Stoichiometry, Structure and Strain Engineering in LiNbO3 and LiTaO3 thin films deposited by PI MOCVD – LiLa

One of the most promising deposition methods for multicomponent films is pulsed injection MOCVD, providing digital deposition/composition control.
In order to optimize the Li content in the film, the method, able to measure Li concentration with precision of 0.1-0.2 %, is required. The indirect methods, used for the single crystals cannot be applied directly due to the presence of strain, size effects or other defects in the films, which influences also the structural, optical and other physical properties. Therefore, in this project the indirect methods for thin films, based on Curie temperature and Raman mode dampings , will be developed. Then, the Li concentration in films will be optimized varying the deposition parameters or by vapor transport equilibration. The twins and domains will be eliminated by studying the twin formation mechanisms, optimizing deposition and post-deposition conditions or by applying static electric field. The electrical and electro-optical properties of stoichiometric single-domain and twin-free films will be studied. The thermal expansion will be tuned by applying strain engineering and TCF of these films will be studied.

The preliminary results obtained on LT and LN film deposition by PI MOCVD are encouraging. The films of high epitaxial quality, consisting of pure LT and LN phase, were deposited. However, all films were twinned. The first measurements of thermal expansion in LN and LT showed that it can changed by SEVERAL TIMES thus, opening a possibility to tune the TCF in thin films.
Raman spectroscopy, being highly sensitive to the symmetry of materials, was used as a countercheck of phase composition and for mapping of distributions of the phases. In the case of LN and LT films it was found that standard XRD analysis of phase composition of textured LN/LT films might be not straightforward and confusing due to close reflection angles of LiNb(Ta)O3 and LiNb(Ta)3O8 phases and different possible origins of asymmetric diffraction peak profiles. Raman spectroscopy, a fast and local probe was applied for investigation and mapping of phase composition in LN & LT films. The wavenumbers of Raman modes of parasitical phases were identified from Raman spectra of LiNb(Ta)3O8 and Li3Nb(Ta)O4 powders.
It was found that Li concentration in the LN films can be varied by changing the Li/Nb precursor ratio in the solution or by changing deposition pressure. The formation of in-plane and out-of-plane twins were reduced by increasing the deposition pressure.

LiNbO3 and LiTaO3 single crystals are employed in a wide variety of optical, acousto-optic, pyroelectric, electrooptic and nonlinear optical applications and even, most recently, as desk-top neutron sources. The LN and LT films promises to be equally interesting, particularly because of the possible applications in electronic, opto-electronic and acoustic devices and their miniaturisation and integration. From technological perspective, the method, which enables producing of stoichiometric, twin and domain free LN or LT films, will attract scientific and industrial attention.
Although this project is essentially a fundamental scientific investigation, the importance of LN and LT in industry is such that we shall need to bear possible applications in mind. We shall investigate in particular the possibility to apply LN and LT films in BAW/SAW and electro-optical devices and to reduce the TCF in LN and LT films by strain engineering for SAW applications.

Publications:
A. Bartasyte, V. Plausinaitiene, A. Abrutis, S. Stanionyte, S. H. Margueron, P. Boulet, T. Kobata, Y. Uesu, and J. Gleize “Identification of LiNbO3, LiNb3O8 and Li3NbO4 phases in thin films synthesized with different deposition techniques by means of XRD and Raman spectroscopy” J. Phys D : Cond. Matter 25, 205901 (2013).
Proceedings:
S. Margueron, A. Bartasyte, V. Plausinaitiene, A. Abrutis, P. Boulet, V. Kubilius, Z. Saltyte, « Effect of deposition conditions on the stoichiometry and structural properties of LiNbO3 thin films deposited by MOCVD” SPIE Proceedings, Vol. 8626: Oxide-based Materials and Devices IV, March 2013

Invited presentation:
- SPIE Photonic West 2013, Oxide-based Materials and Devices IV, San Francisco, USA, «Residual stresses, stoichiometry and clamped thermal expansion in LiNbO3 and LiTaO3 thin films« S. Margueron, A. Bartasyte, V. Plausinaitiene, A. Abrutis, T. Murauskas, P. Boulet, S. Robert, J. Gleize, V. Kubilius, Z. Saltyte.
Oral presentations:
- ISAF-IUS-PFM-IFCS-EFTF, Prague July 2013, “Strain and chemical engineering in LiNbO3 and LiTaO3 thin films” A. Bartasyte, V. Plausinaitiene, A. Abrutis, S. Margueron, T. Murauskas, P. Boulet, Z. Saltyte.
- EMRS spring meeting, Strabourg May 2013, “Strain and chemical engineering in LiNbO3 and LiTaO3 thin films” A. Bartasyte, V. Plausinaitiene, A. Abrutis, P. Boulet, Z. Saltyte and S. Margueron.
- International School of Atomic and Molecular spectroscopy,
a NATO Advanced Study Intitute –Erice, Italy, July 2013 “Chemical and strain engineering of functional oxides” A. Bartasyte, V. Plausinaitiene, A. Abrutis and S. Margueron
Poster presentations:
- ISAF-IUS-PFM-IFCS-EFTF, Prague July 2013, “Identification of LiNb(Ta)O3, LiNb(Ta)3O8 and Li3Nb(Ta)O4 phases in thin films synthesized with different deposition techniques by means of XRD and Raman spectroscopy” A. Bartasyte, V. Plausinaitiene, A. Abrutis, S. Stanionyte, S. Margueron, P. Boulet, T. Kobata, Y. Uesu, and J. Gleize.

LiNbO3 (LN) and LiTaO3 (LT) are the two of the most important crystals, being the equivalent in the field of optics, nonlinear optics and optoelectronics to silicon in electronics. Thus, the studies about epitaxial ferroelectric LN and LT thin films are of great interest because of their potential application as elements in static random access memories, high dielectric constant capacitors, acoustic delay lines, microwave tunable devices, and optical waveguides. Although LN and LT films have been fabricated by different techniques, many electrical and electro-optical properties reported are not comparable for those of LiNbO3 and LiTaO3 single-crystals. Thus, the degradation of physical properties in thin films can be explained by the difficulty to control and to measure the Li concentration within the film. Moreover, the grain boundaries in polycrystalline films and twin structure in epitaxial ones lead to light scattering and large optical losses in waveguide devices fabricated from these films. Thus, high epitaxial quality, single-domain (twin free) and stoichiometric LN and LT thin films are needed.
The aims of this project are challenging:
-Deposition of stoichiometric LN and LT films (with 50 mol% of Li);
- Elimination of twins and ferroelectric domains in films with thicknesses = 1 µm;
-tuning of the thermal expansion of thin films in order to reduce thermal frequency coefficient (TCF) – an important parameter in SAW devices.
One of the most promising deposition methods for multicomponent films is pulsed injection MOCVD, providing digital deposition control. The preliminary results obtained on LT and LN film deposition by PI MOCVD are encouraging. The films of high epitaxial quality, consisting of pure LT and LN phase, were deposited. However, all films were twinned. The first measurements of thermal expansion in LN and LT showed that it can changed by SEVERAL TIMES thus, opening a possibility to tune the TCF in thin films.
In order to optimize the Li content in the film, the method, able to measure Li concentration with precision of 0.1-0.2 %, is required. The indirect methods, used for the single crystals cannot be applied directly due to the presence of strain, size effects or other defects in the films, which influences also the structural, optical and other physical properties. Therefore, in this project the indirect methods for thin films, based on Curie temperature and Raman mode dampings , will be developed. Then, the Li concentration in films will be optimized varying the deposition parameters or by vapor transport equilibration. The twins and domains will be eliminated by studying the twin formation mechanisms, optimizing deposition and post-deposition conditions or by applying static electric field. The electrical and electro-optical properties of stoichiometric single-domain and twin-free films will be studied. The thermal expansion will be tuned by applying strain engineering and TCF of these films will be studied.