Characterisation of electroplated Ni 45 Fe 55 thin films on n-Si (111)

Abstract

Electrodeposition of NiFe films, on hydrogen-terminated n-Si (111)-H from acidic dilute sulphate solution, was studied by electrochemical measurements at room temperature in the presence and absence of saccharin. The electroplating process kinetics was investigated by voltammetric study and the effect of Fe²⁺ concentration on the deposit composition was studied as well with energy dispersive spectrometry analysis. The average composition of the Ni₄₅Fe₅₅ film was obtained for Fe²⁺ concentration in the range of [0.030–0.035] mol L⁻¹ at a current density of −6 mA cm⁻². Correlation between Fe²⁺ concentration in the NiFe deposit and electronic properties was examined by electrochemical impedance spectroscopy. Film roughness depends on Fe²⁺ concentration and a smoother deposit was obtained for the Ni₄₅Fe₅₅ film. Very low coercivity (less than 1.3 Oe) was measured in the Ni₄₅Fe₅₅ film with a nominal thickness of 640 nm. The very soft magnetic properties of the NiFe films provide information about the low level of inhomogeneities present in these films.

Keywords

Electrodeposition EIS corrosion resistance crystal structure coercivity

Introduction

The need for recording heads to write on highly coercive media at high frequencies created additional requirements for new soft magnetic materials with higher magnetic moments. A magnetic recording head should preferably have a low coercivity, high magnetic moment, large electrical resistance, no internal stresses and high corrosion resistance. As a consequence, a combination of suitable magnetic properties with excellent corrosion behaviour would improve the reliability of the devices. Corrosion resistance is considered as an indication of the durability of the film material [1,2]. Due to this combination of properties, NiFe film seems to be an appropriate candidate to meet these technological demands. Since it is possible to change its magnetic properties by controlling its form and composition, different studies were carried out on the Fe _x Ni_1−x alloy, including thin films[3 -7] and nanowires or nanotubes [8,9],

The Permalloy film consisting of 80%Ni and 20% Fe was the nickel–iron alloy traditionally used in the magnetic recording manufacture of those devices [10,11]. Recently significant progress was made to elaborate soft NiFe alloys with higher saturation magnetic moment by increasing Fe content in the deposit. In this context, Ni₄₅Fe₅₅ was introduced as a new material in the fabrication of write-heads by IBM in 1997 [12,13]. Another advantage of this almost equi-atomic FeNi alloy is that in particular conditions of deposition and annealing, a phase transformation can occur. This gives rise to the formation of a tetragonal L₁₀ structure which exhibits remarkable magnetic properties identical to those observed in rare earth-based compounds [14,15]

Direct metal deposition on silicon surfaces was intensively investigated because of the technological importance of the Schottky contacts in the microelectronic industry. The integration of magnetic materials and nanostructures with semiconductor technology is an interesting approach to develop various nanoscale supports such as patterned recording media and highly integrated sensors. Such thin films on silicon are usually deposited through physical methods and lithography techniques to make patterned films [16,17]. Electrodeposition has the attractive features of cost-effectiveness, simplicity in operation and ability to make nanostructures by direct deposition on the nanoporous substrate. Interest on hydrogen-terminated silicon surfaces has grown over the years [18]. Permalloy (Ni₈₀Fe₂₀) electrodeposited directly onto silicon has attracted a special attention [19 -22]. However, to the best of our knowledge, there have so far been only some papers that have been devoted to the electrodeposition of the Ni₄₅Fe₅₅ alloys on Si [23,24]. This is probably owing to the difficulty to obtain uniform nucleation directly on silicon and also to maintain a uniform composition of the deposit in a highly concentrated iron bath because of the known anomalous co-deposition phenomena [10,25].

Organic additives are often used in electroplating operations to moderate deposit growth rates and control film quality. Saccharin (SAC) is the most commonly used one. It has been reported that SAC tends to inhibit the solid-state reaction between Fe and Si and enhance the protection against oxidation under air exposition compared to deposition without the additive. SAC is also essential to obtain compact and smooth deposits with a shiny appearance and good adhesion on the silicon substrate [20,21]. Moreover, SAC was considered as a grain refiner agent [26]. It has been also reported that in order to obtain magnetic alloys with low coercivity, SAC must be added with optimum concentration [27].

Despite the variety of approaches being examined for the NiFe electrodeposition on n-Si(111) surfaces, so far there has not been much work investigating the electronic properties of NiFe films by EIS. Correlation between electronic properties and Fe content in the NiFe deposit is crucial for the successful design of electrodeposition process and its integration in magnetic devices.

The purpose of this work was to perform an investigation on the morphological, structural, electronic and magnetic properties of NiFe film electrodeposited onto n-Si(111)-H substrate, and to show how it is possible to electroplate, from a dilute electrolyte, a very soft Ni₄₅Fe₅₅ magnetic alloy, with a lower Hc value than the ones obtained on the permalloy. It was also important to characterise the corrosion properties of the NiFe alloy in the service environment since normal atmospheric corrosion is unavoidable in many applications where soft magnetic materials are used. Changes in magnetic properties including coercivity and magnetisation according to Fe content were also analysed.

Experimental

Electrodeposition was performed on monocrystalline n-Si (111) with a resistivity of 3–5 Ω cm. The silicon wafers were cut into (1 × 1 cm²) squares that were first degreased in boiling acetone for 10 min and sequentially cleaned ultrasonically for 10 min in ethanol and water. The electrodes were then treated for 30 min with a 3:1 H₂SO₄:H₂O₂ mixture, heated at 80°C in order to remove any trace of impurities. The oxide film was then removed by etching with 1% HF solution for 1 min and thoroughly rinsed with deionised water. Ohmic contacts were formed by applying In/Ga eutectic on the back side of the wafers. The electrodes were mounted on a Teflon holder and the silicon area exposed to the solution was 0.26 cm². Before the electrochemical experiments, the electrode surface was again etched for 1 min in 10% HF solution and rinsed in deionised water. Through this procedure, hydrogen-terminated Si surfaces with wide terraces were prepared [19 -21]. The electrodes were then immediately transferred into the electrolysis bath.

The electroplating bath composition consisted of 0.01–0.06 mol L⁻¹ FeSO₄ 7H₂O, 0.06 mol L⁻¹ NiSO₄ 6H₂O, 0.35 mol L⁻¹ Na₂SO₄ and 0.4 mol L⁻¹ H₃BO₃. The SAC content was 2 g L⁻¹. The electroplating bath was prepared using deionised water having a resistivity of 18 MΩ cm and the pH was adjusted to 3. All experiments were carried out at 298 K in a three-electrode cell using platinum wire as a counter electrode and saturated calomel electrode as a reference electrode. The solution was carefully de-aerated for 30 min with argon bubbling before any electrochemical experiment and the cell was placed inside a Faraday cage to avoid electrical interferences and light incidence.

For electronic and corrosion properties investigation, experiments were carried out at ambient temperature in borate buffer solution (0.075 M Na₂B₄O₇, 10 H₂O + 0.05 M H₃BO₃). Cyclic voltammetry and deposition experiments were achieved with a potentiostat/galvanostat (EG&G PAR 273A). The voltammetric experiments were performed at a scan rate of 20 mV s⁻¹ on the silicon surface. Only the first cycle was considered in each experiment. In order to reduce the effect of any resistance variation on the back contact of the substrate, the galvanostatic deposition was preferred to the potentiostatic mode. For EIS measurements, an Autolab PGSTAT-30 driven by FRA 4.9 Software (Eco Chemie, the Netherlands) was used. An ac amplitude voltage of 10 mV and an applied frequency ranging from 1 kHz to 0.01 Hz have been used. The Equivalent Circuit Software (Equivcrt) is used for the data modelling and curve fitting. This programme is based on non-linear least squares fitting. Note that all potentials are quoted with respect to SCE.

The surface morphology and the average film composition were examined using scanning electron microscopy (SEM), JEOL JSM-6700, combined with energy dispersive spectrometry (EDS). The thicknesses of the different deposits were determined from the cross-section SEM images. The structural characterisation was carried out by X-ray diffraction with a Cu Kα radiation (λ = 1.5418Å). Magnetic characterisation was performed at room temperature using a vibrating sample magnetometer.

Results and discussion

Electrodeposition of NiFe film from sulphate solution on n-Si(111)-H

Voltammetric measurements

To highlight the contribution of SAC additive on the hydrogen evolution reaction on n-Si (111)-H electrode, voltammetric studies were performed in the supporting electrolyte with and without SAC. Figure 1 shows the corresponding voltammograms in the absence (curve (a)) and in the presence (curve (b)) of SAC. The two voltammograms start from −0.15 V towards the negative scan. As can be seen, a weak cathodic current begins at around −0.35 V, which is related to the H⁺ ions reduction. This current increases at −1.12 V and reaches its limiting value at around −1.37 V. Up to −1.4 V, the current rapidly increases owing to the H₂O reduction. During the positive scan, a small anodic current owing to the silicon oxidation is observed up to 0 V. It is also important to mention that in the presence of SAC, hydrogen potential reduction is shifted towards more cathodic values and the limiting current density is well decreased. The effect of SAC on the electronic properties of H-Si(111)-n is the subject of another paper [28].

The voltammetric responses, recorded from the open circuit potential (OCP) in the complete solution containing SAC and different Fe²⁺ concentrations (0.015, 0.03, and 0.045 mol L⁻¹), are displayed in Figure 2. One can notice that the current density did not increase rapidly between −0.45 and −1.1 V. This potential range corresponds to the predominance of the hydrogen evolution reaction as mentioned previously. Consequently, an increase in the interfacial pH can generate hydrolysis reactions with formation of iron and/or nickel hydroxide species.

The co-deposition of NiFe starts up to −1.1 V, this value is quite similar to the one obtained by Gao et al. [19] who observed a large overpotential compared to its reduction potential on a metallic substrate. The change of the slope observed at a potential lower than −1.3 V is probably owing to the increase of the metallic hydroxide formation because of hydrogen evolution and H₂O reduction on the deposited NiFe crystallites.

The effect of the Fe²⁺ ions concentration is well evidenced in the cathodic part of the voltammograms. Indeed, it is clearly seen that the reduction potential shifts towards positive direction as the Fe²⁺ ions concentration increases in the bath, suggesting that it enhances the electrodeposition of NiFe alloy. On the other hand, as reported in the literature [25,29,30], the anomalous phenomenon, in which catalytic iron hydroxide occurs during the reduction process and competes with the nickel hydroxide, is enhanced with increasing Fe²⁺ content. Indeed, the reaction mechanisms proposed in the literature to describe the deposition process of the FeNi alloy are given below: (1) (2) (3)

According to the reactions above, the deposition of the FeNi alloy is preceded by the formation of adsorbed metallic hydroxides. Due to the huge difference between the dissociation constant of the iron and nickel ion hydroxides, the concentration of the FeOH⁺ will be much higher than that of Ni(OH)⁺. Therefore, the adsorption sites for the NiOH_ad should be occupied by the FeOH_ad. Consequently, the deposition of Ni atoms should be partially inhibited.

On the reverse scan of the voltammogram, the reduction of the metallic species occurs on a continuous metallic film showing a crossover potential, which is typical of nucleation and growth processes. In the anodic part of the curves, a single dissolution peak of the NiFe deposit is observed close to −0.6 V, its intensity increases with increasing Fe²⁺ ions concentration of the bath. The silicon oxidation takes place at 0 V.

An estimation of the current efficiency may be obtained from the voltammograms by the calculation of the anodic to cathodic charge ratio (Q_a/Q_c). The obtained values increase from 9 to 32% when increasing the iron ion concentration in the electrolytic bath. This low efficiency may be attributed to the dilute electrolyte used in the present work. According to these results, a more compact film is expected to be obtained at a high iron concentration (higher efficiency).

Characterisation of the deposit

Composition, morphology and crystal structure of the electrodeposited NiFe films

On the basis of this preliminary study, different NiFe alloys with different Fe contents have been synthesised under galvanostatic mode, at a constant current density of −6 mA cm⁻² (the corresponding electrode potential is in the range [−1.1, −1.3 V] depending on the Fe²⁺ concentration in the bath).

The plot of the Fe content (at.-%) in the NiFe alloy as a function of the ion concentration ratio [Fe²⁺]/([Fe²⁺] + [Ni²⁺]) in the solution is shown in Figure 3. The measurements were taken at different positions on the surface samples. As can be seen, there is a slight variation in the film composition (as determined by EDS) especially at the edges of the sample.

The relationship between Fe content in the deposit and the ion concentration ratio [Fe²⁺]/([Fe²⁺] + [Ni²⁺]) is not linear, as indicated by the composition Reference Line (CRL) (shown in the same figure), where the composition in the bath equals to that in the deposit. The fraction of the Fe content in the FeNi alloy is always greater than the ion ratio [Fe²⁺]/([Fe²⁺] + [Ni²⁺]) in the solution. Such behaviour has already been reported in the literature [3] and was attributed to the well-known anomalous co-deposition of the NiFe alloy [10], where the less noble metal deposits preferentially to the nobler one, making the control of the film composition difficult.

As illustrated in the figure, the Fe content in the NiFe deposits increases from about 18 at.-% to about 64 at.-% as the molar ratio of Fe²⁺/Fe²⁺ + Ni²⁺ in the electrolytes varies from 0.14 to 0.3. The Ni₄₅Fe₅₅ alloy can be obtained when the molar ratio is around 0.33, corresponding to the Fe²⁺ ion concentration of 0.030 M in the bath.

One may note here that this is just an average composition obtained in a bulk material, indeed the anomalous co-deposition observed in this iron group alloys and the diffusion control of the Fe deposition may often lead to graded composition materials as already observed by several authors [4], especially in the first stage of deposition (nearly 20 nm thickness) where the alloy is generally iron-rich one. To minimise the effect of this composition variation at the interface, the electrochemical and the magnetic analysis performed in the present work concerned NiFe deposits with a nominal thickness around 600–760 nm.

The morphology of the deposit has been examined as a function of the iron concentration. Figure 4 illustrates SEM micrographs taken on different films obtained by electrodeposition during 300 s at different iron concentrations. One can see that at a low iron concentration ([Fe²⁺] = 0.015 M), the deposit exhibits a lot of holes along the surface. As pointed before, a very weak current efficiency was estimated at low iron concentration, indeed owing to hydrogen evolution, a lot of active sites on the electrode surface are inhibited resulting in the presence of holes in the deposit. When increasing iron concentration ([Fe²⁺] = 0.03 M), corresponding to a nominal thickness of 640 nm and to the Ni₄₅Fe₅₅ composition, the deposit becomes smooth, more uniform and compact. However, for higher iron concentration ([Fe²⁺] = 0.045 M), the deposit is still compact but becomes rough.

The influence of SAC additive on the morphology of the growth is also well highlighted. The presence of SAC in the bath leads to a very smooth film (Figure 4(d)). A rough film is obtained in the absence of SAC.

Figure 5(a) shows XRD patterns of pure Ni film, pure Fe film and NiFe deposits with different Fe content. As we can see, two different behaviours are observed according to the iron content in the deposit. Between 0 at.-% Fe and 55 at.-% Fe, the fcc structure predominates, with the presence of the (111) and the (200) fcc peaks. With increasing Fe content in the deposit, the two fcc peaks shifts towards the small angles indicating an increase in the lattice parameters. It is also interesting to notice that the intensity of the (200) peak decreases indicating a decrease in the fcc phase. For the 55 at.-% Fe content, the (111) fcc peak presents a dissymmetric form on the high-angle side, indicating the probable presence of a slight shoulder owing to the contribution of the bcc (110) peak. Up to 65 at.-% Fe, the (200) fcc peak has completely disappeared and the shoulder owing to the (110) bcc peak is more pronounced in the XRD patterns. Increasing the Fe contents in the deposit to 73 at.-% leads to an increase in the bcc (110) contribution and a decrease in the fcc (111) one, indicating the preponderance of the bcc phase at this range of composition.

Using the Highscore Plus XRD analysis software and by deconvoluting the most intense X-ray diffraction peaks for the different NiFe deposits, we have calculated the lattice constant (a) for each phase (fcc or bcc), the grain size (D) using Scherrer formula and the mean strain ε (%), defined as: (4) The fcc lattice constant value increases from 0.352 to 0.358 nm when increasing the iron content in the deposit, whereas the bcc lattice parameter is almost constant (∼0.285 Å). Otherwise, the strain ε (%) increases from 0.7 to 1.25%. The positive value of indicates that FeNi/Si films are under tensile stresses.

The estimated crystallites grain sizes are in the nanometer scale and they decrease from 15 to 8 nm with the increase of the iron concentration in the deposit.

The dependence of the grain size on the deposit composition of NiFe alloys has already been studied by other authors [5]. It was found that the smallest grain size was obtained in the mixed phase (bcc/fcc) alloy, where the composition of the deposit is around 55% Fe. This is in good agreement with our study.

EIS characterisation

Several factors have a bearing on the soft magnetic alloy choice. One external factor is the protective effectiveness properties. For this reason, the investigations in the direction of knowledge of electrical and protective properties of the NiFe films have been evaluated by means of EIS. For this purpose, the measurements were first performed in borate buffer solution at the OCP on Ni₄₅Fe₅₅ film, electrodeposited in the absence and presence of SAC in the electroplating solution. Typical Nyquist and Bode plots are illustrated in Figure 6.

As can be seen, Nyquist plots (Figure 6(a)) were composed of two depressed semicircles. The low-frequency semicircle is much flattened for the Ni₄₅Fe₅₅ film deposited in the absence of SAC. Phase angle diagrams (Figure 6(b)) reveal dispersion owing to R-CPE combination. The modulus of the phase angle maxima is lower than 70°, clearly indicating a deviation from ideal capacitance behaviour. The observed CPE behaviour may be owing to a distribution of the relaxation times as a result of surface inhomogeneities [31,32] or surface porosity [33].

The electrochemical interface can be divided into two sub-interfaces: electrolyte/deposit and electrolyte/substrate.The equivalent circuit used for modelling is shown in the inset of Figure 6(a). In this model, R_e is the electrolytic resistance, R_p is the pore deposit resistance, which is related to electrolyte penetration. CPE_p is the deposit pseudo-capacitance, which describes the Ni₄₅Fe₅₅ film properties. CPE_dl is the double-layer pseudo-capacitance and R_ct is the charge-transfer resistance, which corresponds to substrate dissolution under the deposit and describes the electrochemical reactions at the substrate/electrolyte interface.

The impedance of a CPE is given by: (5) where α is the CPE exponent and Q the CPE parameter.

Table 1 summarises the values of the circuit parameters obtained by regression for the Ni₄₅Fe₅₅ film on n-Si(111)-H from the fit with the experimental impedance diagrams. It is clearly shown that SAC addition in the electroplating bath has a noticeable effect on circuit element values. The deposit pore resistance values are different for each specimen. A higher value of R_p was found for the deposit formed in the presence of SAC in the electroplating bath. The significantly higher value of R_p indicates lower electrolyte penetration through the pores of the deposit, i.e. lower porosity, and hence higher protective effectiveness of the deposit by producing a more compact and less defective nature of the deposit surface morphology, as confirmed by SEM observations.

Table 1

Fitting results of EIS data obtained at OCP, on Ni₄₅Fe₅₅ deposited on n-Si(111)-H, in the absence and presence of SAC in the supporting electrolyte.

	R_e (Ω cm²)	R_p (kΩ cm²)	Q_p × 10⁻⁴ (S^α Ω⁻¹ cm⁻²)	α _p	R_ct (kΩ cm²)	Q_dl × 10⁻⁶ (S^α Ω⁻¹ cm⁻²)	α _dl
In the absence of SAC	291.2	0.254	1.69	0.53	16.91	8.45	0.80
In the presence of SAC	260.8	11.2	5.52	0.91	61.6	0.77	0.67

According to this preliminary analysis, one may conclude that addition of saccharin to the bath leads to an increase in the protective efficacy of the electrodeposited layer and so to an increase of the life of the material which is imperative for technological applications.

In order to investigate the effect of Fe content on the electronic properties of NiFe films formed in the presence of SAC, impedance spectra were performed in borate buffer solution as a function of Fe concentration: 0, 0.015, 0.03 and 0.045 mol L⁻¹ at the OCP. Figure 7 presents typical Nyquist and Bode diagrams.

In the frequency range of measurements, all impedance spectra have the same features. At least two overlapped semicircles are present with corresponding time constants. The high-frequency semicircle may be associated with the porosity of the deposit, whereas the low-frequency semicircle may describe the properties of the substrate/electrolyte interface. As can be seen, the diameter and magnitude of the low-frequency semicircle increase with Fe concentration.

The same equivalent electrical circuit (previously suggested for Ni₄₅Fe₅₅ deposit) was used for the regression of the observed impedance spectra in the whole investigated frequency domain for all the samples. Table 2 gives the calculated values of the electrical circuit parameters. These values show an apparent Fe content dependence of the circuit components. It is interesting to mention that R_p and R_ct follow the same trend. As Fe concentration is increased, R_p increased, reflecting a lower incorporation of the electrolyte through the deposit and a higher corrosion resistance of the substrate. R_p was very low for the deposit with a high concentration of Ni, indicating the existence of a certain degree of porosity in the deposits.

Table 2

Fitting results of EIS data obtained at OCP on NiFe alloys formed with varying Fe content on n-Si(111)-H. (In the presence of SAC in the electroplating bath).

Fe content	R_e (Ω cm²)	R_p (kΩ cm²)	Q_p × 10⁻⁶ (S^α Ω⁻¹ cm⁻²)	α _p	R_ct (kΩ cm²)	Q_dl × 10⁻⁴ (S^α Ω⁻¹ cm⁻²)	α _dl
0	311.5	0.37	233.0	0.54	9.16	1.10	0.79
0.015	232.7	0.33	29.5	0.64	8.55	0.25	0.91
0.03M	260.8	11.2	5.52	0.91	61.6	0.77	0.67
0.045	286.4	10.47	8.49	0.88	9.26	1.60	0.66

Different values of CPE parameters associated with the deposit were found according to Fe content. As stated before, the CPE behaviour is attributed to a surface heterogeneity, CPE exponent may, therefore, represent a measure of the surface inhomogeneity, so its decrease should be associated with an increase in heterogeneity, which results from the metal surface roughening and porosity. Values of the CPE exponent were between 0.54 and 0.91. As Fe concentration increases, increases, indicating the increase in surface homogeneity and compactness. The dependence of and on the Fe amount may be related to the different microstructural properties of the deposit when varying Fe amount. As reported previously, the smallest grain size was achieved with higher iron content. When grain size decreases, more uniform and more adherent film on the surface of the sample was produced.

Magnetic analysis

Magnetic properties of the NiFe thin films are studied by taking hysteresis measurements (M–H loops) at room temperature with a magnetic field applied parallel to the film plane. Figure 8 shows typical hysteresis loops for three NiFe thin films with different iron concentrations. As expected, the characteristics of NiFe thin films are typical of a very soft magnetic material, with a low coercivity: Hc = 0.8 Oe for the Ni₆₅Fe₃₅ alloy (corresponding to [Fe] = 0.015 mol L⁻¹]); 1.3 Oe for the Ni₄₅Fe₅₅ alloy (corresponding to [Fe] = 0.03 mol L⁻¹]) and 1.4 Oe for the Ni₃₅Fe₆₅ alloy (corresponding to [Fe] = 0.045mol L⁻¹]). As can be noticed, the coercivity of the NiFe films increased with increasing Fe concentration. Coercivity changes in a similar way as the mean internal stress ε and in a reverse way as the crystallite grain size hence in the same way as the grain boundaries density. Internal stress and grain boundaries density are one of the main sources of the coercivity. It was already assumed that magnetic domain walls are pinned at the grain boundaries [34]. The magnetic domain walls are fixed by the pinning sites on the grain boundaries and the walls deform by bending under the action of the magnetic field.

The saturation magnetisation (B_s) was calculated from the formula: (6)

where m_s refers to the saturation magnetic moment (emu) in the magnetisation curve, and V the film volume (cm³). For these calculations, the average film thicknesses d were estimated from the SEM transversal observations. By this way, the calculated B_s values are 1.3 T for the Ni₆₅Fe₃₅ alloy (d = 610 nm), 1.5 T for the Ni₄₅Fe₅₅ alloy (d = 640 nm) and 1.6 T for the Ni₃₅Fe₆₅ alloy (d = 760 nm).

The squareness ratio (defined as the ratio of the remanent magnetisation to the saturation magnetisation) varies from 40% for lower iron concentration to 14% for higher iron concentration; this low value indicates a slight anisotropy in the film plane.

Coercivity (H_C) and squareness (S) are extrinsic properties which are dependent on the film microstructure, including grain size, thickness, porosity, preferred orientation and stress. These properties can be controlled by adjusting electroplating parameters (current density, additives …) [35], whereas saturation magnetisation (B_s) is an intrinsic magnetic property which depends on the film composition.

Accordingly, Ni₄₅Fe₅₅ obtained in our experimental conditions (plating mode, SAC additives) is a magnetically very soft alloy (low coercivity) with a high B_s.

Conclusion

In this work, a dilute sulphate plating bath was used to electroplate Ni₄₅Fe₅₅ alloy directly on a silicon substrate by a galvanostatic mode.

The relationship between Fe content in the electroplated films and Fe²⁺ concentration is not linear and this was attributed to the anomalous co-deposition phenomena. The Ni₄₅Fe₅₅ composition was obtained with 0.06 mol L⁻¹ Ni²⁺ ion concentration and around 0.030 mol L⁻¹ Fe²⁺ ion concentration.

According to the SEM observations, the almost equi-atomic NiFe deposit was very smooth whereas the iron-rich deposit was very rough and the nickel-rich deposit present lot of holes on the surface.

EIS results demonstrated that Ni₄₅Fe₅₅ film produced with SAC possesses a higher protective effectiveness by producing a more compact and smooth surface. Considerable improvement in the corrosion resistance of NiFe deposit with increased Fe content has been also observed. When Fe concentration is increased, lower incorporation of the electrolyte through the deposit and higher corrosion resistance of the substrate were obtained.

Structural analysis performed on the electroplated Ni₄₅Fe₅₅ thin Films revealed that the deposit consists of a mixture of fcc/bcc phase with a small particle grain size (10 nm).The magnetic measurements revealed high soft magnetic properties of the films (B_s = 1.5 T, H_c = 1.3Oe).

Footnotes

Acknowledgements

The authors thank D. Dufeu (Neel Institute, CNRS Grenoble) for the magnetic measurements and F. Robaut (CMTC Grenoble) for quantitative EDS analysis.

Disclosure statement

No potential conflict of interest was reported by the authors.

References

Yoo

, Hernandez

, Pardy

, Electrodeposition of FeCoNi thin films for magnetic-MEMS devices. Electrochim Acta. 2006;51:6346–6352. doi: 10.1016/j.electacta.2006.04.020

Pavithra

, Chitharanjan Hegde

Magnetic property and corrosion resistance of electrodeposited nanocrystalline iron-nickel alloys. Appl Surf Sci. 2012;258:6884–6890. doi: 10.1016/j.apsusc.2012.03.124

Tabakovic

, Inturi

, Thurn

, Properties of Fe_1-xNi_x (0.1<x<0.9) and invar (x=0.64) alloys obtained by electrodeposition. Electrochim Acta. 2010;55:6749–6754. doi: 10.1016/j.electacta.2010.05.095

Katalin

, Attila

, Kálmán

, Near-substrate composition depth profile of direct current-plated and pulse-plated Fe–Ni alloys. Electrochim Acta. 2013;103:179–187. doi: 10.1016/j.electacta.2013.04.063

Myung

, Nobe

Electrodeposited Iron group thin-film alloys. Structure-property relationships. J Electrochem Soc. 2001;148(3):C136–C144. doi: 10.1149/1.1345875

, Wei

, Lou

, Preparation of NiFe films by magnetic electroplating. Surf Eng. 2012;28(1):24–29. doi: 10.1179/1743294411Y.0000000063

Fazli

, Bahrololoom

ME.

Effect of plating time on electrodeposition of thick nanocrystalline permalloy foils. Trans IMF. 2016;94(2):92–98. doi: 10.1080/00202967.2015.1122918

Erschinga

, Isoppob

, Tumeleroc

, Uniformity and homogeneity of Fe_XNi_100−X nanowires electrodeposited in nanoporous alumina. J Phy Chem Solids. 2017;104:124–132. doi: 10.1016/j.jpcs.2017.01.008

Zhang

, Zhang

, Wua

, Comparative study in fabrication and magnetic properties of FeNi alloy nanowires and nanotubes. J Magn Magn Mater. 2013;331:162–167. doi: 10.1016/j.jmmm.2012.11.033

10.

Brenner . Electrodeposition of alloys. New York : Academic Press; 1963.

11.

Andricacos

, Romankiw

LT.

Magnetically soft materials: Their properties and electrochemistry. In: Gerischer

, Tobias

, editors. Advances in electrochemical science and engineering V3. New York : VCH Publishers; 1994.

12.

Robertson

, Hu

HL.

Ching Tsang, High performance write head using NiFe 45/55. IEEE Trans Magn. 1997;33(5):2818–2820. doi: 10.1109/20.617741

13.

Cooper E

, Bonhote

, Heidmann

, Recent developments in high-moment electroplated materials for recording heads. J Res Develop. 2005;49:103–126.

14.

Kojima

, Mizuguchi

, Takanashi

Growth of L₁₀–FeNi thin films on Cu(001) single crystal substrates using oxygen and gold surfactants. Thin Solid Films. 2016;603:348–352. doi: 10.1016/j.tsf.2016.02.040

15.

Frisk

, Lindgren

, Pappas

, Resonant X-ray diffraction revealing chemical disorder in sputtered L₁₀ FeNi on Si (0 0 1). J Phys Condens Matter. 2016;28:406002, 9 pp. doi: 10.1088/0953-8984/28/40/406002

16.

Li X

, Yang

Effects of sputtering conditions on the structure and magnetic properties of Ni-Fe films. Mater Sci Eng B. 2004;106:41–45. doi: 10.1016/j.mseb.2003.07.003

17.

Rook

, Gupta

, Hieronymi

Sputter deposition of 45% Ni - 55% Fe for high moment top pole applications. J Appl Phy. 2000;87:5843–5845. doi: 10.1063/1.372541

18.

Ouir

, Sam

, Fortas

, FeNi alloys electroplated into porous (n-type) silicon. Phys Stat Sol. 2008;c:1–4.

19.

Gao

, Ma

Novogradecz KM, Norton PR, Characterization of permalloy thin films electrodeposited on Si(111) surfaces. J Appl Phys. 1997;81(11):7595–7599. doi: 10.1063/1.365305

20.

Spada

, de Oliveira

, da Rocha

, Thin films of Fe_xNi_1−x electroplated on silicon (100). J Magn Magn Mater. 2004;e891:272–276.

21.

Zarpellon

, Jurca

, Klein

, Electrodeposition de Fe thin films on Si(111) surfaces in the presence of sodium saccharin. Electrochim Acta. 2007;53:2002–2008. doi: 10.1016/j.electacta.2007.08.058

22.

Hou

, Liu

, Li

, Microstructure and Magnetic Properties of Electrodeposited Fe-Based Thin Films on Silicon Substrate Under Heat Treatment. Mater Manuf Process. 2015;0:1–5.

23.

O'Donnell

, Wang

, Kulkarni

, Electrodeposited anisotropic NiFe 45/55 thin films for high-frequency micro-inductor applications. J Magn Magn Mater. 2010;332:1690–1693. doi: 10.1016/j.jmmm.2009.01.038

24.

Roy

, Connell

, Ludwig

, Pulse reverse plating for integrated magnetic on Si. J Magn Magn Mater. 2005;290–291:1524–1527. doi: 10.1016/j.jmmm.2004.11.566

25.

Hessami

, Tobias

CW.

A mathematic model for Anomalous co-deposition of Nickel-Iron on a Rotating Disk Electrode. J Electrochem Soc. 1989;136:3611–3616. doi: 10.1149/1.2096519

26.

Rashidi

, Amadeh

The effect of saccharin addition and bath temperature on the grain size of nanocrystalline nickel coatings. Surf Coat Technol. 2009;204:353–358. doi: 10.1016/j.surfcoat.2009.07.036

27.

Brankovic

SR.

Saccharin effect on properties of 2.4 T CoFe Films. Electrochim Acta. 2012;84:139–144. doi: 10.1016/j.electacta.2012.07.061

28.

Chenna

, Hamadou

, Benbrahim

, Electronic properties of Hydrogen terminated n-Si (111)/electrolyte interface: Effect of saccharin addition. J Electroanal Chem. 2017;802:118–122. doi: 10.1016/j.jelechem.2017.08.053

29.

Vaes

, Fransaer

, Celis

JP.

Cathodic inhibition effects during NiFe and ZnNi alloy deposition. J Electrochem Soc. 2002;149(11):C567–C572. doi: 10.1149/1.1511189

30.

Nakano

, Matsuno

, Oue

, Mechanism of Anomalous Type Electrodeposition of Fe-Ni Alloys from Sulfate Solutions. Mater Trans. 2004;45(11):3130–3135. doi: 10.2320/matertrans.45.3130

31.

Kerner

, Pajkossy

Impedance of rough capacitive electrodes: The role of surface disorder. J Electroanal Chem. 1998;448:139–142. doi: 10.1016/S0022-0728(98)00025-4

32.

Córdoba-Torres

, Mesquita

, Devos

, On the intrinsic coupling between constant-phase element parameters α and Q in electrochemical impedance spectroscopy. Electrochim Acta. 2012;72:172–178. doi: 10.1016/j.electacta.2012.04.020

33.

Jurczakowski

, Hitz

, Lasia

Impedance of porous Au based electrodes. J Electroanal Chem. 2004;572:355–366. doi: 10.1016/j.jelechem.2004.01.008

34.

Van der Zaag

, Johnson

, Ruigrok

JJM

, The grain boundary influence on the coercivity of polycrystalline MnZn-ferrites. J Magn Magn Mater. 1994;129:L137–L140. doi: 10.1016/0304-8853(94)90103-1

35.

Kochar

, Alper

, Kuru

, Magnetic anisotropy and its thickness dependence for NiFe alloy films electrodeposited on polycrystalline Cu substrates. J Magn Magn Mater. 2006;304:e736–e738. doi: 10.1016/j.jmmm.2006.02.207