Insight into the Mechanism of the Fe-Ni Alloys Co-Deposition from Poly-Nuclear Complexes

Electrodeposition of the Fe-Ni alloys in the citrate bath was anomalous co-deposition, but normal co-deposition in the citrate-ammonia bath. The Fe and Ni elements were uniformly distributed in the Fe-Ni deposit, which were highly alloyed. These results cannot be simply explained by the hydroxide deposition model or the competitive adsorption of intermediate model, which proposed in literature. A poly-nuclear complex co-deposition mechanism of Fe-Ni alloy was proposed in the above two baths, which can explain better electro-deposition behavior of the Fe-Ni co-deposition and the experimental phenomena we observed. © The

Fe-Ni alloys have many desirable properties, such as low thermal expansion coefficient, good soft magnetic properties and protective properties. [1][2][3] They have been widely used in computer storage devices, magnetic simulators, magnetic shielding, and solar cells, basal, 4-7 etc. Many methods are used to prepare Fe-Ni alloys, including mechanical-alloying method, 8,9 microwave plasma method 10 and electroplating method. 11 Where, electroplating method is the most cost-effective to prepare thin strip and thin film on a variety of surfaces with varied geometries. The properties of Fe-Ni alloy are affected greatly by its composition which determined by electrodeposition process. Therefore, it is of great significance to make clear the mechanism of Fe-Ni deposition process.
It has been well known that the deposition behavior of Fe-Ni alloys showed a typical feature of anomalous co-deposition, in which the less noble Fe is preferentially deposited on electrode in comparison with the noble Ni. [12][13][14][15][16][17][18][19] But some researchers [20][21][22] reported that anomalous co-deposition of Fe-Ni alloys can be changed to normal co-deposition by adjusting the electroplating conditions, such as the use of reverse pulse electro-deposition, 20 or by increasing the plating temperature to 353K, 21 or by using ionic liquids in the electrolyte solution. 22 The mechanism of Fe-Ni anomalous co-deposition has been studied by many researchers. [23][24][25][26][27][28][29][30] The hydroxide deposition model 23 and the competitive adsorption of intermediate model 30 have been well accepted so far to explain the anomalous co-deposition of Fe-Ni. In hydroxide deposition model, 23 it said that noble Ni deposition was strongly suppressed in the presence of Fe(OH) 2 , which preferentially formed and adsorbed on the cathode. While in competitive adsorption of intermediate model, 30 it said that two-step reaction process existed in deposition of Fe-Ni alloy, first, the univalent cation ion of metal hydroxide (FeOH + , NiOH + ) formed and adsorbed on electrode, then they were further reduced to metal alloy. As FeOH + is more preferentially adsorbed on electrode than NiOH + , Ni deposition was inhibited. This model was confirmed by Baker and West, 31 Harris and Clair. 32 However, all of these models could not explain the reason why anomalous co-deposition of Fe-Ni alloys changed to normal co-deposition by adjusting the electroplating conditions.
In this paper, we reported normal co-deposition of Fe-Ni alloys in the citrate-ammonia bath and proposed a poly-nuclear complexes co-deposition model in Fe-Ni alloys electrodeposition process. It can explain why Fe-Ni showed anomalous co-deposition in some conditions but normal co-deposition when changed conditions. It is of great significance for the further study of the mechanism of electrodeposition of metal alloys. O into the above baths, the corresponding citrate-ammonia baths were generated. The pH values of all baths were adjusted to 7.5 with 20%NaOH and 10% H 2 SO 4 solutions. All plating baths were prepared using deionized water and analytical grade reagents that were used as received without further purification.

Electroplating, composition and microstructure of Fe-Ni alloy.-
Fe-Ni alloys were electroplated on 316 stainless steel plates for 30 min at 7 A/dm 2 , 75 • C in the baths. Under the above experimental conditions, the smooth and good adhesion coatings can be obtained. The peeled deposit was dissolved in HCl (1:1) + H 2 O 2 solution. The content of Fe and Ni were determined by WFX-1C atomic absorption spectrophotometer.
The morphology of the Fe-Ni alloy were characterized by an environmental SEM FEI QUANTA 200 and EDAX attachment produced by EDAX. The X-ray diffraction patterns for Fe-Ni alloys were recorded using a D/max-rA diffractometer with nickel-filtered and graphite-monochromatized Cu K α radiation operating at 50KV and 100 mA over 2θ range of 5∼80 • . EQCM measurement.-For the EQCM, an AT-cut 9 MHz goldcoated electrode quartz crystal(area of 0.29cm 2 ) was used as working electrode, an Ag/AgCl electrode (saturated KCl) served as the reference electrode and all potentials reported versus this scale, and a platinum wire served as the counter electrode.
Mass changes during deposition and dissolution result in frequency shift which can be described by the Sauerbrey's equation: 33 Where, f: the resonant oscillator frequency shift of the quartz crystal which caused by mass change (Hz); f 0 : the resonant frequency of the quartz crystal (Hz); m: the mass change; A: the effective working area of quartz crystal (cm 2 ); μ q : the shear modulus of quartz crystal (GPa); ρ q : the density of quartz. For the same quartz, f 0, A, μ q and ρ q are constant. So K is constant for the given quartz crystal. The resonant frequency was measured using an HP4395A impedance analyzer and the mass change was calculated according to Equation 2. Electrochemical results were performed usingCHI600A  electrochemical workstation (CH Instruments, USA). The electrolyte volume was fixed to 20 ml, the potential scan range was from −0.1V to −1.5V, at 5mV s −1 .
UV-vis measurement.-UV-vis absorbance measurements of the plating solutions were performed by UV-vis spectrometer SHI-MADZU UV-1800 over the range of wavelength 400∼1000nm.

Results and Discussion
The anomalous and normal co-deposition behavior of Fe-Ni alloys.-In the study process of Fe-Ni electrodeposition, we found that Fe-Ni alloy deposited in the citrate bath exhibited anomalous co-deposition, but in the citrate-ammonia bath, it showed normal codeposition. Fig. 1 shows the Fe and Ni content in Fe-Ni alloys obtained by electrodeposition in the citrate and citrate-ammonia baths. In the citrate baths, as the content of Ni in the deposits was less than that in the electrolytes, it was anomalous co-deposition. In the citrateammonia baths, as the content of Ni in the deposits was greater than that in the electrolyte, it was normal co-deposition. Fig. 2 shows the morphology of Fe59.091Ni40.9 alloy obtained in the citrate bath (0.1M FeSO 4 and 0.1M NiSO 4 ). 13μm thickness of Fe-Ni deposit was smooth, no crack and pit. It was tightly adhered to the substrate without gaps. Fig. 3 shows the XRD patterns of Fe, Ni and Fe59.091Ni40.9 alloy electrodeposited from the citrate baths and Fe, Ni and Fe18.7Ni81.3 alloy from citrate-ammonia baths. As can be seen, FeNi (XRD JCPDS data file No. 18-0646) phase was observed in the citrate bath and

Study of Fe-Ni electro-deposition behavior by EQCM.-EQCM
technique was used to study electrochemical behavior and mass changes of Fe-Ni deposition process in different electroplating baths. The electro-deposited Fe was completely dissolved into electrolyte during the anodic stripping process, so the proportionality coefficient K in Equation 2 can be determined by the deposition and dissolution process of Fe. According to the measurement, the proportionality coefficient K value was 1.18ng•Hz −1 . The mass changes calculated from the frequency shift according to Equation 2. The partial current densities of Fe, Ni and Fe-Ni deposits can be derived via Faraday's law from the deposition rate (dm/dt). 34,35 For the alloy deposition, the composition changes during the potential sweep and could not be determined. We reckon Fe59.9Ni40.1 in the citrate bath and Fe18.7Ni81.7 in the citrate-ammonia bath according to the results of Fig. 1.
The mass changes, total current densities and Fe, Ni and Fe-Ni alloy partial current densities versus scanning potential in the citrate and citrate-ammonia electrolyte baths are shown in In Fig. 4A-I, the maximum amount of the Fe deposits were 22.80μg•cm −2 in the citrate bath and 59.53μg•cm −2 in the citrateammonia bath. Obviously, the amount of Fe deposit obtained in the citrate-ammonia was higher than that in the citrate bath, which meant ammonia accelerated the deposition of Fe. The same results were obtained by Osseo-Asare. 34 When swept back to −0.76 V, the Fe deposits began to dissolve in both baths. The mass change was close to zero when the potential swept back to −0.558V in the citrate bath and to −0.385V in the citrate-ammonia bath, which meant Fe deposits had been completely dissolved in the baths.
From the insert graphs of Fig. 4A-II and Fig. 4A-III, in the Fe citrate bath, the total current density increased at −0.98V and the Fe partial current density increased at −1.10V; and in the Fe citrateammonia bath, the total current density increased at −0.99V and the Fe partial current density increased at −1.10 V. The potential difference in Fig. 4A-II and Fig. 4A-III indicated hydrogen evolution or other side reaction occurred. The current efficiency of Fe deposition were 4.43% in the citrate bath and 5.71% in the citrate-ammonia bath. As the deposition potential of Fe did not change in both electrolytes, the deposition mechanism of Fe in both baths may not change. The increase in current efficiency of Fe deposition in the citrate-ammonia bath may cause by the inhibition of side effects.
In Fig. 4B-I, the maximum amount of the Ni deposits were 7.49μg•cm −2 in the citrate bath and 102.08μg•cm −2 in the citrateammonia bath. After the potential swept back to −0.1V, there were still 7.27μg•cm −2 and 101.84μg•cm −2 Ni deposits remained on the electrodes in both baths respectively. These means ammonia can accelerate the deposition of Ni more efficiently than Fe and the Ni deposits did not completely anodic dissolve in baths.
As seen in the insert graphs of Fig. 4B-II and Fig. 4B-III, in the Ni citrate bath, the total current density increased at −0.86V and the Ni partial current density increased at −0.96V; and in the Ni citrateammonia bath, the total current density increased at −0.82V and the Ni partial current density increased at −0.92V. The potential difference in Fig. 4B-II and Fig. 4B-III indicated hydrogen evolution or other side reaction occurred. The current efficiency of Ni deposition were 2.01% in the citrate bath and 10.71% in the citrate-ammonia bath. As the deposition potential of Ni changed from −0.96 V in citrate bath to −0.92V in citrate-ammonia bath, the mechanism of Ni deposition changed. A sharp increase in current efficiency resulted by the change of deposition process.
In Fig. 4C-I, the maximum amount of the Fe-Ni deposits were 10.18μg•cm −2 and 77.17μg•cm −2 respectively. After the potential swept back to −0.1V, there were 9.43 μg•cm −2 and 76.63 μg•cm −2 deposits remained on the electrodes respectively. Increased amount of Fe-Ni alloy deposited was less than Ni, but much higher than Fe by addition of ammonia. The Fe-Ni alloy deposits did not completely anodic dissolve in baths.
In insert graphs of Fig. 4C-II and Fig. 4C-III, the total current and Fe-Ni partial current increased at −1.00V and −1.08V respectively in the citrate bath, and at −0.92V and −1.06V in the citrateammonia bath. The potentials which corresponding to total current change were positive than those which corresponding to Fe-Ni partial current change indicated side reaction occurred. The difference of Fe-Ni alloy deposition potential in both baths was caused by the change of deposition process. In the citrate-ammonia bath, the potential which side reaction happened positive shift 0.08V and the Fe-Ni deposition potential only move 0.02V in a positive direction compared with two baths. The current efficiency of Fe-Ni deposition were 2.03% in citrate bath and 8.92% in citrate-ammonia bath. Much higher current efficiency of Fe-Ni deposition in the citrate-ammonia bath may cause by new electrochemical reactant.
Based on the results of Fig. 4, the following conclusions can be made. The deposition rate of Fe, Ni, and Fe-Ni in the citrate-ammonia bath was accelerated, but the increase rate of each species was not equivalent. The plating velocity order was: Fe<Fe-Ni<Ni. While in the citrate bath, the plating rate order was: Ni<Fe-Ni<Fe. The current efficiency of Fe, Ni, Fe-Ni deposition were improved by adding ammonia. The ranking of amplitude increases was Ni>Fe-Ni>Fe. The change of deposition potential of Ni and Fe-Ni in both baths means the change of deposition processes. Although the extent of the positive shift of potential for side reaction occurrence was larger than that of the Fe-Ni deposition, the current efficiency of Fe-Ni deposition was still greatly improved. This means the reduced complex must be electro-active. Fig. 5 shows the UV-visible absorption spectra of the various baths. In Fig. 5A, in the citrate bath, when increasing the concentration of Ni (II), the absorption intensity increased and the absorption peak gradually blue shifted (comparing curve b and curve g, the absorption peak blue shifted about 26nm, 684nm→658nm). The shift of the absorption peak indicated that, in the Fe-Ni bath, Fe (II) and Ni (II) did not exist as a separate complex but formed new complexes. In the Fe citrate bath, the predominant species was the [Fe 2 (cit) 2 (OH) 2 ] 4− species, which can be calculated by software Hydra/Medusa 36 (see Fig. 6). In the Ni citrate bath, a poly-nuclear complex of the composition [Ni 4 (cit) 3 (OH) 4 ] 5− formed in the weakly alkaline solutions. 37 The shift of the absorption peak in Fig. 6A showed that Fe(II) and Ni(II) in the Fe-Ni citrate bath interacted to form a new poly-nuclear [Fe x Ni y (cit) 3 (OH) 4 ] 5− (x+y = 4), in which the values of x and y in the formula were related to the concentration of Fe(II) and Ni(II) in the baths, as evidenced by the difference in absorbance peaks and values of the different Fe (II)/Ni (II) ratios baths. Compared with curve g and curve h, the absorption peak red shifted about 11 nm (658nm→669nm) and the absorbance increased, which meant that Fe(II) took part in the coordination of [Fe x Ni y (cit) 3 (OH) 4 ] 5− (x+y = 4), and the concentration of [Fe x Ni y (cit) 3 (OH) 4 ] 5− (x+y = 4) increased. At the same time, the deposition potential of Fe-Ni in Fig. 4 was different from that of single Fe and Ni, indicating that a new electrochemical reactant was formed.

UV-visible absorption spectroscopy of different electrolyte baths.-
In Fig. 5B, in the citrate-ammonia bath, similar phenomenon had been observed, with the increasing of the Ni (II) concentration, the absorption intensity increased and the absorption peak blue shifted (comparing curve b and curve g, the absorption peak blue shifted about 30 nm, 642.5nm→612nm). The shift of the absorption peak also indicated that, in the Fe-Ni citrate-ammonia bath, Fe (II) and Ni (II) did not exist as a separate complex but form new complexes. Comparing Fig. 5A and Fig. 5B, it is not difficult to find that, in the citrateammonia bath, the absorption peaks of each bath were blue-shifted due to the coordination field strength NH 3 > OH − , OH − was replaced   by NH 3 . 8,38,39 The shift of the absorption peak in Fig. 5B showed that Fe(II) and Ni(II) in the Fe-Ni citrate-ammonia bath interacted to form a new poly-nuclear complex [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4), in which the values of x and y in the formula were related to the concentration of Fe(II) and Ni(II) in the baths, as evidenced by the vary of absorbance peaks and value of the different Fe (II)/Ni (II) ratios baths. Compared with curve g and curve h the absorption peak red shifted about 18nm (612nm→630nm) but the absorbance almost unchanged, which meant that although Fe (II) took part in the coordination of [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4), the concentration of the [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4) did not increase due to the weak affinity between Fe(II) and NH 3 . Meanwhile, the deposition potential of Fe-Ni in the citrate-ammonia bath was different from that of single Fe and Ni, and not the same as the Fe-Ni citrate bath (Fig. 4), which indicating the formation of the new electrochemical reactant.

New poly-nuclear complex deposition mechanism model.-For
Fe-Ni alloy, the deposition potential was not equal to single Fe or Ni indicates a new species different from individual Fe and Ni formed and participated the electrochemical reactions. However, the present hydroxide deposition and competitive adsorption of intermediate models didn't deal with the Fe-Ni alloy deposition potential and also cannot explain the phenomenon that Fe and Ni elements were highly alloyed and uniformly dispersed in the Fe-Ni coatings. A new poly-nuclear complex deposition mechanism model was proposed for the Fe-Ni alloy electro-deposition process.
In the Fe-Ni citrate bath, the following cathodic reactions occurred:  Fig. 5A. Because the OH − ion favored association with the Fe 2+ more than with the Ni 2+ , the value of x/y was greater than that of c Fe(II) /c Ni(II) in the electrolytes, e.g. c Fe(II) = c Ni(II) = 0.1M, x/y>1.
In the Fe-Ni citrate-ammonia bath, the following cathodic reactions occurred: NH 4 + + e − → NH 3 + H ads [7] Fe x Ni y (cit) 3 (NH 3 ) 4 − (x + y = 4)+8e − → Fe x Ni y +2cit 3 − +4NH 3 [8] Here, the Fe-Ni alloy electro-deposition resulted from the discharge of the poly-nuclear complex [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4), where the values of x and y were related to the concentration of Fe(II) and Ni(II) in the bath, this can be confirmed by the results in Fig. 5B. Due to the NH 3 coordinated better with Ni 2+ than with Fe 2+ , the value of x/y was less than that of c Fe(II) /c Ni(II) in the electrolytes, e.g. c Fe(II) = c Ni(II) = 0.1M, x/y<1. Generally, the poly-nuclear complex has a high degree of entropy value and high internal energy, and the chemical bond is easily broken, so the Fe-Ni poly-nuclear complexes are electro-active and easy to be reduced. These were confirmed by the experimental results showed in Fig. 4.
In summary, we proposed the poly-nuclear complex deposition mechanism model. It was demonstrated that Fe and Ni were simultaneously reduced by cathodic reduction in the ratio of x/y. The content of Fe and Ni in the deposit was determined by the ratios of x and y. As the deposition process proceeded, reduced Fe and Ni elements were highly alloyed and uniformly dispersed in the coatings. In the Fe-Ni citrate bath, the stoichiometry ratio x/y of the poly-nuclear complex [Fe x Ni y (cit) 3 (OH) 4 ] 5− (x+y = 4) was greater than that of the c Fe(II) /c Ni(II) in the electrolytes, then, exhibited the typical anomalous co-deposition. In the Fe-Ni citrate-ammonia bath, the stoichiometry ratio x/y of the poly-nuclear complex [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4) was less than that of the c Fe(II) /c Ni(II) in plating baths, so exhibited the typical normal co-deposition. In the Fe-Ni citrate bath, the charged poly-nuclear complex [Fe x Ni y (cit) 3 (OH) 4 ] 5− (x+y = 4) contained five negative charges. And in the Fe-Ni ammonia citrate bath, the charged poly-nuclear complex [Fe x Ni y (cit) 3 (NH 3 ) 4 ] − (x+y = 4) was with only one negative charge. The latter complex diffused faster to the cathode surface, so the electro-deposition rate of the Fe-Ni alloy in the Fe-Ni citrate-ammonia bath was much faster than that in the Fe-Ni citrate bath.

Conclusions
In this reported study, a new mechanism model for the codeposition of Fe-Ni alloys from poly-nuclear complexes was developed. The results of this effort showed that anomalous and normal co-deposition of Fe-Ni depended on the composition of the electroactive poly-nuclear complex in the plating solution. Simultaneous deposition of Fe and Ni was determined by the composition of the complex. Highly alloyed, uniform Fe-Ni deposits were obtained. The deposition rate in the Fe-Ni ammonia-citrate bath was found to be faster than that in the Fe-Ni citrate bath, which may have been caused by the difference in the negative charge number of poly-nuclear complexes.