. A model for pit initiation based on the reactivity of the oxide grain (or hydroxide) boundaries was presented [31]. This model considers the oxide grain boundary regions as generic inter-granular defective sites separating the different grains and the sites of oxide breakdown due to the lower resistance to ionic transport. The Cl^- competes with OH^- adsorption and accelerates ion transport at the localized defect sites. Therefore, the corrosion begins at these specific sites where the hydroxide film firstly dissolves or thins and some small pits firstly appear along the grain boundaries (Fig. 8(a)). With increasing immersion time, the pits become bigger and continually develop towards the interior of alloys, connecting with each other to form bigger shallow pits, as depicted in Fig. 8(b). Simultaneously, the corrosion products also begin to aggregating and depositing progressively. Finally, the corrosion develops continuously in depth (Fig. 8(c)), the corrosion products cover compactly over the big shallow pits and the pits would be sealed completely over time. The Cl^- ions of small radii migrate towards the bottom of pits to remain charge balance, which can improve the conductivity of solution, resulting in lower potential and accelerated the release of Zn²⁺ as Eq. (1). Otherwise, it is hard for the O₂ molecules of large radii to diffuse into the pits through the corrosion product. The cathodic reaction cannot sustain due to the depletion of O₂ molecules in the pits. The abundant O₂ molecules in the solution could react with water molecules as Eq. (2) over the metal surface. Therefore, galvanic coupling forms between the surface and the pits bottom. That is to say, the Zn matrix of pits acts as an anode and the SBF in contact with the metal surface, which generates hydroxide ions acts as a cathode. The high Cathodic-Anodic area ratio promotes the dissolution of metal [32]. The deposition of corrosion products and dissolution of Zn promotes the action of one another, so the corrosion rate increases significantly in this region as the corrosion proceeds. The pits spread rapidly into the depth, while most of the rest area was hardly affected.

View original graphic| Download| PPT slide

Fig. 8.   Schematic diagram on corrosion of the (a)-(c) original and (d)-(f) LSR-treated samples in SBF.

In this work, a progressively faster localized corrosion of the casting Zn-Zr alloy in SBF is observed and similar results in simulated physiological solution or body have been reported recently [11,21,[33], [34], [35]], which illustrates that the severe localized corrosion is common for the Zn alloy. This corrosion mode may of great harm to implants since a small amount of corroded material could result in serious surface defects and fast decrease of mechanical integrity. Additionally, the Zn₂₂Zr phase in the matrix would be partly exposed due to deep undermining and it can be inferred that some may peeling from the alloy into the body fluid, causing the inflammation or other physical damage when it’s implanted into the body.

The LSR improves the uniformity of microstructure and composition in the remelting layer and hence the susceptibility to pitting corrosion decreases with immersion for a short time. However, there are still some sites of high energy acting as anodes and sites of low energy acting as cathodes, possibly due to the compositional nuance on a micro level. Similarly, the corrosion begins at these sites act as small pits and spreads around over time (Fig. 8(d) and (e)). Nevertheless, the LSR-treated sample undergoes noticeably different corrosion behavior when the corrosion process developed to the later stage. The corrosion pits continue to develop towards the surrounding sites rather than into the depth (Fig. 8(f)), caused by two major reasons: the corrosion-resistant of the Zn₂₂Zr phases and the looser corrosion products. On one hand, the Zn₂₂Zr is not only less prone to corrosion compared with the Zn matrix, but also refined after LSR, which can effectively prevent the aggressive chloride from absorbing on the surface [36]. After a thin layer is eroded, the abundant Zn₂₂Zr precipitation at the bottom of pits inhibit the pits penetrating into the depth. On the other hand, the pristine dendritic Zn₂₂Zr also prevents the corrosion product from depositing continuously and firmly, so it is enough for the O₂ to migrate into the pits easily. Therefore, there is a lack of indispensable condition for serious pitting corrosion. Taken together, the remelting layer undergoes the uniform corrosion with uniform thickness reduction.

4.4. Electrochemical impedance spectroscopy

As mentioned above, the coarse grains transform into uniform tiny dendritic structures with exceedingly small dendrite arm spacing by LSR. Also, it is indicated that the dendrite microstructures are in fact Zn₂₂Zr phases that show higher corrosion resistance as compare to Zn matrixes. Therefore, after an immersion in SBF, homogenous and thin Zn(OH)₂ film can form on the surface of the LSR-treated alloy. This film as well as the dendrites acts as a barrier for accessing more corrosive agents and hence the R_ct increases. While in the case of the original alloy, because of the coarse grains and randomly dispersed second phases, heterogeneous and a bit thicker film may form on the Zn matrix, so R_f is higher than that of the LSR-treated one. On the other hand, local thinning/dissolution will appear at the film grain boundaries, so that corrosive agents can diffuse and reach the substrate and consequently accelerate the corrosion of the substrate, which leads to the appearance of Warburg element in the corresponding equivalent circuit. As a result, after immersion in SBF for a long time, the original sample suffers severe localized pitting corrosion while the LSR-treated alloy is almost safe and only the Warburg element appears, dues to the destruction in thin hydroxide film. Appearance of the third time constant in both Nyquist and Bode plots of original sample after 20 days immersion in SBF further confirms the occurrence of pitting corrosion.

4.5. Application prospect

The LSR can refine the metal microstructures, making transition from bulky grains to equiaxed dendrites, and the corrosion mode also transforms from localized pitting corrosion to uniform corrosion, which provides maximum protection for the mechanical properties of the implants and prevent large particles form peeling off. These results are interpreted with the transition of microstructure that affect the formation of corrosion product and are consistent with the EIS results. From the above analysis, the key of corrosion mode transformation is the highly uniform anticorrosive phases distribution in microscopic. Therefore, it is predicted that this method can apply not only to the Zn-Zr alloy, but also to other Zn alloys that can form anticorrosive intermetallics and fine dendrites. The investigations on degradation mechanisms of LSR treated Zn alloys within extended immersion time (over 6 months) are ongoing. It is hoped that the investigation results can provide useful guidance to optimize the corrosion behavior in a particular application by microstructure transition not only limited to the LSR.

5. Conclusions

The influence of LSR on microstructure and degradation mechanism of Zn-Zr alloy in simulated body fluid was investigated. Based on the results, the following conclusion can be drawn:

(1)The coarse grains of the as-cast Zn-Zr alloy changed into fine hierarchical dendrites with small arm spacing by the LSR. The bulky Zn₂₂Zr phase is dispersed uniformly and the distribution of Zn and Zr elements are homogenized.

(2)The original sample is susceptible to severe pitting corrosion initiates at the grain boundaries and develops into deep corrosion holes with the corrosion proceeds. The corrosion of the LSR-treated sample starts from subtle small area and propagates towards the whole surface, forming a homogenously corroded surface.

(3)The transformation from localized corrosion to uniform corrosion is induced by the change of corrosion microenvironment. The Zn-rich phase in the original alloy has poor corrosion resistance and the aggregated corrosion products result in further localized corrosion. The uniformly dispersed Zn₂₂Zr and loose corrosion products in the LSR-treated specimen prevent the corrosion from penetrating into the depth.