Yuxing Liu, Michele Curioni, Shiyun Dong, Zhu Liu,*

a Department of Materials, School of Natural Sciences, The University of Manchester, Manchester, M13 9PL, UK

b National Key Laboratory for Remanufacturing, Beijing, 100072, China

Abstract In this study, a KrF excimer laser was used to modify the biodegradable Mg-1Ca alloy and the time-evolution degradation behavior of the alloy before and after laser treatment was investigated in simulated body flui (SBF) solution using immersion tests and electrochemical impedance spectroscopy (EIS).A 5 μm melted layer with a homogeneous microstructure and an MgO fil on the surface were achieved by laser radiation.Corrosion observations (hydrogen evolution, morphology and corrosion products) and EIS results revealed an improvement of corrosion resistance after laser treatment for 48 h.It was found a two-layer structure developed after 2 h immersion on both the untreated and laser-treated alloys, but the sequence of forming the two layers was opposite and greatly influence by the laser-treated layer.The time-evolution corrosion processes on the untreated and laser-treated alloys were discussed, providing a better understanding of corrosion behavior of biodegradable Mg alloys modifie by excimer laser.

? 2021 Chongqing University.Publishing services provided by Elsevier B.V.on behalf of KeAi Communications Co.Ltd.

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Peer review under responsibility of Chongqing University

Keywords: Magnesium; Biodegradable implant; Excimer laser; Simulated body fluid Corrosion; EIS.

1.Introduction

Biodegradable magnesium (Mg) alloys are drawing extensive attention as medical implant materials for orthopedic applications since they have comparable mechanical properties to the natural bone and therefore mitigate the effect of “stress shielding''[1,2]generally induced by permanent metallic biomaterials.The degradable nature of Mg derives from its corrosion rate in aqueous solutions.Mg alloys in physiological media degrade at a relatively high rate due to the low potential for magnesium oxidation with respect to the hydrogen evolution reaction combined with the formation of a relatively non-protective corrosion product fil [3,4].However,for biodegradable applications,magnesium alloys should have a tunable dissolution rate that ensures that the implants in the human body maintain their mechanical integrity until the surrounding tissues have healed.In addition, the degradation products should be nontoxic and they should be formed at a rate that is slow enough such as they can be consumed or absorbed by the surrounding tissues.

Compared with other biodegradable Mg alloys alloyed with Al, Mn, Zn and RE, Mg-xCa alloys have excellent biocompatibility and the cations produced during degradation in the human body do not interfere significantl with cell functions[1,5-7], thus Mg-xCa alloys have been considered as promising candidates for medical implants.Meanwhile, the properties of Mg-xCa alloys are greatly influence by the Ca contents.Previous studies [7-11]revealed that Mg-1 wt.% Ca alloy has excellent biological compatibility and adequate mechanical properties.Also, it has the optimum anti-corrosion performance [12,13]in solutions containing Cl-.However,the corrosion rate of Mg-1Ca alloy is still excessive, andefforts to reduce it are needed if the alloy is to be used as implant material.

Among the approaches that can improve corrosion performance of Mg alloys, coatings including fluorid conversion coatings,chemical solution deposition of hydroxyapatite coatings and plasma electrolytic oxidation coatings[7,14-19]have been considered.However, the materials of coatings generally differ from Mg alloys and a specifi selection of such materials is needed for orthopedic applications.An otherwise effective approach to improving corrosion performance of Mg alloys is laser surface processing that can tailor the surface microstructure and composition of the biomaterials without introducing additional chemical species.Thus, laser surface modificatio is drawing increasing attention for orthopedic applications of Mg alloys [20-22].Among the various lasers,the excimer laser is characterized by high energies and short pulses, together with low reflect vity in the UV range; this enables efficien modificatio of thin surface layers, without thermal loading of the underlying materials [23-25].For Mg alloys, excimer laser normally favors the refinemen of the microstructure [24, 27-36], the improvement in mechanical properties [26,27,34]and anti-corrosion performance in NaCl solutions [23,24,26-36].For example, Coy et al.[30].applied a KrF pulsed-excimer laser treatment on a die-cast AZ91D alloy and investigated the microstructure evolution and corrosion behavior of the alloy in 3.5 wt.% NaCl solution.They attributed the improvement of the corrosion resistance after laser treatment to the highly homogeneous and refine melted microstructure on the top surface of the laser-treated alloy.In this respect, excimer laser offers significan benefit and is promising to be used for treating material for orthopedic applications.However, with respect to the degradation behavior of the excimer laser-treated Mg-1Ca alloy in SBF solution,the effects of the excimer laser are not well-understood and the corrosion behavior has not been fully investigated.

To gain a better insight into the time-evolution corrosion process of the biodegradable Mg-1Ca alloy treated by excimer laser, in this work, immersion tests and electrochemical impedance spectroscopy (EIS) measurements, rather than polarization were conducted in a simulated body flui (SBF) solution.For a corrosion system, EIS is a quick, non-destructive and steady-state technique and its responses are generally obtained by applying a small amplitude of alternate current(AC)over a range of applied frequencies [37-39].By fittin EIS data to an equivalent electrical circuit model, the corrosion resistance (impedance) and meaningful physical interpretation of the studied specimen are attained.However, for Mg electrodes,corrosion rates obtained by polarization are not always reliable because ohmic drop is very prominent, causing severe distortions of polarization curves [37,40-41].Moreover,the 'negative difference effect' (i.e.enhanced hydrogen evolution at anodic polarization) [4,42-44]on Mg electrodes further distorts the anodic polarization curves.As a result, an erroneous estimation of corrosion rate is derived if the Tafel extrapolation of the polarization curves is simply used.More importantly,for the excimer laser-treated Mg alloys,the lasertreated layer in a thickness of a few micrometers is easily destroyed by polarization testing, thus it is unlikely to provide reliable corrosion rates and a convincing time-evolution corrosion process.In this study, we firs characterized the modification of microstructure and compositions in the nearsurface region of the Mg-1Ca alloy after a KrF excimer laser treatment.Then we investigated the corrosion performance of the alloy before and after laser treatment using immersion tests and EIS.In addition to the corrosion resistance, the corrosion morphology and corrosion compounds are characterized and compared.Based on the results, the time-evolution of corrosion process is discussed and a better understanding of corrosion behavior of the biodegradable Mg-1Ca alloy modifie by excimer laser is established.

2.Experimental procedure

2.1.Specimens preparation

A cast Mg-1Ca alloy was used in this study.The material was cut into specimens with dimensions of 10 × 10 × 5 mm for excimer laser treatment.Prior to laser treatment, the specimens were ground to a 2400 SiC grit finis and then dried in cold air.Laser treatment was carried out in an ambient environment, using a 80 W Lumonics KrF pulsed-excimer laser with a wavelength of 248 nm and pulse width of 15 ns.For the laser processing, a series of experiments were conducted using laser fluence (J/cm2) of 2.6, 3.3 and 4.0 and numbers of pulses (NOP) at each point of 10, 20 and 30, with a fi ed frequency of 10 Hz and an overlap ratio of 20%.The optimized condition (3.3 J/cm2fluenc and 20 NOP) was determined by the results of relatively uniform melted depth,minimum porosity and absence of cracks.In this work, the optimized laser condition was used to produce the specimens for corrosion studies.

2.2.Microstructural characterization

Materials characterization of the specimens before and after excimer laser treatment was undertaken by scanning electron microscopy (SEM) using a Philips XL30, equipped with energy dispersive X-ray (EDS) and backscattered electron(BSE) detectors.Prior to the SEM observation, the surface of the untreated specimens and the cross-section of the lasertreated specimens were polished after grinding to a 2400 SiC grit finish the surface after laser radiation was directly observed by SEM.The phases of the specimens before immersion were determined by low-angle X-ray diffraction (XRD)using Bruker D8 X-ray diffractometer, with Cu Kαradiation under a 40 mA current, a 40 kV voltage and a 2° angle diffraction.The analysis of XRD peaks and the phase amount calculation were conducted using MDI Jade 6.5 (Materials Data.Inc., USA).

2.3.Corrosion evaluation

Prior to the immersion tests,the specimens of the untreated and laser-treated Mg-1Ca alloy were sealed by lacquer with a1 cm2exposed surface.The specimens were immersed in the SBF solution at the temperature of 36.5 ± 0.5 °C in a water bath.The SBF used in this work with ion concentrations nearly equal to those of human blood plasma.The preparation of SBF solution (1000 mL) is based on the recipe reported by Kokubo et al.[45].with the following reagents and amounts: NaCl (8.035 g), NaHCO3(0.355 g), KCl (0.225 g),K2HPO4·3H2O (0.231 g), MgCl2·6H2O (0.311 g), 1.0MHCl (39 ml), CaCl2(0.292 g), Na2SO4(0.072 g) and Tris(6.118 g).After immersion for a period, the specimens were gently washed in deionized water, then immediately dried in cold air.The surface and cross-sectional images at different immersion time were recorded by the BSE detector of Zeiss SIGMA, equipped with an energy dispersive X-ray (EDS) detector.

The time-evolution corrosion rate of the specimens in the SBF solution was evaluated by corrosion depths.The crosssectional images of the corroded specimens were firstl captured and then spliced by the optical microscope (JVC TKC1380 color video camera).Subsequently, corrosion depths were measured on the panoramic cross-sectional images using Image J software.The corrosion depths were obtained by measuring three points on two different specimens after an immersion duration and the average values were reported.

Corrosion resistance in the SBF solution was assessed by electrochemical impedance spectroscopy using a VersaSTAT 4 potentiostat/frequency response analyzer.The specimens with an exposed area of 1 cm2were immersed into the SBF solution at 36.5 ± 0.5 °C.A three-electrode electrochemical cell was used, with a saturated calomel electrode (SCE) as the reference electrode, a platinum sheet as the counter electrode and the mounted specimens as the working electrode.The open-circuit potential (OCP) of the working electrode was monitored in real-time for 24 h immersion in the SBF solution.The EIS experiments were performed after corresponding OCP tests at different immersion time.A sinusoidal potential perturbation of 10 mV was applied for EIS experiments over the frequency range of 105-10-2Hz.In all the measurements of OCP and EIS, at least three similar results were considered and reported.

3.Results and discussion

3.1.Microstructure and composition

Fig.1 shows the microstructure of Mg-1Ca alloy and its EDS elemental maps.The bright regions belong to intermetallic particles and grain boundaries that have a higher concentration of calcium.In the matrix,α-Mg is the primary phase and the phases in the intermetallic particles and grain boundaries comprise theα-Mg phase andβ-Mg2Ca phase.Moreover, the intermetallic particles distribute in the grains in either small size (＜0.5 μm) or relatively large size (＞6 μm).The EDS results confir the locations of the phases and the particles.

Fig.2 shows the surface and cross-sectional SEM images and the corresponding EDS elemental maps of the laser-treated alloy.In Fig.2a,laser treatment produces a rough rippling structure on the surface, which is typical for metals after excimer laser radiation [28,30,46-47].In addition, small pores(the yellow arrows)were observed on the surface,probably due to the release of hydrogen that was initially dissolved in the as-cast magnesium matrix.During the laser treatment,the hydrogen atoms combined to form hydrogen molecules that escape from the melt pool [30,48].However, due to the rapid solidificatio process, some hydrogen molecules were trapped in the solidifie Mg-1Ca alloy, forming the pores in the laser-treated layer.The insert SEM image in high magnificatio in Fig.2b shows a fin platelet-like structure [49,50]formed by oxidation of the alloy during laser treatment.The elemental maps corresponding to Fig.2a show that the particles and grain boundaries disappear after laser treatment and Ca is deficien on the rippling laser-treated surface compared to that on the untreated alloy in Fig.1.The deficien y may be caused by the dissolution of the Ca-rich particles and grain boundaries in theα-Mg matrix.The oxygen map confirm that the oxidation of the alloy occurs during laser treatment.In Fig.2c, excimer laser produces a ～5 μm layer in thickness with small pores (the red arrows).Moreover, the grain boundaries rich in Ca dissolve in the laser-treated layer, resulting in a homogeneous surface microstructure.However,there is a zone rich in Ca (the yellow arrows) in the lasertreated layer, corresponding to the Ca elemental distribution on the laser-treated surface in Fig.2a.This small amount of Ca in the laser-treated layer implies that the dissolution of Ca is incomplete because the heat generated by laser radiation is insufficien to melt theβphases in some regions.Such regions may dissolve preferentially in the SBF solution because Ca is more active than Mg.Moreover, O is abundant in the laser-treated layer especially on the top of the surface,confirmin the existence of oxides on the laser-treated alloy surface.

Fig.3 shows XRD patterns of the untreated and lasertreated Mg-1Ca alloy.Visible diffraction peaks of theα-Mg phase andβ-Mg2Ca phase are define for the untreated Mg-1Ca alloy.After laser treatment, the diffraction peaks ofβ-Mg2Ca phases are not apparent.By calculating,the fraction ofβ-Mg2Ca phases on the untreated surface(4.225±0.239 wt.%) is over 4 times that on the laser-treated surface (0.928±0.067 wt.%), which confirm the dissolution of calcium in the Mg matrix during laser radiation.In addition, the diffraction peaks of MgO are observed on the lasertreated alloy, which is consistent with the EDS results in Fig.2.Notably, theβ-Mg2Ca phase was reported to have a more negative corrosion potential compared to theα-Mg phase [5,51], thus it acts as an anode during the galvanic corrosion to undergo a high dissolution rate at the expense of theαphase.

3.2.Hydrogen evolution during immersion

Fig.4 shows the appearance of corroding surfaces of the untreated and laser-treated Mg-1Ca alloy with immersion time in the SBF solution.A substantial amount of large hydrogenbubbles releases from the corroding surface of the untreated alloy whereas hydrogen gas bubbles are significantl less on the corroding surfaces of the laser-treated alloy.However, the cloudy hydrogen streams formed by many small gas bubbles are apparent above the laser-treated alloy surface.For both the untreated and laser-treated alloys,the release rate of hydrogen gas decreases with time, indicating the decreasing corrosion rate during the 60 min of immersion.

3.3.Corrosion morphology

3.3.1.Corroded surface morphology

Fig.5 shows the corroded surfaces of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution for 5 and 30 min respectively.For the untreated alloy after 5 min immersion in Fig.5a, theβ-Mg2Ca phases located at the grain boundaries are preferentially corroded and subsequently covered by the platelet-like corrosion products (region A and the corresponding EDS mapping) that mainly consist of Mg oxide/hydroxide.The oxide/hydroxide can prevent corrosion from proceeding along the grain boundaries but accelerate the corrosion of theα-Mg matrix.For the laser-treated alloy in Fig.5b, corrosion is insignifican after 5 min except that some pits (white arrows) develop on the Ca-rich regions, as shown in the corresponding EDS image.After 30 min, the corroded surface of the untreated alloy (Fig.5c) is covered by corrosion products and it has two different regions,i.e.relatively smooth regions and cracked regions.This difference can be attributed to the different compositions of corrosion products in different regions.The EDS results confir that the smooth regions are mainly covered by corrosion products (oxide/hydroxide) rich in Mg and O and the cracked regions are covered by both corrosion products and chemical precipitates (phosphates/carbonates) rich in P, Ca, C and O.Moreover, some dome-shaped compounds (white arrows)develop and are covered by chemical precipitates.Similarly,the laser-treated alloy after 30 min (Fig.5d) is also characterized by the crack chemical precipitates (mainly Ca phosphates/carbonates) that occupy a large area of the corroded surface.

Fig.2.Surface and cross-sectional SEM images of the laser-treated Mg-1Ca alloy and the corresponding EDS elemental distribution.

Fig.3.XRD patterns of the untreated and the laser-treated Mg-1Ca alloy.

Fig.4.Time-evolution of the hydrogen gas release on the untreated and laser-treated Mg-1Ca alloy after immersion in the SBF solution for 60 min.

Fig.6 shows the surface SEM images of the untreated and laser-treated alloys after 1, 2, 48 and 120 h immersion in the SBF solution.After 1 and 2 h immersion, the cracked corrosion products extend over the corroded surfaces for both the untreated and laser-treated alloys, as shown in Fig.6a, b.On the untreated alloy surface in Fig.6a.1, in addition to the cracked regions, some regions in white color are temporarily free of covering by the corrosion products.Such regions are likely to be the remnants of the dome-shaped compounds that only consist of oxide/hydroxide (Fig.A1a.1).After 2 h of immersion, these regions become narrower due to the expansion of the cracked regions.Moreover, the cracked regions after 1 h are mainly covered by oxide/hydroxide(Fig.A1a.1) whereas more and more chemical precipitates (mainly Mg phosphates/carbonates) take over the laser-treated surface after 2 h immersion (Fig.A1b.1).For the laser-treated alloy, after 1 h immersion, the ripping structure disappears and a cracked morphology develops (Fig.6b.2).Some of the dome-shaped features are well-developed and some are developing (white arrows).After 2 h immersion, the dome-shaped features disappear and large cracked chemical precipitates occupy the entire surface.The chemical precipitates that mainly consist of Mg/Ca phosphates/carbonates (the content of Mg is bigger than that of Ca in Fig.A1a.2 and Fig.A1b.2) develop a protective layer for the laser-treated alloy.

Fig.5.SEM top views and the corresponding EDS mapping results of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution after 5 and 30 min.

Fig.6.SEM top views of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution after 1, 2, 48 and 120 h.

After 48 h,there are no substantial changes of the corroded surfaces for both the untreated and the laser-treated alloys,except that cracks cover the corroded surface and chemical precipitates build up on the top surface (Fig.6c, d, Fig.A1c, d).With increasing immersion time, the cracks become broader and deeper and the fragments grow larger (Fig.6d.1, d.2).Compared with the untreated alloy, the corroded surface of the laser-treated alloy is rougher due to the deposition of small sized new-formed precipitates after 120 h immersion.The EDS maps in Fig.A1c, d imply that the precipitates on the corroded surfaces for both untreated and laser-treated alloys are composed of Ca/Mg phosphates/carbonates (the content of Ca is bigger than that of Mg in Fig.A1c, d).

3.3.2.Cross-sectional corrosion morphology

Fig.7 shows the cross-sectional SEM images of the lasertreated alloy after 30 min immersion in the SBF solution and the EDS mapping results of the enlarged region b.A layer(3-4 μm) of chemical precipitates deposits on top of the laser-treated alloy and the thickness of the laser-treated layer remains unchanged, suggesting no dissolution of the lasertreated layer occurs.However, cavities develop at the interface between the precipitate layer and the laser-treated layer(Fig.7a).The formation of the cavities might be attributed to the effect of hydrogen gas that evolves from the regions where the localized corrosion occurs on the laser-treated layer(Fig.7b).The corresponding EDS results confir the elements that constitute the precipitate layer are C, O, Ca, P and a small amount of Mg.Accordingly, the compounds of the precipitates can be Ca/Mg phosphates/carbonates (Ca2+is the main cation).

Fig.8 shows cross-sectional SEM images of the untreated and laser-treated alloys after 1, 2, 48 and 120 h immersion in the SBF solution.After 1 h, corrosion products (oxide/hydroxide) form a discontinuous layer on the untreated alloy and another layer deposits on top of the corrosion products layer after 2 h immersion (Fig.8b.1).The layers develop a two-layer structure (the insert image in Fig.8b.1).The upper layer is relatively compact and the bottom layer is loose and with cavities.In addition, some regions are not covered by chemical precipitates, corresponding to the dome-shaped compounds in Fig.6b.1.For laser-treated alloy, the crosssectional corrosion morphology after 1 h (Fig.8a.2) is similar to that after 30 min in Fig.7b.The thickness of the precipitate layer did not change substantially, but the thickness of the laser-treated layer slightly decreases due to its dissolution.However, there exists an obvious fissur between the layers after 1 h immersion, which can be attributed to the impact of small hydrogen gas bubbles when the laser-treated layer dissolves.After 2 h, the laser-treated layer cannot be observed in Fig.8b.2 because the uniform corrosion under the fissur in Fig.8b.1 consumes the layer.At the same time, a layer beneath the precipitates layer develops and takes the place of the laser-treated layer, leading to the formation of a similar two-structure (the insert image in Fig.8b.2).For both the untreated and laser-treated alloys, the two-layer structure has a similar cross-sectional morphology, except that the thickness of the structure of the laser-treated alloy (5.77 μm) is larger than that of the untreated alloy (4.76 μm).

With increasing time, the two-layer structure grows in thickness (Fig.8c, d).However, the thickness of the individual layers of the two-layer structure is different.After 48 h immersion, the bottom layer is thicker than the upper layer for the untreated alloy whereas the opposite is observed for the laser-treated alloy (Fig.8c.1, c.2).After 120 h, the upper layer is thicker than the bottom layer for the untreated alloy (Fig.8d.1) and the thickness of the two layers for the laser-treated alloy is comparable (Fig.8d.2).In addition, the upper layers in Fig.8c, d are featured by cracks and segments corresponding to the cracked surfaces in Fig.6c, d.Notably, although the two-layer structure on the laser-treated alloy is thicker after 120 h immersion (Fig.8d.2) compared with the untreated alloy (Fig.8d.1), significan gaps develop horizontally inside the two layers, which degrades the corrosion protection of the two-layer structure.The corresponding EDS maps in Fig.A2 distinguish the layers of the twolayer structure that has identical elements for both the untreated and laser-treated alloys after 48 and 120 h immersion.The elements distribution in the two layers from Fig.A2 is consistent with the EDS results of the corroded surface in Fig.A1c,d.The bottom layer contains more Mg and O that constitute Mg oxide/hydroxide and the upper layer comprises a large amount of C, O, P and Ca that constitute Ca/Mg phosphates/carbonates (Ca2+is the main cation).

3.4.Corrosion depth and corrosion rate

From Fig.8, corrosion proceeds uniformly on both the untreated and laser-treated alloys.Fig.9a shows the average corrosion depth with immersion time.It appears that the corrosion depth is independent of the corrosion morphology and increases linearly with the immersion time.The corrosion rate based on corrosion depth is shown in Fig.9b.For both the untreated and laser-treated alloys,corrosion rate declines sharply during the firs 48 h, due to the rapid dissolution of the alloy.Afterward, corrosion rate decreases slowly from 48 h to 120 h, indicating that the dissolution of the alloy slows down due to the thickening and relatively compact two-layer structure.Compared to the untreated alloy, the average corrosion depth and the corrosion rate of the laser-treated alloy are always smaller.Overall, laser treatment favors the improvement of corrosion performance of Mg-1Ca alloy in the SBF solution for the firs 48 h.

Fig.7.SEM and EDS images presenting the cross-sectional microstructure and elements of the laser-treated Mg-1Ca alloy immersed in the SBF solution after 30 min.

Fig.8.Cross-sectional SEM images of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution after 1, 2, 48 and 120 h.

Fig.9.Time-evolution of the average corrosion depth (a) and the corrosion rate (b) of the untreated and the laser-treated Mg-1Ca alloy immersed in the SBF solution.

3.5.Electrochemical impedance spectra

Fig.10 shows EIS spectra of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution for different exposure time up to 120 h.The open-circuit potentials (OCP)monitored for 24 h are shown in Fig.A3.A similar trend is revealed for both the untreated alloy and laser-treated alloy but a more stable OCP value is observed for the laser-treated alloy.Moreover, the OCP values become stable after 12 h but the fina value for the laser-treated alloy (-1.79 V vs.SCE) is nobler than that for the untreated alloy (-1.84 V vs.SCE).Notably, OCP is changing during the EIS measurement period especially before 12 h immersion, corresponding to the changes of the developing layers (Fig.7, 8) withimmersion time.To obtain a reliable EIS evolution trend in this study without big influence of distorted OCP data, the OCP was fi ed and the testing time lasted only for about 5 min in each EIS measurement.From Fig.10, for 1 h immersion, the EIS responses for both the untreated and lasertreated alloys display a well-define high-frequency capacitive loop and a small low-frequency capacitive loop.After 1 h immersion, intermediate-frequency capacitive loops appear and become more and more significant In addition,low-frequency inductive loops appear after 2 h immersion.From the EIS spectra, it can be concluded that the corrosion resistance of the laser-treated alloy is higher than that of the untreated alloy for the firs 48 h immersion.After 120 h, the corrosion resistance of the untreated alloy is similar to that of the lasertreated alloy in Fig.10f.

Fig.10.EIS measurement (scatter plot) and model fi (solid lines) of Mg-1Ca alloy after 0.5 to 120 h immersion at open circuit in the SBF solution.

Fig.11 shows the equivalent electrical circuit used in this study to fi the EIS data.The same circuit was used for both the untreated and laser-treated alloy because of the similarity of the EIS spectra in Fig.10.In Fig.11,Rsis the solution resistance (the bigger values ofRsfor the laser-treatedalloys in Fig.10(a-d)are possibly related to the increased ions concentration and pH near the working electrode and the formation of porous layers during corrosion [52]);Qlaccounts for the layers (air-formed/laser-radiated MgO film or corrosion products or precipitates)formed on alloys;Qdlrepresents the electric double-layer capacitance at the active regions of the alloy-electrolyte interface.Rcprepresents the resistance associated with local environmental changes nearby cathodic regions, such as the presence of corrosion products, depositions or formation of gels [53,54];Rctrepresents the charge transfer resistance associated with the redox reactions under the activation control andWs, the Warburg impedance, represents the diffusion processes nearby the corroding surface.In the fittingWsconsists ofWS-R accounting for the diffusion resistance.The sum of the three resistances (Rcp,RctandRws) represents the cathodic resistance (Rc):

Fig.11.Equivalent electrical circuit used to fi electrochemical impedance responses of Mg-1Ca alloys in the SBF solution.

Moreover,RaandLarepresent the resistive and inductive behavior associated with the anodic reaction of magnesium oxidation at localized corrosion sites.The total resistance (Rt)of the redox reactions is calculated by the parallel resistances ofRaandRc:

The fittin results are given in Table 1.

Fig.12 shows the time-evolution resistances extracted from EIS responses.For both the untreated and the laser-treated alloys, resistances (Rct,RcandRt) always increase with immersion time butRaexperiences a firs decrease for 12 h corrosion and then an increase.For all the resistances, the difference of every corrosion resistance between the untreated and the laser-treated alloy becomes smaller with immersion time and they are comparable after 120 h corrosion.However, all resistances of the laser-treated alloy are bigger than of the untreated alloy.The results in Table 1 and Fig.12 are consistent with the corrosion rates calculated by the corrosion depths (Fig.9b).Significantl ,RcapproximatesRtfor both the untreated and laser-treated alloys, indicating that cathodic resistance greatly influence the total corrosion rate.

Table 1The fittin results of resistances in the EEC, which was used to simulate the impedance data of the untreated and the laser-treated Mg-1Ca alloy after 0.5 to 120 h immersion in the SBF solution.

3.6.Time-evolution of corrosion behavior

Fig.13 illustrates the time-evolution corrosion behavior of the alloys before and after laser treatment in the SBF solution.The excimer laser treatment generates apparent modifi cations of surface morphology, microstructure and phases of the Mg-1Ca alloy (Fig.2, 3 and Fig.13a).Such changes are responsible for the slower corrosion rate at early-stage corrosion (Fig.9b and Fig.12) for the laser-treated alloy, which is consistent with previous studies [30,55-56].In Fig.13a for 1 h immersion,β-Mg2Ca phases in the alloy substrate of the laser-treated alloy are protected by the 5 μm lasertreated layer(Fig.2c),thus the localized corrosion and microgalvanic corrosion is less severe compared with the untreated alloy.For the untreated alloy, the severe degradation of the exposedβ-Mg2Ca phases contributes to a ‘remote' current for Mg dissolution at the corrosion fronts [53].Meanwhile,large hydrogen bubbles release from the regions where theβ-Mg2Ca phase is exposed to the solution (Fig.4).The availability of ‘remote' current accelerates the corrosion on theα-Mg matrix.By comparison, due to the slight localized corrosion, only fin hydrogen bubbles appear on the laser-treated alloy.Also, corrosion proceeds much slower on the lasertreated alloy because the ‘remote' current generated by the exposedβ-Mg2Ca phase is relatively small.After 1 h immersion, the untreated alloy (especially on the exposedβ-Mg2Caphases) is covered by corrosion products mainly containing MgO/Mg(OH)2, and a discontinuous layer of corrosion products develops (Fig.8a.1).By contrast, the laser-treated layer remains and is covered by a precipitates layer of Ca/Mg phosphates/carbonates (Ca2+is the main cation coming from the SBF solution at this moment because the release of Mg2+by electrochemical reactions is insignificant) The precipitates layer on the laser-treated alloy contributes to the initial smaller corrosion rate (Fig.9b and Fig.12).

Fig.12.Time-evolution of (a) charge transfer resistance (Rct), (b) zero frequency resistance (Rt), (c) cathodic behavior resistance (Rc) and (d) anodic behavior resistance (Ra) for the untreated and laser-treated Mg-1Ca alloys immersed in the SBF solution.

For 2 h immersion in Fig.13b, a two-layer structure develops on both the untreated and laser-treated alloys, leading to decreasing corrosion rates (Fig.9b).The upper precipitates layer of the two-layer structure is more compact than the bottom layer (having more cracks and cavities), indicating the upper layer offers more effective corrosion protection.For the laser-treated alloy,although the laser-treated layer was not observed after 2 h, the corrosion rate (Fig.9b) is smaller compared with the untreated alloy for 48 h immersion.This anti-corrosion advantage benefit from the formation of the protective precipitate layer during 1 h immersion and the twolayer structure developed after 2 h.Although the morphology and compositions of corrosion compounds (Fig.A1b.1, A1b.2)of the two-layer structure are similar for the untreated and laser-treated alloys, the formation sequence of the two layers and their characteristics with time evolution (Fig.8a, b)vary.For the untreated alloy, the bottom layer is the firs to develop, and then the top layer forms on the bottom layer.However, for the laser-treated alloy, the upper layer depositson the laser-treated layer, followed by the formation of the bottom layer when the laser-treated layer dissolves.Moreover,according to the EDS maps in Fig.A1b.1 and Fig.A1b.2, the precipitates layer developed on the alloy surfaces is composed of Mg/Ca phosphates/carbonates and Mg2+is the main cation at this moment since a large amount of Mg2+release by rapid Mg metal dissolution.

Fig.13.Schematic diagrams illustrating corrosion process of the untreated and laser-treated Mg-1Ca alloy immersed in the SBF solution for 120 h.

For a longer immersion time (Fig.13c), the two layers grow in thickness leading to a further decrease of corrosion rates (Fig.9b), although large fissure and cavities develop inside the two layers for both the untreated and laser-treated alloys.Specificall , after 48 h immersion, the two-layer structure is almost identical in morphology, thickness and element compositions, leading to comparable corrosion rates for the untreated and laser-treated alloys.However, the bottom layer is thicker than the upper layer for the untreated alloy whereas it is the opposite for the laser-treated alloy after 48 h.The bottom layer is thicker than the upper layer(Fig.8c.1) for the untreated alloy because the dissolution of alloy substrate is severe before 48 h so that the formation of the oxide/hydroxide layer is faster than precipitation of phosphates/carbonates.On the contrary, the dissolution of the substrate alloy of the laser-treated alloy is relatively slow under the protection of the precipitate layer developed during 1 h immersion, thus the formation rate of precipitates outweigh that of oxide/hydroxide, resulting in a thicker upper layer (Fig.8c.2).With immersion time, the chemical deposition of precipitates becomes easier than the formation of oxide/hydroxide on the untreated alloy because the dissolution rate of the underlying alloy is greatly hindered by the layers, leading to the rapid growth of the upper layer that is thicker than the bottom layer after 120 h (Fig.8d.1).By comparison, the bottom layer on the laser-treated alloy after 120 h immersion grows faster than the upper layer due to the underlying alloy dissolution reactions becoming active after the fissure and cavities damage the upper layer (Fig.8d.2).Notably, although the compounds that constitute the upper precipitates layers are also Ca/Mg phosphates and carbonate for long-term immersion (Fig.13c), Ca2+other than Mg2+is the main cation at this moment since the Mg metal dissolution is greatly depressed.

The time-evolution corrosion behavior illustrated in Fig.13 can be verifie by EIS results.The comparableRcandRtfor both the untreated and laser-treated alloys imply that the cathodic reaction that is greatly influence by the layers controls the overall decreasing corrosion rate.However, the fluctuatio ofRathat is associated with the inductive behavior (Fig.12d) indicates that the anodic reaction also contributes to the overall corrosion rate.Specificall , the decreasingRacorresponds to the increasing importance of anodic reaction for the short-term corrosion when the localized corrosion takes place (Fig.5a, b).However, for long-term immersion,Rastarts to increase and its influenc onRtbecomes negligible (Eq.(2)).The reduction in the contribution of the anodic resistance to the overall corrosion resistance can be attributed to the formation of the two-layer structure that protects the alloy substrate from severe localized corrosion.Interestingly,Rafor the laser-treated alloy is smaller thanRafor the untreated alloy after 120 h immersion (Fig.12d), indicating that anodic reaction on the laser-treated alloy remains active,which is consistent with the above discussion about the rapid growth of the bottom layer for the laser-treated alloy.Overall, the excimer laser treatment is effective in protecting Mg alloys from short-term corrosion in the SBF solution.However, more efforts on selecting lasers in terms of wavelength, pulse width and laser processing conditions to achieve homogenized microstructure with big melt depths for long-term corrosion protection of Mg alloys are needed.

4.Conclusions

(1) Excimer laser treatment produces a 5 μm melted layer on the Mg-1Ca alloy.The laser-treated surface is characterized by the ripping structure with tiny pores.During laser treatment,β-Mg2Ca phases dissolve, leading to a homogeneous microstructure without apparent grain boundaries and intermetallic particles in the laser-treated layer.At the same time, oxidation of the alloy occurs, resulting in the formation of MgO phases on the rippling surface.This refine microstructure and the MgO fil contribute to the lower corrosion rate of the laser-treated alloy at the beginning of immersion corrosion.

(2) At the beginning of corrosion, the laser-treated alloy suffers less from localized corrosion due to the dissolution ofβ-Mg2Ca phases and the corresponding microstructural homogeneity in the laser-treated layer.However, for longterm corrosion, corrosion rates are comparable for the untreated alloy and the laser-treated alloy due to a similar two-layer structure developing on the surfaces.Although the two-layer structures have similar corrosion morphology, thickness and element compositions, they form in a different sequence that the bottom layer is the firs to develop and then the upper layer forms on top of the bottom layer for the untreated alloy and it is opposite for the lasertreated alloy.

(3) From EIS responses, the corrosion rate is generally controlled by the cathodic reaction for both the untreated and the laser-treated alloy.The anodic reaction is significan for short-term immersion in which the inductive behavior is apparent due to the active localized corrosion, especially for the untreated alloy.After the two-layer structure is well-developed, the corrosion behavior is controlled by the increased thickness of the layers as well as the uniform corrosion.

Declaration of Competing Interest

None

Acknowledgments

The authors acknowledge the use of the Department of Materials X-ray Diffraction Suite and Electron Microscopy centre at the University of Manchester and are grateful for the technical support and advice by Mr.Gary Harrison and Ms.Xiangli Zhong.

Appendix

.

Fig.A1.EDS elemental maps of the corroded surfaces for untreated alloy and laser-treated Mg-1Ca alloy after 1, 2, 48 and 120 h immersion in the SBF solution.

Fig.A2.Cross-sectional EDS elemental maps of the corroded untreated alloy and laser-treated Mg-1Ca alloy after 48 and 120 h immersion in the SBF solution.

Fig.A3.Time-evolution of open circuit potential (OCP) of the untreated and laser-treated alloys for 24 h immersion.