Resumen de Òxids de metalls de transició, conductors i transparents. Propietats elèctriques i òptiques de capes primes epitaxials i la seva integració en heteroestructures "tot-òxid" fotoabsorbents
Transparent conducting oxides (TCOs) are key elements to many technological devices. Their ability to combine high electrical conductivity and high optical transparency to visible light, make them particularly useful in a myriad of devices such as displays, solar cells, smart windows, etc.
Indium tin oxide (ITO) is so far the most widespread TCO. By Sn-doping, this wide band gap In2O3 semiconductor can reach low resistivity (only about two orders of magnitude above conventional metals) while preserving its transparency. A major drawback of ITO is its high cost as indium, its main component, is a scarce material. Moreover, due to its nature of doped-semiconductor, some physical limits impose that its properties cannot be further improved.
On the other hand, some intrinsic metallic oxides composed of early transition metals also turn out to be transparent. In these materials, the partially filled narrow d band is responsible for high density of free carriers with increased effective mass, thus bringing the reflection edge down to the near-IR region.
In this thesis, we were interested in exploring the properties of metallic oxide thin films grown by pulsed laser deposition (PLD), namely SrVO3 (SVO; 3d1) and SrNbO3 (SNO; 4d1).
As high epitaxial quality is essential to obtain good functional properties, the first step consisted in optimizing the growth parameters for single phase and flat SVO/SNO films, displaying high crystallinity, conductivity and transparency. As anticipated, films need to be grown in ultra-high vacuum (UHV) to stabilize the 4+ oxidation state of V/Nb and using a high substrate temperature (700-800°C) to allow good mobility of the species on the substrate. However, the deposition in UHV and its subsequent highly energetic PLD plasma plume lead to a high concentration of point defects. We have solved this issue by using an inert background gas. Finally, we have studied the impact of epitaxial strain on the electrical conductivity and optical transparency window. All in all, it turned out that optimal films display larger conductivity than ITO, for a similar transparency.
Conventional wisdom would suggest that a low plasma frequency would be due to the electron-electron correlations within the narrow nd1 band. In a systematic analysis of SVO transport data (temperature-dependent resistivity, etc.), we have concluded that the Fermi liquid theory alone cannot account for the carrier mass enhancement. Instead, we have suggested that the 2D-like Fermi surface and the electron-phonon coupling play a major role. In addition, we have shown that the classical rigid band picture, of one free electron evolving in a 3d-t2g band is only a rough approximation, as attested by the observed hybridization of the V 3d and O 2p orbitals. Moreover, strain affects this hybridization by modifying the orbital hierarchy and covalency which could be responsible for the observed strain-dependent resistivity and effective mass.
By appropriate optical measurements, we have also discussed the nature of the plasmonic excitations at plasma frequency in SNO and SVO films. Interestingly, the possibility of exciting volume plasmons in these TCOs gives a glimpse on their potential applications in the field of plasmonics.
Finally, we have tested the suitability of SVO as electrode in photoabsorbing all-oxide heterostructures. In particular, we have successfully observed a photovoltaic effect in LaFeO3-based capacitors and disclosed the important role of the electrode work function on the device performances. As outlook, we have concluded that SVO and SNO, by having distinct work functions, could allow to tune any device properties.
This work demonstrates the suitability and high potential of this whole new category of TCOs as electrode material in all-oxide devices. We are convinced that it opens the way to a plethora of possible devices, photovoltaic-wise or other.
