Corresponding author

Gennady Logvenov

Max Planck Institute for Solid State Research

References

1.
Kawashima, K.; Logvenov, G.; Christiani, G.; Habermeier, H.-U.
Interelation of epitaxiaql strain and oxygen deficiency in La0.7Ca0.3MnO3-d thin films
2.
Baiutti, F.; Christiani, G.; Logvenov, G.
Towards precise defect controlin layered oxide structures by using oxide molecular beam epitaxy

Scientific Facility "Technology"

Precise defect control in complex oxides thin epitaxial films

Authors

K. Kawashima, F. Baiutti,  G. Christiani, H.-U. Habermeier, and G. Logvenov

Departments

Scientific Facility "Technology"

The Scientific Service Group Technology develops advanced epitaxial growth technologies for depositing complex compounds with atomic-layer precision. Using pulsed laser deposition (PLD) systems and a unique oxide molecular beam epitaxy (MBE) system we synthesize thin films, multilayers and superlattices of different complex oxides including cuprates, manganates, nickelates, cobaltates, ruthenates. The combination of both methods enables the accurate fabrication and in-depth exploration of new heterostructure systems with defect control at the atomic level. In this report we summerize our last achievements reported in [1,2].

In [1], we present the interrelation between epitaxial strain and oxygen deficiency in La0.7Ca0.3MnO3-δ (LCMO) thin films prepared using the PLD technique. Epitaxial strain and oxygen deficiency of the samples were systematically modified using three different substrates ((100) SrTiO3 (STO), (001) (LaAlO3)0.3-(Sr2AlTaO6)0.7 (LSAT) and (001) LaSrAlO4 (LSAO)) and four different oxygen pressures during the film growth, ranging from 0.27 mbar to 0.1 mbar. The combination of these two parameters, i.e. the controlled lattice mismatch and the oxygen pressure during the film growth, enabled us to explore the interrelation between the oxygen deficiency in the LCMO films and the lattice strain. It was demonstrated that the oxygen incorporation depends on the epitaxial strain: in particular, oxygen vacancies were induced to accommodate tensile strain whereas the compressive strain suppressed the generation of oxygen vacancies. These conclusions were obtained performing structural and functional measurements. After the LCMO film growth, surface morphology was investigated by atomic force microscopy (AFM). X-ray diffraction (XRD), as well as reciprocal space mapping (RSM) were used to characterize the crystal structure and the epitaxial strain states. Finally, a Quantum Design system SQUID vibrating sample magnetometer (VSM) and a PPMS Quantum Design system were applied to investigate the magnetic and transport properties.

<strong>Fig. 1: </strong>(a) Relation between the surface roughness and the deposition pressures of all the samples. The AFM images of 1000 &Aring; LCMO on LSAT deposited at 0.27 mbar (b) and at 0.1 mbar (c). The AFM image of LCMO on LSAO deposited at 0.1 mbar (d). Zoom Image
Fig. 1: (a) Relation between the surface roughness and the deposition pressures of all the samples. The AFM images of 1000 Å LCMO on LSAT deposited at 0.27 mbar (b) and at 0.1 mbar (c). The AFM image of LCMO on LSAO deposited at 0.1 mbar (d). [less]

Figure 1(a) shows the dependence of the root mean square (RMS) surface roughness measured by AFM versus the oxygen pressure during the growth of all our LCMO films. Irrespective of the substrate, we observed that samples deposited at pressures lower than 0.2 mbar shows RMS values less than 1 nm, whereas a higher oxygen pressure (0.27 mbar) was found responsible for an increase of the RMS values up to 2.7–4.2 nm. We conclude that the increase of the oxygen pressure causes a crossover from a layer-by-layer growth mode to an island growth mode.

The high quality of the crystalline structure was confirmed by the presence of Laue fringes in the XRD θ-2θ scan (Fig. 2) of LCMO films grown on STO and LSAT. Laue fringes were absent when LCMO was grown, under the same conditions, on LSAO substrates. This result indicates that the choice of the substrate influences the crystalline structure of the LCMO film, due to the film-substrate lattice mismatch and to the morphology of the substrate surface.

<strong>Fig. 2: </strong>XRD scans around LCMO (002) peak. The diffraction intensity is normalized to STO(002), LSAT(002) and LSAO(006), respectively. The (006) diffraction peak of LSAO substrate is at 42.91&deg;, which is out of the range shown in the figure. The vertical lines (green) corresponds to the expected positions of fully strained entirely stoichiometric LCMO films, the dotted line (black) represents the position of the bulk LCMO (002) peak. Zoom Image
Fig. 2: XRD scans around LCMO (002) peak. The diffraction intensity is normalized to STO(002), LSAT(002) and LSAO(006), respectively. The (006) diffraction peak of LSAO substrate is at 42.91°, which is out of the range shown in the figure. The vertical lines (green) corresponds to the expected positions of fully strained entirely stoichiometric LCMO films, the dotted line (black) represents the position of the bulk LCMO (002) peak. [less]

The LCMO (002) X-ray diffraction peak shifts from higher to lower angles as the oxygen growth pressure becomes smaller. The change of the diffraction peak position to lower angles is interpreted as an increase of the out-of-plane lattice parameter  as a consequence of the introduction of oxygen deficiency. The shift of the peak position is predominant in the film under tensile strain (grown on STO), and it is suppressed in the film deposited under compressive strain (on LSAO and LSAT). This clearly indicates the relation between strain and oxygen stoichiometry, i.e. oxygen vacancies are induced to accommodate the tensile strain, whereas compressive strain suppresses the formation of oxygen vacancies.

The same tendency was observed in the transport and magnetic measurements. A summary of Curie temperature (TC) and the peak temperature (TP), in which one can observe metal-to-insulator transition, is given in Fig. 3. TC and TP are suppressed in the tensile-strained LCMO film on STO substrate grown at 0.27 mbar because of the induced oxygen vacancies, while the LCMO films on LSAT and LSAO substrates show a sharp transition and high TC and TP. A change of the oxygen deposition pressure from 0.27 mbar to 0.2 mbar does not affect TC and TP in the compressive-strained LCMO film on LSAO substrate, whereas it determines a variation of the transport and magnetic properties for LCMO films grown on STO and LSAT substrates. This is related to the fact that compressive strain suppresses the formation of oxygen vacancies.

<strong>Fig. 3:</strong> Dependence of the Curie temperature <em>T</em><sub>C</sub> and the peak temperature <em>T</em><sub>P</sub> on the oxygen pressure during growth. Zoom Image
Fig. 3: Dependence of the Curie temperature TC and the peak temperature TP on the oxygen pressure during growth.

The results clearly indicate that oxygen stoichiometry is related to the epitaxial strain. Oxygen deficiency has an impact on functional properties of strongly correlated oxide systems, and investigating oxygen stoichimetry in thin films is an essential task. Our achievement supports a deep understanding of oxide thin films.

In [2], we present a new concept of the Atomic-Layer-by-Layer Oxide Molecular Beam Epitaxy (ALL-oxide MBE). The use of ALL-oxide MBE technique allows us for the deposition of atomically smooth single-crystal thin films of various complex oxides, artificial compounds and heterostructures.  In particular, we have deposited La2-xSrxCuO4 and La2-xSrxNiO4 with different doping levels (x), LaNiO3, LaAlO3, LaSrAlO4 and several other complex oxides, and superlattices consisting of ultrathin layers, down to one unit cell thick. In most cases, such films and heterostructures have shown good crystallinity, smooth surfaces and atomically abrupt interfaces. Since defect formation and its distribution have a crucial role in the definition of the functionalities of complex oxides heterostructures, we aim to move from the idea of a "perfect sample", which resembles theoretical models with minimized presence of imperfections, to the concept of the precise defect control, in which defect chemistry is tuned, for example, by an appropriate choice of growth condition, layer stoichiometry and by the strategic introduction of interfaces.  The application of "layer-by-layer engineering" by using ALL-oxide MBE, could allow us to develop new methods for the fabrication of layered oxides and heterostructures with novel functional properties.

In summary, the combination of deposition conditions, epitaxial strain between substrate and oxide films and precise stoichiometry control in both PLD and ALL-oxide MBE methods provide new insights in the mechanisms underlying complex oxides and interface properties, allowing for the synthesis of novel compounds and devices. In particular, defect chemistry plays an important role in the definition of the functionalities and in the possibility of tuning them through a control of defects concentration and distribution.

 
loading content