NANOSTRUCTURED CuCo NANOWIRES. Fedosyuk V.M.

NANOSTRUCTURED CuCo NANOWIRES Fedosyuk V.M. Institute of Solid State Physics and Semiconductors of the Belorussian Academy of Sciences, P Brovki str 19, 220072 Minsk, Belarus E-mail:fedosyuk@ifttp.bas-net.by Introduction Nanogranular films, which represent nanoinclusions of ferromagnetic material (Co, Fe, Ni) in dia - (f.e.cu) or paramagnetic (f.e.re) matrix, are extremely promising at the present time [1-9]. Originally, the interest to inhomogeneous films, formed from alloys of transition metals of the iron group with Cu, arose from the possibility of applying these materials as substitutes for various, currently used, conventional, magnetoresistive sensors which are based on permalloy transducers. In comparison with the latter, these inhomogeneous(or nanogranular) alloy films are far more promising owing to the fact that they possess, under certain conditions, an isotropic giant magnetoresistance (GMR) in addition to significantly lower noise. A variety of techniques has been applied for the preparation of these materials including MBE, sputtering, vapour deposition, and, laterly, laser ablation. Although electrodeposition (ED) has been known for many years, it is only fairly recently that it has been applied with any great success to the production of magnetic nanomaterials, including nanogranular films and nanowires. Template electrodeposition, by which is meant electrodeposition in natural or artificial holes in an insulating layer on a conducting substrate, offers an inexpensive route to the fabrication of patterned metal films, in some cases with structures that cannot be produced by any other method. For example, we show that it is possible to prepare nanowires of Co-Cu granular alloy by template electrodeposition in porous aluminium oxide, and that the GMR of the wires increases, as expected, on annealing. Nanogranular alloys are more straightforward to prepare than other GMR materials, as unlike superlattices and spinvalves they do not require the control of layer thicknesses to sub-nm precision. They may therefore prove to be suitable field sensing elements for magnetic sensors where a low unit cost and a large response are more important than sensitivity to small field changes. Experiment Co-Cu heterogeneous alloy nanowires having lengths of several tens of µm and diameters of 10 or 300 nm were deposited in the pores of commercially available anodic aluminium oxide membranes. Unlike the nuclear track-etched polycarbonate mernbranes used in previous studies of rianowire, these can be used in annealing studies. Prior to sample growth, all membranes were coated with an evaporated Au layer on the side that was not exposed to electrolyte, in order to have a conducting substrate for electrodeposition. Each membrane was subsequently mounted on a Cu plate, with which the Au coating made electrical contact, and all but the area of the membrane in which deposition was desired was masked off with kapton tape. For deposition, the membrane acted as the working electrode in a simple 2-electrode cell in which a Pt plate acted as counter electrode. The electrolyte consisted of 30 g CuSO. 4 5H 2 O (Cu sulphate), 6.6 g H 3 BO 3 (boric acid), 120 g Na 3 C 6 H 5 O. 7 5.5H 2 0 (sodium citrate) and 50 g CoSO. 4 7H 2 O (Co sulphate) per liter of purified H 2 O (conductivity < 10-18 Ω -1 m -1 ). Deposition was carried out at room temperature and electrolyte ph = 5.7, using a galvanostat (EG&G model 363) to control the current density. Sufficient metal was deposited for the wires to emerge from the ends of the membrane pores, forming approximately hemispherical caps. Results and Discussion Following growth, the room temperature magnetoresistance (MR) of the nanowires was measured with one contact made using Ag DAG to the caps at the tops of the nanowires and the

other to the Au layer at the base of the wires. This configuration meant that generally the MR of several nanowires would be measured in parallel. Figure 1 shows the measured MR for Co-Cu nanowires electrodepositedd in an aluminium oxide membrane with quoted thickness 60 µm, pore diameter 200 nm and pore density 10 9 cm -2. In fig. l(a) the applied magnetic field was parallel to the long axis of the wires, and in fig. l(b) it was perpendicular. Since the current was always along the wires, these configurations correspond to measuring the longitudinal and transverse MR, respectively. It is immediately apparent that although the absolute magnitude of the MR is small (less than 1%), in both configurations it is negative, showing that we do indeed have GMR, as expected for a heterogeneous Co-Cu alloy. There appear to be symmetric "shoulders" either side of the central peak in both MR curves, which may be related to details of the magnetization reversal mechanism. For the sample of fig.1, the nominal deposition current density (assuming the quoted pore diameter and density, and that electrodeposition took place in all pores) was 50 ma. cm -2. Figure 1. Percentage magnetoresistance for Co-Cu heterogeneous alloy nanowires electrodeposited in an aluminum oxide membrane with quoted pore diameter 200 nm and pore density 10 9 cm -2, measured at room temperature with the magnetic field applied (a) parallel and (b) perpendicular to the long axis of the wires. Samples were also prepared in membranes with quoted pore diameter 20 nm and density 10 10 cm -2. Figure 2 shows the longitudinal (a) and transverse (b) MR for Co-Cu nanowires grown with a nominal deposition current density of 400 ma. cm -2 in one such membrane. Despite the current density for this sample being a factor of 8 larger than that for the sample of fig. 1, the measured compositions of both samples were similar, i.e. Co 19 Cu 81. The GMR for this sample is also significantly less than 1%. Interestingly, the GMR of electrodeposited continuous Co-Cu heterogeneous alloy films of a given composition was also reported to be insensitive to current density. Figure 2. Percentage magnetoresistance for Co-Cu heterogeneous alloy nanowires electrodeposited in an aluminum oxide membrane with quoted pore diameter 20 nm and pore density 10 10 cm -2, measured at room temperature with the magnetic field applied (a) parallel and (b) perpendicular to the long axis of the wires.

The GMR of continuous Co-Cu heterogeneous alloy films grown by a variety of methods has been shown to increasee on annealing, due to phase separation, which leads to an increase in the number and size of the Co-rich particles. We therefore investigated the effect of annealing on the GMR of our electrodeposited heterogeneous alloy nanowires. Samples were annealed while still in the aluminiumm oxide membranes for 30 minutes at 200 C or 400 C, in a 10-5 torr vacuum to reduce the risk of oxidation. In all cases, the room temperature GMR was found to increase significantly. For example, fig. 3shows the longitudinal (a) and transverse (b) MR for the same, sample as fig. 2 following a 400 C annealing, and it can be seen that the maximum MR in each configuration has more than doubled. A particularly interesting feature of fig. 3 is that the shapes of the MR curves with the applied field (a) parallel and (b) perpendicular to the long axis of the wires are distinctly different. This apparent anisotropy in the magnetization reversal process suggests that the nanowire geometry does influence the magnetic properties of the heterogeneous alloy, possibly because some of the Co-rich particles approach the nanowire diameter (20 nm) in size, or Figure 3. Percentage magnetoresistance for the same sample as Figure 2, but following an annealing at 400 C for 30 minutes, measured with the magnetic field applied (a) parallel and (b) perpendicular to the long axis of the wires possibly because even if the nanowire diameter is much larger than the particle dimensions, it may be less than the range over which magnetic interactions are important. Alternatively, during electrodeposition and subsequent annealing, the nanowire geometry could influence the shapes and distributions of the Co-rich magnetic particles, or could lead to stress-induced anisotropy in these particles. Although porous aluminium oxide membranes do allow annealing studies, they have the disadvantage that it is relatively difficult to dissolve the aluminium oxide to release the nanowires for structural studies. Some nanowires were therefore electrodeposited in nuclear track-etched polycarbonate membranes which can easily be dissolved using chloroform to facilitate characterization by transmission electron microscopy (TEM).

Fig.4. Transmission electron micrograph of Cu- alloy nanowires Co heterogeneous electrodeposited in a nuclear track-etched polycarbonate membrane with quoted pore diameter 10 nm. Fig. 4 is a transmission electron micrograph of Co-Cu heterogeneous alloy nanowires electrodeposited in a nuclear track-etched polycarbonate membrane with thickness 6 µm, quoted pore diameter 0,01 µm and pore density 6 x 10 8 cm -2. Although his membrane thickness and pore diameter and density are rather different to those of the aluminium oxide membranes, nanowires grown in both kinds of membrane showed GMR at room temperature. The figure shows that these heterogeneous alloy nanowires are highly polycrystalline with very small grain sizes. X-ray fluorescence compositional analysis of a number of samples showed that despite significant differences in the nominal current densities, the composition remained approximately constant at ~ Cu 80 Co 20. Continuous Co-Cu films with a similar composition required an electrolyte less concentrated in Co (30g CoSO. 4 7H 2 O I -1 ), but were also grown with rather lower current densities (typically 5 ma cm -2 ). Fig.5 shows M-H curves for nanowires grown in an aluminiumm oxide membrane with quoted pore diameter 0.02 µm. Curves are presented for the applied field (a) parallel and (b) perpendicular to the long axis of the wires. They show only a relatively small amount of hysteresis at 5 K and less at 300 K. If the average composition is Co 20 Cu 80, one can assume that the average magnetization of a single heterogeneous alloy nanowire will be ~1/5 that of Co. This would give a demagnetizing field of ~1.7kOe when the magnetization is saturated perpendicular to the long axis of the wires and ~0 when it is parallel, leading to a shape anisotropy which is clearly seen in fig. 5. Fig. 6 shows the ZFC/FC initial magnetic susceptibilities of heterogeneous alloy nanowires. This sample was grown in the same kind of membrane as the sample in Fig. 6, but with twice the deposition current. ZFC measurements were made after cooling the sample to 5 K in zero field and measuring up to 300 K in a field of 50 Oe parallel to the long axis of the wires. The FC measurements were then made by cooling the sample in the same field. Since the ZFC initial susceptibility in Fig. 7 shows no clear peak, but rather keeps increasing with increasing T, it is apparent that the Co clusters in these heterogeneous nanowires have a range of blocking temperatures which extends at least to room temperature. This is different to what was observed for continuous electrodeposited Co 2o Cu 8o films ], for which there was a peak in the ZFC initial susceptibility at T ~107 K.

Fig. 5. Hysteresis loops measured at 5 K for nanowires growth in aluminum oxide membrane with quoted pore diameter 0.02 µm (a). Applied field parallel to the long axis of the wires (i.e. perpendicular to the membrane in which they are contained). (b) Applied field perpendicular to the long axis of the wires (i.e. parallel to the membrane). The lines guides to the eye, joining measured data points. Fig. 6. Normalized initial magnetic susceptibility of heterogeneous alloy nanowires after cooling in zero magnetic field (ZFC) and in an applied field of 50 Oe (FC). The nanowires were growth in an aluminum oxide membrane with quoted pore diameter 0.02 µm, and were not removed from the membrane prior to measurement. The lines are guides to the eye. Fig. 7 Initial magnetic susceptibility after cooling in zero magnetic field (ZFC) and in an applied field of 50 Oe (FC) for nanowires from the same sample as those of Fig. 6, but annealed at 400 C for 30 min. The nanovires were not removed from the membrane prior to measurement, and the lines are guides to the eye. Another difference between these heterogeneous nanowires and continuous films is the presence of a rise in the ZFC/FC curves in the low-temperature range 5-20 K, which is absent for the continuous films. This feature is also absent from ZFC/FC curves measured for heterogeneous nanowires electrodeposited in pores of larger diameter. A similar feature was previously observed for electrodeposited Co-Re films and attributed to the presence of extremely small clusters [39]. In the case of the heterogeneous nanowires, this explanation is supported by the observation that the sharp rise at low temperatures disappears or is reduced on annealing the sample at 400 C for 30min ( fig. 7), presumably as a result of agglomeration of the extremely small clusters with larger clusters. A comparison of fig. 6 and fig.7 also shows that the shoulder seen in both ZFC curves moves to higher temperatures for the annealed sample, indicative of higher blocking temperatures, consistent with the annealed wires containing larger ferromagnetic particles resulting from agglomeration. The absence of the sharp rise in the FC/ZFC susceptibility for continuous heterogeneous alloy films and larger diameter heterogeneous alloy nanowires suggests that these contain many fewer extremely small clusters, possibly as a consequence of their being grown at lower current densities. In conclusion, we have shown that it is possible to prepare heterogeneous Co-Cu alloy in the form of nanowires using constant current electrodeposition. The room temperature GMR of the as-deposited wires is small, and appears to be rather insensitive to the nominal current density, but further work is needed to establish the dependence of the composition and

microstructure on the electrodeposition conditions. Annealing the nanowires leads to a significant increase in the GMR, presumably as a result of increased phase segregation, and the shapes of the MR curves measured with applied field parallel and perpendicular to the long axis of the wires reveals anisotropy in their magnetic properties. References 1. Fedosyuk V.M. Twinning in Nanostructured Films and Nanowires, BSU, 2009, 523p. 2. Fedosyuk V.M. e,a. Nanomaterials and Nanotechnologies, BSU, 2008, 375p. 3. Fedosyuk V.M. Nanogranular CuCo nanowires. In Handbook of Powders of Non- Ferrows Metals, Elsevier Advanced Technology, 2007, 576 p. 4. Fedosyuk V.M. Nanostructured Films and Nanowires, BSU, 2006, 311p. 5. Fedosyuk, V.M. Structure and mechanism formation of nanogranular films, in "Atomistic aspects of Apitaxial Growth." Editted by M. Kotrla, N.I. Papanicolau, D.D.Vvedenski and L.T.Wille Kluwer Academic Publishers, 2005, vol.65, pp.535-550. 6. Fedosyuk V.M. Nanogranular electrodeposited magnetic cobalt alloys //Encyclopedia of Nanoscience and Nanotechnology, 2004, vol.2, p.895-918. American Scientific Publishers, ed. by N.S.Nalwa 7. Fedosyuk V.M. Nanoclustered Films and Nanowires. In Synthesis,Functional Properties and Applications of Nanostructures Ed. By T.Tsakalakos and I.Ovid ko, Kluwer Acad. Publishers,2003, p.557-578. 8. Fedosyuk V.M. Electrodeposited Nanostructures, Minsk, BSU, 2002, 353p. 9. Fedosyuk V.M. Multilayered Magnetic Structures, Minsk, BSU, 2000, 197 p.