In vitro degradation of ZnO flowered coated Zn-Mg alloys in simulated physiological conditions

Materials Science and Engineering C 70 (2017) 112–120

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

In vitro degradation of ZnO flowered coated Zn-Mg alloys in simulated physiological conditions Marta M. Alves a,⁎, Tomas Prosek b, Catarina F. Santos a,c, Maria F. Montemor a a b c

CQE Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal Technopark Kralupy, University of Chemistry and Technology in Prague, Zizkova 7, 278 01 Kralupy nad Vltavou, Czech Republic EST Setúbal, DEM, Instituto Politécnico de Setúbal, Campus IPS, 2910 Setúbal, Portugal

a r t i c l e

i n f o

Article history: Received 23 June 2016 Received in revised form 22 August 2016 Accepted 26 August 2016 Available online 28 August 2016 Keywords: Functional coatings In vitro degradation Raman mapping Zn-Mg alloys ZnO electrodeposition

a b s t r a c t Flowered coatings composed by ZnO crystals were successfully electrodeposited on Zn-Mg alloys. The distinct coatings morphologies were found to be dependent upon the solid interfaces distribution, with the smaller number of bigger flowers (ø 46 μm) obtained on Zn-Mg alloy containing 1 wt.% Mg (Zn-1Mg) contrasting with the higher number of smaller flowers (ø 38 μm) achieved on Zn-Mg alloy with 2 wt.% Mg (Zn-2Mg). To assess the in vitro behaviour of these novel resorbable materials, a detailed evaluation of the degradation behaviour, in simulated physiological conditions, was performed by electrochemical impedance spectroscopy (EIS). The opposite behaviours observed in the corrosion resistances resulted in the build-up of distinct corrosion layers. The products forming these layers, preferentially detected at the flowers, were identified and their spatial distribution disclosed by EDS and Raman spectroscopy techniques. The presence of smithsonite, simonkolleite, hydrozincite, skorpionite and hydroxyapatite were assigned to both corrosion layers. However the distinct spatial distributions depicted may impact the biocompatibility of these resorbable materials, with the bone analogue compounds (hydroxyapatite and skorpionite) depicted in-between the ZnO crystals and on the top corrosion layer of Zn-1Mg flowers clearly contrasting with the hindered layer formed at the interface of the substrate with the flowers on Zn-2Mg. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Metallic biomaterials are of utmost relevance because their resistance to mechanical stress and tension make them especially suitable for load bearing implants. Inside this class of biomaterials, bioresorbable metals and alloys are growing in interest [1]. In transient woundhealing pathologies, a biomaterial would ideally be absorbed by the organism, eliminating the need of a second surgery for implant removal. Amongst the available resorbable metallic materials, Zn has up surged as a promising candidate [2]. As a micronutrient, Zn has a tolerable upper intake of 40 mg/day, with beneficial effects in wound-healing, namely in skeletal, muscle remodelling and associated inflammatory responses [3]. In elderly, Zn supplementation was described to decrease severity and extent of common colds, blindness and mortality due to cardiovascular diseases [3,4]. Under physiological conditions, metallic Zn has been described to have adequate corrosion rates for tissue healing [2,5]. However, when considering load-bearing implants the poor mechanical strength of Zn limits its use. This drawback can be ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (M.M. Alves).

http://dx.doi.org/10.1016/j.msec.2016.08.071 0928-4931/© 2016 Elsevier B.V. All rights reserved.

solved by alloying Zn with physiological compatible relevant elements, a strategy that results in superior strength and castability [6]. Mg, in particular, is physiologically non-toxic and supports important biological processes, as bones growth and development [7] and is well known as bioresorbable element [2,8]. Whatsoever, the addition of Mg to Zn yielded alloys with a maximum strength and elongation for Mg contents only around 1 wt.% [6,9] with in vitro tests confirming Zn-Mg alloys biocompatibility [9–13]. To further modulate the corrosion behaviour of Zn-Mg alloys the addition of functional coatings can be considered. The addition of ZnO flowered coatings has been reported to favoured the formation of biomimetic compounds, osteoblasts proliferation, collagen secretion and extracellular matrix mineralization [14–16]. Besides the increased biocompatibility envisaged by these coatings, the reported antimicrobial properties of ZnO flowers [14,17] could additionally present an effective strategy to overcome the world-wide increasing implant-associated infections. Therefore, the aim of this work was to functionalize Zn-Mg alloys with ZnO flowered coatings. A detailed evaluation of the in vitro corrosion behaviour, using simulated physiological conditions, was performed by combining electrochemical impedance spectroscopy (EIS), Raman Spectroscopy, scanning electron microscopy (SEM) and Energy dispersive X-Ray (EDS) analyses.

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

2. Materials and methods 2.1. Preparation of the materials The preparation of the alloys was previously described by Prosek et al. [18]. Briefly, Zn was melted and ZnCl2 flux was added at the temperature of 450 °C. Magnesium was set in at the temperature by 50 K higher than the temperature of liquidus of each alloy. The melt was strongly homogenized by a graphite bar and poured into a mould with a quatrefoil section. After solidification, it was machined to a bar with diameter of 20 mm and cut to coin-like samples. The ZnO flowered coatings were obtained by electrodeposition, as previously described by the authors for pure Zn [19]. Briefly, Zn-1Mg and Zn-2Mg samples were mounted in cold curing epoxy resin, polished with SiC paper up to 2500 grit, washed with absolute ethanol and dried in air prior to electrodeposition. The electrodeposition was carried out in a three-electrode electrochemical cell with platinum as counter electrode, saturated calomel electrode (SCE) as reference electrode and the alloys as working electrode. The electrolyte was composed of 50 mM Zn(NO3)2 and 50 mM H3BO3 at pH 6. The electrodeposition was carried out at the constant cathodic potential of − 1.9 V for 20 s in a Voltalab PGZ 100 potentiostat (Radiometer Analytical). The electrodeposited coupons were washed with absolute ethanol and dried at room temperature. 2.2. Electrochemical impedance spectroscopy (EIS) The corrosion behaviour of coated alloys, mounted in the epoxy resin, was studied by submitting the samples to simulated physiological simulated conditions, using a 3-electrodes cell arrangement. A simulated body fluid (SBF) solution (137.5 mM NaCl, 4.2 mM NaHCO3, 3.0 mM KCl, 1.0 mM K2HPO4·3H2O, 1.5 mM MgCl2·6H2O, 2.6 mM CaCl2, 0.5 mM Na2SO4, 50.5 mM (HOCH2)3CNH2 at pH 7.4) [20] was used and the immersed samples were kept at 37 °C for 24 h. The EIS measurements were carried out with an AUTOLAB PGSTAT 302N by applying 10 mV perturbation. The measuring frequency ranged from 105 Hz down to 10−2 Hz. The EIS spectra were periodically recorded at open circuit potential (OCP). A three-electrode electrochemical cell, with a ZnO flowered coated sample as working electrode, SCE as reference electrode and a platinum coil as counter electrode was used. The EIS data analyses, the equivalent circuit modelling and corresponding elements values were calculated with Zview software. 2.3. Physico-chemical characterization The morphologies of the prepared surfaces and corrosion products formed during immersion in SBF were analysed by scanning electron microscopy (SEM), using Hitachi S2400 apparatus, or by FEG-SEM, using a JEOL-JSM7001F apparatus, and the elemental chemical composition by the respective energy dispersive X-ray (EDS). Raman spectra and maps were collected (Horiba LabRAM HR800 Evolution) using the radiation source with a solid-state laser operating at 532 nm with an output power of 20 mW. A spectrograph with a 600 lines/mm grating was used. A 50 or 100 times objective lens focused the laser beam on the samples surface. Spectra were obtained with acquisition time of 10 s and 10 accumulations. 3. Results and discussion 3.1. Electrodeposition of flower-like ZnO coatings Coating surfaces with ZnO flowers has been proposed as an expedite route to functionalise metallic materials [14,19]. In this work flowered coatings were successfully prepared by electrodepositing ZnO on ZnMg alloys. As depicted in Fig. 1, distinct ZnO flowered coatings were obtained on Zn-1Mg and Zn-2Mg. A low density of larger flowers grew on Zn-

113

1Mg whereas a higher density of smaller flowers was observed on Zn2Mg. These results revealed that the flowers density was inversely related to their diameter. These flowers, composed by several branches from where diverse ramifications emerge, are the result of the assembly of several small ZnO lamina-like subunits. Depending on the alloys, distinct flower morphologies were obtained: on Zn-1Mg the ZnO flowers presented thicker branches and average diameters of 45.9 ± 0.8 μm whereas on Zn-2Mg the thinner branched flowers had average diameters of 38.2 ± 0.3 μm. The distinct flowers' distributions and morphologies on the alloys' surfaces were investigated by the currenttransient plots obtained along the electrodeposition time. As depicted in Fig. 2, the current-transient curves show that a lower current density is required to produce the flowered coatings on Zn-1Mg than on Zn-2Mg. In the early beginning of electrodeposition an increased current density can be related with double layer discharge events. On Zn-1Mg a slight current density increase was clearly noticed whereas for Zn-2Mg this effect was much attenuated. After that stage, a decrease in current density can be associated to the nucleation step. On Zn-1Mg a decreased current density occurred until reaching another plateau after c.a. 5 s while for Zn-2Mg a sharper decrease occurred with the stable plateau starting at around 2 s. During this flower seeding process the slight decrease in current density observed for Zn-1Mg suggested a low nucleation density, when compared with that of Zn-2Mg, where a higher nucleation density is related with the sharp current density decrease. These findings are associated with the number of flowers present on the alloys surfaces (Fig. 1). Thereafter a less pronounced decreasing trend in the current density was associated to flowers growth. This trend, more pronounced on Zn2Mg than on Zn-1Mg, is in agreement with the previously reported formation of nanostructured ZnO flowers on pure Zn [19]. Although having small differences in the Mg composition, the impact on the current density distribution on Zn-1Mg and Zn-2Mg was enough to yield distinct ZnO flowered coatings. To confirm the presence of ZnO as part of the flowers composition, EDS maps and Raman spectra of the flowered coating were carried out. As depicted in Fig. 3, the EDS maps revealed that the presence of Zn and oxygen could be assigned to the electrodeposited flowers (Fig. 3a), with the higher intense spots observed in the oxygen maps being speculated to be part of a Zn(OH)2 veil as suggested by Wadowska et al. [21]. Magnesium EDS maps revealed the phase distribution typical of these alloys: Zn + Mg2Zn11 eutectic interdendritic areas and Zn dendrites, as reported elsewhere [18]. The comparison of the magnesium EDS map with the corresponding SEM images depicted the localization of the nucleation points. As outlined in Fig. 3a the nucleation points occurred preferentially in the solid interfaces with the presence of fewer interfaces in Zn-1Mg being consistent with the lower number of flowers formed on this alloy (Fig. 1). Conversely, the higher number of interfaces is in agreement with the high number of flowers formed on Zn2Mg surface (Fig. 1). As depicted in Fig. 3b, the Raman spectra collected from flowers revealed the presence of E2H–E2L (318 cm− 1), E2H (431 cm− 1), 2LA (~ 507 and 533 cm− 1), 2B1L (533 cm− 1) and A1LO (571 cm−1) vibrational modes, typical assigned to ZnO [22]. The strong peak corresponding to E2H photon mode, and related weaker multiphonon mode (E2H–E2L), were assigned to wurtzite ZnO crystal [19, 23]. The presence of A1LO phonon mode, with weak intensity, suggested the presence of well-ordered ZnO crystals. These results evidenced that flower units were formed by wurtzite and well-organized ZnO crystals on both alloys. The additional peak detected at 550 cm−1 in the substrate has been previously ascribed to the surface optical (SO) phonon vibrations. This may result from disorderly arranged areas or boundaries between alloys surface and deposited and/or disordered regions of ZnO crystals [22]. Despite the solid phase distribution in the Zn alloys being a crucial factor for flowers distribution and morphology, as previously discussed, the presence of small amounts of Mg alloying Zn did not to modify the chemical nature of the ZnO crystals units.

114

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

Zn-2Mg

Zn-1Mg

50 μm

50 μm

1 μm

1 μm

10 μm

10 μm

Fig. 1. Scanning electron microscopy (SEM) micrographs of the ZnO flowered coating electrodeposited on Zn-Mg alloys using a constant cathodic potential of −1.9 V vs. SCE.

3.2. In vitro corrosion behaviour To assess the in vitro degradation of these coated materials, an electrochemical impedance spectroscopy (EIS) assay was carried out in simulating physiological conditions. As depicted in Fig. 4, the Nyquist and Bode plots obtained for the flowered coated alloys revealed opposite degradation behaviours, indicating that different corrosion processes were underway for these resorbable materials. For coated Zn-1Mg an early immersion (2 h) resulted in the formation of two overlapped conductive loops in the medium-high frequency region. An additional time constant, related to the onset of the corrosion activity was displaced to the very low frequency range. Twenty four

hours later there was a shift of the medium-high frequencies time constant to the higher frequencies. The presence of a time constant together with an inductive loop in the low frequency region indicated that corrosion was progressing. For coated Zn-2Mg the medium-high frequency time constants observed at 2 and 24 h of immersion could be clearly discriminated. The overall impedance modulus depicted at the lowest frequency decreased from 9.2 × 103 Ω cm2 to 4.2 × 103 Ω cm2 with the immersion time for the flowered coated Zn-1Mg. This may be assigned to an active undergoing corrosion. Contrarily, for Zn-2Mg the resistance increased along the immersion time, with the overall impedance modulus increasing from 1.6 × 103 Ω cm2 to 4.4 × 103 Ω cm2. This increase may be related to the formation of a protective layer. Whatsoever the

ZnO crystal Zn-Mg alloy

Nucleation

Growth

|i| ( mAcm-2)

dl

Zn-2Mg

Zn-1Mg

Time (s) Fig. 2. Current-transient plots for the electrodeposited ZnO flowers on Zn-Mg alloys using a constant cathodic potential of −1.9 V vs. SCE. The different developmental stages of the flowers along the electrodeposition are represented; dl, double layer.

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

Zn-1Mg

115

Zn-2Mg

a)

Normalized intensisty (a.u.)

E2H

SE

Zn

O

Mg

2LA; 2B 1L ZnO crystal

E2H-E2L

b)

Zn-Mg alloy

2LA A1LO SO

300

350

400

450

500

550

600

Wavenumber (cm-1) Fig. 3. Scanning electron microscopy (SEM) micrographs, corresponding EDS maps a) and representative Raman spectra b) of the electrodeposited flowers on Zn-Mg alloys surfaces. Stars represent the nucleation points of flowers on the Zn-Mg alloys surface.

shift on the time constant towards higher frequencies observed in the phase angle plot for both coated alloys can be assigned to the building-up of a corrosion layer (Fig. 4). Despite the envisaged influence of the ZnO flowered coatings towards the formation of Zn/ZnO galvanic couples [24], the presence of Mg in the substrate may be, as well, impacting this corrosion behaviour. An increased corrosion resistance with increasing Mg contents up to 6 wt.% reported for bare Zn-Mg alloys [18,25] is in line with an increased corrosion resistance observed with an increasing Mg content on the ZnO flowered coated alloys. To have a deeper insight of the corrosion behaviour of these coated materials, an equivalent circuit (EC) was assigned to the EIS data (Fig. 4). Fig. 5 depicts the EC that fits the experimental EIS data and Table 1 shows the corresponding numerical fitting results, with chi-squares (χ2) in the order of 10−3 to 10−4 suggesting a good fit for the proposed EC model. This EC consisted on three time constants that could be assigned to different events. In this EC the electrolyte resistance (Rs) was assigned to the SBF solution and the CPE1/R1 elements described the response of the ZnO coating at high frequencies. The admittance of the CPE and the resistance values could be related to the flowers size. A higher admittance of the CPE and resistance values were detected for bigger flowers on Zn-1Mg, while a lower admittance of the CPE and resistance values were determined for smaller flowers on Zn-2Mg. The semi-conductive nature of ZnO is consistent with the CPE behaviour and the increase in R1, observed for Zn-1Mg, suggesting the growth of an insoluble layer of corrosion products on the flowers. The elements CPE2/R2 were necessary to fit the middle frequency range and were assigned to the interfacial layer that, at a later stage, also included the

corrosion products (Fig. 5). This corrosion layer seemed to be formed at the interface coating/substrate being probably composed by insulating Zn corrosion products [25]. The admittance of the CPE of this layer differed between the two alloys, with Zn-1Mg presenting values one order of magnitude above those of Zn-2Mg 24 h after immersion (Table 1). This behaviour suggested the formation of corrosion layers with distinct compositions and protectiveness, with a decreasing resistance observed along the immersion time for coated Zn-1Mg suggesting that this corrosion layer was more unstable. This allowed the access of the electrolyte to the substrate further promoting this material degradation. Contrarily, for Zn2Mg an increasing resistance indicated the formation of more protective layer, which by hindering the access of the electrolyte to the substrate delayed the degradation processes. This opposite corrosion behaviours indicated that different mechanisms were governing the degradation of these coated alloys. If an increased resistance is associated to the deposition of insulating Zn-derived products, it may be assumed that the presence of ZnO on Zn-1Mg favoured the corrosion of Zn through Zn/ ZnO galvanic couple [24]. This agrees with an increased coating resistance (R1) observed for Zn-1Mg, suggesting that the active corrosion process resulted in the deposition of products on the flowers. For Zn2Mg, the corrosion process seemed to occur mainly through the Zn/ Mg galvanic couple [25], as an increased interfacial resistance (R2) suggests the preferential deposition of the resulting products in the interfacial substrate/flowers layer. The third time constant described by CPE3/ R3 in the low frequencies range was associated to diffusion processes with the value of the CPE exponent (n) close to 0.5 expected for mass

116

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

Zn-1Mg

Zn-2Mg

-7.5

-5

-4

Zim (KΩ cm2)

-5

Zim (KΩ Ω cm2)

2h 24 h

-2.5

-3

-2

0 -1 2.5 0

2.5

5

7.5

0

10

103

-65

102

-40

101

-15 10

100 100

101

102

103

2

3

4

5

104

-90

104

|Z|/ Ω cm2

-90

10-1

1

105

Frequency (Hz)

103

-65

102

-40

101

-15

Phaseangle(o)

104

10-2

0

Zre (KΩ cm2) Phaseanlgle(o)

|Z|/ Ω cm2

Zre (KΩ cm2)

100 10-2

10-1

100

101

102

103

104

105

Frequency (Hz)

Fig. 4. Nyquist and Bode plots of the ZnO flowered coated Zn-Mg alloys in simulated physiological conditions.

transfer-controlled processes, which may be occurring during possible Zn pitting events [26]. Overall, the distinct corrosion behaviour observed for these two coated alloys may result in different corrosion products that, when considering resorbable biomedical material, can have a great influence in the upcoming wound-healing events. To evaluate the corrosion layers formed on these coated alloys SEM and EDS analyses were performed. As depicted in Fig. 6 clear morphological alterations on the ZnO flowered coatings were observed due to the immersion in the SBF solution at 37 °C. The micrographs show that a corrosion layer was formed at the flowers with no other visible structures formed in the naked areas. On the flowers of Zn-1Mg an agglomeration of corrosion products was visible on the flowers centre while the peripheral flower tips were still uncovered. Contrary, on the flowers of Zn-2Mg there was a peripheral ring covering the flower tips that was distinct from that of the centre (Fig. 6). This observation is consistent with the differential interfacial layer formed on both coated alloys as previously suggested by the EIS data (Fig. 4). The chemical species present in the electrolyte will dictate the formation of the associated corrosion compounds. The presence of chloride and carbonate in

the SBF solution may lead to the formation of the typically Zn-derived corrosion products such as Zn5Cl2(OH)8·H2O (simonkolleite), Zn5(CO3)2(OH)6 (hydrozincite), ZnCO3 (smithsonite) and ZnO (zincite) [25], while the presence of calcium and phosphates are often assign to the precipitation of a biomimetic layer of Ca5(PO4)3(OH) (hydroxyapatite) [27]. The detection of chloride in the EDS maps corroborates the formation of simonkolleite while that of carbon suggested the formation of the Zn-derived carbonates (hydrozincite, smithsonite) (Fig. 6). The detection of calcium and phosphorus is in agreement with the predicted precipitation of Ca5(PO4)3(OH) (hydroxyapatite) with the previous reported formation of this compound on a ZnO flowered coating further corroborating this hypothesis [14]. When considering bone implants, the formation of this compound may present advantageous in term of biocompatibility as the presence of hydroxyapatite is described to favour angiogenesis and decrease inflammatory reactions, two essential event that support tissue regeneration towards for a successful bonehealing progression [28]. The absence of magnesium in the corrosion layers is consistent with the reported influence of this element in the

Table 1 Equivalent circuit (EC) element values used to fit the EIS data of the ZnO flowered coated Zn-Mg alloys in simulated physiological conditions.

Fig. 5. Equivalent circuit (EC) model used to describe the EIS data of the ZnO flowered coated Zn-Mg alloys in simulated physiological conditions.

χ2 Rs (Ω cm2) CPE1 (Ω−1 cm−2 sn) n1 R1 (Ω cm2) CPE2 (Ω−1 cm−2 sn) n2 R2(Ω cm2) CPE3 (Ω−1 cm−2 sn) n3 R3(Ω cm2)

ZnO flowered coated Zn-1Mg

ZnO flowered coated Zn-2Mg

2h

24 h

2h

24 h

4.5 × 10−3 31.1 4.4 × 10−6 0.90 1.0 × 103 1.3 × 10−5 0.65 4.2 × 103 1.0 × 10−3 0.45 7.5 × 103

2.7 × 10−4 21.8 2.4 × 10−6 0.91 1.9 × 103 2.2 × 10−5 0.61 5.2 × 102 1.0 × 10−5 0.45 2.0 × 103

2.6 × 10−4 29.2 8.8 × 10−7 0.97 5.1 × 101 3.5 × 10−5 0.67 1.1 × 103 6.8 × 10−3 0.50 7.0 × 103

7.3 × 10−4 23.5 1.8 × 10−7 1.05 4.3 × 101 8.2 × 10−6 0.73 2.5 × 103 6.6 × 10−4 0.50 2.8 × 103

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

100 µm

10 µm

1 µm

100 µm

117

10 µm

1 µm

C

P

Cl

C

P

Cl

O

Ca

Zn

O

Ca

Zn

Fig. 6. Scanning electron microscopy (SEM) micrographs, flowers insets and EDS maps of the corroded flowers on Zn-Mg alloys when submitted to simulated physiological conditions.

modulation of local pH rather than by the direct participation in the composition of Zn-derived corrosion products [29,30]. The spatial distribution of the corrosion products depicted by the EDS maps substantiate the distinct morphologies already pinpointed by the SEM images (Fig. 6). On Zn-1Mg flowers the EDS maps indicated

a dense distribution of the corrosion products (carbon, phosphorous, chloride, oxygen and calcium) on the flowers centre. This is in agreement with the suggested mixture of oxides, hydroxides and carbonates with no well-defined interfaces that composed Zn-derived corrosion layers [31]. On the Zn-2Mg flowers a more diffuse distribution of the

Fig. 7. Raman spectra of the corroded flowers on Zn-Mg alloys submitted to physiological simulated conditions. Raman spectra were matched with those deposited in the RRUFF database (http://rruff.info) for simonkolleite (SMK, RRUFF ID: R130616), smithsonite (SMT, RRUFF ID: R040035), hydrozincite (HDZ, RRUFF ID: R050635), skorpionite (SKP, RRUFF ID: R060938) and hydroxyapatite (HAP, RRUFF ID: R060180).

118

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

corrosion products was depicted with the preferentially localization of chloride and calcium on the peripheral region being assign for simonkolleite and hydroxyapatite. A more homogenous distribution for oxygen can be assigned to the ZnO that compose the flowers, with no inference being drawn from the formation of zincite as a corrosiondriving product. Nevertheless, the presence of high oxygen amounts detected on the centre of the Zn-2Mg flowers, together with altered morphological features may suggest the formation of zincite on these flowers centre. This hypothesis is further corroborated by the EIS data (Table 1) where an n closer to 1 suggests the presence of a pure capacitor rather than that of other insulating Zn-derived corrosion products. To further confirm the products formed on the corroded a Raman analysis was additionally conducted. As depicted in Fig. 7, the Raman spectra obtained could be matched with those of hydrozincite (352, 396, 698, 734 and 1070 cm− 1), smithsonite (352 and 1114 cm− 1), simonkolleite (253, 396, 480, 734 and 900 cm−1) and hydroxyapatite (427 and 970 cm−1) deposited in RRUFF database (http://rruff.info). These results are in agreement with the corrosion products already suggested by the elements pinpointed in the EDS maps (Fig. 6). However, a new compound was suggested by these Raman spectra, Ca3Zn2(PO4)2CO3(OH)2 (skorpionite), to which are also assign the two peaks of hydroxyapatite (427 and 970 cm−1), and two other peaks at 427 and 1080 cm−1. The presence of this new Zn-derived compound containing carbon, calcium and phosphorous is still sustained by the EDS data of the corrosion layers (Fig. 6). The presence of Zn-doped Ca derived phosphates was already reported on the corroded surface of bare Zn-Mg alloys [32,33] with the possibly presence of Zn-doped hydroxyapatite analogues further envisaging a favoured bone development [34–36]. To have a deeper insight into the products distribution on the corroded flowers, Raman maps at distinct flowers heights were collected (Fig. 8). From the 3D analyses it is possible to infer that the corroded flowers on Zn-1Mg have circa 15 μm in height while on Zn-2Mg the corroded flowers did went further than the 10 μm in height (Fig. 8b). When analysing the maps collected at the different heights the central distribution of the products on the Zn-1Mg corroded flowers and the peripheral distribution on the Zn-2Mg corroded flowers is further

confirmed. However, a new insight was revealed by this analysis, which is the distribution of the corroded products in-between the ZnO crystals of the corroded flowers on Zn-1Mg while no such distribution was observed on the Zn-2Mg flowers (Fig. 8b). These findings further corroborate the degradation progression suggested by the EIS data (Fig. 4) where an increased R1 in the flowered Zn-1Mg indicated an increased coating resistance while an increased R2 in the flowered Zn1Mg showed an increase interfacial layer resistance (Table 1). The 2D projection of the Raman maps collected at different heights (Fig. 8b) for the characteristic wavenumber of each compound showed a similar distributions of the corrosion products within the same flowered coating (Fig. 8c), with a defined distribution of the corrosion products on the Zn-1Mg flowers contrasting with a diffused distribution detected on the corroded flowers of Zn-2Mg (Fig. 8c). These findings are in line with the previously discussed EDS map results (Fig. 6). All together these experimental data revealed that the ZnO flowered coated Zn-Mg alloys presented distinct degradation progressions in simulated physiological conditions with the precipitation of the corrosion products occurring at the electrodeposited flowers. As represented in Fig. 9, distinct spatial distributions of the corrosion products were attained on the flowers, with the preferential precipitation of the bone analogue compounds (hydroxyapatite and skorpionite) in-between the ZnO crystals and on the top of the flowers of Zn-1Mg contrasting with the preferential peripheral precipitation on the ZnO flowers of Zn-2Mg. 4. Conclusions The presence of small amounts of Mg alloying Zn altered the distributions of electrodeposited ZnO flowers, with the solid interfaces defining the nucleation points. However, the presence of Mg did not modify the chemical nature of the ZnO crystals units that form the flowers. The in vitro characterization of these coated alloys revealed distinct degradation behaviours, with the ZnO flowered coated Zn-2Mg presenting a more resistive behaviour when compared to that of coated Zn-1Mg. In both cases the corrosion progression leads to the formation of

Fig. 8. Representation of Raman maps collection a), 3D distribution b), and 2D projection c) of the products on the corroded flowers of Zn-Mg alloys submitted to simulated physiological conditions.

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

119

Fig. 9. Representation of the ZnO coatings electrodeposited on Zn-Mg alloys and their response under simulated physiological conditions.

smithsonite, simonkolleite, hydrozincite, hydroxyapatite and skorpionite. However, distinct distributions were attained, with the corrosion products being mainly distributed in-between the ZnO crystals and on the top centre of the flowers of Zn-1Mg and preferentially on the flowers periphery on Zn-2Mg.

Acknowledgments The author Marta Alves thanks FCT (SFRH/BPD/76646/2011) for providing financial support. The authors also acknowledge FCT support towards UID/QUI/00100/2013 and Prof. Luís Santos for support with the Raman studies.

References [1] M.F. Montemor, Functional and smart coatings for corrosion protection: a review of recent advances, Surf. Coat. Technol. 258 (2014) 17–37. [2] P.K. Bowen, J. Drelich, J. Goldman, Zinc exhibits ideal physiological corrosion behavior for bioabsorbable stents, Adv. Mater. 25 (2013) 2577–2582. [3] N.W. Solomons, Update on zinc biology, Ann. Nutr. Metab. 62 (2013) 8–17. [4] R.D. Lindeman, Hydration, electrolyte and mineral needs, in: C.W. Bales, C.S. Ritchie (Eds.), Handbook of Clinical Nutrition and Aging, Humana Press, NY 2009, pp. 137–156. [5] X. Liu, J. Sun, Y. Yang, Z. Pu, Y. Zheng, In vitro investigation of ultra-pure Zn and its mini-tube as potential bioabsorbable stent material, Mater. Lett. 161 (2015) 53–56. [6] D. Vojtech, J. Kubásek, J. Serák, P. Novák, Mechanical and corrosion properties of newly developed biodegradable Zn-based alloys for bone fixation, Acta Biomater. 7 (2011) 3515–3522. [7] J.H.F. de Baaij, J.G.J. Hoenderop, R.J.M. Bindels, Magnesium in man: implications for health and disease, Physiol. Rev. 95 (2015) 1–46. [8] A. Myrissa, N.A. Agha, Y. Lu, E. Martinelli, J. Eichler, G.b. Szakács, C. Kleinhans, R. Willumeit-Romer, U. Schafer, A.-M. Weinberg, In vitro and in vivo comparison of binary Mg alloys and pure Mg, Mater. Sci. Eng. C 61 (2016) 865–874. [9] J. Kubásek, D. Vojtech, E. Jablonská, I. Pospísilová, J. Lipov, T. Ruml, Structure, mechanical characteristics and in vitro degradation, cytotoxicity, genotoxicity and mutagenicity of novel biodegradable Zn-Mg alloys, Mater. Sci. Eng. C 58 (2016) 24–35. [10] H. Gong, K. Wang, R. Strich, J.G. Zhou, In vitro biodegradation behavior, mechanical properties, and cytotoxicity of biodegradable Zn–Mg alloy, J. Biomed. Mater. Res. B Appl. Biomater. 103 (2015) 1632–1640. [11] M. Dambatta, N. Murni, S. Izman, D. Kurniawan, G. Froemming, H. Hermawan, In vitro degradation and cell viability assessment of Zn-3Mg alloy for biodegradable bone implants, Proc. Inst. Mech. Eng. H 229 (2015) 335–342. [12] N.S. Murni, M.S. Dambatta, S.K. Yeap, G.R.A. Froemming, H. Hermawan, Cytotoxicity evaluation of biodegradable Zn-3Mg alloy toward normal human osteoblast cells, Mater. Sci. Eng. C 49 (2015) 560–566. [13] E. Mostaed, M. Sikora-Jasinska, A. Mostaed, S. Loffredo, A.G. Demir, B. Previtali, D. Mantovani, R. Beanland, M. Vedani, Novel Zn-based alloys for biodegradable stent applications: design, development and in vitro degradation, J. Mech. Behav. Biomed. Mater. 60 (2016) 581–602.

[14] M. Alves, D. Cunha, C.F. Santos, N. Mira, M.F. Montemor, In vitro corrosion behaviour and anti-Candida spp. activity of Zn coated with ZnO-nanostructured ‘Anastacia’ flowers, J. Mater. Chem. B (2016), http://dx.doi.org/10.1039/C6TB00797J. [15] H. Hu, W. Zhang, Y. Qiao, X. Jiang, X. Liu, C. Ding, Antibacterial activity and increased bone marrow stem cell functions of Zn-incorporated TiO2 coatings on titanium, Acta Biomater. 8 (2012) 904–915. [16] J.K. Park, Y.-J. Kim, J. Yeom, J.H. Jeon, G.-C. Yi, J.H. Je, S.K. Hahn, The topographic effect of zinc oxide nanoflowers on osteoblast growth and osseointegration, Adv. Mater. 22 (2010) 4857–4861. [17] M.F. Khan, M. Hameedullah, A.H. Ansari, E. Ahmad, M.B. Lohani, R.H. Khan, M.M. Alam, W. Khan, F.M. Husain, I. Ahmad, Flower-shaped ZnO nanoparticles synthesized by a novel approach at near-room temperatures with antibacterial and antifungal properties, Int. J. Nanomedicine 9 (2014) 853–864. [18] T. Prosek, A. Nazarov, U. Bexell, D. Thierry, J. Serak, Corrosion mechanism of model zinc–magnesium alloys in atmospheric conditions, Corros. Sci. 50 (2008) 2216–2231. [19] M.M. Alves, C. Santos, M.J. Carmezim, M.F. Montemor, Nanostructured 'Anastacia' flowers for Zn coating by electrodepositing ZnO at room temperature, Appl. Surf. Sci. 332 (2015) 152–158. [20] T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomater. 27 (2006) 2907–2915. [21] M. Wadowska, T. Frade, D. Siopa, K. Lobato, A. Gomes, ZnO nanostructured films electrodeposited at room temperature, ECS Electrochem. Lett. 2 (2013) D40–D42. [22] M.J. Bushiri, R. Vinod, S. Alfredo, S. Juan Angel, Pressure-induced phase transition in hydrothermally grown ZnO nanoflowers investigated by Raman and photoluminescence spectroscopy, J. Phys. Condens. Matter 27 (2015) 385401. [23] S.K. Sharma, A. Rammohan, A. Sharma, Templated one step electrodeposition of high aspect ratio n-type ZnO nanowire arrays, J. Colloid Interface Sci. 344 (2010) 1–9. [24] A. Nazarov, E. Diler, D. Persson, D. Thierry, Electrochemical and corrosion properties of ZnO/Zn electrode in atmospheric environments, J. Electroanal. Chem. 737 (2015) 129–140. [25] P. Volovitch, C. Allely, K. Ogle, Understanding corrosion via corrosion product characterization: I. Case study of the role of Mg alloying in Zn-Mg coating on steel, Corros. Sci. 51 (2009) 1251–1262. [26] M.A. Amin, S.S. Abd El-Rehim, F.D.A. Aarão Reis, I.S. Cole, Metastable and stable pitting events at zinc passive layer in alkaline solutions, Ionics 20 (2014) 127–136. [27] X. Lu, Y. Leng, Theoretical analysis of calcium phosphate precipitation in simulated body fluid, Biomater. 26 (2005) 1097–1108. [28] D.R. Katti, A. Sharma, A.H. Ambre, K.S. Katti, Molecular interactions in biomineralized hydroxyapatite amino acid modified nanoclay: in silico design of bone biomaterials, Mater. Sci. Eng. C 46 (2015) 207–217. [29] N.C. Hosking, M.A. Strom, P.H. Shipway, C.D. Rudd, Corrosion resistance of zinc-magnesium coated steel, Corros. Sci. 49 (2007) 3669–3695. [30] T. Prosek, D. Persson, J. Stoulil, D. Thierry, Composition of corrosion products formed on Zn–Mg, Zn–Al and Zn–Al–Mg coatings in model atmospheric conditions, Corros. Sci. 86 (2014) 231–238. [31] E. Diler, B. Lescop, S. Rioual, G. Nguyen Vien, D. Thierry, B. Rouvellou, Initial formation of corrosion products on pure zinc and MgZn2 examinated by XPS, Corros. Sci. 79 (2014) 83–88. [32] J. Kubasek, D. Vojtech, Zn-based alloys as an alternative biodegradable materials, 21st International Conference on Metallurgy and Materials (Metal 2012) 2012, pp. 1355–1361. [33] C.Z. Yao, Z.C. Wang, S.L. Tay, T.P. Zhu, W. Gao, Effects of Mg on microstructure and corrosion properties of Zn-Mg alloy, J. Alloys Compd. 602 (2014) 101–107. [34] X. Luo, D. Barbieri, N. Davison, Y. Yan, J.D. de Bruijn, H. Yuan, Zinc in calcium phosphate mediates bone induction: in vitro and in vivo model, Acta Biomater. 10 (2014) 477–485.

120

M.M. Alves et al. / Materials Science and Engineering C 70 (2017) 112–120

[35] Z.-S. Tao, W.-S. Zhou, X.-W. He, W. Liu, B.-L. Bai, Q. Zhou, Z.-L. Huang, K.-k. Tu, H. Li, T. Sun, Y.-X. Lv, W. Cui, L. Yang, A comparative study of zinc, magnesium, strontiumincorporated hydroxyapatite-coated titanium implants for osseointegration of osteopenic rats, Mater. Sci. Eng. C 62 (2016) 226–232.

[36] S. Ribeiro, S.C. Sartoretto, R. Resende, M. Uzeda, A.T. Alves, S. Albuquerque Santos, G. Pesce, A. Malta Rossi, J.M. Granjeiro, F. Miguel, M.D. Calasans-Maia, In vivo evaluation of zinc-containing nanostructured carbonated hydroxyapatite, Key Eng. Mater. 696 (2016) 223.