Morphology and composition of the Fe–Ni powders electrodeposited from citrate containing electrolytes

Electrochimica Acta 55 (2009) 535–543

Contents lists available at ScienceDirect

Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta

Morphology and composition of the Fe–Ni powders electrodeposited from citrate containing electrolytes U. Laˇcnjevac, B.M. Jovic´ 1 , V.D. Jovic´ ∗,1 Institute for Multidisciplinary Research, P.O. Box 33, 11030 Belgrade, Serbia

a r t i c l e

i n f o

Article history: Received 9 June 2009 Received in revised form 1 September 2009 Accepted 5 September 2009 Available online 15 September 2009 Keywords: Electrodeposition Powder Fe–Ni Morphology Composition analysis

a b s t r a c t The electrodeposition of the Fe–Ni powders from citrate containing electrolytes for different Ni/Fe ions concentration ratios, using Fe(III) and Fe(II) salts at pH 4.5 and pH 4.0 respectively was investigated by the polarization measurements and cyclic voltammetry. The morphology and composition of the electrodeposited powders were investigated by SEM and EDS analysis. The EDS analysis of the alloy powders confirmed anomalous co-deposition of Fe and Ni from both solutions, with the one obtained using Fe(III) salt being more pronounced. The morphology of electrodeposited powders was found to depend on the Ni/Fe ions concentration ratio, as well as on the valence of Fe ions used. A common characteristic for all powder samples was the presence of cone shaped cavities and nodules, while for the ratio Ni/Fe = 9/1 in both electrolytes pagoda like crystals, corresponding to the FeNi3 single crystal, have been detected. In the case of Fe(III) containing electrolytes current efficiency for powder electrodeposition was very small (about 1–2%) due to the first step in the electrochemical reaction being reduction of Fe(III) into Fe(II), while in the case of Fe(II) containing electrolytes current efficiency for powder electrodeposition varied between about 15% and 8% depending on the Ni/Fe ratio. © 2009 Elsevier Ltd. All rights reserved.

1. Introduction The Fe–Ni alloys are of worldwide economic interest because of their usage in a great variety of products [1]. Due to their magnetic properties these alloys have many applications in the area of memory devices for computers, and they are also resistant to corrosion, receptive to chrome, ductile etc. [1]. Their mechanical and magnetic properties are widely investigated [2–6], while the best known example is Permalloy (Fe–Ni alloy used for soft magnetic read/write heads [7–9]). Most of the investigations concerning their electrodeposition are in connection with the deposition of compact coatings and some of them are discussed below. The parameters influencing the electrodeposition of Fe–Ni alloys, such as the presence of sulfate or chloride solutions, addition of boric acid, citric acid and l-ascorbic acid at different pH values (2 and 3) were investigated in the paper of Kieling [10] and Yin and Lin [11]. Among the anomalous character of Fe and Ni co-deposition these investigations showed that the deposition of Fe–Ni alloys with larger Fe content occurred with lower current efficiency [10]. The addition of saccharin up to 3 g dm−3 in the electrolyte produced smooth and bright deposit of uniform thickness in comparison with

∗ Corresponding author. Tel.: +381 11 3303688; fax: +381 11 3055289. ´ E-mail address: [email protected] (V.D. Jovic). 1 ISE member. 0013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2009.09.012

burnt and torn deposit obtained from a pure chloride electrolyte [12]. Nucleation and growth of Fe–Ni alloys from chloride solution at pH 2 and pH 4 was investigated by current transient measurements, showing that the nucleation mechanism is instantaneous with typical 3D nucleation and growth for all investigated solutions [13]. The electrodeposition of Fe–Ni alloy powders was the subject of only few papers. Zhelibo et al. [14,15] suggested a method for producing very fine Fe–Ni alloy powder by the electrolysis in a two-layer electrolytic bath using a hydrocarbon solvent from an oil refining fraction as an upper organic layer with evaporation at 180 ◦ C and subsequent reduction annealing in a hydrogen atmosphere. Powders were electrodeposited from the simple Fe(II) and Ni(II) salts of different concentrations. The composition of Fe–Ni powders varied between 31 at.% and 50 at.% of Ni, depending on concentration of metal ions and the temperature of the electrolysis. The influence of the reduction annealing temperature [14] and the electrolysis temperature [15] on the formation, chemical and phase composition, structure and magnetic properties of highly dispersed Fe–Ni alloy powders were investigated and the optimal thermal conditions for the production of powders with micron-sized particles were determined [14,15]. By the SEM investigation the Fe powder particles were found to be dendritic, while the Fe–Ni alloy powder particles were fern-like in the case of electrolysis at 40 ◦ C. With the increase of electrolysis temperature to 60 ◦ C particles became significantly larger, while at the temperature of 80 ◦ C they became again smaller and their shape changed to rounded ones.

536

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

Because of low magnification the internal structure of obtained particles could not be identified. The effect of complexing agents (citric and oxalic acid of the concentration of 0.05 M) on the process of Fe–Ni alloy powders was also investigated in the range of pH values between 3.4 and 6.0 [16]. It was shown that complexing agents influence the kinetics of powders electrodeposition, as well as the morphology of the Fe–Ni powders. According to the conclusion of the author finer powders were produced in the presence of citric acid in comparison with those obtained in the presence of oxalic acid [16]. It should be mentioned here that at the magnification used agglomerates could not be seen in the powders and that the surfaces of certain agglomerates were investigated by the SEM analysis. Nodule like particles were present in both cases (Figs. 8–11 of Ref. [16]), except that the dimension of nodules was smaller in the powders electrodeposited from citrate containing electrolyte. Hence, it could be stated only that the dimension of nodules on the surface of powder particles was smaller in the presence of citrate complexing agent. In the absence of complexing agents all powder particles were dendritic and fern-like. Unfortunately no data about the alloy composition were presented [16]. In this study an attempt was made to investigate the process of the Fe–Ni powders electrodeposition from citrate-sulfate and citrate-ammonium chloride containing electrolytes in the presence of either Fe(III) or Fe(II) salts, as well as their morphology and composition. The Fe(III) salts were used to prevent inevitable oxidation of Fe(II) into Fe(III) on the anode during the powders electrodeposition. Our intention was to find out how much the anomalous character of Fe–Ni alloy powders deposition is pronounced in order to obtain powders with desired composition (for example NiFe2 powder which can be used for further oxidation into NiFe2 O4 magnetic material [17]).

2. Experimental The polarization diagrams were recorded in a threecompartment standard electrochemical cell at the temperature of 25 ± 1 ◦ C. The platinum foil counter electrode and the reference – saturated silver/silver chloride, Ag/AgCl – electrode (Eref = 0.20 V vs. NHE) were placed in separate compartments. The latter was connected to the working electrode by a Luggin capillary positioned at the distance of 0.2 cm from the working electrode surface. The working electrode was glassy carbon rod (d = 0.3 cm) sealed in epoxy resin so that only the surface area of the disc of 0.071 cm2 was exposed to the solution and was placed parallel to the counter electrode in a vertical position. Before each experiment working electrode surface was polished down to 0.05 ␮m alumina impregnated polishing cloths, cleaned in an ultrasonic bath for 10 min, thoroughly washed with distilled water and transferred to the electrochemical cell. The polarization measurements were performed by a computercontrolled potentiostat (PAR M273A) using the corrosion software (PAR M352/252, version 2.01) with the sweep rate of 1 mV s−1 . For obtaining polarization curves corrected for IR drop, the current interrupt technique, with the time of current interruption being 0.5 s, was used. All powders were electrodeposited at the room temperature in the cylindrical glass cell (total volume of 1 dm3 ) with cone shaped bottom of the cell in order to collect powder particles. Fe–Ni alloy powders were deposited under galvanostatic conditions on glassy carbon cylinder (d = 0.5 cm, h = 3 cm) at the appropriate limiting current density (see Section 3.1). Ni powder was washed with distilled water and alcohol after deposition. In the case of solutions with Fe(III) salts Fe in the alloy powders was protected from oxidation during the subsequent drying in the air at 100 ◦ C [18]. The powders of Fe–Ni alloys were washed with the solution containing

0.1% of sodium soap Sap G-30 (which contains 78% of total fatty acids) in distilled water [18]. In the case of solutions with Fe(II) salts, after washing powders with distilled water and alcohol, alloy powders were protected from oxidation by subsequent drying in the N2 atmosphere at 95 ◦ C. All solutions were made from analytical grade purity chemicals (NiSO4 , Na3 C6 H5 O7 , Na2 SO4 , Fe2 (SO4 )3 , NiCl2 , NH4 Cl, FeCl3 and FeCl2 ) and distilled water by the following procedure: Na3 C6 H5 O7 was first dissolved, then the pH was adjusted to slightly higher value than desired by corresponding acid; in the next step metal (Fe(III), or Fe(II) or Ni) salts were dissolved and finally supporting electrolyte was added and pH adjusted to the exact value. Concerning stability of solutions it is well known that during the investigation some Fe(II) become oxidized into Fe(III). This should be particularly pronounced during the deposition of powders for 1 h or 2 h. Taking into account that in all cases Fe(II) or Fe(III) made very stable complexes with citrate anions, we did not experience problems in the case of polarization measurements (polarization curves were practically the same after 3–4 measurements), but for any case before each experiment fresh solution has been made and used for investigation, as well as for powder electrodeposition. For better understanding of the electrochemical process some experiments were performed on a rotating disc electrode made of Au (0.312 cm2 ) using Tacussel Controvit rotating system and Gamry potentiostat Reference 600 with the software PHE 200. Before each experiment working electrode surface was polished down to 0.05 ␮m alumina impregnated polishing cloths, cleaned in an ultrasonic bath for 10 min, thoroughly washed with distilled water and transferred to the electrochemical cell. The morphology of the electrodeposited powders was examined using scanning electron microscope (SEM), Tescan VEGA TS 5130MM equipped with an energy-dispersive X-ray spectroscopy (EDS), INCAP enta FET-x3, Oxford Instruments. Accordingly, composition of powders was determined by the EDS analysis. The chemical analysis of the composition of some powder samples was performed by AAS using SPECTRO ICP-OES 17.5 MHz spectrometer. Samples for the analysis of about 30 ␮g were dissolved in 5 cm3 HCl (1:1) at slightly elevated temperature. Each experiment was repeated three times and the average values are presented in the paper. The variation of the results was ±5%.

3. Results and discussion Experiments were performed in two supporting electrolytes: 1 M Na2 SO4 and 1 M NH4 Cl. In sulfate supporting electrolytes NiSO4 , Fe2 (SO4 )3 and FeSO4 salts were used, while in ammonium chloride supporting electrolyte NiCl2 , FeCl3 and FeCl2 salts were used in order to keep the same anions in the solution. It appeared that the supporting electrolyte did not influence the shape of the polarization curves and powder morphology in the presence of either Fe(III) or Fe(II) salts. The only difference was higher current density for powders electrodeposition from the electrolyte containing 1 M NH4 Cl (see Fig. 1) and very small change in the position of polarization curves depending on the supporting electrolyte used. Four different electrolytes, with total metal ions concentration of 0.1 M, were used for the alloy powders electrodeposition in the presence of either Fe(III) or Fe(II) salts: x M Ni(II) + 1 M (Na2 SO4 or NH4 Cl) + 0.2 M Na3 C6 H5 O7 + y M Fe(III) or Fe(II) with x = 0.09 M, 0.075 M, 0.05 M and 0.025 M and y = 0.075 M, 0.05 M, 0.025 M and 0.01 M. In such a way the Ni/Fe ions concentration ratios were 9/1, 3/1, 1/1 and 1/3, while the pH of the solution was kept constant (pH 4.5 in the case of sulfate supporting electrolyte and pH 4.0 in the case of ammonium chloride supporting electrolyte) by adding H2 SO4 or HCl respectively. Polarization diagrams were also

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

537

recorded for pure Ni and pure Fe electrolytes. All powders for microstructure and composition analysis were electrodeposited at the limiting current density (position of the inflection point 䊉 on the polarization diagrams, see Fig. 1a, or line representing jpd in Fig. 1b, slightly below the point 䊉). At a given pH values in the case of Fe(III) salt it was not possible to deposit Fe powder, while in the case of Fe(II) salt it was possible to deposit Fe powder (see further discussion). Because of oxidation of Fe(II) into Fe(III) species on the anode during polarization measurements and powders electrodeposition, electrolytes made of Fe(II) salts were made fresh for each measurement. It should be mentioned here that an attempt was made to deposit Fe–Ni powders from the electrolytes of the same composition but of low pH 2. The powder particles were successfully produced on the cathode, but immediately after falling off from the cathode surface they started dissolving with gas evolution in all investigated electrolytes. Only in the case of very short time of electrolysis and removal of remained powder (which has not yet been dissolved) it was possible to obtain small amount of powder for further analysis. That was the reason why all experiments were performed in the solutions of pH 4 or pH 4.5, since in this solution powder was stable after deposition. 3.1. The polarization measurements

Fig. 1. The polarization diagrams for the Fe, Ni and Fe–Ni alloy powders electrodeposition recorded at a sweep rate of 1 mV s−1 (after IR drop compensation) for different Ni/Fe ions concentration ratios (marked in the figure). (a) Electrolytes containing Fe(III) salts and 1 M Na2 SO4 . (b) Electrolytes containing Fe(II) salts and 1 M NH4 Cl. The current densities used for powders electrodeposition for the SEM and EDS analysis are marked with (䊉) in (a) and with jpd in (b).

The polarization curves recorded in different electrolytes for Fe(III) salts are shown in Fig. 1a (for sulfate supporting electrolyte) and for Fe(II) salts in Fig. 1b (for ammonium chloride supporting electrolyte). The Ni/Fe ions concentration ratio is marked for each curve, while polarization curves for pure powders are marked with Ni and Fe. All polarization curves possess a similar shape. They are characterized by two inflection points, the first one reflecting massive alloy deposition (sharp increase of current density) with simultaneous hydrogen evolution and the second one (marked with 䊉 in Fig. 1) corresponding to the moment when the deposition process is controlled by the rate of hydrogen bubble formation, actually the potential at which the diffusion limiting current density of alloy powder deposition is reached (as explained in our previous papers [19,20] – see also Fig. 2). As can be seen in Fig. 1a (for the Fe(III) salts electrolytes) the potentials for Fe–Ni alloy powders deposition at all investigated Ni/Fe ratios are more positive than those for pure metal powders deposition. As the Ni/Fe ratio decreases (concentration of Fe

Fig. 2. The polarization curve for the electrodeposition of Fe–Ni alloy powders from the electrolyte containing 0.1 M FeCl2 + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl (solid line – jtot ) (), polarization curve for hydrogen evolution (dashed line – jH2 ) () and polarization curve for Fe–Ni powder electrodeposition after subtraction of hydrogen evolution current (dotted line – jall ) (). The corresponding Ni/Fe ions concentration ratios are marked in the figure.

538

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

increases) the polarization curves are placed at slightly more positive potentials. In the case of the electrodeposition from Fe(II) salts polarization diagrams for Fe–Ni alloy powders are placed between those for pure metals (as was the case for Ni–Co alloy powders electrodeposition [19]). Taking into account that the process of Fe–Ni alloy deposition belongs to the type of anomalous co-deposition (according to Brenner’s classification [21]) it is not surprising that the polarization curves for alloy deposition are not placed at the positions expected from the Ni/Fe ratio. In the case of Fe(III) salts they are even more positive than those for pure metals, indicating most likely more pronounced anomalous co-deposition in these electrolytes (see Section 3.3 – Fig. 10). The current efficiency for Fe–Ni powders electrodeposition has been determined by the procedure explained in great details in our previous papers [19,20]. It was found that for the Fe(III) salts electrolytes the current efficiency was very low, 1–2% (polarization curves for powder deposition (jtot ) and for hydrogen evolution (jH2 ) practically overlapped) and it was necessary to deposit powders at least for 2 h in order to obtain amount of powder that could be used for the morphology and composition analysis (SEM, EDS). In the case of Fe(II) salts electrolyte current efficiency at the potentials more negative than the second inflection point (䊉 in Fig. 1) varied between 8% and 15% depending on the Ni/Fe ratio, as shown in Fig. 2 (polarization curve for the Fe–Ni alloy powders electrodeposition (solid line – jtot ) (), polarization curve for hydrogen evolution (dashed line – jH2 ) () and polarization curve for Fe–Ni powder electrodeposition after subtraction of hydrogen evolution current (dotted line – jall ) () – average values for the diffusion limiting current densities for alloy powder deposition being jall = −0.26 A cm−2 for the ratio 1/3 and jall = −0.49 A cm−2 for the ratio 9/1). With the increase of iron concentration (as well as the amount of iron in the deposit – decrease of Ni/Fe ratio) the current efficiency for hydrogen evolution increased which is in accordance with the data obtained for compact Fe–Ni alloy deposits [10]. The polarization curves were recorded starting from the potential of −0.6 V. By comparing polarization curves for Fe powder deposition from Fe(III) (curve a) and Fe(II) (curve b) salts, shown in Fig. 3, it could be seen that certain cathodic current (∼−2.5 mA cm−2 ) has been detected already at the starting potential for curve a (inset of Fig. 3). This current remained constant down to the potential of about −1.0 V and it started rising at the same potential value at which the current recorded in the presence of Fe(II) salts started to rise. A low current efficiency for the Fe–Ni alloy powder deposition from the solution containing Fe(III) salts

Fig. 3. The polarization curves for the electrodeposition of Fe powders from the electrolyte containing 0.1 M FeCl3 + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl (a) and 0.1 M FeCl2 + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl (b).

Fig. 4. The cyclic voltammograms recorder at the sweep rate of 10 mV s−1 and RPM = 1000 onto Au electrode in the electrolytes containing 0.1 M Fe(III) (dotted line) and 0.1 M Fe(II) (solid line) salts. Electrolyte: 0.1 M FeCl3 (or FeCl2 ) + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl.

is the consequence of the first step in the overall reaction being reduction of Fe(III) species into Fe(II) species, which takes place at all potentials more negative than −0.2 V. 3.1.1. CVs of Fe deposition from Fe(III) and Fe(II) salts containing electrolytes In order to investigate this phenomenon more closely RDE electrode was used. The CVs at the sweep rate of 10 mV s−1 and RPM = 1000 were recorded in both solutions in the potential limit from −1.6 V to 0.6 V. The results are shown in Fig. 4 (from 0.2 V to 0.6 V there was no indication of any reaction). In the conditions of convective diffusion, deposition of Fe starts at about −1.3 V in both solutions, with the peaks of Fe dissolution being placed between −0.8 V and −0.5 V. In the presence of Fe(III) salts well defined cathodic wave, typical for the diffusion controlled process, has been detected in the potential range between −0.2 V to −1.2 V. Hence, it is obvious that before the commencement of Fe deposition from Fe(III) salts, reduction of Fe(III) to Fe(II) occurs. The Fe(III) to Fe(II) reduction process has been investigated at different rotation rates. As can be seen in Fig. 5a the current density plateau for this process increases with increasing rotation rate. By plotting the values of the current density plateaus recorded at the potential of −1.0 V (jL ) as a function of the ω1/2 well defined linear relation has been obtained, Levich equation [22], as shown in Fig. 5b. The diffusion coefficient of the species undergoing reduction during this process, calculated from the slope of this dependence, amounts to ∼3.7 × 10−5 cm2 s−1 . In comparison with the diffusion coefficient for Fe(CN)6 3− ions (0.6 × 10−5 cm2 s−1 ) [22] higher value has been obtained, indicating that this process is not a simple diffusion controlled one electron exchange from Fe(III) to Fe(II), since in both cases reduction of complexes (Fe(CN)6 3− and FeC6 H5 O7 ) is considered. Taking into account that the concentration of reacting Fe(III) species (actually FeC6 H5 O7 complex) is high (0.1 M), a mixed activation–diffusion control could be expected. Considering plots presented in Fig. 5c (j−1 vs. ω−1/2 recorded at different potentials marked in the figure) it is obvious that the reduction of Fe(III) species into Fe(II) species is not under pure diffusion control. As can be seen the intercept on the j−1 axis exists at potentials of −1.0 V, −0.9 V and −0.8 V, while at more positive potentials this dependence deviates from linearity, indicating the presence of activation controlled current density [22]. It should be emphasized here that in the case of convection (RPM > 0) no counter peak for the oxidation of these species could be detected at potentials more positive

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

539

Fig. 5. (a) Cathodic CVs for Fe(III) reduction (v = 10 mV s−1 ) recorded at different RPM’s: 1 – RPM = 1000; 2 – RPM = 2000; 3 – RPM = 3000 and 4 – RPM = 4000. (b) jL vs. ω1/2 recorder at the potential of −1.0 V. (c) j−1 vs. ω−1/2 recorder at different potentials (marked in the figure). Electrolyte: 0.1 M FeCl3 + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl.

than 0.0 V. In the case of stagnant electrolyte (RPM = 0) reduction and oxidation peaks of these species are clearly seen on the CVs, as shown in Fig. 6a. The presence of two cathodic peaks and one anodic peak confirms the statement that this process is not a simple diffusion controlled one electron exchange from Fe(III) to Fe(II), as well as not pure diffusion controlled process (activation control is also involved [22]). By plotting current density of oxidation peaks (jp(ox) ) vs. v1/2 [22] well defined linear dependence is also obtained (Fig. 6b), indicating diffusion controlled oxidation, but the value of the diffusion coefficient calculated from the slope of this dependence (assuming that the concentration of reduced species is 0.1 M) was for one order of magnitude lower than expected. Hence, it could be concluded that Fe(III) species reduce in the potential range between −0.2 V and −1.0 V and that under the convective diffusion reduced species were removed from the electrode surface into the bulk of solution and could not be oxidized during the reverse sweep. Their oxidation is possible only in stagnant

electrolyte, where they remain in the vicinity of the electrode surface (Fig. 6a). Considering the stability (formation) constants for all complexes that could be formed in the investigated solutions following data were obtained [23]: Fe3+ + HC6 H5 O7 −2 → FeHC6 H5 O7 + 3+

Fe

3+

Fe

3+

Fe

+ C6 H5 O7

−3

→ FeC6 H5 O7

−

+ OH → FeOH

2+

K 1 = 1.0 × 10

K 1 = 7.4 × 10

−

+ 2OH → Fe(OH)2

Fe3+ + 3OH− → Fe(OH)3

+

Fe

+ OH → FeOH

+

(1) (2) (3)

21

K 3 = 4.7 × 1029

Fe2+ + C6 H5 O7 −3 → FeC6 H5 O7 − −

25

11

K 2 = 1.5 × 10

Fe2+ + HC6 H5 O7 −2 → FeHC6 H5 O7

2+

K 1 = 3.17 × 1012

(4) (5)

K 1 = 1.2 × 103

(6)

K 1 = 3.16 × 1015

(7)

K 1 = 3.6 × 10

5

(8)

Fig. 6. (a) The cyclic voltammograms of Fe(III) species reduction and Fe(II) species oxidation recorded at different sweep rates (marked in the figure in mV s−1 ) in the stagnant electrolyte (RPM = 0) onto Au electrode in the electrolyte containing 0.1 M FeCl3 + 0.2 M Na3 C6 H5 O7 + 1 M NH4 Cl. (b) jp(ox) vs. v1/2 dependence obtained from the anodic current density peaks of oxidation shown in (a).

540

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

Fe2+ + 2OH− → Fe(OH)2

K 2 = 5.9 × 109

Fe2+ + 3OH− → Fe(OH)3 −

(9)

K 3 = 4.7 × 109

Fe2+ + 4OH− → Fe(OH)4 −2

(10)

K 4 = 3.8 × 108

(11)

As can be seen 11 different species could be formed, 5 with Fe3+ ions and 6 with Fe2+ ions. Taking into account their stability constants and concentration of complexing agents it appears that for Fe(III) salts the most stable ones are FeC6 H5 O7 and Fe(OH)3 , while for Fe(II) salts the most stable ones are FeC6 H5 O7 − and Fe(OH)2 . The possibility of Fe(OH)3 formation could be excluded since at pH 4.5 concentration of OH− ions is 3.16 × 10−10 M and the dominant species in the Fe(III) salt electrolyte is complex FeC6 H5 O7 . Hence, let us reconsider the process of Fe powder deposition from the electrolyte containing Fe(III) salt. The reduction of FeC6 H5 O7 starts at about −0.2 V producing Fe(II) species. Taking into account the presence of two cathodic peaks during this reduction process it could not be specified which species were formed, pure Fe2+ ions or complex FeC6 H5 O7 − (the most stable complex with the Fe2+ ions), or both. It is most likely that both species were formed and that Fe deposition occurs by further reduction of these species with simultaneous hydrogen evolution which must be pronounced at potentials more negative than −1.3 V. Since during the Fe deposition pH in the vicinity of the electrode surface is much higher due to massive hydrogen evolution, it is possible that Fe2+ ions react with OH− ions forming Fe(OH)2 which precipitates and prevent deposition of pure Fe powder. On the other side, in the solution containing Fe(II) salt the possibility of Fe(OH)2 formation could also be excluded since at pH 4.0 concentration of OH− ions is 1.0 × 10−10 M and the dominant species is complex FeC6 H5 O7 − . Accordingly, reduction of FeC6 H5 O7 − starts at about −1.3 V producing Fe deposit directly without possibility of eventual Fe(OH)2 formation and in this solution Fe powder has been successfully deposited. Similar experiment, but in the presence of Fe(II)/Fe(III) EDTA solutions of various pH values (3.3, 3.7 and 5.9) in 0.5 M Na2 SO4 , has been performed by Juzeliunas and Jüttner [24]. They also detected redox reaction of Fe(III) EDTA ↔ Fe(II) EDTA complex, while Fe(II) reduction and Fe deposition started at about −1.0 V vs. Ag/AgCl with the peak of Fe dissolution appearing between −1.0 V and −0.7 V. 3.1.2. Complexes that could be formed in the Ni(II) + Fe(II) salts containing electrolytes In the case of the solutions for Fe–Ni alloy powder electrodeposition among the species presented above for Fe, following species could be formed with Ni [23] in the investigated solutions: Ni2+ + HC6 H5 O7 −2 → NiHC6 H5 O7 2+

Ni

2+

Ni

2+

Ni

2+

Ni

+ C6 H5 O7 −

−3

→ NiC6 H5 O7

+ OH → NiOH

+

K 1 = 2.0 × 10

K 1 = 9.3 × 10

−

+ 2OH → Ni(OH)2 −

−

K 1 = 1.3 × 105

+ 3OH → Ni(OH)3

4

K 2 = 3.5 × 10 −

14

(12) (13) (14)

8

K 3 = 2.0 × 10

Fig. 7. SEMs of typical powder particles obtained for electrodeposited Ni powder (a) and Fe powder (b). Ni powder was electrodeposited at the current density of the inflection point (䊉) in both supporting electrolytes. Fe powder was electrodeposited at the current density jpd in the NH4 Cl supporting electrolyte and at the current density of the inflection point (䊉) in Na2 SO4 supporting electrolyte.

Fe(III) and sulfate complexes could be formed with Fe(III) and Ni, but they all have for several orders lower stability constants in comparison with citrates and cannot influence the reaction mechanism.

(15) 11

(16)

As can be seen, again the most stable one is NiC6 H5 O7 − and the same conclusion concerning formation of Ni-hydroxide species in the investigated solutions, as the one given above for Fe could be applied. Hence, Ni deposition occurs by reduction of citrate complex (NiC6 H5 O7 − ) without possibility for Ni-hydroxide species formation. It should be mentioned here that because of the presence of NiC6 H5 O7 − species in the solution, eventual influence of Fe(OH)2 precipitation for the electrolytes containing Fe(III) salts, should be less pronounced, as it is shown in this work. Among above mentioned complexes of Fe and Ni ammonium complexes could be formed with Fe(II) and Ni ions in ammonium chloride solution, chloride complex could be formed with Fe(II) and

3.2. The morphology of the Fe, Ni and Fe–Ni powders The morphology of powder particles electrodeposited onto glassy carbon electrode from different electrolytes is shown in Figs. 7–9. Typical pure Ni powder for both supporting electrolytes, Fig. 7a, is characterized by the presence of flakes of the maximum size of about 50 ␮m covered with nodules of mainly flat surfaces. On some particles formation of a second zone of dendrites (beginning with the formation of small crystals), typical for powder deposition [25] could be detected. Typical particle of Fe powder for both supporting electrolytes is shown in Fig. 7b. The Fe powder particles of the size of about 200 ␮m also contain nodules of flat surfaces, but they are characterized by the presence of cone shaped cavities. The presence of cavities on the Fe powder particles is in accordance

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

541

Fig. 8. (a) SEM of typical cone shaped cavities detected in the Fe–Ni alloy powders electrodeposited from the electrolytes containing both, Fe(III) and Fe(II) salts independently of the Ni/Fe ratio in both supporting electrolytes. (b) SEM of typical cone shaped cavities detected in the Fe–Ni alloy powders electrodeposited from the electrolytes containing Fe(II) salts independently of the Ni/Fe ratio in both supporting electrolytes.

Fig. 9. (a) SEM of typical nodules with flat surfaces detected in the Fe–Ni alloy powders electrodeposited from the electrolytes containing both, Fe(III) and Fe(II) salts independently of the Ni/Fe ratio in both supporting electrolytes. (b) Pagoda like crystals, corresponding to the FeNi3 single crystals, detected in powders with the content of Fe between 60 mol.% and 70 mol.% electrodeposited from the electrolytes containing both, Fe(III) and Fe(II) salts in both supporting electrolytes.

with the fact that more intensive hydrogen evolution occurs during this process [10]. A common characteristic for all Fe–Ni alloy powders, deposited from the electrolytes containing either Fe(III) or Fe(II) salts independently of the Ni/Fe ratio and the supporting electrolyte, is the presence of cone shaped cavities (Fig. 8a) corresponding to the places were the hydrogen bubbles were formed [19,20,25]. Another type of deep cavities, presented in Fig. 8b, has been detected only for the Fe–Ni powders deposited from the electrolyte containing Fe(II) salt. Fig. 8 indicates two types of hydrogen bubbles formation, one present in the electrolytes containing both, Fe(III) and Fe(II) salts (cone shaped ones corresponding to the formation of bigger hydrogen bubbles) and another one characterized with deep cavities (small hydrogen bubbles) present only in the electrolytes containing Fe(II) salts. Another common characteristic for all Fe–Ni alloy powders, deposited from the electrolytes containing either Fe(III) or Fe(II) salts independently of the Ni/Fe ratio and the supporting

electrolyte, is the presence of nodules with flat surfaces, presented in Fig. 9a. Although from the electrolytes containing Fe(III) salts crystals of different shapes were detected on the nodule surfaces, particularly interesting are crystals of the shape of pagoda present in the powder with Ni between 60 mol.% and 70 mol.%, detected in both electrolytes and presented in Fig. 9b. Such crystals were detected in the Fe–Ni powder synthesized in high yield by a simple and facile hydrothermal method without the presence of surfactants [26]. According to this investigation [26] FeNi3 crystals were formed during the described procedure. The products obtained at 120 ◦ C are mixture of FeNi3 and Fe–Ni hydroxides composed of monodispersed microspheres (average diameter of 1.5–2.0 ␮m) identical to those presented in Fig. 9a. With the increase of temperature to 140 ◦ C these microspheres become micropagodas very similar to those shown in Fig. 9b. At higher temperatures (180 ◦ C) these particles transform into perfect 3D FeNi3 dendritic superstructures in certain directions [26]. Hence, comparing the crystals presented in Fig. 9a and b with those obtained by hydrothermal

542

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

method [26] it seems reasonable to ascribe them to the FeNi3 single crystals. It should be noted here that at a given current density of deposition further transformation into dendrites (as in the case of hydrothermal method [26]) was most likely not possible since the powder particles fall off from the electrode surface before the formation of dendrites. Concerning formation conditions, following our previous experience with powders of Ni–Co and Ni–Mo alloys [19,20,25], the formation conditions in the region of potentials more negative than the second inflection point (䊉) do not influence the morphology of powders. Some changes could be observed only in the region between the first and second inflection point, but in that potential region significant amount of the deposit is compact and does not fall off from the electrode surface and accordingly does not represent the powder. 3.3. The EDS analysis of the Fe, Ni and Fe–Ni powders All powder samples were analyzed by EDS in such a way that at least eight powder particles (maximum number 20) on the SEM micrographs are chosen and EDS analysis was performed at 3–10 different positions on each particle. In some cases the analysis was performed at a point of 1 ␮m2 , while in some cases rectangle surface from 120 ␮m2 (10 ␮m × 12 ␮m) to 9600 ␮m2 (80 ␮m × 120 ␮m) has been analyzed. In some cases the composition of the powder particles was uniform all over the particles surface, while in some cases at.% of oxygen was found to vary significantly on the surface of one particle, or on specific part of the analyzed particle. This behavior was independent of the shape of the investigated particle (flat surface, cone shaped cavities, different crystals on the particle surface), as well as of the composition of the powders. Taking into account that the EDS analysis strongly depends on the position and angle of the beam, for powder particles like these presented in this work the EDS analysis results should not be considered as quantitative ones. Hence, in Table 1 are presented average values (obtained by the analysis of significant number of particles for each powder) for the EDS analysis of all powder samples. From the presented results it is most likely that the presence of oxygen could be the consequence of washing and drying procedure. It could also be possible that the oxygen is a result of eventual Fe(OH)2 formation during the powder deposition, as explained in Section 3.1.1 for pure Fe powder deposition. In such

Table 1 The average composition of the Fe–Ni powders electrodeposited from Fe(III) and Fe(II) salts containing electrolytes, obtained by the EDS analysis (in at.%). Ni/Fe

Fe(III) solution O

Ni 9/1 3/1 1/1 1/3 Fe

Fe

Fe(II) solution Ni

O

Fe

Ni solution Ni

O 20

22 34 33 12

34 41 55 83

44 25 12 5

5 8 12 30 20

21 35 58 62 80

Ni 80

74 57 30 8

a case it would be reasonable to expect higher percentage of oxygen in the powders containing higher amount of Fe for Fe–Ni alloy powders deposited from the electrolyte containing Fe(III) salts, while for powders deposited from the electrolyte containing Fe(II) salts percentage of oxygen should be relatively constant for all investigated compositions. As can be seen in Table 1 this is not the case. Hence, it appears that during the deposition of Fe–Ni alloy powders the behavior of pure metals cannot be considered as relevant for the properties of alloy powders, since they form solid solution type alloys which behave in a different way than pure metals. At the same time the presence of oxygen could be the consequence of washing and drying procedure, as mentioned above. In order to determine the type of Fe–Ni alloy powder deposition, the percentages of Fe and Ni, calculated after subtraction of the percentage of oxygen, are used and results are presented in Fig. 10. According to Brenner’s classification [21] anomalous co-deposition of Fe and Ni occurs at all investigated solution compositions, with the less noble metal (Fe) being more readily deposited than the more noble one (Ni), i.e. the percentage of Ni in the powder is much lower than its percentage in the solution (Fig. 10b). According to Fig. 10a linear dependence of the percentage of both metals in the powder as a function of the logarithm of their concentration ratios indicates that their content in the powder depends exponentially on the Ni/Fe ratio. Considering Fig. 10b it could be concluded that the anomalous character is more pronounced in the electrolytes containing Fe(III) salts. This is in accordance with the statement that the formation of some amount of Fe(OH)2 is possible in this electrolyte and assuming “hydroxide suppression” mechanism [27] the anomalous character of Fe and Ni deposition should be more pronounced in this electrolyte. It should be noted here that the

Fig. 10. (a) Dependences of mol.% of Fe in the powder deposited from the electrolyte containing Fe(III) salts (䊉) and Fe(II) salts () and dependences of mol.% of Ni in the powder deposited from the electrolyte containing Fe(III) salts () and Fe(II) salts () as a function of log([Ni]/[Fe]) in solution. (b) Typical Brenner’s presentation of anomalous co-deposition for the powders electrodeposited from the electrolytes containing Fe(III) salts () and Fe(II) salts ().

U. Laˇcnjevac et al. / Electrochimica Acta 55 (2009) 535–543

chemical composition of some powder samples were analyzed by the AAS technique (see Section 2) and that results of this analysis and the EDS analysis (after subtraction of oxygen percentage) were in good agreement.

[2] [3] [4] [5] [6]

4. Conclusions The morphology and composition of electrodeposited Fe–Ni alloy powders depend on the Ni/Fe ions concentration ratio in both electrolytes. Anomalous co-deposition of Fe and Ni has been confirmed by the EDS analysis of alloy powders, being more pronounced in the electrolyte containing Fe(III) salts due to possibility of Fe(OH)2 formation. A common characteristic for all alloy powder samples was the presence of cone shaped cavities and nodules. The possibility of the formation of single crystal FeNi3 phase was found to exist in both electrolytes. It appears that it is better to use electrolyte with Fe(II) salts, independently of the supporting electrolyte, since the current efficiency for Fe–Ni powder electrodeposition is much higher than that from the solution containing Fe(III) salts. Acknowledgement This work was financially supported by the Ministry of Science and Technological Development of the Republic of Serbia through the Project No. 142032G/2006. References [1] P.C. Andricacos, L.T. Romankiw, in: H. Gerischer, C.W. Tobias (Eds.), Advances in Electrochemical Science and Engineering, vol. 3, John Wiley & Sons Inc., New York, 1994, p. 227.

[7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]

543

X.Y. Qin, J.G. Kim, J.S. Lee, Nanostruct. Mater. 11 (1999) 259. J. Ebothe, S. Vilain, J. Phys. D Appl. Phys. 32 (1999) 2342. W.C. Grande, J.B. Talbot, J. Electrochem. Soc. 140 (1993) 669. F. Lallemand, L. Ricq, M. Wery, P. Bercot, J. Pagetti, Surf. Coat. Technol. 179 (2004) 314. H. Li, F. Ebrahimi, Mater. Sci. Eng. A Struct. Mater.: Prop. Microstruct. Process. 347 (2003) 93. K. Ohashi, M. Ito, M. Watanabe, Electrochem. Soc. Proc. 88 (1988) 525. L. Ricq, F. Lallemand, M.P. Girander, J. Pagetti, Surf. Coat. Technol. 138 (2001) 278. N.V. Myung, D.-Y. Park, D.E. Urgiles, T. George, Electrochim. Acta 49 (2004) 4379. V.C. Kieling, Surf. Coat. Technol. 96 (1997) 135. K.-M. Yin, B.-T. Lin, Surf. Coat. Technol. 78 (1996) 205. S.-H. Kim, K.-J. Sohn, Y.-C. Joo, Y.-W. Kim, T.-H. Yim, H.-Y. Lee, T. Kang, Surf. Coat. Technol. 199 (2005) 43. F.R. Bento, L.H. Mascaro, Surf. Coat. Technol. 201 (2006) 1752. E.P. Zhelibo, N.N. Kravets, M.Yu. Gamarkin, S.V. Remez, Powder Metall. Metal Ceram. 34 (1995) 113. E.P. Zhelibo, N.N. Kravets, Powder Metall. Metal Ceram. 36 (1997) 264. C.-M. Chu, J. Chin. Inst. Eng. 34 (2003) 689. S.D. Sartale, C.D. Lokhande, M. Giersig, V. Ganesan, J. Phys.: Condens. Matter. 16 (2004) 773. ´ Lj.J. Pavlovic, ´ B.M. Jovic, ´ M.G. Pavlovic, ´ J. Serb. Chem. Soc. 73 V.M. Maksimovic, (2008) 861. ´ B.M. Jovic, ´ M.G. Pavlovic, ´ Electrochim. Acta 51 (2006) 5468. V.D. Jovic, ´ B.M. Jovic, ´ V. Maksimovic, ´ M.G. Pavlovic, ´ Electrochim. Acta 52 (2007) V.D. Jovic, 4254. A. Brenner, Electrodeposition of Alloys. Principles and Practice, vol. II, Academic Press Inc., New York and London, 1963, p. 413. A.J. Bard, L.R. Faulkner, Electrochemical Methods Fundamentals and Applications, John Wiley & Sons, Inc., USA, 2001. J.A. Daen, Lange’s Handbook of Chemistry, 15th ed., McGraw-Hill Inc., New York, 1999. E. Juzeliunas, K. Jüttner, Electrochim. Acta 43 (1998) 1691. ´ V. Maksimovic, ´ M.G. Pavlovic, ´ K.I. Popov, J. Solid State Electrochem. V.D. Jovic, 10 (2006) 373. X.-M. Zhou, X.-W. Wei, Cryst. Growth Des. 9 (2009) 7. T. Horkans, J. Electrochem. Soc. 126 (1979) 186; T. Horkans, J. Electrochem. Soc. 128 (1981) 45.