- Email: [email protected]
Bi thin films on Cu Substrate
Accepted Manuscript Title: Morphological, physicochemical and magnetic characterization of electrodeposited Mn-Bi and Mn-Bi/Bi thin films on Cu Substrate Author: B. Benfedda N. Benbrahim S. Boudinar A. Kadri E. Chainet F. Charlot S. Coindeau Y. Dahmane L. Hamadou PII: DOI: Reference:
S0013-4686(16)31040-4 http://dx.doi.org/doi:10.1016/j.electacta.2016.05.007 EA 27223
To appear in:
Electrochimica Acta
Received date: Revised date: Accepted date:
27-12-2015 1-5-2016 2-5-2016
Please cite this article as: B.Benfedda, N.Benbrahim, S.Boudinar, A.Kadri, E.Chainet, F.Charlot, S.Coindeau, Y.Dahmane, L.Hamadou, Morphological, physicochemical and magnetic characterization of electrodeposited Mn-Bi and Mn-Bi/Bi thin films on Cu Substrate, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2016.05.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Morphological,physicochemical and magnetic characterization electrodeposited Mn-Bi and Mn-Bi/Bi thin films on Cu Substrate B. Benfedda(1)*, N. Benbrahim (1), S.Boudinar (1 ), A. Kadri (1), E. Chainet (2), F.Charlot Coindeau (3), Y. Dahmane(4), and L. Hamadou(1)
of (3 )
, S.
1 Laboratoire de Physique et Chimie des Matériaux (LPCM), Université M. Mammeri de Tizi- Ouzou (Algérie) 2 Laboratoire d’Electrochimie et de Physico-chimie des Matériaux et des Interfaces (LEPMI), UMR 5631 CNRS-INPG-UJF, BP 75, 38402 Saint-Martin d’Heres Cedex, France 3 C.M.T.C., Grenoble I.N.P. Bât Phelma Campus, Domaine Universitaire, 38402 Saint Martin d’Hères, Grenoble, France 4SPINTEC/CEA Grenoble 17 rue des Martyrs, 38054 GRENOBLE Cedex 9
Corresponding author: BayaBenfedda, E-Mail: [email protected]
Abstract Mn-Bi thin films were electroplated on Cu (111) substrates in an acidic chloride bath. In order to determine the deposition potential of each element, cyclic voltammetry using a rotating disk electrode was performed. Two types of thin films were obtained using two deposition mode: the first one called thin Mn-Bi layers by using a single applied potential and the second one called Mn-Bi/Bi bilayers by using a double pulse potential. Annealing treatments at 300°C for 1 hour under vacuum condition were carried out in order to cause an interdiffusion between manganese and bismuth. The morphological and crystalline structure of the various deposits was investigated by scanning electron microscopy with field effect (SEM-FEG) and by X-ray diffraction analysis (XRD). Magnetic characterizations were also made using a superconducting quantum interference device (SQUID) magnetometer. The morphological and the structural properties of the thin layers and the bilayers are completely different, indicating that the growth process changes according to the plating mode. After annealing a mixed MnBiCu phase with a coercivity of 300 Oe and 400 Oe was observed on the thin layers and the bilayers respectively.
Key words: Mn-Bi, Hard magnetic material, Electrodeposition, Voltammetry, SEM-FEG, EDS, XRD
1. Introduction Because of their many technological applications, the magnetic materials have a considerable economic importance on a worldwide scale. These last years, many studies have been reported and currently activated in order to synthesize new materials able to meet increasingly powerful needs. In this context, various magnetic materials, such as, NiFe, CoFe, CoPt, FePt, MnBi….etc, have been the subject of several works [1-4] due to their potential application in the magnetic recording field. The low-temperature phase (LTP) of MnBi which crystallizes in the Ni-As type hexagonal crystal structure with c axis as the easy direction of magnetization, exhibit a high uniaxial magnetic anisotropy (2×107 erg/cm3) making it very attractive for high-density magnetic recording. In addition, the LTP MnBi is known to show a high polar Kerr rotation at room temperature, which makes it attractive also for the magnetooptical (MO) perpendicular recording application [4-6]. But although MnBi has potential applications in the high density recording field, it is very difficult to obtain high quality MnBi by conventional means, because, according to the Mn-Bi phase diagram [7,8], at the peritectic temperature (719 K) Mn tends to segregate from the Mn-Bi liquid. On the other hand, if the temperature is lowered to below that of the eutectic, the reaction becomes very slow. Generally, works cited in the literature reported on Mn-Bi layers prepared by physical techniques such as melt spinning, magnetron sputtering and vacuum evaporation [4-6, 9]. However, these methods remain expensive as they require heavy equipment and are difficult to implement. In the present study, we chose the electrochemical way which is an important
processing technology for micro fabrication due to its low cost, high yield, low energy requirement, and capability for generating high aspect ratio features. Many works using the electrochemical synthesis methods were already carried out, in particular for the elaboration of different materials such as NiFe, FeCo, CoPt, SnBi , TeBi… etc [11-15], to the best of our knowledge, no work have been reported on the synthesis of the MnBi alloy by the electrochemical way. The electrodeposition of Mn-Bi by electrochemical method is a very delicate process, considering on one hand, the complexity of the manganese deposition kinetics [16-19] and on the other hand, the large difference between the potential related to the Mn2+/Mn and Bi3+/Bi couples. In addition, the process of metals and alloys electrodeposition depends on several parameters which must be controlled, in particular, pH; temperature, concentration of metal ions to be deposited, deposition potential, additives, plating mode… etc. In a previous work [19], we have described preliminary study of principal conditions to deposit Mn-Bi system in chloride bath containing ammonium chloride as additive. In this respect, cyclic voltammetry has been used to determine the nature of the various reactional stages occurring during the electrodeposition of Mn and Bi as a function of ammonium chloride concentration. The thermodynamic study of the electrolysis bath was also carried out to identify the electroactive species. Various characterizations (structural, morphological) were also performed on the MnBi deposited system. In the present work, a series of Mn-Bi thin layers and Mn-Bi/Bi bilayers have been grown on (111) textured Cu substrate using single and double pulse applied potential. It is well known that the initial stage of electrocrystallization, including the nucleus formation and their growth mechanism, directly determines the properties and quality of the thin electrodeposited layer. So, we have studied in a first time the initial stage of the Mn-Bi electrocrystallization on the Cu substrate and on an electrodeposited Bi on the Cu substrate. In a second time, we aimed to
investigate and correlate the structural, morphological and magnetic properties of the electroplated thin films with the plating mode and the annealing treatment. A chloride based electrolysis bath was used. The chosen bath is based so on the studies reported in literature on manganese and bismuth metal electrodeposition. So, manganese deposition can be obtained via manganese sulfate and chloride baths with the corresponding ammonium salts [20-23], whereas bismuth deposition is often synthesized in nitrates and chlorides baths [24-27].
2. Experimental details A vitreous carbon rotating disk electrode (RDE) (0.2 cm2) with a rotation speed fixed at 250 rpm and either Cu/Ta/Si substrates (0.5 cm2) were used as working electrodes in a classical three electrodes electrochemical cell. The counter electrode was a platinum wire immersed in a separate compartment containing solution without electroactive metallic cations. All potential values were measured with respect to a saturated calomel reference electrode (SCE). All experiments were carried out at ambient temperature in a chloride bath deoxygenated by nitrogen gas bubbling. The electrochemical Mn-Bi bath consisted of MnCl2 4H2O (0.4 mol.L1
), BiCl3 (10-2mol.L-1) and NH4Cl (3mol.L-1). The pH was adjusted to 2 by adding HCl
solution. The solution was prepared immediately prior to each experiment using deionised water and analytical grade reagents (Aldrich). The MnBi various deposits were prepared with continuous stirring of the electrolyte solution. The
electrochemical
measurements
were
performed
using
an
EG&G
273A
Potentiostat/Galvanostat controlled by a microcomputer via GPIB interface operated by M352 EG&G software. The composition of the samples was determined by energy dispersive X-ray spectroscopy (EDX). The morphology was examined by scanning electron microscopy (SEM) and the scanning electron microscopy with field effect (SEM-FEG). The crystalline structure was investigated by X-ray diffraction (XRD) with CuK radiation. Magnetic properties have
been also characterized at room temperature using a superconducting quantum interference device (SQUID) magnetometer.
3. Results and discussions 3.1 Electrochemical characterization The voltammetric study was first carried out in two different baths containing each of the two elements separately on a RDE electrode. Figure 1-a shows the voltammogram recorded on a bismuth chloride bath (10-2 mol.L-1 of BiCl3 and 3 mol.L-1 of ammonium chloride NH4Cl, pH = 2). Starting from the rest potential (-0.2 V vs SCE) and sweeping towards the cathodic direction, a plateau of a weak current is observed in the potential range (-0.23 V vs SCE to 0.5 V vs SCE), it is the limiting current region where Bi (III) reduction occurs under mass transport control. In the reverse scan, one anodic peak appears at around -0.2 V vs SCE. It is attributed to the bismuth dissolution according to a more detailed previous study in the same bath [17]. Figure 1-b shows another voltammogram recorded in a manganese chloride bath (0,4 mol.L-1 of MnCl2 and 3 mol.L-1 of ammonium chloride NH4Cl, pH = 2). Starting from the rest potential (-0.8 V vs SCE) and scanning towards the cathodic direction, a low current density is observed. This step is attributed to H+ reduction. Up to -1.3 V vs SCE, a weak reduction plateau followed by a rapid increase in the current density at (-1.57 V vs SCE ), corresponding probably to water reduction and also to the manganese ions reduction is observed. In the reverse scan, one anodic peak related to the manganese dissolution is observed at around -1,42 V vs SCE. The current loop noticed between the forward and the reverse scan results from the nucleation process on the vitreous carbon RDE. The current efficiency was estimated from the ratio between the anodic charge and the cathodic charge calculated from the different voltammograms recorded in the Mn and the Bi
electrolyte respectively. It was found that the efficiency of the manganese deposition is about (12,63%) whereas the one of the bismuth deposition is close to 100% (93,84 %). These values are predictable considering the equilibrium potential of Mn2+/Mn and Bi3+/Bi. One may note here that by this way the current efficiency is under estimated as the deposit is not completely dissolved in the anodic sweeping. Cyclic voltammetry recorded on the RDE electrode in the complete Mn-Bi electrolyte described above is shown on figure 2. As can be seen, starting from the rest potential (E = 0.033 V vs SCE), a plateau of a weak current density is observed between -0.2V vs SCE to – 1.4 V vs SCE, this current was attributed to bismuth reduction followed by hydrogen evolution according to the previous study. As the cathodic potential increases up to -1.4 V vs SCE, the current density increases indicating a probable reduction of Mn2+ ions simultaneously with the bismuth and the proton reduction (hydrogen evolution). In the reverse scan and in the range of cathodic potential, a cross over attributed to the typical nucleation process of the Bi and the Mn grains on the working electrode is observed. On the anodic side, two separate peaks are observed. Peak A observed at -1.42 V vs SCE is attributed to manganese dissolution. Whereas the peak B observed at -0.2 V vs SCE is attributed to bismuth dissolution. It is important to note the relative intensity of the Mn and Bi dissolution peaks comparing to the relative concentration of the two ions species in the electrolysis bath. Peak B is more significant than peak A because the Bi3+ reduction is not much affected by the simultaneous hydrogen evolution as in the case of Mn2+ reduction. Consequently, the current efficiency of bismuth deposition is higher than that of Mn as mentioned above.
Based on the above results we have used two potential values of (-1.65 V vs SCE and -1.75 V vs SCE) for the deposition of Mn-Bi thin layer during 100s and a couple of potential -0.3 V
vs SCE /-1.65 V vs SCE for the deposition of the Mn-Bi/Bi bilayer during 100s at each potential step. The current-time transients obtained on a thin layer and a bilayer deposited on the Cu substrate are shown on figure 3-a and 3-b respectively. One can observe that the current-time transient obtained at -1.65 V vs SCE (Fig. 3-a) is not constant; the observed change in the slope highlights the reduction of various species, namely manganese, hydrogen and bismuth. The curve b (Fig.3-b) related to a double chronoamperometry shows two different responses in the current density according to the imposed potential. The first response reveals a very stable current density which characterizes the discharge of Bi3+ ions on the Cu substrate. The second response shows a less stable current density due to the reduction of various species including Mn2+ reduction as on curve a. One can also note in this figure, the difference in the current densities magnitude (then in the total charge) for the same applied potential of -1.65 V vs SCE (The corresponding calculated total charge per area from the two figures 3-a and 3-b are 15.56 C/cm2 and -12.76 C/ cm2 respectively): in fact, in the thin layer Mn is deposited directly on a copper substrate while in the bilayer it is deposited on a substrate already coated with small amount of bismuth, the latter being thus occupied a number of active sites on the copper substrate that has resulted in the decrease in the deposition rate of manganese .
3.2 Electrocrystallisation study Before making further characterization on the deposited thin films, it seemed important to study the electrocrystallisation mechanism at the early stages of deposition under the same conditions used in the electroplating thin films. A series of current transients measurements are carried at different applied potentials during short time. Figure 4(a and b) show the
experimental current transients I(t) recorded on Cu substrate and on Bi (electroplated at -0.3V vs SCE during 100 s) on a Cu substrate at three applied potentials (-1.65V vs SCE, -1.7Vvs SCE and -1.75Vvs SCE). All the transients exhibit a same behavior, an increase in the current density until a maximum value Imax reached at a time tmax corresponding to the discharge of the electrochemical double layer and the formation of new phase on the electrode surface. The current maximum increases, and shifts towards shorter times when the potential increase. When t tmax it can be observed on the Figure 4(a) that the current decreases slightly and then stabilizes with time corresponding to the growth of the formed nuclei. The current-time transient registered on the Bi/Cu substrate (Fig 4-b) shows also dependence with a potential, but the current density does not increase with time. In order to obtain general information about the nucleation and the growth mechanism of MnBi system the current transients are compared to the Scharifker and Hills (SH) tridimensional theoretical model [28]. In this model tow cases can be mentioned: the instantaneous nucleation and the progressive nucleation. Instantaneous nucleation indicates that the nuclei formed instantaneously on the electrode surface, ie all sites are activated simultaneously. Progressive nucleation corresponds to gradually formation of nuclei on the active sites (the nucleation sites are activated progressively). The expressions for instantaneous and progressive nucleation are given by Eqs.(1) and (2) respectively.
i i max
i i max
2 t 1.9542 1 exp 1.2564 t t max t max
2 t 1.2254 1 exp 2.3367 t t max t max
2
2
Eq. (1)
2
Eq.(2)
Where Imax is the maximum current density and tmax is the time corresponding to Imax.
Figure 5 (a,b) shows the experimental non-dimensional plots (I/Imax)2 vs (t/tmax) compared to the theoretical model curves. The analysis of the experimental current transients according to the (S-H) model shows a good agreement with theoretical curves for instantaneous nucleation up to tmax. After tmax, the curves shifts from the instantaneous theoretical model and exhibit a higher current at longer times, this deviation can be attributed simultaneously to the partial kinetic control of the growth and to the hydrogen evolution at this range of potentials, this behavior was already reported in the literature [29-31]. In a previous study Boudinar and Co.[32] show that the nucleation and growth mechanism of Mn-Bi co-deposited in a sulfatenitrate bath on a copper foil substrate takes place in a tridimensional progressive nucleation and shifts to the instantaneous nucleation at more negative potentials when the hydrogen evolution became important. On the other hand, the electrochemical nucleation mechanism of Bi is very sensitive to the applied potential and to the ions concentration, at lower potentials and lower concentrations the nucleation mechanism of Bi can be described by a 3D progressive nucleation and shifts to the tridimensional instantaneous nucleation with increasing the Bi3+ concentration and the applied potential as reported by several authors [31,33-34]. It should be noted also that at these range of negative potentials, the electrochemical mechanism process of Mn-Bi may deviate from the diffusion control because of the presence of many other electrochemical and chemical reactions such as:
2𝐻2 𝑂 + 2𝑒 → 𝐻2 + 2(𝑂𝐻)− 2𝐻+ + 2𝑒 → 𝐻2 𝑀𝑛2+ + 2(𝑂𝐻)− → 𝑀𝑛(𝑂𝐻)2
Eq. (3)
Eq. (4)
Eq. (5)
Indeed, the strong hydrogen evolution in the investigated range of potential (eq.3 and Eq.4) has caused a variation in the interfacial pH which may generate the formation of hydroxides (eq.5). This phenomenon has significantly affected the nucleation process.
At very short time according to Scharifker and Hills model the current as function of time can be described for the instantaneous nucleation by the following relation ;
i (t) = zFc1/2 D3/2 N0 (
8πcM −1/2 1/2 ρ
)
t
Eq (6)
Where z corresponds to the number of electrons involved in the electrochemical reaction, F the Faraday’s constant, c is the ions concentration in the electrolyte, D the diffusion coefficient, N0 the nuclei density, M the molecular weight and ρ the density. So another diagnostic can be performed to characterize the nucleation mode in the early stage of electrodeposition by the representation of I vs. t1/2 given by the equation (eq. 6) above. The plot of (-I) vs. t1/2 (Figure 6-a and b) shows a good degree of linearity which confirms that nuclei are formed instantaneously on the electrode surface at first deposition time. From the imax and tmax values, the diffusion coefficient and nuclei density can be calculated [28]. For the instantaneous nucleation mode, D and N0 can be expressed as follows:
i2
t
max max D = 0.1629(zFc) 2
Eq. (7)
and
Eq. (8)
8πcM −1/2
N0 = 0.065 (
ρ
)
zFc
(
imax tmax
)
2
The basic requirement in the analysis of transients according to the relations for a three dimensional nucleation is that the product i2m tm remains constant, or does not significantly change with the applied potential [28]. The experimental data deduced from the analysis of the transients in this work and shown in table 1satisfied theses conditions well. We can also note that the average value of the diffusion coefficient increases from 1.8 10-6 cm2 s-1 for the Mn-Bi deposited on Cu to 2.8 10-6 cm2 s-1 for the Mn-Bi deposited on Bi/Cu. Unfortunately estimation of the nuclei density was not possible as the density of the deposit is not known at this early stage of deposition. However, from the equation (Eq 8) the produce imax2.tmax2 can be expressed as a function of (1/N0). As it is shown in table 1, imax2.tmax2 decreases with increasing the over potential; this means that the nuclei density increases with the over potential. Furthermore if we compare between the two depositions modes, we observe that the nuclei density is lower in the Mn-Bi/Bi/Cu films. This is in agreement with the discussion above related to the decrease in the number of the active sites in the bilayer (section 3.1). In conclusion, although the non dimensional plots do not allow to discern between the nucleation in the two deposition methods (especially after tmax); the difference in the kinetic parameters between the two depositions modes at the early stage of deposition will certainly induce a difference in the thin films morphologies, this will be discussed in the section below.
3.3 Morphological characterization before annealing treatment In this section we present the morphological analysis carried out on the deposited samples under different conditions. First of all in order to complete the electrocristallization study, preliminary SEM characterizations at the early stage of the MnBi thin layers and MnBi /Bi
bilayers deposition are performed during 1s. We present as example the SEM images of the thinner deposits obtained at the deposition potential of -1.65V vs SCE (Fig.7). As we can see, the copper substrate is completely covered by the thin deposit even at the very short deposition time; this confirms the instantaneous character of the electroctrocristallization process at the early stage of deposition. Furthermore, we note also a net difference in the two layers morphologies at a very short deposition time. The same behavior was observed in the different applied potentials. As the deposition time increases to 100s (Fig.8-9-10), different morphologies were achieved depending on the deposition potential and the plating mode: for a less cathodic potential (- 0.3 V vs SCE) (Fig.8), the deposit presents a dendritic aspect composed of Bi alone according to the EDS analysis (EDS spectrum not shown). Its nucleation occurs in three dimensions. One can also note that the substrate surface is not completely covered, indicating that bismuth grows preferentially on initial Bi germs than on the copper substrate. Similar results were reported in various works [25-27]. At less negative potential (-1.65 and -1.75 Vvs SCE), the deposits show a completely different aspect (Fig.9), SEM picture 1 and 2). The two films have heterogeneous appearance consisting of large Mn crystallites of about 1 µm in diameter on which are incorporated a spongy deposit of bismuth. Surface substrates seem more covered with the manganese deposit (especially as the deposit potential is more electronegative) indicating that manganese crystallizes more easily on copper. The cross section SEM analysis (picture 3 and 4) shows the Cu/Ta/Si interface on which is incorporated the Mn-Bi deposit. The average thickness of the crystallites increases from 1.4 m to 2.1m as the potential is more electronegative. The SEM-FEG analysis of the film surface (Fig. 9-b picture 1) and the corresponding cross section analysis (Fig. 9-b picture 2) highlights clearly the details of the heterogeneous deposit. The bismuth nucleation seems to be affected by the electrodeposition potential as the dendritic
aspect is not observed, this may be attributed to hydrogen incorporation in the deposit at this range of potential. The heterogeneous chemical composition of the deposit is evidenced by EDS analysis (EDS spectrum not shown) where two regions are distinguished: a clear one with Bi the predominant species and a grey one with Mn the predominant species. Fig.10 presents the morphological aspect of Mn-Bi/Bi bilayer deposited at -0.3 V vs SCE (during 100s) and -1.65 V vs SCE (during 100s) respectively. As we can see, the deposit obtained seems now to be distributed more uniformly on the surface, covering entirely the copper substrate. The manganese crystallite grain size decreased considerably to around 0.3 µm in diameter, while the bismuth deposit occurs with a dendritic form (Fig.10 picture 1 and 2). The cross section SEM-FEG analysis show more details of the deposit morphology (Fig. 10, picture 3 and 4). As we can see we distinguish three regions: the first one just at the interface with the copper consists of very fine and discontinuous bismuth grains (the corresponding EDS spectrum not shown), the second one with a columnar shape (and a thickness of about 450 nm) is composed of manganese and the third one with a dendritic shape (and a thickness about 727 nm) is composed of bismuth.. This particular morphology can be explained as follows: during the first impulsion at -0.3 V vs SCE, bismuth is deposited in a dendritic form by leaving empty areas on the copper substrate. When the second impulsion at -1.65 V vs SCE is applied manganese accompanied with bismuth are deposited simultaneously and heterogeneously, manganese will occupy the empty area left by the first Bi deposit on the copper substrate, instead of spreading out horizontally as in the case of Figure 9, it will spread vertically giving finally this columnar form, while the bismuth will deposit on the first dendritic bismuth grains and will grow perpendicularly to the surface to exceed the manganese layer. This interpretation is coherent with the electrochemical analysis performed in the section 3.1 above.
One may note here that whatever the deposition mode the obtained deposit thicknesses are in the range of the micrometer which means that the deposition efficiency is not so low.
3.4 Morphological characterization after annealing treatment In order to induce a possible inter-diffusion between Mn and Bi phases, annealing was performed under high vacuum (10-6 Torr) during 1 hour for both thin Mn-Bi layers and MnBi/Bi bilayers. We present in this section the SEM observations obtained for a 300 °C annealing temperature. The different images of the thin Mn-Bi layers and the corresponding EDS analyzes of each region are shown on Fig.11-a, 11-b and 11-c respectively. Figure 11-a indicates that the morphology of the deposit has changed with annealing: although manganese crystallites are not observable, the heterogeneous appearance of the deposit is still visible notably with the coalescence of bismuth grains giving circular nodules with an average diameter of 1.3 µm. It should be noted here that the annealing temperature (300°C) is higher than the melting temperature of bismuth, so it was already in a liquid state during the annealing treatment(performed for 1 hour). The circular nodules observed on the deposit result then probably from cooled bismuth droplets. The SEM images presented on Fig.12 show also a significant and more effect of annealing on the Mn-Bi/Bi bilayers morphology. We note that, the deposit has a uniform appearance along the entire surface. In this case, the grains have a strong coalescence comparatively to those not annealed. We note also, the absence of bismuth dendrite on the surface of the deposit. This behavior may be probably due to a better interdiffusion reaction between manganese and bismuth elements, which induces a homogeneous deposit with sufficiently reduced grain size.
3.5 X-rays analysis
Figures (13,14) display the XRD patterns of the Bi, Mn-Bi thin layers and Mn-Bi /Bi bilayers. The obtained results show that the crystal orientations change considerably according to the imposed potential and the deposition mode. For a deposition potential of -0.3VvsSCE, as shown on Fig.13, the spectrum (b) reveals the characteristic peak of bismuth in its rhombohedral structure. Fig.14 shows the XRD pattern of the Mn-Bi thin layer and Mn-Bi/Bi bilayers obtained at various potential before annealing. In the case of the Mn-Bi thin layers and for deposition potential of -1.65 V vs. SCE and -1,75 V vs. SCE (Fig.14, spectrum b and c), the XRD spectra reveal the presence of the (012) bismuth peaks in its rhombohedral structure and the characteristic peaks of manganese phases αMn(111) (body centred cubic system) and -Mn (411) (body centred tetragonal system). The spectrum d of the same figure 14 shows the XRD patterns of the Mn-Bi /Bi bilayer, as we can see, only the peaks characteristic of bismuth, textured in the (012) direction are present. No diffraction peaks of manganese were observed. This behavior may be explained by the absence of Mn crystallization on Bi element as it was the case on copper. We note also that no phase of MnBi alloy was observed in the three spectra. These results can be explained by the absence of the interdiffusion reaction between the two elements at room temperature. After the annealing treatment at different temperatures (100°C- 300°C), no diffraction peaks were observed until a temperature of 300°C. The corresponding XRD patterns are shown on the figure 15. Concerning the thin layer deposits, the obtained results display many new peaks (Fig.15.a and b) attributed especially to the ternary Mn3Bi4Cu4 alloy (located at 2θ = 25.36° and 29.5°) in its FCC phase [35]. The presence of copper element in this phase is due to its interdiffusion from the substrate during annealing. The other peaks are attributed to manganese oxide ( at 2θ =35° and 40.5°) and to bismuth oxides (at 2θ = 27.98°, 72.75°, 36.90°, 46.28°, 47.9° and 48.74°) . They are probably due of the atmospheric oxygen during the samples transfer.
In the case of the bilayers (Fig.15.c), the spectrum obtained after annealing reveals also new diffraction peaks, located at 2θ = 25.36°, 29.5° and assigned to the Mn3Bi4Cu4 phase in the FCC structure. The peaks related to bismuth have completely disappeared. In addition, minor peaks of manganese were observed after annealing. It is important to note that some peaks, located at 2θ = 35°and 41°, related to manganese oxide were also observed in this annealed bilayer, but their intensities are lower than those observed in the thin layers. Otherwise, no peaks related to the bismuth oxide were observed.
3.6 Magnetic characterization In this study, preliminary magnetic characterizations using a SQUID magnetometer were also investigated at room temperature on the Mn-Bi thin layers and the Mn-Bi/Bi bilayers annealed at 300°C under vacuum condition. Before any discussion one may note here that in this MnBi system, the atomic magnetic moment is carried with manganese element (which is antiferromagnetic in its alpha or gamma cubic phase), whereas the bismuth is diamagnetic but as it crystallizes in an uniaxial structure (hexagonal or rhombohedral) it is responsible in the appearance of magnetic anisotropy in the hexagonal MnBi phase. The magnetic measurements obtained for the magnetic field applied parallel to the film plane and for the deposits annealed at low temperatures (T<300°C) didn’t show any ferromagnetic signal, we just observed the diamagnetic signal of the copper substrate. For an annealing temperature of 300°C a ferromagnetic signal appears, the corresponding magnetization curves are reported on figure 16. All the hysteresis curves showed the presence of a low magnetic signal with characteristics typical to the one of a hard magnetic material. However the values of the coercive field (around 300-400 Oersted) remains lower than that required for a hard magnetic material as MnBi. This behavior is related to the incorporation of the copper from
the substrate in the annealed deposit. Indeed, in a former work [36] Y. Chen et al. have reported on the properties of the MnBiY (Y = Al, Ag, Au, Cu, In, Zn, Pt) thin films deposited by evaporation under vacuum condition onto a glass substrate and annealed to 300-400 °C. It was found that Ag, Cu and In are reactive metals with small size, so they can easily penetrate into the MnBi layer and form a MnBiY alloy with a change of the NiAs type symmetry. This modification in the crystal structure of the MnBi alloy induces a great modification in its magnetic properties. In another work, J.Chen et al. [37] have reported that the magnetic properties of the MnxBiCu films can be improved by varying the Mn concentration in the film. The perpendicular magnetic anisotropy, the remanence, the squareness and the coercivity increased with increasing Mn concentration in the MnxBiCu alloy and these values tend to saturate at x=2. The other important point observed on Fig.16 concerns the difference in the magnitude of the saturation magnetization between the thin layers and the bilayers. Indeed the saturation magnetization is lower in the bilayer elaborated at -0.3V vs SCE(100s) and -1.65 Vvs SCE(100s) (curve a) than in the thin layer elaborated at -1.65 V vs SCE (100s) (curve b) while the XRD data (Fig.15) indicate almost the same peak intensities of the MnBiCu alloy in the different deposits. This is probably related to the fact that the amount of non-reacted manganese is lower and more diluted in the bilayer than in the thin layer even if the deposition potential is the same, as it was already mentioned above in the chronoamperometric analysis (Fig.3). These non-reacted manganese atoms contribute to the total magnetic moment. Increasing the cathodic potential to -1.75V v SCE in the thin layer induces an increase in the Mn concentration in the film and then an increase in the saturation magnetization (curve c). In a previous work, Terzief et al.[38] who studied the magnetic behavior of diluted Mn atoms in liquid bismuth have reported that the magnetic moments per
atom of Mn, turn out to be nearly independent of the concentration of the dilute atoms but the magnetic susceptibility of the alloys increases with increasing manganese concentration. Note that, the preliminary magnetic measurements reported in this work for the first time on a Mn-Bi electrodeposited system has not been fully investigated. The electrodeposition of the MnBi alloy and the study of its magnetic properties still remains an area of active research in order to optimize the plating parameters giving the best magnetic properties (strong coercivity, very large Kerr effect).
4 Conclusions In summary, this study showed that the initial stage of the Mn-Bi electroplated in a chloride bath is governed by a three dimensional instantaneous nucleation whatever the deposition mode and the applied potential. However, the morphological and the structural properties of the electrodeposited thin films are strongly dependent on the applied potential and the deposition mode. Indeed, on the thin Mn-Bi layers and for a less cathodic potential, only the bismuth element is present in the deposit with a dendritic morphological aspect in the rhombohedral structure. For more cathodic potential (E ≤ -1.65 V vs SCE), the electrodeposited films contain large crystallites of manganese on which grows a spongy deposit of Bi. The annealing treatment performed at 300°C under vacuum conditions has induced a little interdiffusion between manganese and bismuth with the formation of a Mn3Bi4Cu4 alloy and has also provoked a big coalescence of bismuth giving circular nodules with an average diameter of 1,3 µm. In the case of the Mn-Bi/Bi bilayers, the morphological analysis revealed a lower grain size compared to those obtained in the thin layers. Before annealing only the characteristic peaks
of bismuth are present. After annealing at 300°C the deposit exhibit a very compact and homogeneous surface without cracks and voids. Mixed Mn3Bi4Cu4 alloy was also observed. A coercivity around 300-400 Oe has been measured in the different deposits after annealing.
Acknowledgments The authors thank Claire Meyer (InstitutNéel - CNRS Grenoble) for SQUID measurements and Amirouche Saifi (UMMTO) for SEM observations.
References [1]E. I. Cooper, C.
Bonhote, J. Heidmann, Y. Hsu, P. Kern, J. W. Lam,
M.Ramasubramanian, N. Robertson, L. T. Romankiw and H. Xu, IBM J. Res. & Dev. Vol. 49 N° 1 Jan. (2005) 103 [2] Y. Dahmane, L.Cagnon, J.Voiron, S.Pairis,M.Bacia, L.Ortega, N.Benbrahim and A.Kadri, J. Phys.D. App. Phys.39 (2006) 4523 [3] L. Cagnon, Y. Dahmane, J. Voiron, S. Pairis, M. Bacia, L. Ortega, N. Benbrahim, A. Kadri, Jour. of Magn. andMagn. Mat. 310 (2007) 2428
[4] U. Rudiger, H. Berndt, A, Schirmeisen, P.Fumagalli and G.Guntheradt, J. Appl. Phys, 78 (9) (1995) 5391 T. Zhuand, Y. J. Wang, J. Phys. D: Appl. Phys. 32(1999) 2609 [5] S. Saha, R. T. Obermyer, B. j. Zande, V. K. Chandhok, S.Simizu and S.G.Sankar, J. Appl. Phys, Volume 91, N° 10 (2002) 8525 [6] K. Kang, L. H. Lewis, Y.F. Hu, Oiang Li , A.R. Moodenbaugh and Young-Suk Choi, J. Appl. Phys.99 (2006) 08N703-1 [7] M. Hanson, Constitution of Binary Alloys (Mc Graw-Hill, New (1986). York, 1958), p. 318.
[8] T. Chen, J. Appl. Phys. 45, 2358 (1974)] [9] K. Kang, A.R. Moodenbaugh and L. H. Lewis, Appl.Phys.Lett.90, (2007) 153112-1 [10] B. Li ,W. Liu , X. G. Zhao , W.J.Gong , X.T. Zhao , H.L.Wang , D. Kim , C.J. Choi and Z.D. Zhang, J. Magn. Magn. Mat. 372(2014)12 [11]A. Brenner,“Electrodeposition of alloys”Vol.1 Academic Press, New York (1963). [12] M. Cortes, A. Serra, E. Gomez1and E. Valles, ElectrochimicaActa 56 (2011) 8232 [13] Fernando M.F. Rhen, J.M.D. Coey, J. Magn. Magn. Mater (2004) 272 [14]Min-Suk Suh, Chan-Jin Park and Hyuk-Sang Kwon, Surface & Coatings Technology 200 (2006) 3527 [15] Yong X.Gan,James Sweetman, Joseph G.Lawrence, Mater. Lett. 64 (2010) 449 [16] P. Diaz-Arista, R. Antano-Lopez, Y.Meas, R.Ortega, E.Chainet, P.Ozil and G.Trejo, Electrochem. Acta, 51 (2006) 4393 [17] C. Savall, C. Rebere, D. Sylla, M. Gadouleau, Ph. Refait and J. Creus, Mater Sci. Eng. A 430 (2006) 165 [18] J. Gonga, G. Zangari, J. Electro. Soc.149 (2002) C 209 [19]B. Benfedda, N. Benbrahim, A. Kadri, E. Chainet, F. Charlot and S.Coindeau, Electrochim. Acta 56 (2011) 1275 [20] R.S.Dean. « Electrolytic Manganese and its Alloys » , The Ronald Press Co, New York (1952) [21] P. Diaz-Arista, G. Trejo, Surf.Coat. Technol. 201 (2006) 3359 [22] P. Wei, O. E. Hileman Jr, M. RezaBateni, X. Deng, A. Petric, Surf. Coat. Technol. 201 (2007) 7739 [23] J. Wei Wu, D. Christopher Johnson, Y. Jiang, Randall, S. Gemmen, X. Liu, Electrochem. Acta 54 (2008) 793 [24] S. Sandnes, M. E. Williams, U. Bertocci, M. D. Vaudin and G.R. Stafford, Electrochim. Acta 52 (2007) 6221
[25] S. Jiang, Y. H. Huang, F. Luo, N. Du and C. H. Yan, Inorg. Chem. Commun. 6 (2003) 781 [26] S .I. Cordoba de Torresi; I. A. Carlos, J. Electroanal. Chem. 414 (1996) 11 [27] H. Sato, T. Homma, H. Kudo, T. Izumi, T. Osaka and S. Shoji, J. Electroanal. Chem. 584 (2005) 28 [28] B. Scharifker, G. Hills, Electrochim. Acta. 28, (1983) 879-889. [29] D. Grujicic and B. Pesic, Electrochim. Acta. 47, (2002) 2901-20912. [30] C. Han, Q. Liu, and D. G. Ivey, Electrochim. Acta. 54 (2009) 3419 -3427. [31] S.B. Sadale, P.S. Patil, Solid State Ionics. 167, (2004) 273-283. [32] S. Boudinar,N. Benbrahim, B. Benfedda, A. Kadri,aE. Chainet, L. Hamadou, J.Electrochem.Soc, 161 (5) (2014) D227-D234. [33] M. Yang, Z.Hu, J. Electroanal. Chem. 583, (2005) 46-55. [34] N.Vladislavic, S.Brinic, Z. Grubac, M. Buzul, J.Electrochem.Sci, 9 (2014) 6020-6032. [35] G. Bryan Street, J.C. Suits and Kenneth Lee, Solid State Communications, Vol. 14 (1974) 33-36. A. Szytula and H. Biriczycka, Solid State Communications, Vol. 38, (1981) 41-43. [36] Y. Chen, Y.C.P Luo, Z.T.Guan, Q.Y. Lu, Y.J. Magn. Magn. Mat.115 (1992) 55 [38] J. Chen, K. Wierman, R. D. Kirby, and D. J. Sellmyer, IEEE Trans. Magn. vol. 31, no. 6 (1995) 3334 [39] P. Terzieff, S. Uporov, V. Sidorov ; A.Mikula, J. Magn. Magn. Mater. 321 (2009) 3747
Figures captions Figure 1 a) Cyclic voltammetry on a vitreous RDE electrode in a bismuth chloride bath: 10-2 M BiCl3, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 b) Cyclic voltammetry on a vitreous RDE electrode in a manganese chloride bath: 0,4 M MnCl2, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 Figure 1-a 14
(a) 12 10 8
j / mA.cm
-2
6 4 2 0 -2 -4 -6 -0,4
-0,2
0,0
0,2
0,4
E / V.vs SCE
Figure 1-b 10
(b)
5
j / mA.cm
-2
0
-5
-10
-15
-20 -2,0
-1,5
-1,0
-0,5
E / V.vs SCE
0,0
0,5
Figure 2: Cyclic voltammetry on a vitreous RDE electrode in 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 Figure 2 30
Bi dissolution 20
B
Mn dissolution A
10
j/ mA.cm
-2
0
-10
-20
-30
-40 -1,5
-1,0
-0,5
E/ V.vs SCE
0,0
Figure 3: Current time transients for : (a) a thin Mn-Bi layer deposited at -1.65Vvs.SCE and (b) Mn-Bi/Bi bilayer deposited at EBi = -0.3Vvs.SCE and EMn = -1.65Vvs.SCE. Figure 3-a
-0,09
(a) -0,10 -0,11
j / mA.cm
-2
-0,12 -0,13 -0,14 -0,15 -0,16 -0,17 0
20
40
60
80
Time (s)
Figure 3-b
(b)
0,00
-0,02
j/ mA.cm
-2
-0,04
-0,06
-0,08
-0,10
-0,12
-0,14 0
50
100
Time (s)
150
200
100
Figure 4: Chronoamperometric transients at early stage of deposition for Mn-Bi electrodeposited on (a) Cu/Ta/Si and (b) Bi/Cu/Ta/Si. Figure 4.a
Figure 4-b
Figure 5: Non dimensional plots (I /Imax)2 vs (t /tmax) of the chronoamperometric curves in the Fig.4: a) Mn-Bi on Cu/Ta/Si, b) Mn-Bi on Bi/Cu/Ta/Si Figure 5-a
Figure 5-b
Figure 6: (-I ) vs.t1/2 plots for the initial stages of Mn-Bi electrodeposited on (a) Cu/Ta/Si and (b) Bi/Cu/Ta/Si. Figure 6-a
Figure 6-b
Figure 7: SEM images of MnBi thin layer (a) and MnBi/Bi bilayer (b) at early stage of deposition (deposition time: 1 s, deposition potential: -1,65 V vs SCE). Figure 7
a)
b)
Figure 8: SEM-FEG of Bi thin layer electrodeposited at E = -0.3 Vvs.SCE from acidic chloride bath, pH = 2, 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl
Figure 8
Figure 9 a) SEM images of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE (picture 1) and E= -1.75 V vs.SCE (picture 2) and the corresponding cross section observations (picture 3 and 4 respectively) b) SEM-FEG images of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE (picture 1) and the corresponding cross section analysis (picture 2). Figure 9-a 1
3
Thin layer: E= - 1.65V
2
Thin layer: E= - 1.75V
4
Figure 9-b
1
Mn
2
Bi Mn Cu
Ta
Si
Cu
Figure 10: a) SEM images of Mn-Bi/Bi bilayer (picture 1 and 2) and the corresponding SEMFEG cross section analysis (picture 3 and 4) electrodeposited at EBi = -0.3 V vs. SCE; EMn = 1.65 V vs. SCE from acidic chloride bath, pH = 2, 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl. Figure 10
1)
3) 3 2 Cu
Ta Si
2)
4)
1
Figure 11 a) SEM images of Mn-Bi thin layers electrodeposited from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions. b,c) EDS spectra of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE; from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions, pH = 2, 0.4M MnCl2, 10-2 M BiCl3, 3M NH4Cl, recorded in a white region (b) and a grey region (c) on the deposit
Figure 11-a
Figure 11-b
White region
Figure 11-c Grey region
Figure 12: SEM images of Mn-Bi/Bi bilayers electrodeposited from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions.
Figure 12
A)
Figure13: XRD analysis patterns on Bi thin layer obtained at E = -0.3Vvs.SCE, 0.4M MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si substrate, pH = 2 (a) Substrate, (b) Bi thin layer Figure 13 8000
.
substrat
.
4000
Bi (202)
Bi (104)
Bi (012)
Intensity
6000
(b)
. 2000
(a)
20
25
30
35 2 (Deg.)
40
45
50
Figure 14: XRD analysis patterns on Mn-Bi thin layer and on Mn-Bi/Bi bilayers before 0.4MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si
annealing obtained respectively in
substrate, pH = 2((a) Substrate, (b) Mn-Bi thin layer at -1,65V vs.SCE, (c) Mn-Bi thin layer at -1,75V vs.SCE, (d) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE)
Bi(012)
2500
Bi(110)
Intensity
2000
gamma Mn(200) or alpha Mn(332)
Bi(104)
gamma Mn(111) or alpha Mn(411)
Figure 14
1500
1000 (d)
(c) 500
(b) (a)
0 20
25
30
35
2 (Deg.)
40
45
50
Figure 15: XRD analysis patterns on Mn-Bi thin layers and Mn-Bi/Bi bilayers after annealing at 300°C for 1 hour under vacuum conditions, 0.4MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si substrate, pH = 2, (a) Mn-Bi thin layer at -1,65V vs.SCE, (b) MnBi thin layer at 1,75V vs.SCE, (c) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE) Figure 15
.
MnO2 * Bi203 + Bi4Cu4Mn3
gamma Mn(111) or alpha Mn (411)
-
Bi
.
-
3000
*
+
1000
* (a) 20
+
25
+
30
* * ** 35
(Deg.)
*
*
*
*
-
*
+
-
(b)
+
-
+
-
(c)
-
Intensity
2000
40
45
50
Figure 16: Magnetization curves obtained at room temperature of the Mn-Bi/Bi bilayers and Mn-Bi thin layers annealed at 300°C under vacuum conditions, (a) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE ; (b) Mn-Bi thin layer at -1,65V vs.SCE; (c) MnBi thin layer at 1,75V vs.SCE. Figure 16
0,0008 0,0006
(a) (b) (c)
0,0004
M / emu.cm
-3
0,0002 0,0000 -0,0002 -0,0004 -0,0006 -0,0008 -1500
-1000
-500
0
H / Oe
500
1000
1500
Table1. Electrochemical parameters obtained according to Scharifker and Hills model (t0 is the induction time). t0 (s)
3
2
2
-6
-imax
Mn-Bi/ Cu/Si
-1.65
26.20
4.69
0.21
4.48
3.07
13.78
2.01
-1.7
33.40
2.38
0.13
2.25
2.50
5.63
1.63
-1.75
41.30
1.74
0.06
1.68
2.87
4.83
1.87
-1.65
21.90
10.03
1.02
9.01
4.32
38.97
2.82
-1.7
27.00
6.71
1.09
5.62
4.09
22.99
2.67
-1.75
40.60
3.52
0.65
2.87
4.61
13.23
3.01
mA
tmax- t0
2
-E (V/SCE)
Mn-Bi/Bi/ Cu/Si
tmax (s)
3
Electrodeposition mode
(s)
10 .imax .(tmax- t0) 2
(mA .s)
10 .imax .(tmax- t0) (mA.s)
2
10 .DInst 2
-1
(cm .s )
S0013-4686(16)31040-4 http://dx.doi.org/doi:10.1016/j.electacta.2016.05.007 EA 27223
To appear in:
Electrochimica Acta
Received date: Revised date: Accepted date:
27-12-2015 1-5-2016 2-5-2016
Please cite this article as: B.Benfedda, N.Benbrahim, S.Boudinar, A.Kadri, E.Chainet, F.Charlot, S.Coindeau, Y.Dahmane, L.Hamadou, Morphological, physicochemical and magnetic characterization of electrodeposited Mn-Bi and Mn-Bi/Bi thin films on Cu Substrate, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2016.05.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Morphological,physicochemical and magnetic characterization electrodeposited Mn-Bi and Mn-Bi/Bi thin films on Cu Substrate B. Benfedda(1)*, N. Benbrahim (1), S.Boudinar (1 ), A. Kadri (1), E. Chainet (2), F.Charlot Coindeau (3), Y. Dahmane(4), and L. Hamadou(1)
of (3 )
, S.
1 Laboratoire de Physique et Chimie des Matériaux (LPCM), Université M. Mammeri de Tizi- Ouzou (Algérie) 2 Laboratoire d’Electrochimie et de Physico-chimie des Matériaux et des Interfaces (LEPMI), UMR 5631 CNRS-INPG-UJF, BP 75, 38402 Saint-Martin d’Heres Cedex, France 3 C.M.T.C., Grenoble I.N.P. Bât Phelma Campus, Domaine Universitaire, 38402 Saint Martin d’Hères, Grenoble, France 4SPINTEC/CEA Grenoble 17 rue des Martyrs, 38054 GRENOBLE Cedex 9
Corresponding author: BayaBenfedda, E-Mail: [email protected]
Abstract Mn-Bi thin films were electroplated on Cu (111) substrates in an acidic chloride bath. In order to determine the deposition potential of each element, cyclic voltammetry using a rotating disk electrode was performed. Two types of thin films were obtained using two deposition mode: the first one called thin Mn-Bi layers by using a single applied potential and the second one called Mn-Bi/Bi bilayers by using a double pulse potential. Annealing treatments at 300°C for 1 hour under vacuum condition were carried out in order to cause an interdiffusion between manganese and bismuth. The morphological and crystalline structure of the various deposits was investigated by scanning electron microscopy with field effect (SEM-FEG) and by X-ray diffraction analysis (XRD). Magnetic characterizations were also made using a superconducting quantum interference device (SQUID) magnetometer. The morphological and the structural properties of the thin layers and the bilayers are completely different, indicating that the growth process changes according to the plating mode. After annealing a mixed MnBiCu phase with a coercivity of 300 Oe and 400 Oe was observed on the thin layers and the bilayers respectively.
Key words: Mn-Bi, Hard magnetic material, Electrodeposition, Voltammetry, SEM-FEG, EDS, XRD
1. Introduction Because of their many technological applications, the magnetic materials have a considerable economic importance on a worldwide scale. These last years, many studies have been reported and currently activated in order to synthesize new materials able to meet increasingly powerful needs. In this context, various magnetic materials, such as, NiFe, CoFe, CoPt, FePt, MnBi….etc, have been the subject of several works [1-4] due to their potential application in the magnetic recording field. The low-temperature phase (LTP) of MnBi which crystallizes in the Ni-As type hexagonal crystal structure with c axis as the easy direction of magnetization, exhibit a high uniaxial magnetic anisotropy (2×107 erg/cm3) making it very attractive for high-density magnetic recording. In addition, the LTP MnBi is known to show a high polar Kerr rotation at room temperature, which makes it attractive also for the magnetooptical (MO) perpendicular recording application [4-6]. But although MnBi has potential applications in the high density recording field, it is very difficult to obtain high quality MnBi by conventional means, because, according to the Mn-Bi phase diagram [7,8], at the peritectic temperature (719 K) Mn tends to segregate from the Mn-Bi liquid. On the other hand, if the temperature is lowered to below that of the eutectic, the reaction becomes very slow. Generally, works cited in the literature reported on Mn-Bi layers prepared by physical techniques such as melt spinning, magnetron sputtering and vacuum evaporation [4-6, 9]. However, these methods remain expensive as they require heavy equipment and are difficult to implement. In the present study, we chose the electrochemical way which is an important
processing technology for micro fabrication due to its low cost, high yield, low energy requirement, and capability for generating high aspect ratio features. Many works using the electrochemical synthesis methods were already carried out, in particular for the elaboration of different materials such as NiFe, FeCo, CoPt, SnBi , TeBi… etc [11-15], to the best of our knowledge, no work have been reported on the synthesis of the MnBi alloy by the electrochemical way. The electrodeposition of Mn-Bi by electrochemical method is a very delicate process, considering on one hand, the complexity of the manganese deposition kinetics [16-19] and on the other hand, the large difference between the potential related to the Mn2+/Mn and Bi3+/Bi couples. In addition, the process of metals and alloys electrodeposition depends on several parameters which must be controlled, in particular, pH; temperature, concentration of metal ions to be deposited, deposition potential, additives, plating mode… etc. In a previous work [19], we have described preliminary study of principal conditions to deposit Mn-Bi system in chloride bath containing ammonium chloride as additive. In this respect, cyclic voltammetry has been used to determine the nature of the various reactional stages occurring during the electrodeposition of Mn and Bi as a function of ammonium chloride concentration. The thermodynamic study of the electrolysis bath was also carried out to identify the electroactive species. Various characterizations (structural, morphological) were also performed on the MnBi deposited system. In the present work, a series of Mn-Bi thin layers and Mn-Bi/Bi bilayers have been grown on (111) textured Cu substrate using single and double pulse applied potential. It is well known that the initial stage of electrocrystallization, including the nucleus formation and their growth mechanism, directly determines the properties and quality of the thin electrodeposited layer. So, we have studied in a first time the initial stage of the Mn-Bi electrocrystallization on the Cu substrate and on an electrodeposited Bi on the Cu substrate. In a second time, we aimed to
investigate and correlate the structural, morphological and magnetic properties of the electroplated thin films with the plating mode and the annealing treatment. A chloride based electrolysis bath was used. The chosen bath is based so on the studies reported in literature on manganese and bismuth metal electrodeposition. So, manganese deposition can be obtained via manganese sulfate and chloride baths with the corresponding ammonium salts [20-23], whereas bismuth deposition is often synthesized in nitrates and chlorides baths [24-27].
2. Experimental details A vitreous carbon rotating disk electrode (RDE) (0.2 cm2) with a rotation speed fixed at 250 rpm and either Cu/Ta/Si substrates (0.5 cm2) were used as working electrodes in a classical three electrodes electrochemical cell. The counter electrode was a platinum wire immersed in a separate compartment containing solution without electroactive metallic cations. All potential values were measured with respect to a saturated calomel reference electrode (SCE). All experiments were carried out at ambient temperature in a chloride bath deoxygenated by nitrogen gas bubbling. The electrochemical Mn-Bi bath consisted of MnCl2 4H2O (0.4 mol.L1
), BiCl3 (10-2mol.L-1) and NH4Cl (3mol.L-1). The pH was adjusted to 2 by adding HCl
solution. The solution was prepared immediately prior to each experiment using deionised water and analytical grade reagents (Aldrich). The MnBi various deposits were prepared with continuous stirring of the electrolyte solution. The
electrochemical
measurements
were
performed
using
an
EG&G
273A
Potentiostat/Galvanostat controlled by a microcomputer via GPIB interface operated by M352 EG&G software. The composition of the samples was determined by energy dispersive X-ray spectroscopy (EDX). The morphology was examined by scanning electron microscopy (SEM) and the scanning electron microscopy with field effect (SEM-FEG). The crystalline structure was investigated by X-ray diffraction (XRD) with CuK radiation. Magnetic properties have
been also characterized at room temperature using a superconducting quantum interference device (SQUID) magnetometer.
3. Results and discussions 3.1 Electrochemical characterization The voltammetric study was first carried out in two different baths containing each of the two elements separately on a RDE electrode. Figure 1-a shows the voltammogram recorded on a bismuth chloride bath (10-2 mol.L-1 of BiCl3 and 3 mol.L-1 of ammonium chloride NH4Cl, pH = 2). Starting from the rest potential (-0.2 V vs SCE) and sweeping towards the cathodic direction, a plateau of a weak current is observed in the potential range (-0.23 V vs SCE to 0.5 V vs SCE), it is the limiting current region where Bi (III) reduction occurs under mass transport control. In the reverse scan, one anodic peak appears at around -0.2 V vs SCE. It is attributed to the bismuth dissolution according to a more detailed previous study in the same bath [17]. Figure 1-b shows another voltammogram recorded in a manganese chloride bath (0,4 mol.L-1 of MnCl2 and 3 mol.L-1 of ammonium chloride NH4Cl, pH = 2). Starting from the rest potential (-0.8 V vs SCE) and scanning towards the cathodic direction, a low current density is observed. This step is attributed to H+ reduction. Up to -1.3 V vs SCE, a weak reduction plateau followed by a rapid increase in the current density at (-1.57 V vs SCE ), corresponding probably to water reduction and also to the manganese ions reduction is observed. In the reverse scan, one anodic peak related to the manganese dissolution is observed at around -1,42 V vs SCE. The current loop noticed between the forward and the reverse scan results from the nucleation process on the vitreous carbon RDE. The current efficiency was estimated from the ratio between the anodic charge and the cathodic charge calculated from the different voltammograms recorded in the Mn and the Bi
electrolyte respectively. It was found that the efficiency of the manganese deposition is about (12,63%) whereas the one of the bismuth deposition is close to 100% (93,84 %). These values are predictable considering the equilibrium potential of Mn2+/Mn and Bi3+/Bi. One may note here that by this way the current efficiency is under estimated as the deposit is not completely dissolved in the anodic sweeping. Cyclic voltammetry recorded on the RDE electrode in the complete Mn-Bi electrolyte described above is shown on figure 2. As can be seen, starting from the rest potential (E = 0.033 V vs SCE), a plateau of a weak current density is observed between -0.2V vs SCE to – 1.4 V vs SCE, this current was attributed to bismuth reduction followed by hydrogen evolution according to the previous study. As the cathodic potential increases up to -1.4 V vs SCE, the current density increases indicating a probable reduction of Mn2+ ions simultaneously with the bismuth and the proton reduction (hydrogen evolution). In the reverse scan and in the range of cathodic potential, a cross over attributed to the typical nucleation process of the Bi and the Mn grains on the working electrode is observed. On the anodic side, two separate peaks are observed. Peak A observed at -1.42 V vs SCE is attributed to manganese dissolution. Whereas the peak B observed at -0.2 V vs SCE is attributed to bismuth dissolution. It is important to note the relative intensity of the Mn and Bi dissolution peaks comparing to the relative concentration of the two ions species in the electrolysis bath. Peak B is more significant than peak A because the Bi3+ reduction is not much affected by the simultaneous hydrogen evolution as in the case of Mn2+ reduction. Consequently, the current efficiency of bismuth deposition is higher than that of Mn as mentioned above.
Based on the above results we have used two potential values of (-1.65 V vs SCE and -1.75 V vs SCE) for the deposition of Mn-Bi thin layer during 100s and a couple of potential -0.3 V
vs SCE /-1.65 V vs SCE for the deposition of the Mn-Bi/Bi bilayer during 100s at each potential step. The current-time transients obtained on a thin layer and a bilayer deposited on the Cu substrate are shown on figure 3-a and 3-b respectively. One can observe that the current-time transient obtained at -1.65 V vs SCE (Fig. 3-a) is not constant; the observed change in the slope highlights the reduction of various species, namely manganese, hydrogen and bismuth. The curve b (Fig.3-b) related to a double chronoamperometry shows two different responses in the current density according to the imposed potential. The first response reveals a very stable current density which characterizes the discharge of Bi3+ ions on the Cu substrate. The second response shows a less stable current density due to the reduction of various species including Mn2+ reduction as on curve a. One can also note in this figure, the difference in the current densities magnitude (then in the total charge) for the same applied potential of -1.65 V vs SCE (The corresponding calculated total charge per area from the two figures 3-a and 3-b are 15.56 C/cm2 and -12.76 C/ cm2 respectively): in fact, in the thin layer Mn is deposited directly on a copper substrate while in the bilayer it is deposited on a substrate already coated with small amount of bismuth, the latter being thus occupied a number of active sites on the copper substrate that has resulted in the decrease in the deposition rate of manganese .
3.2 Electrocrystallisation study Before making further characterization on the deposited thin films, it seemed important to study the electrocrystallisation mechanism at the early stages of deposition under the same conditions used in the electroplating thin films. A series of current transients measurements are carried at different applied potentials during short time. Figure 4(a and b) show the
experimental current transients I(t) recorded on Cu substrate and on Bi (electroplated at -0.3V vs SCE during 100 s) on a Cu substrate at three applied potentials (-1.65V vs SCE, -1.7Vvs SCE and -1.75Vvs SCE). All the transients exhibit a same behavior, an increase in the current density until a maximum value Imax reached at a time tmax corresponding to the discharge of the electrochemical double layer and the formation of new phase on the electrode surface. The current maximum increases, and shifts towards shorter times when the potential increase. When t tmax it can be observed on the Figure 4(a) that the current decreases slightly and then stabilizes with time corresponding to the growth of the formed nuclei. The current-time transient registered on the Bi/Cu substrate (Fig 4-b) shows also dependence with a potential, but the current density does not increase with time. In order to obtain general information about the nucleation and the growth mechanism of MnBi system the current transients are compared to the Scharifker and Hills (SH) tridimensional theoretical model [28]. In this model tow cases can be mentioned: the instantaneous nucleation and the progressive nucleation. Instantaneous nucleation indicates that the nuclei formed instantaneously on the electrode surface, ie all sites are activated simultaneously. Progressive nucleation corresponds to gradually formation of nuclei on the active sites (the nucleation sites are activated progressively). The expressions for instantaneous and progressive nucleation are given by Eqs.(1) and (2) respectively.
i i max
i i max
2 t 1.9542 1 exp 1.2564 t t max t max
2 t 1.2254 1 exp 2.3367 t t max t max
2
2
Eq. (1)
2
Eq.(2)
Where Imax is the maximum current density and tmax is the time corresponding to Imax.
Figure 5 (a,b) shows the experimental non-dimensional plots (I/Imax)2 vs (t/tmax) compared to the theoretical model curves. The analysis of the experimental current transients according to the (S-H) model shows a good agreement with theoretical curves for instantaneous nucleation up to tmax. After tmax, the curves shifts from the instantaneous theoretical model and exhibit a higher current at longer times, this deviation can be attributed simultaneously to the partial kinetic control of the growth and to the hydrogen evolution at this range of potentials, this behavior was already reported in the literature [29-31]. In a previous study Boudinar and Co.[32] show that the nucleation and growth mechanism of Mn-Bi co-deposited in a sulfatenitrate bath on a copper foil substrate takes place in a tridimensional progressive nucleation and shifts to the instantaneous nucleation at more negative potentials when the hydrogen evolution became important. On the other hand, the electrochemical nucleation mechanism of Bi is very sensitive to the applied potential and to the ions concentration, at lower potentials and lower concentrations the nucleation mechanism of Bi can be described by a 3D progressive nucleation and shifts to the tridimensional instantaneous nucleation with increasing the Bi3+ concentration and the applied potential as reported by several authors [31,33-34]. It should be noted also that at these range of negative potentials, the electrochemical mechanism process of Mn-Bi may deviate from the diffusion control because of the presence of many other electrochemical and chemical reactions such as:
2𝐻2 𝑂 + 2𝑒 → 𝐻2 + 2(𝑂𝐻)− 2𝐻+ + 2𝑒 → 𝐻2 𝑀𝑛2+ + 2(𝑂𝐻)− → 𝑀𝑛(𝑂𝐻)2
Eq. (3)
Eq. (4)
Eq. (5)
Indeed, the strong hydrogen evolution in the investigated range of potential (eq.3 and Eq.4) has caused a variation in the interfacial pH which may generate the formation of hydroxides (eq.5). This phenomenon has significantly affected the nucleation process.
At very short time according to Scharifker and Hills model the current as function of time can be described for the instantaneous nucleation by the following relation ;
i (t) = zFc1/2 D3/2 N0 (
8πcM −1/2 1/2 ρ
)
t
Eq (6)
Where z corresponds to the number of electrons involved in the electrochemical reaction, F the Faraday’s constant, c is the ions concentration in the electrolyte, D the diffusion coefficient, N0 the nuclei density, M the molecular weight and ρ the density. So another diagnostic can be performed to characterize the nucleation mode in the early stage of electrodeposition by the representation of I vs. t1/2 given by the equation (eq. 6) above. The plot of (-I) vs. t1/2 (Figure 6-a and b) shows a good degree of linearity which confirms that nuclei are formed instantaneously on the electrode surface at first deposition time. From the imax and tmax values, the diffusion coefficient and nuclei density can be calculated [28]. For the instantaneous nucleation mode, D and N0 can be expressed as follows:
i2
t
max max D = 0.1629(zFc) 2
Eq. (7)
and
Eq. (8)
8πcM −1/2
N0 = 0.065 (
ρ
)
zFc
(
imax tmax
)
2
The basic requirement in the analysis of transients according to the relations for a three dimensional nucleation is that the product i2m tm remains constant, or does not significantly change with the applied potential [28]. The experimental data deduced from the analysis of the transients in this work and shown in table 1satisfied theses conditions well. We can also note that the average value of the diffusion coefficient increases from 1.8 10-6 cm2 s-1 for the Mn-Bi deposited on Cu to 2.8 10-6 cm2 s-1 for the Mn-Bi deposited on Bi/Cu. Unfortunately estimation of the nuclei density was not possible as the density of the deposit is not known at this early stage of deposition. However, from the equation (Eq 8) the produce imax2.tmax2 can be expressed as a function of (1/N0). As it is shown in table 1, imax2.tmax2 decreases with increasing the over potential; this means that the nuclei density increases with the over potential. Furthermore if we compare between the two depositions modes, we observe that the nuclei density is lower in the Mn-Bi/Bi/Cu films. This is in agreement with the discussion above related to the decrease in the number of the active sites in the bilayer (section 3.1). In conclusion, although the non dimensional plots do not allow to discern between the nucleation in the two deposition methods (especially after tmax); the difference in the kinetic parameters between the two depositions modes at the early stage of deposition will certainly induce a difference in the thin films morphologies, this will be discussed in the section below.
3.3 Morphological characterization before annealing treatment In this section we present the morphological analysis carried out on the deposited samples under different conditions. First of all in order to complete the electrocristallization study, preliminary SEM characterizations at the early stage of the MnBi thin layers and MnBi /Bi
bilayers deposition are performed during 1s. We present as example the SEM images of the thinner deposits obtained at the deposition potential of -1.65V vs SCE (Fig.7). As we can see, the copper substrate is completely covered by the thin deposit even at the very short deposition time; this confirms the instantaneous character of the electroctrocristallization process at the early stage of deposition. Furthermore, we note also a net difference in the two layers morphologies at a very short deposition time. The same behavior was observed in the different applied potentials. As the deposition time increases to 100s (Fig.8-9-10), different morphologies were achieved depending on the deposition potential and the plating mode: for a less cathodic potential (- 0.3 V vs SCE) (Fig.8), the deposit presents a dendritic aspect composed of Bi alone according to the EDS analysis (EDS spectrum not shown). Its nucleation occurs in three dimensions. One can also note that the substrate surface is not completely covered, indicating that bismuth grows preferentially on initial Bi germs than on the copper substrate. Similar results were reported in various works [25-27]. At less negative potential (-1.65 and -1.75 Vvs SCE), the deposits show a completely different aspect (Fig.9), SEM picture 1 and 2). The two films have heterogeneous appearance consisting of large Mn crystallites of about 1 µm in diameter on which are incorporated a spongy deposit of bismuth. Surface substrates seem more covered with the manganese deposit (especially as the deposit potential is more electronegative) indicating that manganese crystallizes more easily on copper. The cross section SEM analysis (picture 3 and 4) shows the Cu/Ta/Si interface on which is incorporated the Mn-Bi deposit. The average thickness of the crystallites increases from 1.4 m to 2.1m as the potential is more electronegative. The SEM-FEG analysis of the film surface (Fig. 9-b picture 1) and the corresponding cross section analysis (Fig. 9-b picture 2) highlights clearly the details of the heterogeneous deposit. The bismuth nucleation seems to be affected by the electrodeposition potential as the dendritic
aspect is not observed, this may be attributed to hydrogen incorporation in the deposit at this range of potential. The heterogeneous chemical composition of the deposit is evidenced by EDS analysis (EDS spectrum not shown) where two regions are distinguished: a clear one with Bi the predominant species and a grey one with Mn the predominant species. Fig.10 presents the morphological aspect of Mn-Bi/Bi bilayer deposited at -0.3 V vs SCE (during 100s) and -1.65 V vs SCE (during 100s) respectively. As we can see, the deposit obtained seems now to be distributed more uniformly on the surface, covering entirely the copper substrate. The manganese crystallite grain size decreased considerably to around 0.3 µm in diameter, while the bismuth deposit occurs with a dendritic form (Fig.10 picture 1 and 2). The cross section SEM-FEG analysis show more details of the deposit morphology (Fig. 10, picture 3 and 4). As we can see we distinguish three regions: the first one just at the interface with the copper consists of very fine and discontinuous bismuth grains (the corresponding EDS spectrum not shown), the second one with a columnar shape (and a thickness of about 450 nm) is composed of manganese and the third one with a dendritic shape (and a thickness about 727 nm) is composed of bismuth.. This particular morphology can be explained as follows: during the first impulsion at -0.3 V vs SCE, bismuth is deposited in a dendritic form by leaving empty areas on the copper substrate. When the second impulsion at -1.65 V vs SCE is applied manganese accompanied with bismuth are deposited simultaneously and heterogeneously, manganese will occupy the empty area left by the first Bi deposit on the copper substrate, instead of spreading out horizontally as in the case of Figure 9, it will spread vertically giving finally this columnar form, while the bismuth will deposit on the first dendritic bismuth grains and will grow perpendicularly to the surface to exceed the manganese layer. This interpretation is coherent with the electrochemical analysis performed in the section 3.1 above.
One may note here that whatever the deposition mode the obtained deposit thicknesses are in the range of the micrometer which means that the deposition efficiency is not so low.
3.4 Morphological characterization after annealing treatment In order to induce a possible inter-diffusion between Mn and Bi phases, annealing was performed under high vacuum (10-6 Torr) during 1 hour for both thin Mn-Bi layers and MnBi/Bi bilayers. We present in this section the SEM observations obtained for a 300 °C annealing temperature. The different images of the thin Mn-Bi layers and the corresponding EDS analyzes of each region are shown on Fig.11-a, 11-b and 11-c respectively. Figure 11-a indicates that the morphology of the deposit has changed with annealing: although manganese crystallites are not observable, the heterogeneous appearance of the deposit is still visible notably with the coalescence of bismuth grains giving circular nodules with an average diameter of 1.3 µm. It should be noted here that the annealing temperature (300°C) is higher than the melting temperature of bismuth, so it was already in a liquid state during the annealing treatment(performed for 1 hour). The circular nodules observed on the deposit result then probably from cooled bismuth droplets. The SEM images presented on Fig.12 show also a significant and more effect of annealing on the Mn-Bi/Bi bilayers morphology. We note that, the deposit has a uniform appearance along the entire surface. In this case, the grains have a strong coalescence comparatively to those not annealed. We note also, the absence of bismuth dendrite on the surface of the deposit. This behavior may be probably due to a better interdiffusion reaction between manganese and bismuth elements, which induces a homogeneous deposit with sufficiently reduced grain size.
3.5 X-rays analysis
Figures (13,14) display the XRD patterns of the Bi, Mn-Bi thin layers and Mn-Bi /Bi bilayers. The obtained results show that the crystal orientations change considerably according to the imposed potential and the deposition mode. For a deposition potential of -0.3VvsSCE, as shown on Fig.13, the spectrum (b) reveals the characteristic peak of bismuth in its rhombohedral structure. Fig.14 shows the XRD pattern of the Mn-Bi thin layer and Mn-Bi/Bi bilayers obtained at various potential before annealing. In the case of the Mn-Bi thin layers and for deposition potential of -1.65 V vs. SCE and -1,75 V vs. SCE (Fig.14, spectrum b and c), the XRD spectra reveal the presence of the (012) bismuth peaks in its rhombohedral structure and the characteristic peaks of manganese phases αMn(111) (body centred cubic system) and -Mn (411) (body centred tetragonal system). The spectrum d of the same figure 14 shows the XRD patterns of the Mn-Bi /Bi bilayer, as we can see, only the peaks characteristic of bismuth, textured in the (012) direction are present. No diffraction peaks of manganese were observed. This behavior may be explained by the absence of Mn crystallization on Bi element as it was the case on copper. We note also that no phase of MnBi alloy was observed in the three spectra. These results can be explained by the absence of the interdiffusion reaction between the two elements at room temperature. After the annealing treatment at different temperatures (100°C- 300°C), no diffraction peaks were observed until a temperature of 300°C. The corresponding XRD patterns are shown on the figure 15. Concerning the thin layer deposits, the obtained results display many new peaks (Fig.15.a and b) attributed especially to the ternary Mn3Bi4Cu4 alloy (located at 2θ = 25.36° and 29.5°) in its FCC phase [35]. The presence of copper element in this phase is due to its interdiffusion from the substrate during annealing. The other peaks are attributed to manganese oxide ( at 2θ =35° and 40.5°) and to bismuth oxides (at 2θ = 27.98°, 72.75°, 36.90°, 46.28°, 47.9° and 48.74°) . They are probably due of the atmospheric oxygen during the samples transfer.
In the case of the bilayers (Fig.15.c), the spectrum obtained after annealing reveals also new diffraction peaks, located at 2θ = 25.36°, 29.5° and assigned to the Mn3Bi4Cu4 phase in the FCC structure. The peaks related to bismuth have completely disappeared. In addition, minor peaks of manganese were observed after annealing. It is important to note that some peaks, located at 2θ = 35°and 41°, related to manganese oxide were also observed in this annealed bilayer, but their intensities are lower than those observed in the thin layers. Otherwise, no peaks related to the bismuth oxide were observed.
3.6 Magnetic characterization In this study, preliminary magnetic characterizations using a SQUID magnetometer were also investigated at room temperature on the Mn-Bi thin layers and the Mn-Bi/Bi bilayers annealed at 300°C under vacuum condition. Before any discussion one may note here that in this MnBi system, the atomic magnetic moment is carried with manganese element (which is antiferromagnetic in its alpha or gamma cubic phase), whereas the bismuth is diamagnetic but as it crystallizes in an uniaxial structure (hexagonal or rhombohedral) it is responsible in the appearance of magnetic anisotropy in the hexagonal MnBi phase. The magnetic measurements obtained for the magnetic field applied parallel to the film plane and for the deposits annealed at low temperatures (T<300°C) didn’t show any ferromagnetic signal, we just observed the diamagnetic signal of the copper substrate. For an annealing temperature of 300°C a ferromagnetic signal appears, the corresponding magnetization curves are reported on figure 16. All the hysteresis curves showed the presence of a low magnetic signal with characteristics typical to the one of a hard magnetic material. However the values of the coercive field (around 300-400 Oersted) remains lower than that required for a hard magnetic material as MnBi. This behavior is related to the incorporation of the copper from
the substrate in the annealed deposit. Indeed, in a former work [36] Y. Chen et al. have reported on the properties of the MnBiY (Y = Al, Ag, Au, Cu, In, Zn, Pt) thin films deposited by evaporation under vacuum condition onto a glass substrate and annealed to 300-400 °C. It was found that Ag, Cu and In are reactive metals with small size, so they can easily penetrate into the MnBi layer and form a MnBiY alloy with a change of the NiAs type symmetry. This modification in the crystal structure of the MnBi alloy induces a great modification in its magnetic properties. In another work, J.Chen et al. [37] have reported that the magnetic properties of the MnxBiCu films can be improved by varying the Mn concentration in the film. The perpendicular magnetic anisotropy, the remanence, the squareness and the coercivity increased with increasing Mn concentration in the MnxBiCu alloy and these values tend to saturate at x=2. The other important point observed on Fig.16 concerns the difference in the magnitude of the saturation magnetization between the thin layers and the bilayers. Indeed the saturation magnetization is lower in the bilayer elaborated at -0.3V vs SCE(100s) and -1.65 Vvs SCE(100s) (curve a) than in the thin layer elaborated at -1.65 V vs SCE (100s) (curve b) while the XRD data (Fig.15) indicate almost the same peak intensities of the MnBiCu alloy in the different deposits. This is probably related to the fact that the amount of non-reacted manganese is lower and more diluted in the bilayer than in the thin layer even if the deposition potential is the same, as it was already mentioned above in the chronoamperometric analysis (Fig.3). These non-reacted manganese atoms contribute to the total magnetic moment. Increasing the cathodic potential to -1.75V v SCE in the thin layer induces an increase in the Mn concentration in the film and then an increase in the saturation magnetization (curve c). In a previous work, Terzief et al.[38] who studied the magnetic behavior of diluted Mn atoms in liquid bismuth have reported that the magnetic moments per
atom of Mn, turn out to be nearly independent of the concentration of the dilute atoms but the magnetic susceptibility of the alloys increases with increasing manganese concentration. Note that, the preliminary magnetic measurements reported in this work for the first time on a Mn-Bi electrodeposited system has not been fully investigated. The electrodeposition of the MnBi alloy and the study of its magnetic properties still remains an area of active research in order to optimize the plating parameters giving the best magnetic properties (strong coercivity, very large Kerr effect).
4 Conclusions In summary, this study showed that the initial stage of the Mn-Bi electroplated in a chloride bath is governed by a three dimensional instantaneous nucleation whatever the deposition mode and the applied potential. However, the morphological and the structural properties of the electrodeposited thin films are strongly dependent on the applied potential and the deposition mode. Indeed, on the thin Mn-Bi layers and for a less cathodic potential, only the bismuth element is present in the deposit with a dendritic morphological aspect in the rhombohedral structure. For more cathodic potential (E ≤ -1.65 V vs SCE), the electrodeposited films contain large crystallites of manganese on which grows a spongy deposit of Bi. The annealing treatment performed at 300°C under vacuum conditions has induced a little interdiffusion between manganese and bismuth with the formation of a Mn3Bi4Cu4 alloy and has also provoked a big coalescence of bismuth giving circular nodules with an average diameter of 1,3 µm. In the case of the Mn-Bi/Bi bilayers, the morphological analysis revealed a lower grain size compared to those obtained in the thin layers. Before annealing only the characteristic peaks
of bismuth are present. After annealing at 300°C the deposit exhibit a very compact and homogeneous surface without cracks and voids. Mixed Mn3Bi4Cu4 alloy was also observed. A coercivity around 300-400 Oe has been measured in the different deposits after annealing.
Acknowledgments The authors thank Claire Meyer (InstitutNéel - CNRS Grenoble) for SQUID measurements and Amirouche Saifi (UMMTO) for SEM observations.
References [1]E. I. Cooper, C.
Bonhote, J. Heidmann, Y. Hsu, P. Kern, J. W. Lam,
M.Ramasubramanian, N. Robertson, L. T. Romankiw and H. Xu, IBM J. Res. & Dev. Vol. 49 N° 1 Jan. (2005) 103 [2] Y. Dahmane, L.Cagnon, J.Voiron, S.Pairis,M.Bacia, L.Ortega, N.Benbrahim and A.Kadri, J. Phys.D. App. Phys.39 (2006) 4523 [3] L. Cagnon, Y. Dahmane, J. Voiron, S. Pairis, M. Bacia, L. Ortega, N. Benbrahim, A. Kadri, Jour. of Magn. andMagn. Mat. 310 (2007) 2428
[4] U. Rudiger, H. Berndt, A, Schirmeisen, P.Fumagalli and G.Guntheradt, J. Appl. Phys, 78 (9) (1995) 5391 T. Zhuand, Y. J. Wang, J. Phys. D: Appl. Phys. 32(1999) 2609 [5] S. Saha, R. T. Obermyer, B. j. Zande, V. K. Chandhok, S.Simizu and S.G.Sankar, J. Appl. Phys, Volume 91, N° 10 (2002) 8525 [6] K. Kang, L. H. Lewis, Y.F. Hu, Oiang Li , A.R. Moodenbaugh and Young-Suk Choi, J. Appl. Phys.99 (2006) 08N703-1 [7] M. Hanson, Constitution of Binary Alloys (Mc Graw-Hill, New (1986). York, 1958), p. 318.
[8] T. Chen, J. Appl. Phys. 45, 2358 (1974)] [9] K. Kang, A.R. Moodenbaugh and L. H. Lewis, Appl.Phys.Lett.90, (2007) 153112-1 [10] B. Li ,W. Liu , X. G. Zhao , W.J.Gong , X.T. Zhao , H.L.Wang , D. Kim , C.J. Choi and Z.D. Zhang, J. Magn. Magn. Mat. 372(2014)12 [11]A. Brenner,“Electrodeposition of alloys”Vol.1 Academic Press, New York (1963). [12] M. Cortes, A. Serra, E. Gomez1and E. Valles, ElectrochimicaActa 56 (2011) 8232 [13] Fernando M.F. Rhen, J.M.D. Coey, J. Magn. Magn. Mater (2004) 272 [14]Min-Suk Suh, Chan-Jin Park and Hyuk-Sang Kwon, Surface & Coatings Technology 200 (2006) 3527 [15] Yong X.Gan,James Sweetman, Joseph G.Lawrence, Mater. Lett. 64 (2010) 449 [16] P. Diaz-Arista, R. Antano-Lopez, Y.Meas, R.Ortega, E.Chainet, P.Ozil and G.Trejo, Electrochem. Acta, 51 (2006) 4393 [17] C. Savall, C. Rebere, D. Sylla, M. Gadouleau, Ph. Refait and J. Creus, Mater Sci. Eng. A 430 (2006) 165 [18] J. Gonga, G. Zangari, J. Electro. Soc.149 (2002) C 209 [19]B. Benfedda, N. Benbrahim, A. Kadri, E. Chainet, F. Charlot and S.Coindeau, Electrochim. Acta 56 (2011) 1275 [20] R.S.Dean. « Electrolytic Manganese and its Alloys » , The Ronald Press Co, New York (1952) [21] P. Diaz-Arista, G. Trejo, Surf.Coat. Technol. 201 (2006) 3359 [22] P. Wei, O. E. Hileman Jr, M. RezaBateni, X. Deng, A. Petric, Surf. Coat. Technol. 201 (2007) 7739 [23] J. Wei Wu, D. Christopher Johnson, Y. Jiang, Randall, S. Gemmen, X. Liu, Electrochem. Acta 54 (2008) 793 [24] S. Sandnes, M. E. Williams, U. Bertocci, M. D. Vaudin and G.R. Stafford, Electrochim. Acta 52 (2007) 6221
[25] S. Jiang, Y. H. Huang, F. Luo, N. Du and C. H. Yan, Inorg. Chem. Commun. 6 (2003) 781 [26] S .I. Cordoba de Torresi; I. A. Carlos, J. Electroanal. Chem. 414 (1996) 11 [27] H. Sato, T. Homma, H. Kudo, T. Izumi, T. Osaka and S. Shoji, J. Electroanal. Chem. 584 (2005) 28 [28] B. Scharifker, G. Hills, Electrochim. Acta. 28, (1983) 879-889. [29] D. Grujicic and B. Pesic, Electrochim. Acta. 47, (2002) 2901-20912. [30] C. Han, Q. Liu, and D. G. Ivey, Electrochim. Acta. 54 (2009) 3419 -3427. [31] S.B. Sadale, P.S. Patil, Solid State Ionics. 167, (2004) 273-283. [32] S. Boudinar,N. Benbrahim, B. Benfedda, A. Kadri,aE. Chainet, L. Hamadou, J.Electrochem.Soc, 161 (5) (2014) D227-D234. [33] M. Yang, Z.Hu, J. Electroanal. Chem. 583, (2005) 46-55. [34] N.Vladislavic, S.Brinic, Z. Grubac, M. Buzul, J.Electrochem.Sci, 9 (2014) 6020-6032. [35] G. Bryan Street, J.C. Suits and Kenneth Lee, Solid State Communications, Vol. 14 (1974) 33-36. A. Szytula and H. Biriczycka, Solid State Communications, Vol. 38, (1981) 41-43. [36] Y. Chen, Y.C.P Luo, Z.T.Guan, Q.Y. Lu, Y.J. Magn. Magn. Mat.115 (1992) 55 [38] J. Chen, K. Wierman, R. D. Kirby, and D. J. Sellmyer, IEEE Trans. Magn. vol. 31, no. 6 (1995) 3334 [39] P. Terzieff, S. Uporov, V. Sidorov ; A.Mikula, J. Magn. Magn. Mater. 321 (2009) 3747
Figures captions Figure 1 a) Cyclic voltammetry on a vitreous RDE electrode in a bismuth chloride bath: 10-2 M BiCl3, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 b) Cyclic voltammetry on a vitreous RDE electrode in a manganese chloride bath: 0,4 M MnCl2, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 Figure 1-a 14
(a) 12 10 8
j / mA.cm
-2
6 4 2 0 -2 -4 -6 -0,4
-0,2
0,0
0,2
0,4
E / V.vs SCE
Figure 1-b 10
(b)
5
j / mA.cm
-2
0
-5
-10
-15
-20 -2,0
-1,5
-1,0
-0,5
E / V.vs SCE
0,0
0,5
Figure 2: Cyclic voltammetry on a vitreous RDE electrode in 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl, Vb = 20 mV.s-1, = 250 rpm, pH = 2 Figure 2 30
Bi dissolution 20
B
Mn dissolution A
10
j/ mA.cm
-2
0
-10
-20
-30
-40 -1,5
-1,0
-0,5
E/ V.vs SCE
0,0
Figure 3: Current time transients for : (a) a thin Mn-Bi layer deposited at -1.65Vvs.SCE and (b) Mn-Bi/Bi bilayer deposited at EBi = -0.3Vvs.SCE and EMn = -1.65Vvs.SCE. Figure 3-a
-0,09
(a) -0,10 -0,11
j / mA.cm
-2
-0,12 -0,13 -0,14 -0,15 -0,16 -0,17 0
20
40
60
80
Time (s)
Figure 3-b
(b)
0,00
-0,02
j/ mA.cm
-2
-0,04
-0,06
-0,08
-0,10
-0,12
-0,14 0
50
100
Time (s)
150
200
100
Figure 4: Chronoamperometric transients at early stage of deposition for Mn-Bi electrodeposited on (a) Cu/Ta/Si and (b) Bi/Cu/Ta/Si. Figure 4.a
Figure 4-b
Figure 5: Non dimensional plots (I /Imax)2 vs (t /tmax) of the chronoamperometric curves in the Fig.4: a) Mn-Bi on Cu/Ta/Si, b) Mn-Bi on Bi/Cu/Ta/Si Figure 5-a
Figure 5-b
Figure 6: (-I ) vs.t1/2 plots for the initial stages of Mn-Bi electrodeposited on (a) Cu/Ta/Si and (b) Bi/Cu/Ta/Si. Figure 6-a
Figure 6-b
Figure 7: SEM images of MnBi thin layer (a) and MnBi/Bi bilayer (b) at early stage of deposition (deposition time: 1 s, deposition potential: -1,65 V vs SCE). Figure 7
a)
b)
Figure 8: SEM-FEG of Bi thin layer electrodeposited at E = -0.3 Vvs.SCE from acidic chloride bath, pH = 2, 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl
Figure 8
Figure 9 a) SEM images of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE (picture 1) and E= -1.75 V vs.SCE (picture 2) and the corresponding cross section observations (picture 3 and 4 respectively) b) SEM-FEG images of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE (picture 1) and the corresponding cross section analysis (picture 2). Figure 9-a 1
3
Thin layer: E= - 1.65V
2
Thin layer: E= - 1.75V
4
Figure 9-b
1
Mn
2
Bi Mn Cu
Ta
Si
Cu
Figure 10: a) SEM images of Mn-Bi/Bi bilayer (picture 1 and 2) and the corresponding SEMFEG cross section analysis (picture 3 and 4) electrodeposited at EBi = -0.3 V vs. SCE; EMn = 1.65 V vs. SCE from acidic chloride bath, pH = 2, 0.4M MnCl2, 10-2M BiCl3, 3M NH4Cl. Figure 10
1)
3) 3 2 Cu
Ta Si
2)
4)
1
Figure 11 a) SEM images of Mn-Bi thin layers electrodeposited from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions. b,c) EDS spectra of Mn-Bi thin layer electrodeposited at E = -1.65 V vs.SCE; from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions, pH = 2, 0.4M MnCl2, 10-2 M BiCl3, 3M NH4Cl, recorded in a white region (b) and a grey region (c) on the deposit
Figure 11-a
Figure 11-b
White region
Figure 11-c Grey region
Figure 12: SEM images of Mn-Bi/Bi bilayers electrodeposited from acidic chloride bath after annealing at 300°C for 1 hour under vacuum conditions.
Figure 12
A)
Figure13: XRD analysis patterns on Bi thin layer obtained at E = -0.3Vvs.SCE, 0.4M MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si substrate, pH = 2 (a) Substrate, (b) Bi thin layer Figure 13 8000
.
substrat
.
4000
Bi (202)
Bi (104)
Bi (012)
Intensity
6000
(b)
. 2000
(a)
20
25
30
35 2 (Deg.)
40
45
50
Figure 14: XRD analysis patterns on Mn-Bi thin layer and on Mn-Bi/Bi bilayers before 0.4MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si
annealing obtained respectively in
substrate, pH = 2((a) Substrate, (b) Mn-Bi thin layer at -1,65V vs.SCE, (c) Mn-Bi thin layer at -1,75V vs.SCE, (d) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE)
Bi(012)
2500
Bi(110)
Intensity
2000
gamma Mn(200) or alpha Mn(332)
Bi(104)
gamma Mn(111) or alpha Mn(411)
Figure 14
1500
1000 (d)
(c) 500
(b) (a)
0 20
25
30
35
2 (Deg.)
40
45
50
Figure 15: XRD analysis patterns on Mn-Bi thin layers and Mn-Bi/Bi bilayers after annealing at 300°C for 1 hour under vacuum conditions, 0.4MnCl2, 10-2 M BiCl3, 3M NH4Cl on Cu/Ta/Si substrate, pH = 2, (a) Mn-Bi thin layer at -1,65V vs.SCE, (b) MnBi thin layer at 1,75V vs.SCE, (c) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE) Figure 15
.
MnO2 * Bi203 + Bi4Cu4Mn3
gamma Mn(111) or alpha Mn (411)
-
Bi
.
-
3000
*
+
1000
* (a) 20
+
25
+
30
* * ** 35
(Deg.)
*
*
*
*
-
*
+
-
(b)
+
-
+
-
(c)
-
Intensity
2000
40
45
50
Figure 16: Magnetization curves obtained at room temperature of the Mn-Bi/Bi bilayers and Mn-Bi thin layers annealed at 300°C under vacuum conditions, (a) Mn-Bi/Bi bilayer at (-0,3V vs.SCE, -1,65V vs.SCE ; (b) Mn-Bi thin layer at -1,65V vs.SCE; (c) MnBi thin layer at 1,75V vs.SCE. Figure 16
0,0008 0,0006
(a) (b) (c)
0,0004
M / emu.cm
-3
0,0002 0,0000 -0,0002 -0,0004 -0,0006 -0,0008 -1500
-1000
-500
0
H / Oe
500
1000
1500
Table1. Electrochemical parameters obtained according to Scharifker and Hills model (t0 is the induction time). t0 (s)
3
2
2
-6
-imax
Mn-Bi/ Cu/Si
-1.65
26.20
4.69
0.21
4.48
3.07
13.78
2.01
-1.7
33.40
2.38
0.13
2.25
2.50
5.63
1.63
-1.75
41.30
1.74
0.06
1.68
2.87
4.83
1.87
-1.65
21.90
10.03
1.02
9.01
4.32
38.97
2.82
-1.7
27.00
6.71
1.09
5.62
4.09
22.99
2.67
-1.75
40.60
3.52
0.65
2.87
4.61
13.23
3.01
mA
tmax- t0
2
-E (V/SCE)
Mn-Bi/Bi/ Cu/Si
tmax (s)
3
Electrodeposition mode
(s)
10 .imax .(tmax- t0) 2
(mA .s)
10 .imax .(tmax- t0) (mA.s)
2
10 .DInst 2
-1
(cm .s )