Complex impedance spectroscopy of Mn-doped zinc oxide nanorod films

Solid State Communications 151 (2011) 1182–1187

Contents lists available at ScienceDirect

Solid State Communications journal homepage: www.elsevier.com/locate/ssc

Complex impedance spectroscopy of Mn-doped zinc oxide nanorod films M.K. Sharma a , R.N. Gayen b , A.K. Pal b , D. Kanjilal c , Ratnamala Chatterjee a,∗ a

Magnetics and Advanced Ceramics Laboratory, Department of Physics, Indian Institute of Technology Delhi, New Delhi 110016, India

b

Department of Instrumentation Science, USIC Building, Jadavpur University, Calcutta 700 032, India

c

Inter University Accelerator Centre, Aruna Asaf Ali Marg, New Delhi 110067, India

article

info

Article history: Received 7 December 2010 Received in revised form 18 March 2011 Accepted 27 April 2011 by D.D. Sarma Available online 20 May 2011 Keywords: A. ZnO nanorods B. Mn doping D. Impedance spectroscopy D. X-ray photoelectron spectroscopy

abstract The frequency-dependent properties of Mn-doped (3–5 at.%) aligned zinc oxide (Mn–ZnO) nanorods, synthesized by hybrid wet chemical route onto glass substrates, were investigated by bias-dependent impedance spectroscopy. No peak of Mn cluster/secondary phases was detected in the X-ray diffraction traces of the samples. XPS studies show the presence of oxygen vacancies in Mn–ZnO nanorods and Mn in Mn2+ and Mn4+ charge states. Although X-ray diffraction/X-ray photoelectron spectroscopy does not give any indication of the presence of metal clusters in the samples, bias-dependent impedance spectroscopy demonstrates significant sensitivity to the formation of Mn clusters in Mn–ZnO nanorods. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction There has been great excitement in reporting ferromagnetism with Curie temperature well above room temperature in transition metal (TM) doped diluted magnetic semiconductors (DMSs) or insulating oxides for advanced spintronic applications [1]. Transition metal doping into ZnO, TiO2 , SnO2 , or HfO2 showed room temperature ferromagnetism [2–5], and the presence of defects or cation/oxygen vacancies [6–8] in materials in thin film form was associated with the magnetism observed in the systems concerned. A key method for realizing a good DMS structure is to substitute the magnetic ions in the matrix of metal oxides and accordingly suppress the formation of magnetic clusters/secondary phases. Thus an important aspect of research in this field is to identify whether or not the large magnetization behavior in the sample is due to magnetic clusters. The existence of magnetic clusters in host materials is often difficult to detect by X-ray diffraction (XRD) or transmission electron microscopy (TEM), due to the fact that the size of the clusters is of the order of subnanometers. The sub-nanometer sized clusters also go undetected in temperature-dependent magnetization measurements as the blocking temperature (TB ) decreases with decreasing size of

∗

Corresponding author. E-mail addresses: [email protected], [email protected] (R. Chatterjee). 0038-1098/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ssc.2011.04.028

clusters, and thus may exceed the low temperature measurement limit of SQUID magnetometer systems [9]. It is known that the electrical properties of materials depend on different contributions from various components of the material. Overall properties of the host dielectric medium would be modulated by intra-grain and inter-grain defects, if any, and through electrode processes at higher temperatures (>800 K). The relaxation time (τ = RC ) or frequency response of complex impedance is known to provide evidence to identify various electrical relaxation mechanisms, arising from grains, grain boundaries, or macroscopic inhomogeneities in the form of equivalent circuit analysis [10,11]. Huang et al. [12] have recently shown that biasdependent impedance spectroscopy may be a useful tool to detect the presence of magnetic clusters in transition-metal-doped ZnO films. The ferromagnetic behavior of the Mn–ZnO nanorods under study may arise due to (i) dilute magnetic structure, (ii) lattice defects, (iii) secondary phase, or (iv) clustering of transition metal dopants. The aim of this work is to use complex impedance spectroscopy as a tool for probing the cause of ferromagnetism observed in these Mn–ZnO nanorods. 2. Experimental details Vertically aligned ZnO nanorods were deposited by wet chemical route onto glass substrates on which ZnO seed particles were pre-deposited by the sputtering technique. Thus, this technique is essentially a hybrid technique, where one may easily

M.K. Sharma et al. / Solid State Communications 151 (2011) 1182–1187

manipulate the size and the distribution of ZnO seed crystallites. It should be clarified that although ZnO grows preferentially (in nanorod form) over the ZnO seed particles, during the chemical deposition process, modulating growth of ZnO over the whole substrate is expected. However, due to the presence of ZnO seed particles, the ZnO is shown to preferentially deposit as nanorods on these seed sites. The thickness of the chemically deposited ZnO film on the rest of the substrate would be negligible as compared to the nanorod length. Details of the synthesis of the above ZnO nanorods are reported elsewhere (Ref. [13] and reference therein). Mn was evaporated on the ZnO nanorods at a system pressure ∼10−6 mbar for three different durations so as to deposit Mn layer with thickness ∼3 nm, ∼6 nm and ∼10 nm keeping the rate of evaporation fixed at ∼0.2 nm/s. The amount of manganese was controlled and recorded by a quartz crystal thickness monitor. The ZnO nanorods containing manganese, thus obtained, were then subjected to rapid thermal annealing (RTA) in argon atmosphere for dispersing Mn in ZnO nanorods. The temperature of annealing was 773 K and duration was 3 min. After annealing, the final composition of the films as estimated by X-ray photoelectron spectroscopy data, were found to be ∼3 at.%, ∼4 at.% and ∼5 at.% for 3 nm, 6 nm and 10 nm Mn-deposited ZnO nanorod films, respectively. An FEI Quanta 200 scanning electron microscopy (SEM) was used to record the surface morphology at an operating voltage of 25 kV in back scatter mode. A Rigaku MiniFlex X-ray diffraction (XRD) using Cu Kα radiation was used to obtain the structural information. The XPS spectra were recorded using a VG Microtech at a base pressure ∼10−10 mbar. The XPS setup used here is incorporated with a dual anode Mg–Al X-ray source with a hemispherical analyzer and a channeltron detector with a resolution 1.3 eV. Monochromatic Mg K α radiation (1253.6 eV) was used at 300 W for our XPS measurements. Data analysis was done using VGX900 software incorporated with the system. The complex impedance spectroscopy was carried out using Agilent 4294 A impedance analyzer with two point contacts in a frequency range from 100 Hz to 20 MHz and a fixed oscillating voltage 500 mV under dc bias voltage Vdc from 0 to 1.5 V. The geometry in which the measurement is done (i.e., two probes attached to the top of the sample) takes care of the impedance contributions from complete sample including nanorods and layer deposited on substrate during chemical deposition. By applying a dc bias voltage the relaxation contribution from different structural origin can be clearly identified. 3. Results and discussion The SEM pictures of ∼3 at.%, ∼4 at.% and ∼5 at.% Mn-doped ZnO nanorods are shown in Fig. 1. XRD traces of the as-deposited and annealed Mn–ZnO nanorods are shown in Fig. 2(a)–(f), respectively for all samples. XRD trace show that the vertically aligned ZnO nanorods are polycrystalline and preferentially [002] direction oriented. It may be noted that the characteristic peak for Mn (111) present in the XRD traces of the as-deposited films were masked due to the presence of very strong (002) peak of ZnO nanorods and hence in Fig. 2 the truncated (002) peak of ZnO is shown so as to reveal the (111) Mn peak located at 2θ ∼ 22.1°. It may be observed that no distinct peaks for Mn are observed in ZnO nanorods containing ∼3 at.% Mn. As Mn concentration increases, Mn (111) peak at 2θ ∼ 22.1° evolve more clearly in the XRD traces (Fig. 2(c) and (e)). It should be noted that after RTA, none of the samples show Mn (111) peak, indicating homogeneously distributed Mn in the ZnO matrix without affecting the observed vertical alignment of the ZnO (see Fig. 2(b), (d) and (f)). Lattice constants, a and c, of the Mn–ZnO nanorod films were computed using computer software package ‘XPowder’ [14] for the annealed

1183

a

b

c

Fig. 1. SEM pictures of Mn–ZnO nanorod films with different amount of Mn: (a) Mn ∼ 3 at.%; (b) Mn ∼ 4 at.%; and (c) Mn ∼ 5 at.%. Table 1 Calculated lattice parameters (in nm), percentage of Mn2+ , Mn/Zn ratio and Zn/O ratio for Mn–ZnO nanorod films with different amount of Mn: (i) Mn ∼ 3 at.%; (ii) Mn ∼ 4 at.%; and (iii) Mn ∼ 5 at.%. Sample

Lattice constants

ZnO

a c a c a c a c

Mn–ZnO (Mn ∼ 3 at.%) Mn–ZnO (Mn ∼ 4 at.%) Mn–ZnO (Mn ∼ 5 at.%)

∼ 0.324 ∼ 0.519 ∼ 0.326 ∼ 0.519 ∼ 0.326 ∼ 0.519 ∼ 0.325 ∼ 0.519

Percentage of Mn2+

Mn/Zn

Zn/O

3

0.133

0.428

4

0.194

0.547

5

0.244

0.525

films and are shown in Table 1. It may be seen that the lattice constants agreed well with the bulk values (a = 0.32495 nm; c = 0.52069 nm) [15]. Results included in this communication hereinafter would relate to the rapid thermal annealed Mn–ZnO nanorod films, unless otherwise stated. The binding states of the compositional elements of Mn–ZnO nanorods were characterized by XPS studies. Fig. 3(a)–(c) show the typical core level peaks corresponding to Zn 2p, O 1s and Mn 2p of the Mn–ZnO nanorods with increasing amount of Mn incorporated in them. The Mn 2p3/2 peak of Mn–ZnO nanorods located at 643.2 eV could be deconvoluted in two peaks located at ∼641.4 and 643.3 eV, which could be attributed to the existence Mn2+ and Mn4+ , respectively [16]. It is well known that, for one kind of element, ion with high valency has a larger binding energy. Generally, peaks for the metallic Mn and Mn4+ ion are located at

M.K. Sharma et al. / Solid State Communications 151 (2011) 1182–1187

36

48

60

(004)

(103)

(102)

(002)

(102)

(103)

(004)

72 24 2θ (degree)

36

48

60

(004)

(103)

(102)

(101)

(100)

(002)

f

(004)

(103)

(110)

(102)

(100)

(101)

(002)

Mn (111)

(101)

(100)

(103)

(110)

(102)

(101)

(100)

e

24

d

(004)

(002)

Mn (111)

(101)

(004)

(103)

(110)

(102)

(101)

(100)

c

(002)

b

(002)

a

(100)

1184

72

Fig. 2. XRD traces of the as-deposited Mn–ZnO nanorod films with different amount of Mn (a) Mn ∼ 3 at.%; (c) Mn ∼ 4 at.%; and (e) Mn ∼ 5 at.%. Fig. 2(b), (d) and (f) show the corresponding XRD traces after subjecting them to RTA.

O1s

Zn2p1/2

O1s

Zn2p1/2

528

1020 1030 1040 1050 1060 Binding energy (eV)

1010

534 530 532 Binding energy (eV)

O1s

Zn2p1/2

1020 1030 1040 Binding energy (eV)

Intensity (a.u.)

Intensity (a.u.)

Zn2p3/2

1050

Mn2p

Intensity (a.u.)

Intensity (a.u.)

Intensity (a.u.)

Zn2p3/2

c

639 642 645 648 Binding energy (eV)

528 530 532 534 Binding energy (eV)

1020 1030 1040 1050 1060 Binding energy (eV)

b

Mn2p

Intensity (a.u.)

Intensity (a.u.)

Intensity (a.u.)

Zn2p3/2

528 534 530 532 Binding energy (eV)

636

Intensity (a.u.)

a

636

639 642 645 648 Binding energy (eV)

Mn2p

639 642 645 648 Binding energy (eV)

Fig. 3. XPS spectra for RTA annealed Mn:ZnO nanorod films showing core level peaks corresponding to Zn 2p, O 1s and Mn 2p for: (a) Mn:ZnO (Mn ∼ 3 at.%); (b) Mn:ZnO (Mn ∼ 4 at.%) and (c) Mn:ZnO (Mn ∼ 5 at.%).

M.K. Sharma et al. / Solid State Communications 151 (2011) 1182–1187

637.7 eV and 642.4 eV as indicated in Ref. [17]. We have computed the peak areas of the different valence bonds of Mn and the percentages of Mn2+ in each sample and the results are shown in Table 1. The O 1s peaks with their deconvolution results for ZnO nanorods with different amount of Mn incorporated in them are also shown in Fig. 3. The peaks for O 1s could be deconvoluted in three peaks centered at ∼529.7, ∼531.3, and ∼533.2 eV. The high binding energy component centered at 533.2 can be ascribed to the chemisorbed oxygen of the surface hydroxyl, −CO3 , absorbed H2 O or absorbed O2 [18], which cannot be easily removed. The component with low binding energy centered at 529.7 eV could be attributed to the O2− ions in the ZnO lattice. The intensity of this peak would indicate the amount of oxygen atoms in the wurtzite structure of the hexagonal Zn2+ ion array. The medium binding energy centered at 531.3 eV could be related to the O2− ions in the oxygen-deficient regions within the matrix of ZnO, whose intensity partly would represent the variation in the concentration of oxygen vacancies. The Zn 2p spectrum (Fig. 3 exhibited two peaks at ∼1021.8 eV for Zn 2p3/2 and at ∼1044.8 eV for Zn 2p1/2 indicating the possibility of existence of multi-component Zn. From the ion radius difference between Mn2+ (0.66 Å) and Zn2+ (0.60 Å), Mn with larger ionic radius replacing Zn of the ZnO lattice could result in increasing the binding energy of Zn. This explains the position of the peak centered at 1021.8 eV being at higher energy than that of Zn in the bulk ZnO (1021.4 eV). The effective concentrations of Mn2+ incorporated into the films are found to be ∼3 at.%, ∼4 at.% and ∼5 at.% corresponding to Mn layer thickness of 3 nm, 6 nm and 10 nm, respectively for the ZnO nanorod samples. Relative amounts of Mn:Zn and Zn:O as computed above are shown in Table 1. It can be noted that the Raman spectroscopy measurements on these samples (not shown here) also indicate that the Mn substitutes at Zn site after annealing and wurtzite hexagonal structure of ZnO remains same after Mn doping [13]. In impedance spectroscopy, the frequency-dependent measurements Z (f ) = R(f ) + iX (f ), where Z (f ) is complex impedance of which R(f ) is the real part and X (f ) is the imaginary part) can be resolved into contributions from grain, grain boundary/interface or other contributions like electrode effect. It should be clarified that the ZnO nanorods (including the Mn-substituted ZnO nanorods) themselves are polycrystalline. Thus for these samples, result of impedance spectroscopy can be expressed in terms of the resistance (Rog ) and capacitance (Cog ) of grains present in polycrystalline ZnO in nanorod form and on film (including the Mnsubstituted ZnO) and Rogb , Cogb of grain boundaries/interfaces in the sample. The contribution from Mn-substituted ZnO nanorods cannot be separated, possibly due to the dilute nature of substitution. However, if some of the Mn ions do not replace Zn and instead are deposited as clusters in the grain boundary/interface regions, added contributions to the total impedance of the sample are expected in the form of resistance of metallic grains Rmg , and contributions from the resistance and capacitance of metal–oxide interface, namely, Rmo and Cmo . The real and imaginary parts of complex impedance, Z (f ) = R(f )+iX (f ), as a function of frequency for ∼3 at.%, ∼4 at.% and ∼5 at.% Mn-doped ZnO nanorod samples are shown in Fig. 4. In all these plots for Mn-doped ZnO samples, two clear steps (first one 5 × 102 < f < 5 × 104 and another in the range 5 × 106 < f < 2 × 107 ) are observed in the real part of complex impedance versus frequency plots. In contrast, the imaginary part of the impedance shows characteristic peaks at the frequency of about 5 × 103 and 5 × 106 Hz. These features in −X (f ) clearly indicate that multiple relaxation mechanisms are effective in our samples. Nyquist plots of our complex impedance (CI) data as per Eq. (1) (for ZnO nanorods (pristine)) and Eq. (2) for Mn-doped ZnO

1185

Fig. 4. The complex impedance spectra as a function of frequency for Mn–ZnO nanorod films (a) Mn–ZnO (Mn ∼ 3 at.%); (b) Mn–ZnO (Mn ∼ 4 at.%); and (c) Mn–ZnO (Mn ∼ 5 at.%). Table 2 Fitting results of the relaxation time constants for: (a) ZnO (Mn ∼ 0); (b) Mn–ZnO (Mn ∼ 3 at.%); (c) Mn–ZnO (Mn ∼ 4 at.%) and (d) Mn–ZnO (Mn ∼ 5 at.%) ZnO nanorod films. Sample

τog (s)

ZnO Mn–ZnO (Mn ∼ 3 at.%) Mn–ZnO (Mn ∼ 4 at.%) Mn–ZnO (Mn ∼ 5 at.%)

τogb (s)

τmo (s)

−7

1.15 × 10 1.33 × 10−7

−4

3.16 × 10 3.60 × 10−4

– 5.20 × 10−4

1.53 × 10−7

5.24 × 10−4

7.10 × 10−4

1.85 × 10−7

6.82 × 10−4

8.63 × 10−4

nanorods are shown in Fig. 5. For ZnO nanorods, two sets of parallel RC components in series and for Mn-doped ZnO nanorods three sets of parallel RC components in series together with a single resistance element (see Fig. 5), have been employed to model the CI spectra and corresponding equivalent circuits (EC) composed of resistance (R) and capacitance (C ) elements are shown in the insets. Z = R + iX = (1/Rogb + iωC ogb )−1 + (1/Rog + iωC og )−1 ,

(1)

Z = R + iX = Rmg + (1/Rmo + iωC mo )−1 + (1/Rogb

+ iωC ogb )−1 + (1/Rog + iωC og )−1 ,

(2)

The solid curves (i.e., superposition of the three dash curves from Eq. (2)) in Fig. 5 are obtained by the best fit of the Nyquist plots of –X vs R. The fitted curves reveal good agreement of the EC model with the CI spectrum. It is noticed that Rmg (∼102 ) is an order of magnitude smaller compared to Rmo , Rogb and Rog (∼103 ). Thus Rmg contributes much less to the overall value of the complex impedance. The relaxation time (τ = RC ) is an intrinsic characteristic property of the materials and is independent of the sample geometry. Owing to the distinct RC contributions, the oxide grain, oxide grain boundary and metal–oxide interface show different relaxation time constants. Thus, the contribution from various RC components clearly identified by frequency-dependent impedance spectroscopy [19], indicates a smaller but definite contribution from metal clusters (represent by subscript ‘mg’ here). It is expected that the grain interior exhibit faster relaxation behavior than grain boundary for semiconducting materials such as zinc oxide. The relaxation time constant due to oxide grain boundary (τogb ) is larger than the time constant due to oxide grain (τog ) by at least three orders of magnitude (see Table 2). For as grown ZnO nanorods and ∼3, ∼4 and ∼5 at.% Mn-doped ZnO nanorod samples, the time constants due to zinc oxide grain τog are of the order of 10−7 s, as fitted by the above equivalent circuit

1186

M.K. Sharma et al. / Solid State Communications 151 (2011) 1182–1187

b

104

104

-X (ohm)

-X (ohm)

a

103

102

103

102

103

104

104

103

R (ohm)

R (ohm)

c

d

104

-X (ohm)

-X (ohm)

104

103

103

102

102

103

101 103

104

R (ohm)

104

R (ohm)

Fig. 5. The Cole–Cole plots for: (a) ZnO; (b) Mn–ZnO (Mn ∼ 3 at.%); (c) Mn–ZnO (Mn ∼ 4 at.%); and (d) Mn–ZnO (Mn ∼ 5 at.%) nanorod films. The experimental data (open circle) are best fitted by the equivalent circuit equation (solid curves). The dash curves describe the contribution from the oxide grain, grain boundary and metal-cluster–oxide interfaces.

a

b 103 -X (ohm)

-X (ohm)

103

102

103

102

104

103

R (ohm)

104 R (ohm)

d

c

103 -X (ohm)

-X (ohm)

103

102

103

104 R (ohm)

102

104

103 R (ohm)

Fig. 6. The Cole–Cole plots for: (a) ZnO; (b) Mn–ZnO (Mn ∼ 3 at.%); (c) Mn–ZnO (Mn ∼ 4 at.%); and (d) Mn–ZnO (Mn ∼ 5 at.%) nanorod films at 0, 0.5, 1.0 and 1.5 V dc bias voltage (open symbols) and fitting results (solid curves).

models. Another mechanism with relatively slower relaxation is attributed to the presence of oxide grain boundary. For all the samples, the time constants due to oxide grain boundary τogb are in the range of ∼3–7 × 10−4 s. The assignment of the oxide grain interior and oxide grain boundary contribution is consistent with the ‘‘brick-layer’’ model for polycrystals [19]. The third electrical process appears in ∼3, ∼4 and ∼5 at.% Mn-doped ZnO nanorod samples with slowest relaxation (τmo ∼ 5–9 × 10−4 s) is likely associated with metal (cluster)–oxide interface.

To explore and identify possible relaxation mechanism in these nanorods, an additional dc bias voltage (Vdc ) has been applied to CI measurements. Excellent agreement of the equivalent circuit models and experiments has been confirmed with Vdc from 0 to 1.5 V. Fig. 6 shows the bias-dependent impedance spectra of as grown ZnO and ∼3 at.%, ∼4 at.% and ∼5 at.% Mn-doped ZnO nanorods with fitted solid curves. Fig. 7 shows the R and C fitting parameters plotted as a function of Vdc . The Rogb and Rmo decreases with increasing bias voltage Vdc while as expected, the zinc oxide

M.K. Sharma et al. / Solid State Communications 151 (2011) 1182–1187

a

1187

in the form of very small Mn nanoparticles in the grain boundary region, along with substituted Mn in ZnO matrix. Although no secondary phase was detected in XRD/SEM/XPS or Raman spectra, the CI measurements could give a qualitative estimate and show the definite presence of some metal clusters, even though in the form of very small Mn nanoparticles in the samples. 4. Conclusion

b

c

The aligned Mn–ZnO nanorods grown by hybrid wet chemical route on glass substrates do not show presence of Mn cluster/secondary phase in any of the films in our XRD data. XPS studies indicated that incorporated Mn was in Mn2+ and Mn4+ states. None of these usually used methods could identify the presence of any Mn clusters in the samples. However, clear presence of clusters/secondary phase in Mn–ZnO nanorod films were indicated in the bias-dependent impedance spectroscopy data analysis. The fitting parameters show that oxide grain boundary and metal–oxide resistance decreases with increasing biasing voltage whereas oxide grain contribution remains constant. On the other hand, oxide grain and oxide grain boundary capacitance remain unchanged with biasing voltage while metal–oxide capacitance decreases with biasing. These observations confirm that there are metal clusters in the samples that respond positively with biasing. Similarly grain boundaries of ZnO or Mn–ZnO nanorods/films with possible defects settled in them also show biasing effect; while as expected oxide grains remain unaffected by biasing voltage. Acknowledgments

d

MKS and RNG wish to thank the Council of Scientific and Industrial Research (CSIR) and the University Grants Commission, Government of India, respectively, for granting them fellowships. The authors would like to thank the DST SQUID facility at the Indian Institute of Technology, Delhi. References

Fig. 7. The variations of the fitting parameters for resistance and capacitance as a function of the dc bias voltage for: (a) ZnO (b) Mn–ZnO (Mn ∼ 3 at.%); (c) Mn–ZnO (Mn ∼ 4 at.%); and (d) Mn–ZnO (Mn ∼ 5 at.%) nanorod films.

grain contribution remains constant. The oxide grain boundary and metal–oxide interfaces may contain some defects that can provide conducting channels when carriers are injected from electrodes on application of bias; resulting in the trap states of carriers leading to the observation of decrease in Rogb and Rmo . On the other hand, Cog and Cogb remain almost unchanged with increasing Vdc from 0 to 1.5 V. The independence of Cog and Cogb on Vdc suggests that the oxide grain and grain boundary still effectively act as simple capacitors in spite of the increase of injected carriers by additional dc bias. In contrast, the appearance of Rmo and Cmo for Mn-doped ZnO nanorods confirms the formation of Schottky barriers caused by metal cluster–oxide interfaces. The bias-dependent impedance spectroscopy demonstrates significant sensitivity to the formation of Mn clusters in ZnO. It can also be observed from Figs. 6 and 7 that Mn doping mainly increases the grain boundary resistance together with keeping the grain-bulk resistance almost unchanged. This suggests the presence of some excess Mn probably existing

[1] W. Prellier, A. Fouchet, B. Mercey, J. Phys.: Condens. Matter 15 (2003) R1583. [2] K. Ueda, H. Tabata, T. Kawai, Appl. Phys. Lett. 79 (2001) 988. [3] Y. Matsumata, M. Urakami, T. Shono, T. Hasegawa, T. Fukumura, M. Kawasaki, P. Ahmet, T. Chikyow, S.-Ya Koshihara, H. Koinuma, Science 291 (2001) 854. [4] S.B. Ogale, R.J. Choudhary, J.P. Buban, S.E. Lofland, S.R. Shinde, S.N. Kale, V.N. Kulkarni, J. Higgins, C. Lanci, J.R. Simpson, N.D. Browning, S. Das Sarma, H.D. Drew, R.L. Greene, T. Venkatesan, Phys. Rev. Lett. 91 (2003) 077205. [5] N.H. Hong, J. Sakai, N. Poirot, A. Ruyter, Appl. Phys. Lett. 86 (2005) 242505. [6] P.V. Radovanovic, D.R. Gamelin, Phys. Rev. Lett. 91 (2003) 157202. [7] D.A. Schwartz, D.R. Gamelin, Adv. Mater. 16 (2004) 2115. [8] N.H. Hong, N. Poirot, J. Sakai, Appl. Phys. Lett. 89 (2006) 042503. [9] R. Goswami, G. Kioseoglou, A.T. Hanbicki, O.M.J. van’t Erve, B.T. Jonker, G. Spanos, Appl. Phys. Lett. 86 (2005) 032509. [10] A.K. Jonscher, Nature (London) 264 (1977) 673. [11] J. Han, P.Q. Mantas, A.M.R. Senos, J. Eur. Ceram. Soc. 20 (2000) 2753. [12] J.C.A. Huang, H.S. Hsu, Appl. Phys. Lett. 87 (2005) 132503. [13] M.K. Sharma, R.N. Gayen, A.K. Pal, D. Kanjilal, Ratnamala Chatterjee, Manuscript Communicated. [14] http://www.xpowder.com. [15] D.C. Kundaliya, S.B. Ogale, S.E. Lofland, S. Dhar, C.J. Metting, S.R. Shinde, Z. Ma, B. Varughese, K.V. Ramanujachary, L.S. Riba, T. Venkatesan, Nature Material 3 (2004) 709. [16] J.F. Moulder, Handbook of X-ray Photoelectron Spectroscope, Perkin-Elmer Corporation, Eden Prairie, 1979, p. 226. [17] Y.B. Lin, J.P. Xu, W.Q. Zou, L.Y. Lv, Z.H. Lu, F.M. Zhang, Y.W. Du, Z.G. Huang, J.G. Zheng, J. Phys. D: Appl. Phys. 40 (2007) 3674. [18] J.H. Li, D.Z. Shen, J.Y. Zhang, D.X. Zhao, B.S. Li, Y.M. Lu, Y.C. Liu, X.W. Fan, J. Magn. Magn. Mater. 302 (2006) 118. [19] J.R. Macdonald, Impedance Spectroscopy Emphasizing Solid Materials and System, Wiley, New York, 1987, p. 40, Chap. 4.