See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/228372072
Nanoscale patterning of functional perovskitetype complex oxides by pulsed laser deposition through a nanostencil ARTICLE in APPLIED SURFACE SCIENCE · MAY 2010 Impact Factor: 2.71 · DOI: 10.1016/j.apsusc.2010.01.103
CITATIONS
READS
8
102
5 AUTHORS, INCLUDING: Cristian Victor Cojocaru
Catalin Harnagea
National Research Council Canada
Institut national de la recherche scientifique
24 PUBLICATIONS 176 CITATIONS
117 PUBLICATIONS 2,672 CITATIONS
SEE PROFILE
SEE PROFILE
Alain Pignolet
Federico Rosei
INRS - Institut National de la Recherche Sci…
Institut national de la recherche scientifique
134 PUBLICATIONS 1,937 CITATIONS
275 PUBLICATIONS 5,019 CITATIONS
SEE PROFILE
SEE PROFILE
Available from: Cristian Victor Cojocaru Retrieved on: 15 January 2016
This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy Applied Surface Science 256 (2010) 4777–4783
Contents lists available at ScienceDirect
Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc
Nanoscale patterning of functional perovskite-type complex oxides by pulsed laser deposition through a nanostencil Cristian-Victor Cojocaru 1, Riad Nechache, Catalin Harnagea, Alain Pignolet *, Federico Rosei * INRS – E´nergie, Mate´riaux et Te´le´communications, Universite´ du Que´bec, 1650 Boulevard Lionel-Boulet, Varennes, QC J3X 1S2, Canada
A R T I C L E I N F O
A B S T R A C T
Article history: Available online 4 February 2010
We present studies on parallel nanoscale patterning of piezoelectrics/ferroelectrics via deposition through a nanostencil. Unlike other processes reported for oxide nanostructuring, we selectively deposit the material, directly, by interposing a nanosieve between the substrate and the deposition source. We show that this selective deposition can be realized even with materials as complex as perovskite oxides, both at room temperature and at high temperature. We elaborate on and analyze the performance of the nanostenciling approach for the growth of barium titanate BaTiO3 on strontium titanate SrTiO3(1 0 0). The patterned structures of ferroelectric materials are characterized by X-ray diffraction and imaged locally by scanning probe microscopy in piezoresponse mode to individually probe their functionality. ß 2010 Elsevier B.V. All rights reserved.
Keywords: Parallel patterning Nanostencil lithography Pulsed laser deposition (PLD) Perovskite complex oxides Ferroelectric nanostructures Barium titanate
1. Introduction Following the trend of the semiconductor industry with respect to the continuing miniaturization of integrated devices, functional oxide ceramic thin films are experiencing a similar evolution from microtechnology towards nanotechnology. Increasing efforts are undertaken to nanostructure and pattern thin films of functional oxides and several recent reviews have debated the advantages and difficulties associated with either the topdown or the bottom-up approaches to pattern electroceramic thin films [1–5]. Although some of the alternative patterning techniques to photolithography have already demonstrated the potential to deliver structures with smallest features close to 100 nm, their integration within a more general and complex fabrication scheme is very challenging. The requirement of perfect registration is unavoidable for a process to be used in future micro- or nano-electronics. As structure dimensions become progressively smaller, the new functional materials being envisioned often exhibit pronounced size effects which represent a significant deviation of the properties of low-dimensional structures with respect to their bulk properties [6,7]. The ‘‘size effect’’ in ferroelectrics, induced by a reduction in geometrical dimensions, has been shown to result, among other phenomena, in a reduced remnant polarization and dielectric permittivity, in changes in the domain structure, a
* Corresponding authors. E-mail addresses:
[email protected] (A. Pignolet),
[email protected] (F. Rosei). 1 Present address: NRC-Industrial Materials Institute, Boucherville, QC, Canada. 0169-4332/$ – see front matter ß 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2010.01.103
decrease of the phase transition temperature and an increase of the coercive field [8]. Fundamental studies of patterned functional oxide ceramics are further required for these new materials, especially in the nanoscale regime. These studies are crucial to establish experimentally the useful critical size for each material, i.e. the smallest size at which it would still provide the desired functionality [9–15]. In the particular case of ferroelectric materials, theoretical predictions suggest that, due to an increasing role of the depolarization field and the weakening of long range cooperative interactions, which are the driving force for ferroelectricity, the latter is supposed to disappear below a critical size. These calculations suggest a material-dependent critical thickness and correlation volume (e.g. 2.4 nm for a BaTiO3 thin film between two metallic SrRuO3 electrodes) [16]. Other theoretical studies on size effects estimated the minimum volume at which the polarization and the ferroelectric properties should vanish (critical volume) to be about 1000 nm3 [17]. Experimental studies demonstrated that ferroelectricity exists in ultrathin (4 nm) Pb(Zr,Ti)O3 (PZT) epitaxial films [18] and ferroelectricity was observed in BaTiO3 (BTO) films about 5 nm thick [19]. Such size effects in general are poorly understood. The top-down methods that are currently used in nanotechnology provide high-precision positioning and size control yet are often limited in resolution or are not suitable for complex oxide materials. For example, a conventional patterning process, based on usual resist lithography followed by etching of the oxide film is fraught with severe problems, when applied to ferroelectric thin films: contamination and side-wall redeposition can alter the polarization switching (even for features as large as micrometers) and constitutes challenging issues to be overcome in the patterning of complex oxides in general and of ferroelectric thin
Author's personal copy 4778
C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783 Table 1 Deposition conditions for the patterns analyzed in this work.
films in particular. Alternatively, bottom-up strategies, either by physical or chemical self-patterning of complex oxides [20,21] hold the promise of realizing much smaller features than those achievable with top-down techniques. Self-assembly governed by growth mechanisms similar to those found for Ge islanding on Si(0 0 1) surfaces [22], overcome the low throughput and processing damage of conventional patterning methods [12]. Nevertheless, they suffer from poor registration and to achieve a certain organization requires prior substrate patterning, which in turn involves an additional top-down processing step [23]. Though appealing for scientific purposes, bottom-up approaches are presently still ineffective, somewhat unreliable and not yet ripe to be implemented as new technologies. We present here studies on parallel nanostructuring and patterning of ferroic oxides, mainly piezoelectrics/ferroelectrics, via pulsed laser deposition (PLD) through a nanostencil. The original aim was to assess whether nanostenciling can become a general-purpose nanoscale patterning technique that offers both high resolution and unique flexibility for any combination of deposited material/substrate. Unlike most of the other techniques reported for oxide nanostructuring, we do not process or modify the substrate, but selectively deposit the material, directly, by interposing the nanosieve between the substrate and the deposition source. Simultaneously, there is an inherent control of the locations where the nucleation starts and where the structures’ growth will further take place, if the stencil is accurately positioned with respect to the substrate. A direct copy of the apertures opened in the stencil is realized on the substrate by ‘‘forcing’’ the material to pass through the former providing thus a technique for parallel fabrication of ordered arrays of islands. A similar approach has been investigated independently by other groups [24–26]. In particular, we compare patterns of BaTiO3 (a compound less sensitive to variations in by depositing parameters) obtained by both room temperature deposition (and crystallized by subsequent annealing) and by depositing at high temperature. The patterned structures of ferroelectric materials are characterized by X-ray diffraction (XRD) and probed locally by scanning force microscopy in piezoresponse mode for ferroelectricity.
All the samples were prepared on SrTiO3 single crystalline substrates (0 0 1)-oriented, using the same number of laser ablation pulses (N = 5000) and stencils with the same aperture diameter (350 nm). Two samples were fabricated with stencils having periodic apertures spaced with 1.6 mm pitch. After deposition the samples were cooled down to room temperature, naturally, in a 150 mTorr O2 background pressure. Once the deposition was completed the stencils were unclamped and simply ‘‘lifted off’’ (separated) from the substrate. The properties of a continuous epitaxial film (100 nm thick) deposited in high O2 atmosphere were also investigated. X-ray diffraction in grazing incidence (GIXRD) and Bragg– Brentano (u–2u) configuration was used to investigate the crystalline phases present in the patterned structures and their quality. The lattice parameters of the epitaxially patterned samples were studied more in detail by X-ray reciprocal space mapping (RSM), a technique typically used in epitaxial films to determine the strain state in thin films. Local ferroelectric testing was performed using Piezoresponse Force Microscopy (PFM) as described earlier [27]. We applied a small AC voltage (typically 0.5 V at 29 kHz) between the conductive AFM tip and the sample bottom electrode. Hysteresis loops were acquired by positioning the AFM tip over the center of an individual nanostructure, then the local piezoresponse signal was recorded while a superimposed DC bias voltage was swept between 10 V and +10 V. The induced local mechanical displacement was detected, using long integration times for the lock-in amplifier connected to the AFM. The out-of-plane piezoelectric coefficient d33 was calibrated according to the procedure developed by Harnagea and Pignolet [28].
2. Experimental
3. Results and discussion
PLD patterning experiments were conducted using laser interference lithography (LIL)-based stencil masks, with built-in 500 nm thick Si3N4 nanosieves with circular apertures. These stencils have hexagonal arrays of pores (350 nm in diameter) with a periodicity of either 1.6 mm or 700 nm and are patterned on 12– 14 free-standing, low stress (LS-SiN) membranes (2 mm in length and 100 mm in width each, 100 mm apart). The Si3N4 membranes are prepared on single crystal Si(1 0 0) wafers 380 mm thick and the stencil’s dimension is 5 mm 5 mm with an active area of 2 mm 2.7 mm. The stencils were mechanically clamped and temporarily fixed onto Nb-doped (1 0 0)-oriented SrTiO3 substrates. The assembly substrate-stencil was mounted in the vacuum chamber of a PLD system, in front of a dense BaTiO3 stoichiometric ceramic rotating target. A KrF Lumonics PM-800 excimer laser (radiation with l = 248 nm, pulse duration = 15.4 ns) was employed for ablation with a 458 incidence angle on the target and a laser fluence of 2 J/cm2. A set of depositions were carried out at room temperature (RT), in vacuum at 1 10 5 mbar (7.5 10 3 mTorr), with a laser repetition rate ranging from 5 to 10 Hz and a target–substrate distance of 6.5–7 cm. We further pursued the nanostenciling of BaTiO3 directly at high temperature, in low O2 background gas conditions according to the parameters provided in Table 1. This was possible only thanks to the high thermal resistance and stability of the Si3N4 membranes at temperatures up to 800 8C.
3.1. Stencil lifetime
ID
Temperature
Pressure [O2] (mTorr)
BTO pattern I BTO pattern II BTO pattern III
620 8C 620 8C RT
7.5 10 10 7.5 10
3
3
Post-deposition annealing N/A N/A RTA–700 8C
Rapid and parallel fabrication of ordered BTO nanostructures was achieved at RT in a single deposition step, over the whole sieve areas. Fig. 1(a) displays an SEM micrograph detail from a LS-Si3N4 nanosieve with pores of 350 nm in diameter and a 700 nm pitch used during the depositions and Fig. 1(b) shows the well-ordered, dome-shaped, as-deposited structures, obtained via stenciling through the latter. The SEM micrograph in Fig. 1(c) reveals structures with base enlarged to 385–390 nm and thus suggesting an overall base broadening of 35–40 nm, i.e. 10% larger than the nominal value of the apertures. We found that, for the sieves with 350 nm diameter circular holes, the transfer efficiency (defined as the ratio between the height of the structures and the thickness of a film deposited in the same conditions) is above 80%. Subsequent depositions using the same stencils revealed that for a total, nominal thickness of the deposited BTO equivalent to 3.5 times the aperture diameter, the lateral size of the replicated structures shrank by 30% (Fig. 1(d)). The choice of the physical vapor deposition technique used in combination with nanostenciling plays an important role and the control of the interaction of the material to be deposited, with the membrane apertures, remains a matter of further examination. Identifying efficient cleaning recipes for the stencils, such as
Author's personal copy C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783
4779
Fig. 1. SEM micrograph of a LS-SiN sieve (a) and ordered arrays of BaTiO3 structures obtained by deposition through it (b); (c) and (d) SEM micrographs of BaTiO3 patterned structures obtained at room temperature and 7.5 10 3 mTorr O2 with the same stencil used in consecutive depositions. A 30% shrinkage in structures lateral size is observed after 8 depositions.
selective etching of the material deposited on the sieves, or coating the stencil with protective monolayers to prevent or hinder clogging, also represent issues of high interest; however, this lies outside the scope of the investigations presented herein. For the experiments related to this work, we repeatedly used the same stencil with different target materials: e.g. other oxides such as lead zirconate titanate (PZT) or bismuth ferrite (BFO) or even with metals such as Pt. Gradually, the sieves suffered structural damages due to repeated mechanical clamping on the substrates and the stencil collapsed before we observed a complete clogging of the holes. 3.2. Morphology of the BTO patterns A comparison of the morphology of the patterned structures is shown in Fig. 2. According to SEM and AFM analyzes, the patterns obtained by depositing under the lowest background pressure (vacuum) exhibit the least lateral broadening. The ablated species, having the longest mean free path, are not scattered by the time they reach the apertures of the membrane. A different situation is observed for the sample grown under 10 mTorr of O2, for which we observe a rather large broadening of the structures. The bottom part of the nanostructure is distorted in one direction which reduces the distance between neighboring structures (Fig. 2(b)). For the case of the sample processed by Rapid Thermal Annealing (RTA) we observe a shrinkage in lateral size similar to the one observed after a conventional thermal annealing process; however the crystallites are not very well defined indicating an incomplete crystallization (Fig. 2(c) and (d)). 3.3. Structural characterization of the BTO patterns Even though the amount of material in the islands is rather small to perform usual u–2u XRD scans, we succeeded in recording
an XRD pattern from the patterned areas using long time/step values for data acquisition. In Fig. 3, XRD spectra for the samples listed in Table 1, along with a diffraction pattern from a bare Nb:SrTiO3 substrate are represented together. The parasitic peaks present in all recorded XRD patterns are attributed to the Kb line corresponding to the SrTiO3 substrate peaks and to the line corresponding to tungsten Wa contamination from the X-ray tube. No foreign phases were detected. The additional reflections visible at 2u = 388 and 528 correspond to an unidentified contamination which we attribute to the STO substrate, since these peaks are present in the XRD patterns of the bare substrates. As ‘‘standard’’ in this sequence of measurements, we used the continuous film deposited in 100 mTorr O2 for which the (0 0 l) peaks, corresponding well to the values given by the library reference pattern (PDF card #00-005-0626), clearly indicate the achievement of epitaxial BaTiO3 films. A detailed 2u scan in the range 20–258 displays the (0 0 1) peak of the BaTiO3 film at the expected position of 22.268, while for the samples grown at lower O2 pressure, a shift in the peak position towards lower angles is observed. This corresponds to a shift toward larger out-of-plane inter planar distances of the BTO films, gives a first indication that the patterned structures are strained by the substrate to try to accommodate the lattice in-plane parameter of tetragonal bulk BaTiO3 (3.994 A˚) to that of the SrTiO3 substrate (3.905 A˚). A close inspection of the samples annealed by RTA shows that no peaks are visible. This can be explained by the fact that the XRD peaks are so weak, that they are buried in the instrument background noise due to several reasons: (i) the structures formed are not epitaxial and thus the intensity of the XRD reflections corresponding to various orientations is drastically reduced, (ii) due to the very small grain size (20–30 nm), the width of the XRD reflections is significantly broadened, making it difficult to identify
Author's personal copy 4780
C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783
Fig. 2. SEM micrographs and AFM section profiles for the ordered BaTiO3 arrays obtained in conditions described in Table 1. (a) BTO pattern I [620 8C, 7.5 10 BTO pattern II [620 8C, 10 mTorr O2]; (c) and (d) BTO pattern III before [RT, 7.5 10 3 mTorr] and after RTA annealing [RTA at 700 8C/min].
the peaks and finally (iii) the crystallization into the perovskite phase is not complete and a further optimization of the annealing parameters may be needed (e.g. longer annealing times or higher temperature for RTA). Moreover the granular aspect of the
3
mTorr]; (b)
structures revealed by AFM and SEM suggests that there is no apparent improvement brought by an RTA process in terms of yielding better quality crystalline structures compared to a conventional thermal annealing.
Author's personal copy C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783
4781
Fig. 3. (a) X-ray diffraction spectra (u–2u) recorded from a bare SrTiO3 substrate, BTO epitaxial film and from an area with only BTO patterned arrays, denominated BTO pattern I, II, III and (b) X-ray diffraction spectra (u–2u) detail recorded in the range 20–258.
The lattice parameters of the epitaxial BTO patterned samples were investigated more in detail by X-ray reciprocal space mapping (RSM), a technique typically used in epitaxial films to determine if the film is fully strained (pseudomorphic), partially strained (or relaxed), or fully relaxed. The out-of-plane and inplane lattice parameter of BaTiO3 structures are determined from the in-plane reciprocal lattice parameter Qx and out-of-plane reciprocal lattice parameter Qy values (Fig. 4) using the following formulas: din = 2l/Qx and dout = 2l/Qy with l = 1.5406 A˚ the wavelength of the Cu Ka X-ray radiation. For the film, the position of BaTiO3 (1 0 3) reciprocal lattice point (r.l.p.) with respect to that of STO (1 0 3) indicates a relaxed film and calculations of the exact lattice parameters suggests a small distortion of the unit cell. When the deposition was performed in a small amount of O2 (BTO pattern II) the BaTiO3 structures were found to be almost completely relaxed, whereas strained structures appear when the deposition takes place in vacuum (BTO pattern I) and the (1 0 3) r.l.p. of BTO is shifting towards that of STO (1 0 3). The values measured from the RSM graphs are summarized in Table 2. Analyzing the values obtained for the in-plane and out-ofplane lattice parameters we observed an increased c/a ratio for the sample deposited in vacuum. This ratio is decreasing towards the
Fig. 4. X-ray reciprocal space mapping (RSM) performed for: BaTiO3 film (a) and patterned samples (b) and (c), in order to calculate the in-plane and out-of-plane lattice parameters.
BTO bulk values once the O2 pressure in the deposition chamber is increased (Fig. 5). At lower pressure, due to the higher energy of the ablated particles, more oxygen vacancies are produced in the films
Author's personal copy 4782
C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783 Table 2 Lattice parameters as measured from RSM maps.
Fig. 5. BaTiO3 (c/a) ratio versus O2 pressure for the samples given in Table 2.
(structures) leading to an additional in-plane compressive strain. This compressive strain results in the out-of-plane lattice parameter being elongated and thus an enhanced tetragonality of the structures fact confirmed also by XRD measurements as shown in Fig. 3. Low O2 pressure during the BaTiO3 growth is known to induce the formation of many oxygen vacancies [29] leading to significant structural changes. In fact, we observed that the volume of the BaTiO3 unit cell for the patterned samples is larger than its bulk value and decreases with the increase of oxygen pressure. A similar effect was reported for lead titanate (PbTiO3) films [30]. 3.4. Ferroelectric characterization of individual BTO cells The functionality of the patterned BaTiO3 has been investigated via PFM and the results were discussed in a previous report [4]. The
Sample ID
In-plane lattice parameter, a (A˚)
Out-of-plane lattice parameter, c (A˚)
BTO film BTO pattern I BTO pattern II BTO reference STO reference
4.00 3.974 4.00 3.994 3.905
4.02 4.191 4.067 4.038 3.905
relatively weak PFM signal (d33 = 7 pm/V compared to d33 = 80– 90 pm/V in BaTiO3 single crystal) was attributed to the fine grain size of the nanostructures [31,32]. Ferroelectric properties of individual BTO structures were assessed using scanning probe microscopy in piezoresponse mode. Local piezoresponse hysteresis loops were acquired from patterned BaTiO3 islands (Fig. 6) as described in the experimental section. The piezoelectric response of the BTO III samples was below the detection level of our microscope, albeit in very few instances we could detect a weak hysteresis buried in the noise. Therefore, we conclude that the samples annealed using RTA are incompletely crystallized. In contrast, the samples patterned at high temperature exhibit a distinct ferroelectric behavior, as shown in Fig. 6. The piezoresponse hysteresis loops recorded clearly indicate that ferroelectricity is retained in the BaTiO3 structures stenciled in a O2 low-pressure regime [33,34] The loops are shifted vertically (towards positive values) indicating a preferential polarization state known as imprint. This pre-polarization phenomenon was reported for instance in as-processed epitaxial ((Pb,La)(Zr,Ti)O3) thin film capacitors and the imprint was linked with the oxygen pressure during cooling [35]. Therefore, we infer that the imprint in our BTO structures is mainly due to oxygen vacancy-associated defect dipoles that enhance charge trapping near the substrate– ferroelectric interface. By comparing the hysteresis loops obtained
Fig. 6. PFM domain images and local piezoresponse hysteresis loops recorded from samples BTO pattern I and BTO pattern II prepared at 620 8C in (a) vacuum and (b) 10 mTorr O2, respectively.
Author's personal copy C.-V. Cojocaru et al. / Applied Surface Science 256 (2010) 4777–4783
from samples deposited at different oxygen pressures, we observe a stronger imprint for structures deposited under lower background pressures. Additionally, the squareness of the loop in Fig. 6(a) is more pronounced than that of the loop in Fig. 6(b), pointing to a more reliable switching process, despite the presence of vacancies. In contrast, the structures deposited at higher oxygen pressure (Fig. 6(b)) exhibit a pronounced tilt, associated with a significant linear polarization, suggesting a higher dielectric contribution to the hysteresis loop. One would expect that a higher tetragonality of the BTO unit cell would result in improved piezoelectric and ferroelectric properties, namely higher values for the piezoelectric coefficients in the case of the structures from the sample BTO pattern I. Such differences, however, were not observed in our samples, as both piezoresponse loops showed similar 2d33 values of about 1.75 pm/V. One possibility is that the improvement originating in increased tetragonality is compensated by the higher density of defects (oxygen vacancies and grain boundaries) for the BTO pattern I. Due to technical limitations, we did not measure the permittivity of the nanosize-grained islands. Nevertheless, the magnitude of their piezoresponse is comparable to that of ceramics having similar grain size [31,32]. Since one single island is composed of few hundred nanograins, we expect a comparable relative permittivity of our islands, in the order of 1000. Nevertheless, it is well known that in barium titanate the Curie Temperature has a strong dependence on pressure; according to [36] a pressure of 20 kbar (=2 GPa) decreases the Tc down to room temperature. In our PFM experiment, we estimate that the tip exerts a stress around 1 GPa on the region of the sample underneath. Therefore, a plausible explanation of the reduced piezoelectric response of our structures is due to the huge stress applied during the measurement, which reduces the spontaneous polarization. 4. Conclusions and perspectives We showed that the combination of PLD and stenciling can be carried out at both room and high temperature regimes. We also showed that this patterning method is compatible with high vacuum deposition technology, even when applied to complex oxides, such as perovskites. Since the typical synthesis of thin film perovskites is carried out in an oxygen background atmosphere, a vacuum-based deposition was thought to drastically alter the functional properties of the films due to the appearance of oxygen vacancies. However, for the synthesis of stenciled BTO structures a low-pressure of oxygen is suitable for an accurate replication of the stencil’s apertures and leads to enhanced tetragonality of the unit cells, allowing fabrication of structures with properties similar to those synthesized in a higher partial oxygen pressure, despite a supposedly higher density of defects. Moreover being a resist free process, damages related to etching are avoided and nanostenciling conserves the desired functionality of individual nanostructures, namely ferroelectricity (demonstrated by PFM). In the future, we will work on using this approach to synthesize periodic submicron arrays of multiferroic structures. The fabrication process described here is parallel and suitable for patterning complex multicomponent functional materials. In addition to its use to deliver patterns of mesoscopic structures, a more elaborate exploitation of shadow masks (e.g. controlling their position, rotation, tilting angle with respect to the substrate) can be
4783
envisioned for use in a combinatorial approach. Such an approach would potentially lead to the fast synthesis of entire nanoscale libraries of nanostructured materials of different compositions and shapes that can further be investigated from their structural, compositional, or functional point of view. Acknowledgements We acknowledge financial support from the Canada Foundation for Innovation. C.V.C. is grateful to NSERC of Canada for a personal Post-Doctoral Fellowship. F.R. is grateful to FQRNT and the Canada Research Chairs Program for partial salary support. F.R. and A.P. are supported by Discovery grants (NSERC). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28]
[29] [30] [31] [32] [33] [34] [35] [36]
J. Baborowski, J. Electroceram. 12 (2004) 33. C.R. Martin, I.A. Aksay, J. Electroceram. 12 (2004) 53. M. Alexe, C. Harnagea, D. Hesse, J. Electroceram. 12 (2004) 69. C.V. Cojocaru, F. Ratto, C. Harnagea, A. Pignolet, F. Rosei, Microelectron. Eng. 80 (2005) 448. C.V. Cojocaru, C. Harnagea, A. Pignolet, F. Rosei, IEEE Trans. Nanotechnol. 5 (2006) 470. D. Damjanovic, Rep. Prog. Phys. 61 (1998) 1267. F. Rosei, J. Phys.: Condens. Matter 16 (2004) S1373. A. Gruverman, A. Kholkin, Rep. Prog. Phys. 69 (2006) 2443. S. Clemens, S. Ro¨hrig, A. Ru¨diger, T. Schneller, R. Waser, Small 2 (2006) 500. A. Ru¨diger, T. Schneller, A. Roelofs, S. Tiedke, T. Schmitz, R. Waser, Appl. Phys. A 80 (2005) 1247. S. Clemens, T. Schneller, R. Waser, A. Ru¨diger, F. Peter, S. Kronholz, T. Schmitz, S. Tiedke, Appl. Phys. Lett. 87 (2005) 142904. M. Dawber, I. Szafraniak, M. Alexe, J.F. Scott, J. Phys.: Condens. Matter 15 (2003) L667. L.M. Eng, H.J. Gu¨ntherodt, G. Rosenman, A. Skliar, M. Oron, M. Katz, D. Eger, J. Appl. Phys. 83 (1998) 5973. F. Peter, A. Ru¨diger, R. Dittmann, R. Waser, K. Szot, B. Reichenberg, K. Prume, Appl. Phys. Lett. 87 (2005) 082901. H. Han, K. Lee, W. Lee, M. Alexe, D. Hesse, S. Baik, J. Mater. Sci. 44 (2009) 5167. J. Junquera, P. Ghosez, Nature 422 (2003) 506. S. Li, J.A. Eastman, Z. Li, C.M. Foster, R.E. Newnham, L.E. Cross, Phys. Lett. A 212 (1996) 341. T. Tybell, C.H. Ahn, J.M. Triscone, Appl. Phys. Lett. 75 (1999) 856. Y.S. Kim, D.H. Kim, J.D. Kim, Y.J. Chang, T.W. Noh, J.H. Kong, K. Char, Y.D. Park, S.D. Bu, J.G. Yoon, J.S. Chung, Appl. Phys. Lett. 86 (2005) 102907. I. Szafraniak, C. Harnagea, R. Scholz, S. Bhattacharya, D. Hesse, M. Alexe, Appl. Phys. Lett. 83 (2003) 2211. H. Nonomura, M. Nagata, H. Fujisawa, M. Shimizu, H. Niu, K. Honda, Appl. Phys. Lett. 86 (2005) 163106. C.V. Cojocaru, A. Bernardi, J.S. Reparaz, M.I. Alonso, J.M. MacLeod, C. Harnagea, F. Rosei, Appl. Phys. Lett. 91 (2007) 113112. S. Clemens, T. Schneller, A. van der Hart, F. Peter, R. Waser, Adv. Mater. 17 (2005) 1357. A.F. Rodrı´guez, L.J. Heyderman, F. Nolting, A. Hoffmann, L.M. Doeswijk, M.A.F. van den Boogaart, J. Brugger, Appl. Phys. Lett. 89 (2006) 142508. P. te Riele, A. Janssens, G. Rijnders, D.H.A. Blank, J. Phys.: Conf. Ser. 59 (2007) 404. P.M. te Riele, G. Rijnders, D.H.A. Blank, Appl. Phys. Lett. 93 (2008) 233109. C. Harnagea, C.V. Cojocaru, R. Nechache, O. Gautreau, F. Rosei, A. Pignolet, Int. J. Nanotechnol. 5 (2008) 930. C. Harnagea, A. Pignolet, in: M. Alexe, A. Gruverman (Eds.), Nanoscale Characterisation of Ferroelectric Materials–Scanning Probe Microscopy Approach, Springer, Berlin, 2004, , p 45 (Chapter 2). T. Zhao, F. Chen, H. Lu, G. Yang, Z. Chen, J. Appl. Phys. 87 (2000) 7442. C.H. Park, D.J. Chadi, Phys. Rev. B 57 (1998) R13961. L. Mitoseriu, C. Harnagea, P. Nanni, A. Testino, M.T. Buscaglia, V. Buscaglia, M. Viviani, Z. Zhao, M. Nygren, Appl. Phys. Lett. 84 (2004) 2418. M.T. Buscaglia, M. Viviani, V. Buscaglia, L. Mitoseriu, A. Testino, P. Nanni, Z. Zhao, M. Nygren, C. Harnagea, D. Piazza, C. Galassi, Phys. Rev. B 73 (2006) 064114. C.V. Cojocaru, C. Harnagea, F. Rosei, A. Pignolet, M.A.F. van den Boogaart, J. Brugger, Appl. Phys. Lett. 86 (2005) 183107. X. Du, Q.M. Wang, U. Belegundu, A. Bhalla, K. Uchino, Mater. Lett. 40 (1999) 109. J. Lee, R. Ramesh, Appl. Phys. Lett. 68 (1996) 484. M.E. Lines, A.M. Glass, Principles and Applications of Ferroelectrics and Related Materials, Clarendon Press, Oxford, 1977.
