This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
IP Address: 140.113.38.11
This content was downloaded on 26/04/2014 at 07:22
Please note that terms and conditions apply.
Microstructure and magnetic properties of oxidized titanium nitride thin films in situ grown by
pulsed laser deposition
View the table of contents for this issue, or go to the journal homepage for more 2013 J. Phys. D: Appl. Phys. 46 075002
(http://iopscience.iop.org/0022-3727/46/7/075002)
J. Phys. D: Appl. Phys. 46 (2013) 075002 (7pp) doi:10.1088/0022-3727/46/7/075002
Microstructure and magnetic properties
of oxidized titanium nitride thin films
in situ grown by pulsed laser deposition
S C Chen
1, K Y Sung
1, W Y Tzeng
1, K H Wu
1, J Y Juang
1, T M Uen
1,
C W Luo
1, J-Y Lin
2, T Kobayashi
1and H C Kuo
31Department of Electrophysics, National Chiao-Tung University, Hsinchu, Taiwan, Republic of China 2Institute of Physics, National Chiao-Tung University, Hsinchu, Taiwan, Republic of China
3Department of Photonics and Institute of Electro-Optical Engineering, National Chiao-Tung University, Hsinchu, Taiwan, Republic of China
E-mail:[email protected]
Received 6 September 2012, in final form 18 December 2012 Published 23 January 2013
Online atstacks.iop.org/JPhysD/46/075002
Abstract
Different oxidation states of titanium nitride thin films, including pure TiN(h 0 0), TiN1−xOx(h 0 0), Ti2O3(0 0 l) and pure anatase TiO2(0 0 l), were prepared by pulsed laser
deposition with various oxygen pressures (PO2)using a TiN target. Elaborative evolutions of
the crystal and electronic structures of the obtained films were examined systematically by x-ray diffraction and x-ray absorption spectroscopy. We found that the Ti2O3(0 0 l) film, which was prepared at oxygen pressures PO2= 10−4Torr, exhibited the maximum room temperature
ferromagnetism (RTFM) behaviour. The bound magnetic polaron model is used to clarify the origin of RTFM in these films.
(Some figures may appear in colour only in the online journal)
1. Introduction
Titanium dioxide (TiO2) thin films have been studied and used extensively [1–6] since they possess remarkable optical, electronic, chemical and mechanical properties such as excellent optical transmittance (>85%) in the visible and near infrared wavelength ranges, high index of refraction (n≈ 2.35 at 550 nm), large dielectric constant (εr ≈ 105 at 4.2 K) and low loss tangent (tan δ ≈ 10−7 at 4.2 K) [7,8], high chemical stability and mechanical durability. The widespread applications of TiO2 films include (a) anti-reflecting and protective coating on optical elements; (b) capacitors or gates in microelectronic devices; (c) photocatalyst and catalytic devices; (d) optical waveguide in integrated optics and (e) suitable template layers for growing high-Tcsuperconducting YBa2Cu3O7 (YBCO) films, CrO2 and LaSrMnO3 films for microwave, biepitaxial junctions and spintronics applications [9–11].
Recently, transition metal-doped and oxygen-deficient TiO2 films have been demonstrated to exhibit room-temperature ferromagnetic (RTFM) behaviour and thus have
attracted extensive interest. In particular, the observation of RTFM in the undoped TiO2 films has been termed as d0-magnetism, or magnetism without unpaired d-electrons and is due mostly to defects and/or oxygen vacancies [12–15]. Numerous experimental and theoretical works have been reported ever since to explore the origins of the RTFM in these films. Recently, there were proofs supporting the notion that the existence of local ionic magnetic moment (Ti3+/Ti2+ or doped magnetic elements) accompanied by oxygen vacancies might be the origin for the observed magnetic order [12–15], albeit controversial results are not rare and the genuine mechanism responsible for the observed RTFM is still in extensive debate. Therefore, it is crucial to prepare films with various oxygen vacancy concentrations to systematically delineate the effect of oxygen vacancy on the RTFM manifested in the transition metal-free TiO2films.
The oxygen-deficient TiO2thin films have been prepared with several in situ or ex situ methods, include (1) in situ growth of TiO2 thin films by pulsed laser deposition (PLD) using the synthesized TiO2target under various oxygen partial pressures during deposition [16–18]; (2) in situ growth of
J. Phys. D: Appl. Phys. 46 (2013) 075002 S C Chen et al
TiOxNythin films by direct-current (dc) magnetron sputtering
using a metallic Ti target under various pressure ratios of oxygen and nitrogen [19]; (3) ex situ nitridation of TiO2 films, i.e. incorporating nitrogen into the anatase or rutile phase of TiO2 [20]; (4) ex situ oxidation of TiN films, i.e. incorporating oxygen into TiN [21–24]; (5) ex situ introducing structural disorder, defects and/or columnar amorphization into TiO2films by swift heavy ion irradiation [25]. However, no systematic investigations on the magnetic properties of these films have been carried out.
Previously, epitaxial single-phase rutile or anatase TiO2 thin films were successfully prepared on (1 0 0)-SrTiO3(STO) substrates with in situ PLD by our group [26]. It was found that, for films deposited on STO(1 0 0) substrates directly using a rutile TiO2 single crystal target, pure anatase TiO2(0 0 1) films were obtained even when the substrate temperature (Ts)
was higher than 1000◦C. On the other hand, pure rutile TiO2(1 1 0) films were obtained by in situ oxidation of TiN films immediately after they were obtained by PLD. The oxidation temperature was higher than 700◦C with the oxygen pressures (PO2) being kept at 5 Torr. It is apparent that
the specific phases and the preferred orientation of the films obtained under various conditions were mainly determined by the subtle compatibility between the surface and crystalline structures of the substrate (STO), TiN and TiO2 [26]. In this work, in order to manipulate the oxygen vacancies in titanium oxy-nitride (TiNxOy) films, we have deliberately
varied the oxygen partial pressure to deposit different TiNxOy
films, including pure TiN(h 0 0), TiN1−xOx(h 0 0), Ti2O3(0 0 l) and pure anatase TiO2(0 0 l) on the STO(1 0 0) substrate by in situ PLD using a TiN target. The evolution of crystalline and electronic structures has been systematically studied by x-ray diffraction (XRD) and x-ray absorption spectroscopy (XAS) measurements. Moreover, the observed changes in film crystalline and electronic structures exhibit intimate correlations with the manifested magnetic properties of these films. We found that the corundum structure of Ti2O3(0 0 l) films exhibited the most pronounced room temperature ferromagnetism (RTFM). We infer that formation of the transit structure (Ti2O3)may have generated significant amount of oxygen vacancies needed to trigger RTFM. The interactions between the magnetic ions of Ti3+ and electrons bound by oxygen vacancies are the fundamental ingredients for forming bound magnetic polaron (BMP) and the percolation of the BMPs at high enough densities would lead to RTFM.
2. Experiments and discussion
2.1. Sample preparation
A KrF excimer laser operating at a repetition rate of 5 Hz with an energy density of 4 J cm−2 was used. The target was a hot-pressed TiN (99.9%, purity) pellet. The distance between the target and substrate is 5 cm. The Tswas monitored by a thermocouple attached to the substrate holder and was kept at 700◦C during all deposition processes. The background pressure in the chamber was 2 × 10−7Torr at
Figure 1. (a) The XRD pattern of samples deposited with various
oxygen pressures. (b) Surface morphologies of samples measured by AFM.
room temperature. As we reported previously [26], the best TiN films were obtained under the background pressure and at Ts = 700◦C, and pure rutile TiO2 films could be obtained by oxidizing the TiN films at PO2 > 3.0× 10−1Torr after
deposition. Therefore, in order to systematically investigate the evolution of both the crystalline and electronic structures of the TiNxOyfilms deposited in situ, the system was operated
at Ts = 700◦C with PO2 being varied in the range 0–
0.25 Torr. The thickness of each sample was ∼120 nm as determined by an alpha-step profilometer. The areas of all samples are the same as that of the ∼0.5 × 0.5 cm2 STO substrates used for deposition. The crystalline structure of the films was measured by XRD, using Cu Kα radiation. The surface morphology of the films was examined by means of atomic force microscopy (AFM). The electronic structure of the TiNxOyfilms was investigated by XAS, using
the 6 m high-energy spherical grating monochromatic (6 m-HSGM) beam line at National Synchrotron Radiation Research Center (NSRRC), Taiwan, Republic of China. Moreover, the magnetic property of these samples was measured by a Quantum Design® superconducting quantum interference device (SQUID).
2.2. Crystalline structure
Figures 1(a) and (b) show the results of XRD and AFM measurements for samples prepared at various oxygen pressures, respectively. As shown in figure 1(a), epitaxial TiN(h 0 0) films was grown at Ts = 700◦C and without introducing oxygen (PO2 = 0 Torr) into the chamber during
PLD. The 2θ≈ 42.60◦diffraction peak assigned to TiN(2 0 0) in figure1(a) corresponds very well to the lattice parameter of 4.24 Å of TiN [26]. When the PO2 was increased slightly
(from 2× 10−6 to 5 × 10−5Torr), it can be seen that the TiN(2 0 0) diffraction peak starts to shift to higher diffraction angles and the peak width becomes broader with increasing
PO2, as well. This can be understood as follows. With the
presence of oxygen during deposition, the nitrogen in pure TiN will be replaced by oxygen due to the higher activity of oxygen. Since the crystal structure of TiN and TiO is the same (B1, rock-salt structure), it is quite natural to conceive that the films are basically consisting of TiN–TiO solid solution, i.e. titanium oxynitride TiN1−xOx. However, the ionic radius of
oxygen is smaller than that of nitrogen ion, which in turn might 2
Figure 2. M–H curves measured by SQUID at room temperature (300 K) for the STO substrate and the samples deposited at PO2(a) from 0
to 5× 10−5Torr, (b) from 1× 10−4to 2.5× 10−1Torr.
lead to the local lattice constant reduction [27]. Consequently, although the diffraction peak appears to remain the same as that of pure TiN(2 0 0), the peak evidently shifts to larger diffraction angles with increasing PO2. The surface morphology in these
films, as demonstrated in AFM images (the top two images of the left column of figure1(b)), shows the atomically smooth surface with sparsely distributed particulates distributed over the entire image. The root mean square (rms) roughness of the surface was estimated to be about 2 nm, suggesting that the films are indeed remaining as single-phased TiN-TiO solid solution at this stage.
As we further increased the PO2 from 5× 10−5 to 1×
10−4Torr, it is evident that the peak of TiN1−xOx/TiO (2 0 0)
disappears gradually and a small, new peak at 2θ ∼ 39.6◦ starts to emerge. This new peak can be indexed to the diffraction peak of Ti2O3(0 0 6) (corundum structure). It is noted that the appearance of the Ti2O3(0 0 6) peak was also observed by Xu et al [28], when using PLD to prepare the anatase TiO2 films with a Ti target under various PO2’s [28].
The relatively faint intensity and broadening appearance of this peak indicates that the obtained Ti2O3 films should have rather defective and disordered microstructure. However, as will be shown later, this film is, in fact, possessing the most pronounced RTFM among all the samples investigated in this study. This implies that the electronic state of the Ti ions and the large amount of oxygen vacancies in this film may have offered the most favourable environment to give rise to RTFM. When PO2was further increased to 1× 10−3Torr, we
found that a diffraction peak corresponding to the metastable anatase TiO2(0 0 l) structure appeared, which then evolved into a high intensity and sharp XRD peak corresponding to pure anatase TiO2(0 0 4) diffraction at PO2 = 1 × 10−2Torr. By
further increasing PO2 to 2.5× 10−1Torr, the films became
amorphous TiO2. As revealed by AFM images (figure1(b)), when the growth orientation of the films changes from (h 0 0) to (0 0 l) within the PO2 = 1 × 10−4–1× 10−2Torr range,
the surface morphology of the films also changes to become more granular with high-density of precipitations over the surface, presumably due to the formation of Ti2O3and anatase TiO2. The particulates on the surface then disappear at even higher PO2, reflecting the formation of amorphous TiO2.
All of these surface morphology observations revealed by AFM are quite consistent with the XRD results described above.
2.3. Magnetic property
Figure 2 shows the magnetization (M) versus applied magnetic field (H ) curves measured by SQUID at room temperature (300 K) for all samples. The substrate STO and the samples deposited at PO2 values ranging from 0
to 5 × 10−5Torr (i.e. TiN1−xOx films) display essentially
diamagnetic behaviour (figure2(a)). On the other hand, as shown in figure2(b), for the film deposited at PO2= 10−4Torr,
which comprises mainly the corundum Ti2O3(0 0 l) structure, a well-defined ferromagnetic hysteresis loop is clearly demonstrated, indicating the existence of pronounced RTFM. The RTFM property, nevertheless, diminishes gradually with further increasing PO2 and disappears completely when PO2
reaches up to 1× 10−1Torr at which the TiO2films becomes amorphous. These observations strongly suggest that the manifestation of RTFM in these d0oxides must be intimately related to the detailed crystalline and electronic structures of the material.
2.4. Electronic structure
The advantage of XAS is its sensitivity to chemical properties and electronic structure of the samples under study. More specifically, analysis of the obtained XAS spectra allows one to discern the unique information on the crystal field strength and symmetry, hybridization, as well as the valence of the specific ion of interest, in this case Ti ions. Thus, in order to elucidate the correlations between the observed magnetic properties and the variation of electronic structures in these TiNxOyfilms deposited under different oxidizing atmospheres,
we have systematically measured the O K-edge, Ti L2,3-edge, and N K-edge XAS spectra of all the samples.
Figures3(a) and (b) show the O K-edge and Ti L2,3 -edge XAS spectra taken in total electron yield (TEY) mode for the as-deposited TiNxOy thin films, respectively. The O
K-edge XAS spectra reflect the partial density of unoccupied O p states and map, via hybridization, bands of primary Ti character [29]. The spectra of region I (530–536 eV) is attributed to O 2p states hybridized to Ti 3d states. The characteristic XAS peaks for the standard TiO2 powder are shown in figure3(c) for comparison. The degenerate Ti 3d band splits into t2g (corresponds to 530.3 eV peak) and eg (532.9 eV peak) bands due to crystal field effects [30]. This
J. Phys. D: Appl. Phys. 46 (2013) 075002 S C Chen et al
Figure 3. The XAS spectra of (a) O K-edge, (b) Ti L2,3-edge for the samples prepared at various oxygen pressures. The characteristic XAS peaks of (c) O K-edge, (d) Ti L2,3-edge for standard TiO2powder. The incident direction of x-ray is along the normal line of the sample surface. Details explained in the text.
splitting is very sensitive to the coordination number and to the extent of the hybridization. Closer inspection reveals that the t2gand egsplitting is slightly smaller for samples deposited at lower PO2 (PO2 <5× 10−5Torr), which could be attributed
to weaker Ti 3d–O2p interactions caused by the presence of oxygen vacancies and other defects in TiN1−xOx and Ti2O3
films. On the other hand, peaks C1 and D1 in region II (>537 eV) are attributed to O 2p states hybridized to Ti 4sp bands [31,32]. This region exhibits larger dispersion effects and is more sensitive to long-range order [29]. Similar to the peaks in region I, the spectra in this region demonstrate that the t2g and egsplitting becomes smaller and the Ti 4sp region is shifted towards lower energies for the samples deposited at lower PO2 (<5× 10−5Torr). This indicates that the Ti–O
interaction in TiN1−xOxis much weaker than that in Ti2O3and
anatase TiO2due to lack of long-range order. Finally, we also noted that the O K-edge XAS signals are existent even in the TiN powder and the as-deposited TiN thin films, suggesting that a native oxidation layer may exist at the surface of these TiN samples [29].
The XAS spectra of Ti L2,3-edge display a considerably more complex structure, which is caused by the combination of atomic interaction and crystal field effects [33,34]. Figure3(b) shows the XAS spectra of Ti L2,3-edge for the TiNxOysamples
and figure3(d) exhibits the characteristic XAS peaks for the standard TiO2 powder. The region L3 and L2 correspond to O 2p3/2-Ti 3d and O 2p1/2-Ti 3d transitions, respectively. For
both L2and L3edges, the crystal field splits the 3d band into t2g (A2, C2) and eg(B2, D2) bands. Since the Ti egorbitals point directly towards the 2p orbitals of the surrounding O atoms, the egband is very sensitive to the local environment. Moreover, the eg-related peak of the L3-edge further splits into two peaks, labelled as B2 and B2 in figure 3(d). Peak B2, with intensity smaller than that of B2, is the fingerprint of the anatase TiO2 [35]. As shown in figure 3(c), when the films deposited at PO2 < 5× 10−5Torr, the XAS spectra
of TiN1−xOx films exhibit the same features to that of TiN,
which indicates that both have the same cubic NaCl crystal structure. Perhaps, the most relevant changes in the spectra are occurring in B2 peak. The broadening of the peaks and the disappearance of the shoulder B2may due to the structure distortions and the chemical changes for the films deposited at PO2 < 5× 10−5Torr. On the other hand, the splitting of
(A2, B2)peaks and (C2, D2)peaks due to crystal field effects becomes much more distinguishable and peak B2also appears when PO2>1× 10−4Torr [36].
To further delineate the substitution of nitrogen by oxygen in all the samples, we also measured the N K-edge XAS in fluorescence yield (FY) mode. Figure4(a) shows the XAS spectra of N K-edge. The signals in region (397–404 eV) are attributed to N 2p states hybridized to empty Ti 3d bands [37]. The t2g and egpeaks split from Ti 3d caused by crystal field effects for the standard TiN sample. The spectra for the films deposited at PO2 < 5× 10−5Torr exhibit the same features
Figure 4. (a) The XAS spectra of N K-edge for samples deposited under different oxygen pressures. (b) The XPS spectrum of
Ti2O3(PO2= 1 × 10−4Torr).
except the intensity decreases gradually. This indicates clearly that these TiN1−xOxfilms present a closed chemical similarity
with TiN. The most relevant changes in the spectra occur in the position and magnitude of the egpeak. The egcharacteristic peak for TiN locates at 400.7 eV, which appears to shift to 401.4 eV for the sample prepared at PO2 = 1 × 10−4Torr.
This peak grows even further and completely dominates the spectrum when PO2 >1× 10−3Torr and has been generally
attributed to unbounded nitrogen [38], that is, the nitrogen atoms still remaining and occupying interstitial positions in the TiO2matrix. For the sample deposited at PO2 >1×10−2Torr,
it is evident that all the N K-edge signals included this peak disappeared completely, indicating that the nitrogen atoms migrate towards the surface and are thermally desorbed.
From what is described above, it is apparent that both the XRD and XAS results gave consistent account on the crystalline phases and the associated electronic structures evolution of the TiNxOythin films deposited at various oxygen
pressures. Briefly, it can be summarized as follows: (1) for the films deposited at PO2 < 5× 10−5Torr, the nitrogen in
pure TiN will be replaced by oxygen and form the TiN–TiO solid solution, i.e. titanium oxynitride TiN1−xO1. Then the films gradually turns into titanium monoxide (TiO) phase. Since both TiN and TiO are of cubic NaCl structure, the orientation of the films remains the same. (2) For the films deposited at PO2 between 5× 10−5 and 1× 10−4Torr, the
crystal structure changes from the NaCl structure (TiO) to the corundum structure (Ti2O3)with accompanying changes in preferred orientations. (3) When PO2 further increased to
1× 10−3Torr, the metastable anatase TiO2 (0 0 l) structure starts to emerge and an intensive and sharp diffraction peak corresponding to pure anatase TiO2(0 0 4) is observed at PO2=
1× 10−2Torr. Finally, the crystalline phase TiO2transforms to amorphous by further increasing PO2up to 2.5× 10−1Torr.
The question remaining to be answered is how do such crystalline phases and associated electronic structures correlate with the observed RTFM manifestations? As shown in figure 1(a), the relative faint and broad XRD peak of Ti2O3(0 0 6) indicates that a large amount of oxygen vacancies as well as structural defects maybe introduced when the phase transition involves two phases with drastically different
crystal structures, such as during the transition between the NaCl structure (TiO) and corundum structure (Ti2O3) with
PO2 ≈ 5 × 10−5–1× 10−4Torr. Nevertheless, it is also
the regime where the most pronounced RTFM was observed. Thus, in order to further explore the possible correlations among the RTFM, amount of oxygen vacancies, and the valence of Ti ions in a particular sample, the Ti 2p x-ray photoelectron spectroscopy (XPS) was performed. Figure4(b) displays the Ti 2p XPS spectrum of the Ti2O3film deposited at
PO2 = 1 × 10−4Torr. It is clear that the measured profile
is composed of four peaks (labelled as peaks 1–4). The split energy between peaks 2 and 4 is about 5.8 eV, which corresponds very well to the characteristic Ti 2p3/2(458.6 eV) and Ti 2p1/2(464.4 eV) spin doublet of Ti4+and is consistent with the value measured by Murata et al [39]. On the other hand, it is not trivial to precisely distinguish the origin of peaks 1 (457.4 eV) and 3 (463.2 eV). These two peaks may correspond to 2p3/2 and 2p1/2 of Ti3+ or Ti2+ and/or core level peaks of Ti ions bound to oxygen vacancies [40–43]. It is reasonable to conjecture that both Ti3+ions and oxygen vacancies may exist simultaneously in this unique Ti2O3films. In this case, the origin of RTFM can be explained by the BMP model [44,45]. In the undoped Ti oxides, Ti3+and Ti2+ions can provide the local magnetic moment similar to that provided by the 3d transition metals (Mn, Co and Fe, etc) doped in TiO2 [46–49]. Meanwhile, the oxygen vacancies will induce slight lattice distortion in the anatase structure and act as both electron donors and electron traps. Then the localized electrons exhibit exchange interaction with the d-shell of Ti3+ or Ti2+ ions within a localization radius, leading to the formation of BMP with a large net magnetic moment. If the density of the oxygen vacancy is large enough and the ‘region’ of individual BMP overlaps, the exchange interaction between percolated BMPs would give rise to the ferromagnetic behaviour. It is noted that, within this scenario, for the samples prepared with
PO2<5× 10−5Torr, although there also exist Ti
3+or Ti2+in the TiN1−xO1films, the crystalline structure remains the same cubic NaCl structure and no significant vacancy generation is introduced. As a result, no global RTFM was observed. On the other hand, for the samples prepared at PO2 = 1 × 10−4–
J. Phys. D: Appl. Phys. 46 (2013) 075002 S C Chen et al
anatase TiO2, respectively. In this case, a large amount of oxygen vacancies, structural defects accompanied with Ti3+ or Ti2+ions exist in these thin films and macroscopic RTFM is observed. At PO2 > 1× 10−2Torr, stoichiometric TiO2
films with Ti4+are formed and the films become non-magnetic, again.
3. Conclusion
In summary, we have prepared TiNxOy thin films by PLD
with various PO2using a TiN target. Elaborative evolutions of
the crystalline and electronic structures of these films were examined systematically by XRD and XAS measurements. The results indicated that with increasing oxygen partial pressure introduced into the PLD system during deposition, the films evolved sequentially from TiN, TiN–TiO solid solution, TiO, corundum Ti2O3, anatase TiO2 and finally became amorphous at the highest PO2 (∼0.25 Torr) practiced. The
magnetic property of these samples measured by SQUID revealed that there exist strong correlation between the RTFM behaviour and the crystalline and electronic structures of the obtained TiNxOy films. We found that only the Ti2O3(0 0 l)
and oxygen-deficient anatase TiO2films, which were prepared with PO2 = 1 × 10−4 to 1× 10−2Torr, exhibited the RTFM
behaviour. The XAS and XPS results suggest that the origin of RTFM behaviour is intimately related to the generation of a large amount of oxygen vacancies and magnetic Ti3+or Ti2+ ions under the deposition conditions where dramatic structure transformation occurred. The interactions between Ti3+/Ti2+ and the electrons bounded by the oxygen vacancies produce high density of BMPs. The percolation of the BMPs thus leads to the observed macroscopic RTFM behaviour. In contrast, for the films deposited at relative low (PO2 <1× 10−4Torr)
or high pressures (PO2 > 1 × 10−2Torr), due to either
insufficient oxygen-defects or the absence of Ti3+/Ti2+ ions, only diamagnetic behaviour was observed.
Acknowledgments
This work is financially sponsored by the National Science Council (Grant No NSC98-2112-M-009-006-MY3) and the Ministry of Education of Taiwan, R.O.C., under MOE ATU program operated at NCTU.
References
[1] Weinberger B and Garber R 1995 Appl. Phys. Lett.66 2409
[2] Vorotilov K, Orlova E and Petrovsky V 1992 Thin Solid Films
207 180
[3] Brown W D and Grannemann W W 1978 Solid-State Electron.
21 837
[4] Glassford K M and Chelikowsky J R 1992 Phys. Rev. B
46 1284
[5] Perry A and Pulker K 1985 Thin Solid Films124 323
[6] Chambers S A 2000 Surf. Sci. Rep.39 105
[7] Parker N J, Kharel P, Powell J R, Smith P A, Evans P D and Porch A 1999 IEEE Trans. Appl. Supercond.9 1928
[8] Zuccaro C, Ghosh I, Urban K, Klein N, Penn S and Alford N M 1997 IEEE Trans. Appl. Supercond.7 3715
[9] Lin P I, Liu C W, Hsieh C C, Wu K H, Juang J Y, Uen T M, Lin J Y and Gou Y S 2001 Japan J. Appl. Phys.40 L377
[10] Liu S J, Lin J Y, Juang J Y, Wu K H, Uen T M and Gou Y S 2002 Appl. Phys. Lett.80 4202
[11] Liu S J, Chen S F, Juang J Y, Lin J Y, Wu K H, Uen T M and Gou Y S 2003 Japan J. Appl. Phys.42 L287
[12] Rumaiz A K, Ali B, Ceylan A, Boggs M, Beebe T and Shah S I 2007 Solid State Commun.144 334
[13] Hong N H, Sakai J, Poirot N and Briz´e V 2006 Phys. Rev. B
73 132404
[14] Yoon S D, Chen Y, Yang A, Goodrich T L, Zuo X, Arena D A, Ziemer K, Vittoria C and Harris V G 2006 J. Phys.: Condens. Matter18 355
[15] Pammaraju C D and Sanvito S 2005 Phys. Rev. Lett.94 217205
[16] Durand H A, Brimaud J H, Hellman O, Shibata H, Makita Y, Gesbert D and Meyrueis P 1995 Appl. Surf. Sci.86 122
[17] Garapon C, Champeaux C, Mugnier J, Panczer G, Marchet P, Catherijot A and Jacquier B 1996 Appl. Surf. Sci.96 836
[18] Kim J H, Lee S and Im H S 1999 Appl. Surf. Sci.151 6
[19] Wei X, Skomski R, Balamurugan B, Sun Z G, Ducharme S and Sellmyer D J 2009 J. Appl. Phys.105 07C517
[20] Drera G, Mozzati M C, Galinetto P, Diaz-Fernandez Y, Malavasi L, Bondino F, Malvestuto M and Sangaletti L 2010 Appl. Phys. Lett.97 012506
[21] Wittmer M, Naser J and Melchior H 1981 J. Appl. Phys.
52 6659
[22] Braic M, Balaceanu M, Vladescu A, Kiss A, Braic V, Epurescu G, Dinescu G, Moldovan A, Birjega R and Dinescu M 2007 Appl. Surf. Sci.253 8210
[23] Chiaramonte T, Cardoso L P, Gelamo R V, Fabreguette F, Sacilotti M, Marco de Lucas M C, Imhoff L, Bourgeois S, Kihn Y and Casanove M-J 2003 Appl. Surf. Sci.
212–213 661
[24] Vaz F, Cerqueira P, Rebouta L, Nascimento S M C, Alves E, Goudeau Ph and Riviere J P 2003 Surf. Coat. Technol.
174–175 197
[25] Thakur H, Kumar R, Thakur P, Brookes N B, Sharma K K, Singh A P, Kumar Y, Gautam S and Chae K H 2011 J. Appl. Phys.110 083718
[26] Hsieh C C, Wu K H, Juang J Y and Uen T M 2002 J. Appl. Phys.92 2518
[27] Zajac A T, Radecka M, Zakrzewska K, Brudnik A, Kusior E, Bourgeois S, Macro de Lucas M C and Imhoff L 2009 J. Power Sources194 93
[28] Xu Y and Shen M R 2009 Appl. Phys. A94 275
[29] Soriano L, Abbate M, Fuggle J C, Prieto L, Jimenez C, Sanz J M, Galan L and Hofmann S 1993 J. Vac. Sci. Technol. A11 47
[30] Magnuson M, Mattesini M, Li S, H¨oglund C, Beckers M, Hultman L and Eriksson O 2007 Phys. Rev. B76 195127
[31] Brydson R, Sauer H, Engel W, Thomas J M, Zeitler E, Kosugi N and Karuda H 1989 J. Phys: Condens. Matter
1 797
[32] de Groot F M F, Figueiredo M O, Basto M J, Abbate M, Petersen H and Fuggle J C 1992 Phys. Chem. Miner.19 140
[33] de Groot F M F, Fuggle J C, Thole B T and Sawatzky G A 1990 Phys. Rev. B41 928
[34] de Groot F M F, Fuggle J C, Thole B T and Sawatzky G A 1990 Phys. Rev. B42 5459
[35] Kr¨uger P 2010 Phys. Rev B81 125121
[36] Crocombette J P and Jollet F 1994 J. Phys: Condens. Matter
6 10811
[37] Pfl¨uger J, Fink J, Crecelius G, Bohnen K P and Winter H 1982 Solid State Commun.44 489
[38] Esaka F, Furuya K, Shimada H, Imamura M, Matsubayashi N, Sato H, Nishijima A, Kawana A, Ichimura H and Kikuchi T 1997 J. Vac. Sci. Technol. A15 2521
[39] Murata M, Wakino K and Ikeda S 1975 J. Electron Spectrosc.
6 459
[40] Kumar P M, Badrinarayanan and Sastry M 2000 Thin Solid Films358 122
[41] Sanjines R, Tang H, Berger H, Gozzo F, Margaritondo G and Levy F 1994 J. Appl. Phys.75 2945
[42] Li Y, Hwang D S, Lee N H and Kim S J 2005 Chem. Phys. Lett.404 25
[43] Wang X P, Yu Y, Hu X F and Gao L 2000 Thin Solid Films
371 148
[44] Coey J M D, Venkatesan M and Fitzgerald C B 2005 Nature Mater.4 173
[45] Kaminiski A and Sarma S D 2002 Phys. Rev. Lett.88 247202
[46] Patel S K S, Gajbhiyea N S and Dateb S K 2011 J. Alloys. Compounds509 427
[47] Patel S K S and Gajbhiye N S 2011 Solid State Commun.
151 1500
[48] Yan W, Sun Z, Pan Z, Liu Q, Yao T, Wu Z, Song C, Zeng F, Xie Y, Hu T and Wei S 2009 Appl. Phys. Lett.
94 042508
[49] Patel S K S and Gajbhiye N S 2012 Mater. Chem. Phys.
