Design and Growth of Luminescent Thin Films by Atomic Layer Deposition
- Author
- Michael Getz
- Year
- 2017
- Abstract & Cover
- Source of Information
- Henrik Sonsteby
- University
- University of Oslo (Oslo, Norway)
- External Link
- Read Thesis
Most solar cells exploit the ultraviolet (UV) region of our solar spectrum poorly, while being excellent in the near-infrared (NIR) range. The goal of this thesis has been to create luminescent materials able to enhance the efficiency of such solar cells by converting UV light to visible or NIR photons. The materials have been designed by the ALD method and analysed with respect to various optical and structural properties The luminescent properties are designed using different structural units that absorb UV and units that emit in the visible or NIR. The ALD technique has proven highly suitable in design and distribution of such units from homogeneous mixtures to clusters and layers. As absorbers, both titanium phosphate and vanadates have been used, while lanthanides like Eu3+ and Yb3+ have been used as emitters In the first article we demonstrate that as-deposited amorphous europium titanium phosphate can convert UV light to red, and exhibits unique properties like tuneable reversible reduction in luminescence intensity under UV or X-ray illumination, as well as an emission spectrum whose shape and intensity is highly dependent on the post-deposition annealing temperature and the Eu:Ti ratio in the samples. While the material appears to be a poor choice for enhancing solar cell efficiencies, due to low quantum efficiency and poor photostability under UV-excitation, the system is a good model system of a luminescent amorphous material and is important for understanding energy transfer in thin films made by ALD. Working with the amorphous Eu-Ti-P-O system made me aware that achieving high energy transfer rates from the absorber to the emitter is easier if the host exhibits a reasonable quantum efficiency in the absence of the emitter. A crystalline host material would thus have a better chance of enabling high light-conversion efficiencies. Several host structures with metal cations in high oxidation states were considered due to these ions being easily reduced, which is an indication of a strong M-O charge transfer absorption. I eventually selected YVO4 due to it being a well-known host material for lanthanides as it has a low-symmetry metal site that can easily be substituted with lanthanides like Yb3+. The Y1-xYbxVO4 system also has the potential to convert one UV photon to two NIR photons, which in theory enables quantum efficiencies exceeding 100%. (VO4)3- exhibits a strong charge-transfer absorption in the UV, and the absorbed energy can be transferred to two nearby Yb3+, from which NIR emission takes place. Silicon solar cells utilize the energy in UV-photons poorly, but are excellent at utilizing NIR photons, making the system promising with respect to enhancing their efficiency. In the second article we examine the x = 1 case, i.e. the YbVO4 system, and demonstrate controlled growth of this system with ALD. We establish that YbVO4 crystallization can be aided by an excess of V2O5 acting as a flux, reducing the onset of crystallisation by over 200 °C, and that the NIR emission increases with increased crystallinity. In the third article we determine the optimum Yb3+ concentration with respect to UV to NIR conversion by varying x, which at an excitation wavelength of 325 nm is determined to be ~8 mol% Yb3+. We observe several strong indications of downconversion taking place. The energy transfer rate from (VO4) 3- to Yb3+ in a sample with ~2 mol% Yb3+ was measured to be 34%, resulting in a theoretical upper-limit quantum efficiency of 134%, though the actual quantum efficiency of the sample was determined to be ~15% due to strong quenching of the (VO4)3- emission. The quenching can be reduced by increasing crystallinity, though additional research is required before we can establish whether a film with quantum efficiency exceeding 100% is possible. It is our interpretation that such a material could be realized in the future and Y1-xYbxVO4 should consequently be useful for enhancing the efficiency of solar cells. In the third article we also demonstrate that it is possible to design a crystalline YVO4- YbVO4 layered material with a sub-nanometre layer thickness. This means that either or both of these materials may be used as building blocks when designing advanced optical structures by ALD. This could allow for fine control of optical properties by replacing the Yb3+ in some or all of the layers with various other lanthanides, and then tuning the optical properties by adjusting lanthanide concentrations by varying the amount of layers of the various types, thus enabling emission spectra engineering. This could be useful for applications where a combination of certain wavelengths is desired, like in artificial lighting. The work can be considered a significant contribution to the currently ongoing search for materials that can harvest the light from the sun, while also contributing to understanding how deposition and annealing parameters affect the optical and structural properties of thin films deposited by ALD. The work can also be considered a contribution to the field of luminescence as a whole and several of the findings could be useful for improving existing or future applications.
