Microwave Absorption Properties of Cobalt Nanowires Fabricated by Pulse Electrodeposition

Transcription

1 PIERS ONLINE, VOL. 6, NO. 1, Microwave Absorption Properties of Cobalt Nanowires Fabricated by Pulse Electrodeposition Wenbing Chen, Mangui Han, and Longjiang Deng State Key Laboratory of Electronic Thin Films and Integrated Devices University of Science and Technology of China, Chengdu , China Abstract In this work, cobalt nanowires with a preferred growth orientation have been fabricated by a pulse electrodeposition method. The imaginary part of the permeability spectra for the nanowire/paraffin composite samples exhibit a strong absorption peak at 6.1 GHz and two minor peaks at above 10 GHz. It is determined that the peak at 6.1 GHz is attributed to the natural resonance mechanism and the other two peaks are caused by eddy current effect. We have fitted the permeability spectra attributed to natural resonance using the LLG equation. Calculation based on the Kittel equation substantiates our explanation. The electromagnetic wave reflection loss values of the nanowire/paraffin composite indicate that cobalt nanowires composites can be used for microwave absorption. 1. INTRODUCTION One of the urgent problems plaguing the modern world is the widespread electromagnetic pollution (EMP), brought about by the growing development of wireless telecommunication. To combat this problem, considerable efforts have been devoted to explore novel microwave absorption materials that can meet requirements like strong absorption in a wide frequency range and light weight [1 4]. Low-dimensional ferromagnetic nanowires provide one solution to the EMP problem [5, 6]. One facile way to manufacture metal nanowires is to utilize electrodeposition and porous template such as anodic aluminum oxide (AAO) and polycarbonate membrane. Nanowires deposited in these templates have several advantages. First of all, their diameters usually are smaller than the skin depth, ensuring the propagation of microwaves and the repression of eddy current effect. In addition, the dielectric nature of the templates helps reduce the electric conductivity, which is also desirable for microwave absorption. Moreover, as pointed out by Encinas and co-workers [5], the resonance frequency of nanowire array can be tuned over a large range by varying the porosity of the template, the geometry of nanowires, the type of the magnetic material, and the magnetization states of nanowires. In this work, cobalt nanowires have been fabricated within the AAO templates through a pulse electrodeposition method. The nanowires were released from the template and mixed with paraffin wax in order to investigate their microwave absorption properties in the range of GHz. Their behaviours have been discussed by combining the nanowires morphology, crystal structure, static magnetic property, and classic theories such as the LLG and Kittel equations. 2. EXPERIMENTAL DETAILS The AAO templates with nominal pore size of 200 nm and thickness of 60 µm were purchased from Whatman Corp. R. A typical three-electrode cell was used for electrodeposition. Pulsed electrodeposition was carried out by employing two voltage steps alternatively throughout the experiment: 1.5 V versus a saturated calomel electrode (SCE) for 5 seconds followed by 0.4 V versus SCE for 1.5 seconds. When the voltage was 1.5 V, cobalt was deposited in the AAO template; when the voltage was 0.4 V, the deposition was halted and Co ions were restored in the depleted region near the Au electrode. The deposition-relaxation cycle may help improve the crystal quality and homogeneity of the deposited nanowires [7, 8]. The electrolyte consisted of 100 g/l CoSO 4 7H 2 O and 30 g/l H 3 BO 3. The ph of the electrolyte was adjusted to 3.0 with HCl. The crystal structure of the as-prepared cobalt nanowires within the AAO template was investigated using X-ray powder diffraction (XRD, Cu K α radiation). The static magnetic properties of the as-prepared nanowire array were characterized by a vibrating sample magnetometer (VSM). To investigate the morphology and microwave properties of the cobalt nanowires, the as-prepared sample was dissolved with a 4 wt% NaOH solution for 24 hours and then rinsed with de-ionized water repeatedly until the solution was neutralized. The morphology of the liberated nanowires

2 PIERS ONLINE, VOL. 6, NO. 1, was observed by a scanning electron microscopy (SEM) equipped with energy dispersive x-ray spectrometer (EDS). Then the cobalt nanowires were randomly dispersed in paraffin wax with the nanowires weight ratio of 50% in the composites. The nanowire/paraffin composite samples were pressed into a toroidal shape with outer diameter of 7 mm and inner diameter of 3 mm for microwave measurement. The complex relative permittivity and permeability of the composite samples were measured by an Agilent Vector Network Analyzer 8720 in the range of GHz. All experiments were conducted at room temperature. 3. RESULTS AND DISCUSSIONS Figures 1 and shows the morphology of the cobalt nanowires after dissolving the AAO template. The diameter of the nanowires is around 200 nm and the length of the nanowire is around 6 µm. It is clear that after dissolving the AAO template completely, cobalt nanowires are released from the template, see Fig. 1. The XRD data shows that Co nanowires in the AAO template have a preferential growth direction along the [110] direction, implying that the magnetocrystalline easy axis of the cobalt nanowires the [002] direction is perpendicular to the wire length [9, 10]. Figure 1: SEM image of cobalt nanowires in and : shows the side view of the Co nanowire array after partial dissolution of the AAO template; is the well dispersed nanowires after dissolving the AAO template completely; (c) shows the magnetic hysteresis loops of Co nanowires array. Illustrated in Fig. 1(c) are the normalized magnetic hysterersis loops of the cobalt nanowire array in the AAO template. The measurements were made with the external magnetic field applied perpendicular and parallel to the wire length. It is evident that the easy axis of the nanowires is normal to the wire length, corroborating the XRD results. However, the normalized remanence (M r /M s ) for both loops is smaller than 0.1 and the coercivities are low (276 Oe when magnetic field is perpendicular to the wire length and 220 Oe when magnetic field is parallel to the wire length), which suggests that the large shape anisotropy associated with nanowires is greatly undermined. For cobalt nanowires, the shape anisotropy Kshape is estimated to the as large as J/m 3 by using the equation [11]. K shape = µ 0 (N x N z )M 2 s /2, (1) where µ 0 is permeability of vacuum, M s ( A/m) is the saturation magnetization of cobalt; N x, N z are the demagnetizing factors in the directions normal and parallel to the wire length, respectively. The shape anisotropy would usually align the magnetic moments along the wire length and lead to a square shape of hysteresis loop when the magnetic field is applied parallel to the wire length. The reason we have not observed an obvious anisotropy behaviors is that there are (c)

3 PIERS ONLINE, VOL. 6, NO. 1, other factors affecting the alignment of magnetic moments, such as magnetocrystalline anisotropy and magnetostatic interaction [12, 13]. The frequency (f) dependence of relative complex permittivity and permeability for the nanowire/paraffin composite sample is shown in Figs. 2 and. It can be observed that the imaginary part of permittivity (ε ) exhibits one major peak at around f = 5 GHz and one minor peak at around f = 10 GHz. The inset in Fig. 2 demonstrates that the ε ε relation is mainly composed of a typical Cole-Cole semicircle, indicating that the permittivity spectra conform to the Debye relaxation model [14]. The permeability spectra of the nanowire/paraffin composite samples are shown in Fig. 2. The real part (µ ) firstly increases in the GHz region and then decreases in the GHz regime, after which the value of µ shows minor fluctuations at around f = and GHz. In the meantime, the imaginary part (µ ) shows a major peak at around f = 6.1 GHz and two minor peaks at around f = 12 and 16 GHz. Generally, magnetic loss is attributed to hysterersis, domain-wall displacement, natural resonance, and eddy current effect [15]. Hysterersis could be neglected when a weak field is applied and domain-wall displacement can be ruled out since it only happens in the MHz region, not in the GHz arena. Now consider the eddy current effect and natural resonance mechanism. If the magnetic loss only results from eddy current loss, the value of C should C = µ µ 2 f 1 = 2πµ 0 σd 2 /3 (2) be a constant when the frequency is varied [16]. In the Equation (2), σ refers to the electric conductivity, d is the diameter of nanowires. Fig. 3 shows the value of C as a function of frequency. As can be seen, when f < 10 GHz, C changes drastically, but when f > 10 GHz, C remains approximately a constant. Thus the peak of µ at f = 6.1 GHz is not caused by eddy-current loss. In other words, this peak originates only from natural resonance mechanism. Meanwhile, the other two minor peaks at around f = 12 and 16 GHz are ascribed to eddy current effect, which may be brought about by the large diameter of single cobalt nanowire approaching the skin depth. This indicates the importance of reducing the diameter of nanowires in order to suppress eddy current loss in the GHz region. The dynamic permeability spectrum stemming from natural resonance could be fitted by the Landau-Lifshitz-Gilbert (LLG) equation. After including the shape anisotropy, we can solve the LLG equation and obtain µ and µ as a function of frequency: µ = 1 + µ i f m f y [f x f y (1 + α 2 )(2πf) 2 ] + α 2 (2πf) 2 f m (f x + f y ) [f x f y (1 + α 2 )(2πf) 2 ] 2 + α 2 (2πf) 2 (f x + f y ) 2, (3) µ = µ i 2πfαf m f y (f x + f y ) 2πfαf m [f x f y (1 + α 2 )(2πf) 2 ] [f x f y (1 + α 2 )(2πf) 2 ] 2 + α 2 (2πf) 2 (f x + f y ) 2, (4) f m = γm s, f x = (N x N z )f m, f y = (N y + N k N z )f m, (5) where γ ( ma 1 S 1 ) is the Gilbert gyromagnetic ratio, α is the damping constant, µ i is the initial permeability, f is the frequency. N x, N y, N z are the demagnetizing factors in the X, Y, Z direction. The value of M s is chosen as A/m, which is close to the saturation magnetization of bulk cobalt and cobalt nanowires reported in previous works [17]. Exploiting the above formula, µ f spectrum is fitted firstly and the fitted parameters are then used to calculate the µ f spectrum. The fitted results are displayed in Fig. 2. The best agreement between experiment results and calculation could be obtained when the resonance frequency f res = 6.1 GHz and K 1 = J/m 3, α = 0.098, µ i = Deserving special mention is the value of K 1, which is very close to the value of shape anisotropy and large enough to compete with it. The Kittel equation is employed to evaluate the fitted resonance frequency and related parameters [18]. When the easy axis is perpendicular to the wire length, the Kittel equation gives: f res = γ ((Nx N z )M s H k )(N x N z )M s (6) 2π where H k = 2K 1 /µ 0 M s is the effective magnetocrystalline field. The calculated resonance frequency is f res = 6.03 GHz, which matches well with the experimental value and the LLG solution. This justifies our proposal that the peak of µ at f = 6.1 GHz is due to natural resonance and this value is a result of the competition between shape anisotropy and other factors such as magnetocrystalline anisotropy and magnetostatic interaction. Using the experimental data and the transmission line

4 PIERS ONLINE, VOL. 6, NO. 1, theory, the reflection loss R (db) of the nanowire/paraffin composite samples is calculated [2], as shown in Fig. 3. As shown, the minimum reflection loss value is R = 23.5 db around 6.5 GHz for nanowire/paraffin sample with a thickness of 5 mm and weight ratio of 50%. It is worth noting that the eddy current effect is also suppressed in this frequency regime. It is clear that cobalt nanowires are one of promising candidates as a microwave absorber. Figure 2: The measured relative complex permittivity spectra, the inset in shows the relation between ε and ε (Cole-Cole semicircle). The measured and fitted relative complex permeability spectra. NR denotes natural resonance, ECE denotes eddy current effect. Figure 3: The value of µ µ 2 f 1 as a function of frequency in, and the reflection loss of Co nanowires composites with different thickness in. 4. CONCLUSIONS Cobalt nanowires with preferred growth orientation were prepared by pulsed electrodeposition. It is found that there is a strong absorption peak at 6.1 GHz in the imaginary part of their permeability spectra, which is believed arising from natural resonance. The reflection loss value for the nanowire/paraffin composite sample with thickness of 5 mm and weight ratio of 50% can reach 23.5 db at around 6.5 GHz while the eddy current effect is also suppressed, indicating its potential application as microwave absorbers. ACKNOWLEDGMENT This work is financially supported by NSFC (No ) and NSFC-RS (Royal Society, UK) international jointed program (No ). REFERENCES 1. Han, M. and L. Deng, Appl. Phys. Lett., Vol. 90, , Han, M., Y. Ou, D. Liang, and L. Deng, Chin. Phy. B, Vol. 18, 1601, Liu, Q. and D. Zhang, Appl. Phys. Lett., Vol. 93, , Dong, X. L., X. F. Zhang, and H. Huang, Appl. Phys. Lett., Vol. 92, , Encinas, A., L. Vila, M. Darques, J. M. George, and L. Piraux, Nanotechnology, Vol. 18, , Gao, B., Q. F. Liu, F. S. Li, J. Feng, and D. S. Xue, J. Phys. D: Appl. Phys., Vol. 41, , 2008.

5 PIERS ONLINE, VOL. 6, NO. 1, Zhang, J., G. A. Jones, S. E. Donnelly, and G. H. Li, J. Appl. Phys., Vol. 101, , Ursache, A., J. T. Goldbach, T. P. Russell, and M. T. Tuominen, J. Appl. Phys., Vol. 97, 10J322, Han, X. H., Q. F. Liu, Y. Ren, R. L. Liu, and F. S. Li, J. Phys. D: Appl. Phys., Vol. 42, , Li, D. D., R. S. Thompson, G. Bergmann, and J. G. Lu, Adv. Mater., Vol. 20, 4575, Qunadjela, K., R. Ferré, J. L. Maurice, L. Piraux, and S. Dubois, J. Appl. Phys., Vol. 81, 5455, Li, F. S., T. Wang, L. Y. Ren, and J. R. Sun, J. Phys.: Condens. Matter, Vol. 16, 8053, Encinas-Oropesa, A., M. Demand, and L. Piraux, Phys. Rev. B, Vol. 63, , Fang, J. X. and Z. W. Yin, Physics of Dielectric Materials, 45, Science, Beijing, Liao, S. B., Ferromagnetism, 6, 127, 139, Science, Beijing, Wu, M. Z., Y. D. Zhang, S. Hui, and T. D. Xiao, Appl. Phys. Lett., Vol. 80, 4404, Goglio, G., S. Pignard, A. Radulescu, and L. Piraux, Appl. Phys. Lett., Vol. 75, 1769, Kittel, C., Phys. Rev., Vol. 73, 155, 1948.