Optical Materials 29 (2007) 760–765 www.elsevier.com/locate/optmat
Fen–V2O5 nano-composites: Room temperature magneto-optical and radar absorption properties M. Maaza a
a,*
, O. Nemraoui b, C. Sella c, A. Gibaud d, T.B. Seda e, A.C. Beye
f,g
Nanosciences Laboratories, Materials Research Group, iTHEMBA LABS, P.O. Box 722, Somerset West 7129, Faure, South Africa b Physics Department, University of Zululand, Private Bag X1001, South Africa c Laboratoire d’Optique des Solides, Universite Pierre-Marie Curie, Paris 75006, France d Laboratoire Surface and Interface, Universite du Maine, Le Mans, France e Physics and Astronomy Department, Western Washington University, Bellingham, Washington, DC 08544, USA f Princeton Materials Institute, Princeton University, Princeton, NJ 08544, USA g Groupe de Physique des Solides et Sciences des Materiaux, Universite Cheikh Anta-Diop de Dakar, Fann, Senegal Received 24 February 2005; accepted 1 December 2005 Available online 3 February 2006
Abstract Room temperature magneto-optical and radar absorption measurements were carried out on Fen–V2O5 nano-composites prepared by RF co-sputtering. The relationship between the Fe atomic content and the state of the matrix precursor as well as the metal nano-particles’ size are discussed with emphasis on the interfacial oxidation of the Fe nano-particles. The magneto-optical and radar absorption responses were found to be effective above the Fe nano-particles percolation threshold nC. Using three different complementary techniques, this critical value was found to be about 23% atomic. From absorption viewpoint, these Fen–V2O5 nano-composites could be adequate radar absorbers at the specific radar frequency of 9.45 GHz. Below the critical atomic concentration value of 23%, the radar absorption is weak and relatively broad. 2005 Elsevier B.V. All rights reserved. Keywords: Nano-composites; Percolation threshold; Magneto-optics; Radar absorption; Core-shell nano-materials
1. Introduction Nano-composite structures, sometimes called granular cermets, exhibit an interesting state of matter intermediate between the bulk crystalline and the amorphous state. They consist of metallic nano-particles embedded in a second immiscible dielectric host matrix. Among other features, their linear and nonlinear optical properties depend upon the volume fraction of the metallic nano-particles. From fundamental viewpoint, these nano-composites provide likely artificial candidates to substantiate numerous phenomena linked to percolation processes [1–5]. From technological aspect, they exhibit various optical applications
*
Corresponding author. Tel.: +27 21 843 1145; fax: +27 21 843 3543. E-mail address:
[email protected] (M. Maaza).
0925-3467/$ - see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.optmat.2005.12.005
in particular as efficient and selective UV–VIS absorbers in solar energy thermal conversion [6,7]. In the NIR spectral region, their optical properties are frequency independent, exhibiting an anomalous maximum optical IR absorption in the percolation range. This large IR absorption is attributed to surface plasmon modes broadened towards IR wavelengths due to the fractal nature of the metallic nano-particles network [8]. More precisely, this substantial IR absorption is caused by a confinement effect of the conduction electrons in the fractal finite size metallic particles. If these latter nano-particles hosted within the dielectric matrix consist of noble metals, their strong interaction with light through the resonant excitations of the collective oscillations of the conduction electrons, induces high order optical nonlinearities. Since these metallicinsulator nano-composites in general are macroscopically isotropic, their second nonlinear optical response is
M. Maaza et al. / Optical Materials 29 (2007) 760–765
negligible while their optical nonlinear response is mainly of the third order. This latter is enhanced substantially at the vicinity of the plasmon frequency [9–11]. Such a strong enhancement of the third nonlinear optical susceptibility v(3)(x) was found to be induced by the enhanced local electric field within the metallic nano-particles as alluded to previously by Ricard et al. [12]. If the metallic nano-particles hosted within the dielectric matrix are in addition ferromagnetic or paramagnetic, these nano-composites could exhibit a giant magneto-resistance effects as demonstrated in Co–Al2O3 and Fe–SiO2 systems by Fujomori et al. [13] and Xu et al. [14], respectively. This recent discovery has induced a new interest in the field of this variety of metal-dielectric nano-composites in the field of spintronics due to their spin/tunneling dependence transport properties. Likewise, it was observed experimentally that ferromagnetic-dielectric nano-composites of NiFe(n)–SiO2(1n) display an enhanced Hall resistivity close to percolation threshold nC 0.55, 4 orders of magnitude greater than that of a homogeneous ferromagnetic metal [15]. Above this unexpected enhanced Hall transport properties, it has been shown that the standard percolation theory as well as the effective medium approximation cannot explain the behavior of such a transport phenomena in ferromagnetic metal-dielectric nano-composites. The disagreement between the theory and experiments is the most impressive in the case of the so-called giant extraordinary Hall effect in percolating systems such as NiFe(n)–SiO2(1n) [15]. The reported extraordinary Hall effect is correlated to conduction electrons in the ferromagnetic component. So, this disagreement might be due to the presence of an additional scattering mechanism in the ferromagnetic/paramagnetic metal-dielectric nano-composites which was not previously taken into account. Brouers et al. have considered an additional scattering term related to boundary scattering at the interfaces between the ferromagnetic/paramagnetic metallic nano-particles and the dielectric host matrix [16]. According to this previous anomalous spin dependent responses of ferromagnetic/paramagnetic nano-sized inclusions in dielectric host matrix, this species of nano-structures can be viewed as a new class of inhomogeneous magneto-optical nano-materials. Their optical and magneto-optical properties are much more complex to eluci-
761
date than the homogeneous ferromagnetic materials [17]. The ferromagnetic-dielectric coupled nature of these nano-composites could be exploited in radar absorption applications. Indeed, as shown by Che et al., nano-sized Fe encapsulated in carbon nanotubes exhibit an enhancement in microwave absorption [18]. Similarly, the investigations of Fe–carbon and FeB/carbon nano-composites by Liu et al. have showed, that these magnetic-amorphous carbon nano-structures display almost constant relative permittivity in the range of 0.05–20.05 GHz, while the imaginary part of the relative permeability exhibited a wide peak in the 1–9 GHz range. Both resin nano-composites exhibited good electro-magnetic wave absorption properties ‘‘RL < 20 dB’’ in the 4.4–8.3 GHz and in the 7.5– 16 GHz ranges [19]. Vendange et al. [20] and Shen et al. [21,22] have extended such radar absorbing nano-composites to the standard granular nano-composites such as M–Fe3O4 with M = Fe, Co and Ni. It was found that Fe–Fe3O4 is the likely candidate for radar absorption. This communication reports the room temperature magnetooptical and radar absorption properties of Fen–V2O5 below and above the percolation threshold nC of about 23%. 2. Experimental and discussion Different Fen–V2O5 nano-composites were grown by radio-frequency co-sputtering on float glass as well as onto quartz substrates with a very low surface root mean square roughness rRMS of 0.5 and 1.8 nm, respectively (10 · 10 · 2.5 mm3 and / 10 mm · 1 mm in size, respectively). The target consisted of a pure circular pressed powder of V2O5 of 130 mm diameter ‘‘Johnson–Matthey 99.97%’’ on which small circular Fe plates of 5 mm in diameter were placed in an hexagonal array so to ensure a superior chemical plasma homogeneity during the co-sputtering process. This target configuration allows to synthesize nano-composites whose chemical composition could be varied over a wide atomic range, depending on the number of metallic discs used. A rotating sample holder provided nano-structured samples with a good chemical homogeneity both in the basal and transversal directions. Before the sputtering phase, the Fen–V2O5 target was submitted to a pre-sputtering phase during a period of about 30 min to remove the oxide layer from the surface of the Fe pellets,
Table 1 Summary of room temperature chemical, morphological, structural, electrical, magneto-optical & radar absorption characteristics of the Fe–V2O5 nanocomposites below and above the percolation threshold Fe pellets
19
42
55
61
85
Fe (%) Atomic concentration Film thickness (nm) Matrix structure h/Fei (nm) ± 0.7 nm hR/hS (kX/square) Resistance (kX/square) Radar absorption (%) at 9.45 GHz Absorption width (MHz) at 9.45 GHz
17.38 51.2 a-V2O5 1.9 0.00 52 – –
19.82 64.0 a-V2O5 2.6 0.00 56 – –
23.00 70.7 a-V2O5 3.8 0.00 9.9 20.4 2.8
25.10 97.0 c-V2O5 Obloids 0.67 2.5 31.2 3.3
32.66 113.1 c-V2O5 Textured Percolated network 0.63 1.88 82.6 3.6
109 34.77 130.6 c-V2O5 ! 0.71 1.08 87.2 4.5
127 38.10 147.3 c-V2O5 0.81 1.03 100 6.1
762
M. Maaza et al. / Optical Materials 29 (2007) 760–765
Fig. 1. High resolution scanning electron microscopy cross-section of a non-percolated Fen–V2O5 nano-composites. The dashed lines are an eye guide only to support the columnar structure of the V2O5 host matrix.
if any. During sputtering deposition, the argon pressure and the deposition rate are kept at the values of about 7 · 103 Torr and 5.0 nm/min, respectively. The deposited nano-composites were yellow in color to brown–black in appearance, depending on the Fe concentration. The Fe atomic volume fraction ‘‘n’’ within the different Fen–V2O5 nano-composites was deduced from electron microprobe analysis as reported in Table 1. It varies within the range of 17.38–38.10%. Likewise, Table 1 reports their thicknesses derived from 2D Tencor profilometry measurements. The nano-composites thickness varies within the range of 51.21–47.3 nm ± 0.8 nm. The different synthesized Fen–V2O5 nano-composites showed a uniform thickness homogeneity and a relatively low surface roughness. Fig. 1 reports a typical high resolution scanning electron microscopy cross-section of a non-percolated Fe–V2O5 nano-structure (Fe23.0%–V2O5). One can clearly distinguish the Fe based nano-particles which are generally quasispherical in shape. A close analysis of the cross-section reveals an obvious columnar structure of the host V2O5 matrix. The basal thickness of these V2O5 columns is about 8–11 nm approximately. The dashed lines delimiting neighboring columns of Fig. 1 are an eye guide only. This columnar growth morphology seems to be inherent to these metal-dielectric nano-composites synthesized by radiofrequency specifically as in the case of Pt-Al2O3 and Aun–VO2 [23,24]. The morphological investigations with both high resolution scanning and electron transmission microscopies showed that for an initial low Fe concentration, the nano-composites were amorphous with Fe or Fe oxide nano-particles in the range of 2–4 nm in average diameter as conveyed in Table 1. Fig. 2 reports room temperature X-ray diffraction ‘‘XRD’’ patterns of the different Fen–V2O5 nano-composites. These XRD measurements were carried out on a Phillips XRD type unit with a Ni˚ ’’ over the angular range of filtered CuKa1 ‘‘k = 1.545 A
Fig. 2. Room temperature X-ray diffraction patterns of the different Fen– ˚ ). V2O5 nano-composites (filtered Ni CuKa1 = 1.54 A
M. Maaza et al. / Optical Materials 29 (2007) 760–765
10–90. The total accumulation time was identical for all concerned samples i.e., of about 3 h 30 min. As depicted by Fig. 2, the Fen–V2O5 nano-composites corresponding to n 6 23% exhibit amorphous type patterns suggesting that both Fe based nano-particles and the host matrix V2O5 are in an amorphous state. For larger values of n, the observed Bragg intense peaks fit with an ordered dielectric matrix V2O5 indicating its high degree of crystallinity (ASTM CARD 18-0251). Excluding the Fe32.66%–V2O5 nano-composite, the XRD patterns above 23.00% reveal a significant texture effects. As a first deduction one could interpret the prominent change in the crystallographic behavior of the Fen–V2O5 above the value n 23% as a possible signature of the percolation process, a concentration from which the Fe based nano-particles begin to generate a 3D spatial network. To ascertain such a possible assumption, room temperature transport measurements were carried out with a two contacts probe on all Fen– V2O5 nano-composites. The corresponding experimental DC electrical resistance per square results are reported in Fig. 3 as well as in Table 1. As for the previous XRD measurements, it is apparent that two areas are concerned. Below n 6 23%, the electrical resistance per square is of the order of 55 kX (insulating behavior) while it decreases significantly in a sharp way above this critical threshold to reach values of about 1 kX (quasi-metallic behavior). Thus, the significant drop of the electrical resistance around n 23% Fe atomic concentration, implies obviously to consider this value as the percolation threshold. Therefore, one should expect that both magneto-optical and radar absorption properties will be affected significantly at this Fe atomic concentration threshold. The magneto-optical Faraday rotation measurements were carried out at room temperature with a fixed laser wavelength of 632.8 nm (collimated and linearly polarized) impinging at a normal incidence onto the nano-composites deposited onto float-glass and quartz substrates. The polarizer–analyzer axis was initially fixed at 45. The external magnetic field, perpendicular to the sample surface i.e., parallel to the laser beam propagation direction, was varied in the range of 0–13 · 103 Oe. Before each Faraday rota-
Fig. 3. Room temperature electrical resistance of the different Fen–V2O5 nano-composites versus the initial Fe pellets. This latter could be converted to Fe atomic percentage using Table 1.
763
Fig. 4. Room temperature normalized Faraday rotation hR/hS of the different Fen–V2O5 nano-composites versus the initial Fe pellets. This latter could be converted to Fe atomic percentage using Table 1. kLASER ¼ 632:8 nm, HINC = 90, polarizer–analyzer axis was fixed at 45. EXC
tion measurement, room temperature transmission spectrum of each sample is taken both with and without the external magnetic field. A large number of scans are averaged in each case to increase the signal/noise ratio in particular for the lowest Fe concentrations (Table 1). Fig. 4 depicts the concentration dependence of the normalized room temperature Faraday optical rotation hF. This normalized hF is the ratio of the Faraday rotation value at saturation hS/hR Faraday rotation value at remanence hR. This normalized value is very low, almost 0 for Fe atomic concentration 623.00% while it increases smoothly for a higher Fe concentration content n: 25.10–38.10%. The insignificant Faraday rotation values exhibited by lower Fe concentration nano-composites (623.0%) could be explained merely by considering the well established coreshell structure of the Fe nano-particles with an oxidized Fe shell as shown in the inset of Fig. 4. In addition to the high surface/volume ratio, due to its oxygen affinity, Fe nano-particles have a tendency to adopt such a coreshell morphology. The shell, which acts as a passivation barrier, is generally a stable Fe oxide [25]. The spatial extension of this passivation iron oxide shell depends on the synthesis techniques and decisively upon the synthesis conditions; its thickness is characteristically 0.9–3.0 nm [26,27]. If one considers the average size of the Fe nanoparticles derived from the transmission electron microscopy measurements of Table 1, one could perceive that the average values of h/Fei are equivalent, at certain extent, to the oxidized shell thickness for concentration smaller than 23%. In light of this, one should deduce that the Fe nano-particles within the V2O5 dielectric matrix are mainly iron oxide of Fe2O3 type. As this oxide family is generally antiferromagnetic [28] with a lower magnetic moment per atom relatively to pure Fe, one could a priori explain the insignificant Faraday rotation values below the explicit value of 23% as hF is proportional to the magnetization at saturation of the magnetic nano-sized inclusions:
764
M. Maaza et al. / Optical Materials 29 (2007) 760–765
p hF 2pDcMS e/c, D is nano-composite thickness, MS is magnetization at saturation, e is dielectric constant and c is celerity of light in vacuum. As a partial conclusion, one could deduce that the current Fen–V2O5 nano-composites are sensitive from magneto-optical viewpoint only if the Fe concentration is above 23%. Consequently, one should expect that this limit will influence the radar absorption response too. To support this core-shell structure of the Fe based nano-particles, Mossbauer spectroscopy investigations were performed on the different samples. While the Mossbauer spectra for all samples with n P 23.00% display identical and rich patterns, the nonpercolated nano-composites i.e., with n 6 23.00% present alike but plain patterns. Fig. 5 reports two representative spectra for n 6 23.00% and n > 23.00%. The Mossbauer spectra corresponding to nano-composites with n = 38.10% and 23.00% are reported in Fig. 5a and b, respectively. While the highest percolated sample (n = 38.10%) spectrum corroborates mainly with pure iron with a minor a-Fe2O3 contribution (Fig. 5a), the parameters of the best fitted spectra of the non-percolated samples (n 6 23.00%) indicate the presence only of the doublet signature of pure a-Fe2O3 (Fig. 5b). There is no signature of c-Fe2O3 or any other phase of iron present in these non-percolated samples. More precisely, the isomer shift d and the quadrupole splitting D are about +0.35 and 0.71 mm/s, respectively consistent with nano-scaled a-Fe2O3; the micro-sized values are of the order of d +0.37 and D 0.20. These Mossbauer results are coherent with the XRD, electrical resistivity and magneto-optical results in the qualitative sense regarding the percolation threshold. Likewise, it seems that they sustain the hypothesis of an core-shell structure of the Fe nano-particles; more specifically Fe and a-Fe2O3 as core and shell, respectively.
Fig. 6. Normalized radar absorption at 9.45 GHz of the different Fen– V2O5 nano-composites versus normalized frequencies in the 9.45 GHz region.
Following these previous morphological and magnetooptical investigations, room temperature radar absorption measurements were performed on the different Fen–V2O5 nano-composites with a specific emphasis at the well targeted frequency of 9.45 GHz. The different absorption values were normalized relatively to the largest measured value corresponding to the most concentrated sample in iron, namely Fe38.10%–V2O5 nano-composite (Table 1). As shown in Fig. 6, higher is the Fe concentration, larger is the radar absorption structure factor (RAS) at the targeted frequency of 9.45 GHz. Even if the variation with the Fe atomic concentration is almost linear in the form of RAS 5.36n100.17 (v2 99.3%), the relative radar absorption is indeed substantial above the percolation threshold. This could be explained by the existence of rich iron nano-sized areas, substantiated by Mossbauer measurements as well as the fractal-like geometry of these nano-composites above the percolation threshold [29,30]. While the rich Fe nanoregions will affect the magnetic permeability of the percolated nano-composites, the interconnected configuration, generally a fractal type [29], with micro/meso porosity promotes extensively non-specular scattering effects as in other typical systems, for example PMMA-In2O3 and glass/carbon black composites [31,32]. The width at half maximum of the radar absorption peaks seems to increase slightly with the Fe concentration. The modeling of the n dependence of both the intensity of the radar absorption as well as its width at half maximum is being investigated. 3. Conclusion
Fig. 5. Typical Mossbauer spectra: (a) below (Fe17.38%–V2O5) and (b) above (Fe38.10%–V2O5) percolation threshold.
Microwave absorbing nano-composites consisting of Fe nano-particles embedded in a dielectric host matrix of
M. Maaza et al. / Optical Materials 29 (2007) 760–765
V2O5 were investigated by different techniques, namely transmission electron microscopy, X-ray diffraction, electrical resistivity, Mossbauer spectroscopy, magneto-optical and radar absorption. From morphological viewpoint, the Fe nano-particles seems to corroborate with a core-shell structure with an inner core rich in iron while the shell is typically a-Fe2O3. The metal iron content nFe, was varied within the atomic range of 17.3–38.1% to maximize the Fe core spatial extension within the V2O5 host matrix. Due to the ferromagnetic-dielectric configuration, the Fen–V2O5 nano-composites have been found to exhibit a slight magneto-optical effect but a substantial radar absorption at 9.5 GHz above the Fe percolation threshold nC of about 23.0% atomic.
[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
Acknowledgements This research program was financially supported by the International Liaison Office of the National Research Foundation, the French Centre National pour la Recherche Scientifique. We are grateful to the Rand Afrikaans Universiteit and the University of the Witwatersrand Research Council for their support. We are grateful to Mrs. A.M. Dawe and Prof. M.J. Witcomb from the University of the Witwatersrand for their technical assistance. We are indebted to the African Laser Centre for their valuable financial support.
[21] [22] [23] [24]
[25] [26]
References [27] [1] B. Abeles, P. Sheng, M.D. Coutts, Y. Arie, Adv. Phys. 24 (1975) 407. [2] J. Lafait, D.B. Tanner (Eds.), ETOPIM 2, Proceedings of the 2nd International Conference on Electrical Transport and Optical Properties of Inhomogeneous MediaPhysica A 175 (1989) 1. [3] J. Lafait, A. Baghdadi, P. Gadenne, Opt. Commun. 65 (1988) 17. [4] F. Brouers, J.P. Clerc, G. Firaud, J.M. Laugier, Z.A. Randriamantany, Phys. Rev. B 47 (1993) 666. [5] P. Sheng, Introduction to Wave Scattering, Localization and Mesoscopic Phenomena, Academic Press, San Diego, 1995, p. 1. [6] C.G. Granqvist, Appl. Opt. 20 (1981) 2606. [7] T. Robin, B. Souillard, Opt. Commun. 71 (1989) 15. [8] S. Berthier, J. Peiro, S. Fagnent, P. Gadenne, Physica A 241 (1997) 1.
[28] [29] [30]
[31] [32]
765
D. Ricard, Ph. Roussignol, C. Flytzanis, Opt. Lett. 10 (1985) 511. O. Levy, D.J. Bergman, D.G. Stroud, Phys. Rev. B 52 (1995) 3184. C. Flytzanis, Prog. Opt. 29 (1992) 2539. F. Hache, D. Ricard, C. Flytzanis, U. Kriebig, Appl. Phys. A 47 (1988) 347. H. Fujimori, S. Mitani, S. Ohnuma, J. Magn. Magn. Mater. 156 (1996) 311. Y. Xu, B. Zhao, X. Yan, J. Appl. Phys. 79 (1996) 6137. A.B. Pakhomov, X. Yan, B. Zhao, Appl. Phys. Lett. 67 (1995) 3497. F. Brouers, A. Granovsky, A. Sarychev, A. Kanditsov, Physica A 241 (1997) 284. E.A. Ganshina, A. Granovsky, V.S. Gushin, in: SMM-12, Proc. Soft Magnetic Materials Conference, Krakov, 1995, p. 146. R.C. Che, L.M. Peng, X.F. Duan, Q. Chen, X.L. Liang, J. Sol–Gel Sci. Technol. 11 (3) (1998) 299. J.R. Liu, M. Itoh, T. Horikawa, M. Itakaru, N. Kuwano, K. Machida, J. Phys. D: Appl. Phys. 37 (2004) 2737. V. Vendange, E. Tronc, Ph. Colomban, in: Towards Microwave Absorbing Metal-Oxide Nano-composites: The Problem of MetalMatrix Reaction, Report ONERA, National Office for Aerospace Research, Chatillon, France, 2004, p. 1. O.C. Shen, J. Ding, W. Lezhong, Adv. Mater. 16 (5) (2004) 401. O.C. Chin, J. Ding, W. Lezhong, J. Sol-Gel Sci. Technol. 11 (3) (1998) 299. M. Maaza, O. Nemraoui, C. Sella, A.C. Beye, B. Baruch-Barak, Opt. Commun. 254 (1–3) (2005) 188. M. Maaza, A. Gibaud, C. Sella, B. Pardo, F. Dunsteter, J. Corno, F. Bridou, G. Vignaud, A. Desert, A. Menelle, Eur. Phys. J. B 7 (1999) 339. E. Bonetti, L. Savini, A. Deriu, G. Albanese, J. Moya, J. Magn. Magn. Mater. 262 (2003) 132. A.H. Morris, K. Haneda, X.Z. Zhou, in: G.C. Hadjipanayis, R.W. Siegel (Eds.), Nanophase Materials: Surface Magnetism of Nanometer Particles, Kluwer Academic Publishers, 1994, p. 515. A.H. Morrish, in: J.L. Dormann, D. Fiorani (Eds.), Studies of Magnetic Properties of Fine Particles, Elsevier, Amsterdam, 1992, p. 181. G.A. Fergusson Jr., M. Hass, Phys. Rev. 112 (1958) 1130. V.M. Shalaev (Ed.), Electromagnetic Properties of Small-Particle Composites, Phys. Rep. 272 (1996) 61. V.M. Shalaev, Optical nonlinearities of fractal composites, in: Vladimir M. Shalaev (Ed.), Optical Properties of Nanostructured Random Media, Springer, Berlin, 2002, p. 1. Y. Shan, Y. Zhou, Y. Cao, Q. Xu, H. Ju, Z. Wu, Mater. Lett. 58 (10) (2004) 1655. J.-H. Oh, K.-S. Oh, C.-G. Kim, C.-S. Hong, Composites B: Eng. 35 (1) (2004) 49.
