Selective dissolution of amorphous Zr–Cu–Ni–Al alloys

Corrosion Science 94 (2015) 350–358

Contents lists available at ScienceDirect

Corrosion Science journal homepage: www.elsevier.com/locate/corsci

Selective dissolution of amorphous Zr–Cu–Ni–Al alloys L. Mihaylov a, L. Lyubenova a, Ts. Gerdjikov a, D. Nihtianova a,b, T. Spassov a,⇑ a b

Faculty of Chemistry and Pharmacy, University of Sofia ‘‘St. Kl. Ohridski’’, 1 James Bourchier str., 1164 Sofia, Bulgaria Institute of Mineralogy and Crystallography, BAS, Bulgaria

a r t i c l e

i n f o

Article history: Received 20 September 2014 Accepted 13 February 2015 Available online 27 February 2015 Keywords: A. Zirconium A. Glass C. de-alloying B. SEM B. TEM

a b s t r a c t Amorphous Zr–Ni–Cu–Al ribbons were dealloyed by electrochemical selective dissolution. As a result three-dimensional microporous structures were obtained at different conditions (electrochemical potential, temperature, dissolution time). The smallest porous were obtained at low temperatures and potentials. To study the effect of precursor microstructure on the porous structure the as-cast alloys were annealed at two different temperatures – above Tg to cause structural relaxation of the amorphous material and at temperature above Tx to produce crystalline alloy. It is shown that only the entirely amorphous (as-cast and low-temperature annealed) alloys form homogeneous microporous structures as a result of the selective dissolution. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Nanoporous and microporous metals are intensively studied in the last years [1–6]. Due to their high surface area, good electroand thermal-conductivity, and outstanding mechanical stability, they are considered for various applications in catalysis, sensing devices, filtering, biotechnology, etc. Dealloying is one of the routes to produce porous materials via selective dissolution of the less noble metal [7–9]. The remaining more noble elements undergo reorganization into a nanoporous or microporous structure with bicontinuous ligaments. The pore size in the dealloyed structure is dependent on alloy composition. Sun et al. [10] have shown for the first time the formation of a nanoporous structure by dealloying the more noble component. They demonstrate that nanoporous Ni films can be formed by a process involving electrodeposition of a homogeneous, single-phase NixCu1x binary alloy followed by electrochemical etching of Cu from the alloy. The formation of a passive oxide film on nickel in sulfamate solutions allows the selective electrochemical etching of copper. These results show a general strategy for fabrication of nanoporous structures by electrochemical dealloying where the more active component is passivated and hence is kinetically rather than thermodynamically stable. This strategy is important, for example, in the fabrication of nanoporous structures from transition metal elements. For producing attractive for various applications micro- and nanoporous structures different types of alloys have been used: ⇑ Corresponding author. Tel.: +359 2 81 61 236; fax: +359 2 96 25 438. E-mail address: [email protected]fia.bg (T. Spassov). http://dx.doi.org/10.1016/j.corsci.2015.02.031 0010-938X/Ó 2015 Elsevier Ltd. All rights reserved.

two-phase crystalline alloys [7,11–19] single phase solid solutions [10,20–22] as well as fine nanocrystalline and amorphous alloys [4–6,14,22–39]. To form uniform nanoporosity upon dealloying, an alloy system is required to be a monolithic phase because the nanoporosity is formed by a self-assembly process through surface diffusion, not by the simple excavation of one phase from a pre-separated multiphase system [28]. Lu et al. [17] reported that two-phase nanocrystalline 62Cu–38Zr (wt.%) films were de-alloyed in HCl solution by electrochemical etching of the zirconium component, resulting in the formation of porous copper with uniform diameter of approximately 500 nm. In contrast, no porous copper could be obtained from coarse-grain 62Cu–38Zr alloys under the same conditions. The dealloying of the duplex-phase Cu–20Zr alloys and corresponding sputtered Cu–20Zr films (wt.%) in hydrochloric acid solution was investigated using electrochemical and chemical techniques [7]. The results show that the dissolution mechanisms for the cast alloy and the films were strongly dependent on the electrochemical activities of Zr and Cu in each phase. Supersaturated nanocrystalline (16–19 nm) Cu50Ag50 and Cu70Ag30 solid solutions, synthesized by high-energy ball milling, were subjected to electrochemical dealloying [20]. The electrochemical selective dissolution reveals a quasi-critical potential of Cu dissolution in Cu70Ag30 that is shifted 40 mV positively, compared to the dissolution potential of pure Cu. The electrochemical dealloying of rapidly solidified nanocrystalline Mg–Cu alloys in a 0.2 M NaCl aqueous solution results in nanoporous copper [14]. The microstructure of the as-dealloyed Mg60Cu40 alloy was found to be similar to that of the as-dealloyed Mg67Cu33 alloy, but the size of Cu nanoparticles was reduced (180 ± 30 nm).

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

In comparison with crystalline alloys, multicomponent metallic glasses are monolithic in phase with a homogeneous composition and structure down to subnanoscale. Due to these characteristics, amorphous alloys are ideal for studying the role of minor alloying elements on the process of selective dissolution. Nanoporous palladium was synthesized by electrochemically dealloying multicomponent amorphous Pd30Ni50P20 [28] and Pd42.5Cu30Ni7.5P20 [38] metallic glass ribbons. A porous Pd structure with two length-scale nanopores at 50 and 5 nm [28] and about 2 nm [38] could be observed. Nanoporous gold is synthesized by electrochemically dealloying a multicomponent Au35Si20Cu28Ag7Pd5Co5 [28] and Au40Cu28Ag7Pd5Si20 [37] metallic glasses. Ribbons were obtained with the Au42Cu29Ti8Si21 and Au44Cu31Ti4Si21 compositions close to the parting limit for dealloying [26]. Dealloying was found to be slow, without to extend to the whole thickness of the ribbon during the time of the experiments. Although the morphology of gold aggregates differs on the two sides, they are composed of scattering domains of the same size, of the order of 25 nm. A nanoporous gold–palladium alloy has also been successfully fabricated by electrochemically dealloying a multicomponent glassy Au30Si20Cu33Ag7Pd10 ribbon [30]. A uniform nanoporous structure consisting of quasi-periodic nanoporous channels and gold ligaments with the characteristic length of 50 nm were formed. The SEM–EDS spectrum demonstrates that 3–5 atom% Pd remain in the Au ligaments and less noble elements, such as Si and Cu, have been completely dissolved during the dealloying process. Fabrication of a nanometersized porous Ti-based metallic glass by application of selective dealloying technique to the Y20Ti36Al24Co20 two-phase amorphous ribbon alloy, consisted of Ti43.3Y3.7Al15.3Co37.7 and Y38.8Ti12.8Al37.1Co11.3 [23]. Three-dimensional interconnected porous Ti-based metallic glass with pore sizes in the range of 15–155 nm has been obtained by selective dissolution of the Y-rich amorphous phase. By transforming the surface of the Ti45Y11Al24Co20 phase separated alloy from smooth toward rough with nano-pores in an oxidized state, the passivation behavior of the glassy alloy in simulated body fluid condition was remarkably improved [31]. The average pores size was about 40 nm, as the size of the pores during the dealloying process was increased with increasing immersion time. Nanoporous copper with a pore size of 8–55 nm was obtained through dealloying from amorphous Ti60Cu40xAgx (x = 1, 2 at.%) under free immersion conditions in 0.03 M HF solution [36]. Ti oxide nano-porous structure on the surface of Ti–Cu amorphous alloy was synthesized using a de-alloying method as well [33]. The mean diameter of nanopores was about 50 nm and the thickness of pore walls is about 100 nm. The nano-porous surface was composed of Ti and Cu oxides. The depth of the nano-porous structure is about 500 nm. The effect of minor alloying elements on dealloying and the resulting nanoporous structure of two similar amorphous melt spun alloys, Al70Cu18Mg12 and Al73Cu16Mg8Ni3 was evaluated [9]. The chemical and structural homogeneity of amorphous alloys enables the formation of a uniform dealloyed layer throughout the material without containing irregularities in the nanoporous structure related to interphase interfaces and grain boundaries. A crystalline, Cu-rich, nanoporous structure with a pore diameter of 10–30 nm was formed as a result of the selective dissolution of Al and Mg and the reorganization of the remaining Cu. The aim of the present work was to obtain 3-dimensional microporous metallic structure by selective dissolution of amorphous Zr-based alloys and to study the influence of the primary alloy microstructure and electrochemical conditions on the morphology and structure of the final porous material (pore and ligament size). Based on our previous experience with binary Zr-based glasses [40] Zr–Ni–Cu–Al system was chosen as a

351

precursor for preparing microporous metallic structure first because it is a challenging material for selective dissolution not containing real noble metal and second due to the very good glass forming ability of this system and therefore good opportunity to produce in a maximum degree homogeneous amorphous alloys. The last issue was one of the aims of the study because in general it is not known what kind of heterogeneities into the amorphous material could be the reason to initiate a selective dissolution process either chemically or electrochemically. 2. Experimental Amorphous Zr–Ni–Cu–Al ribbons were prepared by rapid solidification from the melt (melt spinning) with a 25 m/s surface velocity of the copper quenching disk (with a diameter of 20 cm) under argon atmosphere [41]. The morphology, microstructure and chemical composition of the as-cast and heat treated alloys were examined by SEM (JEOL-5510), TEM (JEOL-2100, 200 kV) and energy dispersive X-ray analysis (EDX). The chemical composition of the alloys was found to correspond to the target composition, assigned during the synthesis of the ribbons. ImageJ software was used for determining the pore size distribution at different dealloying conditions. The largest size of the pores was considered for this analysis. The microstructure of the samples was characterized also by X-ray diffraction (XRD) using Cu Ka radiation. The thermal behavior was studied by differential scanning calorimetry (DSC 7, Perkin-Elmer). Cyclic polarization measurements in the potential range from 1.0 to 1.4 V (vs. SHE) were carried out to determine the corrosion potentials of the alloys studied. The electrochemical selective dissolution of the alloys was realized at different constant potentials in the range 400–1100 mV (vs. SHE). The electrolyte 0.1 M HCI + NaF was determined after optimization of the electrolyte type and concentration for obtaining efficient selective dissolution. Three-electrode cell with Ag/AgCl as a reference electrode and counter electrode prepared from Pt was used for the electrochemical experiments. The studied metallic ribbons (being the working electrode) with a measured single surface area of 1 cm2 were polished with 1 lm diamond paste and fixed on a PVC sample holder. For all analyses the free surface side of the ribbon was used, although the difference in the dealloying behavior of both ribbon sides (free surface side and wheel side) were found to be insignificant. PARSTAT 2273 electrochemical system was used for all electrochemical studies. After the dissolution experiments the samples were treated for a short time in an ultrasonic bath for cleaning the ribbon surface from some of the solid phase corrosion products. The composition of the solid corrosion products of dealloying were analyzed as well. To prove the overall ductility of the de-alloyed ribbon a simple bend test has been performed, where the metal ribbon was bent into a U-shape. 3. Results and discussion X-ray powder diffraction pattern of the as-quenched Zr67.5Cu15Ni10Al7.5 alloy, shown in Fig. 1, reveals its amorphous structure. TEM and selected area electron diffraction confirm that the ribbon is entirely amorphous, Fig. 2. Being a known easy glass former [41] this alloy reveals clear glass transition, well separated from the crystallization onset, Fig. 3, with a temperature range between the crystallization temperature Tx and the glass transition temperature Tg over 50 °C. Potentiodynamic polarization scans with different sweep rates of the melt-spun Zr-based amorphous ribbons were used to select the appropriate potential for selective dissolution, Fig. 4. On the

352

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

-1 -2 -3

2

log (i, A/cm )

Intensity (a.u.)

♦ − Zr2Ni • − Cu

(a)

♦

•

♦

(b)

-4 -5 -6

Al

Zr

Cu Ni

-7

20

30

40

50

60

70

Alloy

80

2Θ (degree)

-1,2

Fig. 1. X-ray diffraction patterns of as-quenched (a) and selectively dissolved (b) Zr67.5Cu15Ni10Al7.5 alloy.

-0,8

-0,4

0,0

0,4

0,8

1,2

1,6

E(V) vs SHE Fig. 4. Potentiodynamic polarization curves of amorphous Zr67.5Cu15Ni10Al7.5, pure Ni, Cu, Al and Zr.

Table 1 Chemical composition (in at.%) of the studied as-cast Zr-based alloy (A) and of the ligament after dealloying (B).

A B

10 1/nm

Zr

Cu

Ni

Al

65.5 3.5

15 95.5

12 1.0

7.5 0.0

0,07 a

0,06 -2

Current (A.cm )

b

0,05 c

0,04 d

0,03 0,02 0,01

10 nm

0

50

Fig. 2. TEM micrograph and selected area electron diffraction of as-quenched Zr67.5Cu15Ni10Al7.5 alloy.

100

150

200

Time (s)

endo >

Fig. 5. Time dependence of the corrosion current at different constant potentials of amorphous Zr67.5Cu15Ni10Al7.5: (a) 1100 mV; (b) 800 mV; (c) 600 mV; and (d) 400 mV (vs. SHE).

Δ − Zr 2 Ni_t ∇ − Zr 2 Ni_c

5 mW

♦ − Ni 2 Zr_c − Ni 3 Zr_h • − Al 2 Zr3_t

Intensity (a.u.)

c

Heat flow (mW)

b

a

a b ∇

•

∇

Δ Δ

♦ •

c

Tg

300

350

400

450

500

Temperature (ºC) Fig. 3. DSC curves of as-quenched amorphous (a), annealed at 400 °C (b) and at 470 °C (c) Zr67.5Cu15Ni10Al7.5 alloy.

30

40

50

60

70

2Θ (degree) Fig. 6. X-ray diffraction patterns of as-quenched (a), annealed at 400 °C (b) and at 470 °C (c) Zr67.5Cu15Ni10Al7.5 alloy.

353

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

a

b

1 m

1 m

d

c

1 m

1 m

Fig. 7. SEM micrographs of selectively dissolved Zr67.5Cu15Ni10Al7.5 ribbons at different potentials: (a) 400 mV; (b) 600 mV; (c) 800 mV; and (d) 1100 mV (vs. SHE). Dissolution time – 200 s.

a

500 m

b

500 m Fig. 8. SEM micrographs of selectively dissolved Zr67.5Cu15Ni10Al7.5 ribbons at the very beginning (a) and at later stage (b) of the dissolution process at 1100 mV (vs. SHE).

same figure the polarization curves of pure nickel, copper, aluminum and zirconium are also shown. Thus, the corrosion current, icorr  12.6 lA cm2, and potential Ecorr  0.48 V of the studied alloy were defined. The corrosion potential of the alloy is practically between those of Zr and the rest of the more noble metals, very close to that of Al. The critical passivating current density (ip) was found to be comparable to that of the Zr2Ni amorphous alloy obtained in our previous study [40]. Varying the type and the concentration of the electrolyte as well as the electrode potential optimal conditions for selective dissolution of Zr from the Zr-based alloys were obtained. It was found that in the range of potentials 400–1100 mV (vs. SHE) in 0.1 M HCl + NaF the process of dissolution is selective, as mainly Zr is leached out together with Al and some Ni. EDAX analysis of the selectively dissolved (dealloyed) zones shows nearly pure Cu in the expense of the other less-noble elements, Table 1. The time dependence of the corrosion current at different constant potentials (400–1100 mV vs. SHE), selected in the potential range in which Zr is oxidizing, is presented in Fig. 5. Detailed TEM (electron diffraction) and EDX analyses of the corrosion products show clearly the formation of ZrF4, Na3AlF6 and Na3ZrF7. All these compounds are poorly soluble in water and were easily removed with a short time ultrasonic treatment of the dealloyed samples. Potentiostatic selective dissolution experiments were also carried out with the same alloy, but preliminary annealed at different temperatures in order to achieve different microstructural states of the material. The annealing was made under Ar atmosphere at two temperatures for short time (1 min). The first temperature of 400 °C was chosen exactly above the Tg of this alloy aiming to achieve structural relaxation of the as-cast amorphous alloy (see Fig. 3) and the second, 470 °C, was selected after the second maximum of the exothermic crystallization peak in order to produce crystalline material. The effect of the applied annealing on the alloy microstructure could be seen from the XRD (Fig. 6) and DSC (Fig. 3) analyses of the heat treated samples. As a result, the annealed at

354

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

a

b

1 m

1 m

c

d

1 m

1 m

Fig. 9. SEM micrographs of Zr67.5Cu15Ni10Al7.5 ribbons at the very beginning of the dealloying at different potentials: (a) 400 mV; (b) 600 mV; (c) 800 mV; and (d) 1100 mV (vs. SHE).

60

70

400 mV

60

number of pores

50

number of pores

600 mV

40 30 20 10

50 40 30 20 10 0

0 0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,2

0,8

0,4

pore size (μm)

0,6

0,8

1,0

1,2

1,4

pore size (μm)

90 800 mV

80

60

1100 mV

number of pores

number of pores

70 60 50 40 30 20

50 40 30 20 10

10

0

0 0,2

0,4

0,6

0,8

pore size (μm)

1,0

1,2

0,0

0,2

0,4

0,6

0,8

pore size (μm)

Fig. 10. Pore size distribution of the selectively dissolved at different conditions Zr67.5Cu15Ni10Al7.5 alloys.

1,0

355

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

0,06

a

o

-2

Current (A.cm )

a

25 C

0,05

o

0,04

-10 C

0,03 0,02 0,01 0,00

1 m

0

50

100

150

200

250

Time (s)

b

b

1 m

1 m

180

c

160 Fig. 11. SEM micrographs of dealloyed relaxed amorphous (a) and nanocrystalline (b) Zr67.5Cu15Ni10Al7.5 alloy.

number of pores

the lower temperature ribbon was proven to be entirely amorphous and the alloy heat treated for 1 min at 470 °C appeared to be (nano)crystalline with a presence of about 30% amorphous phase, estimated from the X-ray diffraction pattern (Fig. 6) and confirmed by the DSC analysis of the annealed sample (Fig. 3). Although there is some little difference in the shape of the DSC crystallization peaks of the annealed at 400 °C sample the enthalpy of crystallization is equal to this of the as-cast ribbon, 65.5 J/g. The corrosion current for the as-cast amorphous ribbon increases very fast in the beginning (for 30–40 s), followed by a stage with a continuously decreasing speed of the current rise, reaching a nearly constant value of 45–60 mA/cm2 after 200 s (depending on the applied potential). Compared to the as-cast the preliminary heat treated alloys show somehow higher corrosion current, as that of the relaxed amorphous alloy has maximum value of about 100 mA/cm2. The morphology of the alloys after the potentiostatic selective dissolution at different potentials (400–1100 mV vs. SHE), clearly indicates a selective dissolution process, which has taken place in the studied amorphous metallic ribbons, Figs. 7 and 8. From the SEM micrographs it is evident that the dissolution process in both as-cast amorphous and relaxed amorphous (annealed at Tx > T P Tg) ribbon starts from certain active centers, which surface concentration is almost the same for the amorphous alloys, N = 2.2  104 cm2, Fig. 8. Once generated the partially dissolved zones start to ‘‘grow’’, forming circles in two dimensions (semispheres in three dimensions), which touch each other at certain time of selective dissolution. Although somewhat higher density of the selectively dissolved zones in the areas with some ribbon surface defects has been observed the overall picture reveals

140 120 100 80 60 40 20 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0

pore size (μm) Fig. 12. Time dependence of the corrosion current at 10 °C (a), SEM micrograph (b) and pore size distribution (c) of selectively dissolved at 10 °C Zr67.5Cu15Ni10Al7.5 ribbons at 800 mV (vs. SHE).

homogeneous distribution of the selectively dissolved regions, as the size of these zones is more or less the same. The last result reveals that during the ‘‘growth’’ of the zones new ones are almost not forming. The rather regular semispherical shape of the partially dissolved zones, especially for the as-cast amorphous alloy, is a clear indication for their three dimensional ‘‘growth’’ with equal rate. Dividing the radius of the zones by the time of dissolution at certain potential an average ‘‘growth’’ rate of the selectively dissolved zones of about 0.5 lm/s was determined. This result entirely confirmed the observations made in our previous study on the selective dissolution of amorphous Zr2Ni alloys [40]. It was also interesting to study in detail the morphology and the size distribution of the pores and of the ligaments into the selectively dissolved zones at the very beginning of the dealloying

356

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

a

a

1 m

b

60

b number of pores

50

Zr 2 Ni

40 30 20 10

Cu

10 1/nm 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 Fig. 13. TEM micrograph and selected area electron diffraction of Zr67.5Cu15Ni10Al7.5 after selective electrochemical dissolution at 800 mV for 200 s (taken from the edge of the ligament).

pore size (μm) Fig. 14. SEM micrograph and pore size distribution of Zr74.5Cu10Ni8Al7.5 after selective electrochemical dissolution (800 mV vs. SHE for 200 s).

process, Fig. 9. No substantial difference in the average pore size compared to the advanced stages of the dissolution process (see Fig. 7) could be observed. Thus, it can be concluded that in the amorphous alloys the dissolution process starts from a kind of defects in the otherwise homogeneous disordered structure, and develops uniformly in 3 dimensions. The pore size distribution, determined from the SEM micrographs of the selectively dissolved ribbons (Fig. 7) dealloyed at different conditions is shown in Fig. 10. The metallic ribbons dealloyed at lower potentials (400 mV) show smaller pores compared to those obtained at higher potentials. Generally, the pores are relatively large, most probably due to the low concentration of copper, which atoms have to rearrange, forming ligaments after removing mainly the less noble metallic atoms in the alloy. Not a big difference could be observed in the pore size distribution between the as-cast amorphous alloy and of the alloys in relaxed amorphous state. Completely different is, however, the behavior during selective dissolution of the nanocrystalline alloy (50 nm), prepared by crystallization of the amorphous ribbon, Figs. 11 and 6. In this case 3-dimensional porous structure (with well defined ligaments) could not be obtained. Isolated Cu-rich nanocrystals are deposited (weakly attached to the surface) on the surface of the ribbon. This result is, however, expected, because although nanocrystalline the alloy consists of crystallites of mainly two phases and the one with the higher chemical potential is dissolving first, leaving in this way the other nanocrystals unattached. Another experimental parameter, having potential influence on the selective dissolution process and subsequently on the pore structure is the temperature of the electrolyte during the dissolution. To prove this influence a distinctly lower T (10 °C) for the dissolution was selected as well. The effect of the T is revealed on the SEM micrographs as well as from the pore size distribution, both presented in Fig. 12. Although somehow smaller pores (about

400 nm) compared to those of the ribbon dealloyed at room temperature (25 °C), generally it cannot be concluded that the effect of the atomic mobility reduction on the pores structure is substantial. The average pore size, obtained from the pore size distribution curves, was used to determine the surface diffusion coefficient Ds, according to the equation [42]:

Ds ¼

dðtÞ4 kT ; 32cta4

ð1Þ

where k is Boltzmann constant, c is surface energy, t is the dealloying time, d(t) is the pore size at t, T is the temperature (298 K) and a is the lattice constant. The surface energy of Cu (1.79 J m2) was selected for the purpose of the calculation. Surface diffusion coefficients of Cu thus obtained are in the range 1  1012– 1.5  1012 m2/s and vary a little, depending on the initial alloy microstructure and the temperature of the dissolution process. To have additional proof for the selectivity of the electrochemical dissolution TEM analysis has been performed with the amorphous alloy after the potentiostatic electrochemical experiment (Fig. 13). TEM micrograph and SAED pattern of the selectively dissolved alloys reveal that in the very thin areas close to the pore edge nanocrystals (2–5 nm) of pure Cu are present, Fig. 13. In the ticker parts of the ligaments nanocrystals of Zr2Ni and Ni2Zr could also be observed. The size of the nanocrystals (Ni2Zr and Cu) is in the order of about 2–10 nm. The crystalline phases could also be registered by XRD, taken from the partially selectively dissolved amorphous samples. To prove whether the copper content of the alloy is critical for obtaining 3D porous structure by selective dissolution of amorphous Zr–Ni–Cu–Al ribbons similar analysis was realized with amorphous Zr74.5Ni8Cu10Al7.5. The SEM analysis shows not a big

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

difference in the pore size distribution, compared to the Cu-richer amorphous alloy, Fig. 14. It is however, necessary to be emphasized that the selective dissolution process of both amorphous alloys results in 3D porous structure without complete dissolution of the less-noble than Cu metals. The dealloyed material contains certain amount of Ni, Al and Zr especially in the thicker ligaments parts. Thus, the final porous material contains nanocrystals of Cu and Ni–Zr phases and residual amorphous phase. Such phase composition combines enhanced surface properties (chemical activity) and increased structural stability of the porous metal and could be considered as a desired microstructure for different applications. 4. Conclusion Varying a number of parameters of the electrochemical selective dissolution (electrolyte, electrochemical potential, temperature) 3-dimensional microporous structures have been obtained from amorphous Zr-based alloys precursors. Three different types of samples with the same composition have been studied: as-cast amorphous, annealed at temperature slightly above Tg (relaxed amorphous state) and nanocrystalline with some amorphous phase, prepared by crystallization of the amorphous alloy. Smallest average pore size of about 400 nm (ligament size – 150 nm) was reached at T = 10 °C and 400 mV (vs. SHE) for the as-cast amorphous ribbons. The ligament composition after the dissolution process revealed mainly Cu and Zr–Ni phases. Substantial influence of the electrochemical potential on the pores structure and size was not detected, while the influence of the temperature of dissolution was found to be stronger. Considering the initial microstructure of the alloy to be dealloyed it was clearly proved that only the amorphous state (as-cast and relaxed) lead to homogeneous 3-dimensional porous structure, whereas the selective dissolution of the nanocrystalline sample results in the formation of Cu-rich isolated crystalline grains. The surface diffusion coefficient of Cu has also been estimated. Acknowledgements The work has been supported by Bulgarian Scientific Research Fund under Grant DCVP 02-2/2009 (Union). The authors are grateful to the FP7 project BeyondEverest. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.corsci.2015.02. 031. References [1] J. Zhang, C.M. Li, Nanoporous metals: fabrication strategies and advanced electrochemical applications in catalysis, sensing and energy systems, Chem. Soc. Rev. 41 (2012) 7016–7031. [2] J. Weissmüller, R.C. Newman, H.-J. Jin, A.M. Hodge, J.W. Kysar, Nanoporous metals by alloy corrosion: formation and mechanical properties, MRS Bull. 34 (2009) 577–586. [3] W.B. Liu, S.C. Zhang, N. Li, J.W. Zheng, S.S. An, Y.L. Xing, A general dealloying strategy to nanoporous intermetallics, nanoporous metals with bimodal, and unimodal pore size distributions, Corros. Sci. 58 (2012) 133–138. [4] Z. Dan, F. Qin, Yu Sugawara, I. Muto, N. Hara, Dependency of the formation of Au-stabilized nanoporous copper on the dealloying temperature, Microporous Mesoporous Mater. 186 (2014) 181–186. [5] Z.M. Wang, X.C. Chang, W.L. Hou, J.Q. Wang, Selective dissolution sensitive to minor alloying in CuZr-based metallic glasses, Corros. Sci. 76 (2013) 465–473. [6] N. Tuan, J. Park, J. Lee, J. Gwak, D. Lee, Synthesis of nanoporous Cu films by dealloying of electrochemically deposited Cu–Zn alloy films, Corros. Sci. 80 (2014) 7–11. [7] H.-B. Lu, Y. Li, F.-H. Wang, Dealloying behaviour of Cu–20Zr alloy in hydrochloric acid solution, Corros. Sci. 48 (2006) 2106–2119.

357

[8] G. Adamek, J. Jakubowicz, Microstructure of the mechanically alloyed and electrochemically etched Ti–6Al–4V and Ti–15Zr–4Nb nanocrystalline alloys, Mater. Chem. Phys. 124 (2010) 1198–1204. [9] T. Aburada, J.M. Fitz-Gerald, J.R. Scully, Synthesis of nanoporous copper by dealloying of Al–Cu–Mg amorphous alloys in acidic solution: the effect of nickel, Corros. Sci. 53 (2011) 1627–1632. [10] L. Sun, C.-L. Chien, P.C. Searson, Fabrication of nanoporous nickel by electrochemical dealloying, Chem. Mater. 16 (2004) 3125–3129. [11] S. Brittman, A.J. Smith, S. Milenkovic, A.W. Hassel, Copper nanowires and silver micropit arrays from the electrochemical treatment of a directionally solidified silver–copper eutectic, Electrochim. Acta 53 (2007) 324–329. [12] A.W. Hassel, B. Bello-Rodriguez, S. Milenkovic, A. Schneider, Electrochemical production of nanopore arrays in a nickel aluminium alloy, Electrochim. Acta 50 (2005) 3033–3039. [13] H. Yasuda, I. Ohnaka, S. Fujimoto, N. Takezawa, A. Tsuchiyama, T. Nakano, K. Uesugi, Fabrication of aligned pores in aluminum by electrochemical dissolution of monotectic alloys solidified under a magnetic field, Scr. Mater. 54 (2006) 527–532. [14] C. Zhao, X. Wang, Z. Qi, H. Ji, Z. Zhang, On the electrochemical dealloying of Mg–Cu alloys in a NaCl aqueous solution, Corros. Sci. 52 (2010) 3962–3972. [15] H.-Y. Lee, M. Demura, Y. Xu, D.-M. Wee, T. Hirano, Selective dissolution of the c phase in a binary Ni(c)/Ni3Al(c0 ) two-phase alloy, Corros. Sci. 52 (2010) 3820– 3825. [16] S. Milenkovic, V. Dalbert, R. Marinkovic, A.W. Hassel, Selective matrix dissolution in an Al–Si eutectic, Corros. Sci. 51 (2009) 1490–1495. [17] H.-B. Lu, Y. Li, F.-H. Wang, Synthesis of porous copper from nanocrystalline two-phase Cu–Zr film by dealloying, Scr. Mater. 56 (2007) 165–168. [18] A.K. Gupta, P.H. Marois, D.J. Lloyd, Review of the techniques for the extraction of second-phase particles from aluminum alloys, Mater. Charact. 37 (1996) 61–80. [19] R. Mao, S. Liang, X. Wang, Q. Yang, B. Han, Effect of preparation conditions on morphology and thermal stability of nanoporous copper, Corros. Sci. 60 (2012) 231–237. [20] T. Spassov, L. Lyubenova, Y. Liu, S. Bliznakov, M. Spassova, N. Dimitrov, Mechanochemical synthesis, thermal stability and selective electrochemical dissolution of Cu–Ag solid solutions, J. Alloys Compd. 478 (2009) 232–236. [21] X. Lu, E. Bischoff, R. Spolenak, T.J. Balk, Investigation of dealloying in Au–Ag thin films by quantitative electron probe microanalysis, Scr. Mater. 56 (2007) 557–560. [22] J.-K. Chang, S.-H. Hsu, I.-W. Sun, W.-T. Tsai, Formation of nanoporous nickel by selective anodic etching of the nobler copper component from electrodeposited nickel–copper alloys, J. Phys. Chem. C 112 (2008) 1371–1376. [23] J. Jayaraj, B.J. Park, D.H. Kim, W.T. Kim, E. Fleury, Nanometer-sized porous Ti-based metallic glass, Scr. Mater. 55 (2006) 1063–1066. [24] D. Frankel, S. Milenkovic, A.J. Smith, A.W. Hassel, Nanostructuring of NiAl–Mo eutectic alloys by selective phase dissolution, Electrochim. Acta 54 (2009) 6015–6021. [25] X.H. Chen, X.C. Zhang, Y. Zhang, G.L. Chen, Fabrication and characterization of metallic glasses with a specific microstructure for micro-electro-mechanical system applications, J. Non-Cryst. Solids 354 (2008) 3308–3316. [26] F. Scaglione, A. Gebert, L. Battezzati, Dealloying of an Au-based amorphous alloy, Intermetallics 18 (2010) 2338–2342. [27] X. Wang, W. Wang, Z. Qi, C. Zhao, H. Ji, Z. Zhang, Electrochemical catalytic activities of nanoporous palladium rods for methanol electrooxidation, J. Power Sources 195 (2010) 6740–6747. [28] J. Yu, Y. Ding, C. Xu, A. Inoue, T. Sakurai, M. Chen, Nanoporous metals by dealloying multicomponent metallic glasses, Chem. Mater. 20 (2008) 4548– 4550. [29] L. Battezzati, F. Scaglione, Dealloying of rapidly solidified amorphous and crystalline alloys, J. Alloys Compd. 509 (2011) S8–S12. [30] X.Y. Lang, H. Guo, L.Y. Chen, A. Kudo, J.S. Yu, W. Zhang, A. Inoue, M.W. Chen, Novel nanoporous Au–Pd alloy with high catalytic activity and excellent electrochemical stability, J. Phys. Chem. C 114 (2010) 2600–2603. [31] J. Jayaraj, J.M. Park, P.F. Gostin, E. Fleury, A. Gebert, L. Schultz, Nano-porous surface states of Ti–Y–Al–Co phase separated metallic glass, Intermetallics 17 (2009) 1120–1123. [32] S.L. Zhu, J.L. He, X.J. Yang, Z.D. Cui, L.L. Pi, Ti oxide nano-porous surface structure prepared by dealloying of Ti–Cu amorphous alloy, Electrochem. Commun. 13 (2011) 250–253. [33] X. Luo, R. Li, L. Huang, T. Zhang, Nucleation and growth of nanoporous copper ligaments during electrochemical dealloying of Mg-based metallic glasses, Corros. Sci. 67 (2013) 100–108. [34] G. Lan, Z. Xie, Zh. Huang, S. Yang, X. Zhang, Y. Zeng, J. Jiang, Amorphous alloy: promising precursor to form nanoflowerpot, Adv. Mater. Sci. Eng. (2014) (Article ID 263681). [35] Z. Dan, F. Qin, Yu. Sugawara, I. Muto, N. Hara, Effects of the initial microstructure of Ti–Cu alloys on final nanoporous copper via dealloying, J. Alloys Compd. 557 (2013) 166–171. [36] Z. Dan, F. Qin, A. Makino, Yu Sugawara, I. Muto, N. Hara, Fabrication of nanoporous copper by dealloying of amorphous Ti–Cu–Ag alloys, J. Alloys Compd. 586 (2014) S134–S138. [37] P. Rizzi, F. Scaglione, L. Battezzati, Nanoporous gold by dealloying of an amorphous precursor, J. Alloys Compd. 586 (2014) S117–S120. [38] Z. Dan, F. Qin, T. Wada, S. Yamaura, G. Xie, Yu Sugawara, I. Muto, A. Makino, N. Hara, Nanoporous palladium fabricated from an amorphous Pd42.5Cu30Ni7.5P20

358

L. Mihaylov et al. / Corrosion Science 94 (2015) 350–358

precursor and its ethanol electro-oxidation performance, Electrochim. Acta 108 (2013) 512–519. [39] F. Scaglione, P. Rizzi, F. Celegato, L. Battezzati, Synthesis of nanoporous gold by free corrosion of an amorphous precursor, J. Alloys Compd. 615 (2014) S142–S147. [40] L. Mihailov, M. Redzheb, T. Spassov, Selective dissolution of amorphous and nanocrystalline Zr2Ni, Corros. Sci. 74 (2013) 308–313.

[41] D. Zander, L. Lyubenova, R. Janlewing, U. Koester, Microstructural design by controlled crystallization and alloying of Zr–Cu–Ni–Al based metallic glasses, J. Metastable Nanocryst. Mater. 15–16 (2003) 23–30. [42] E.G. Seebauer, C.E. Allen, Estimating surface diffusion coefficients, Prog. Surf. Sci. 49 (1995) 265–330.