Journal Name Cite this: DOI: 10.1039/c0xx00000x www.rsc.org/xxxxxx Dynamic Article Links ► ARTICLE TYPE This journal is © The Royal Society of Chemistry [year] [journal], [year], [vol], 00–00 | 1 Luminescent YbVO 4 by Atomic Layer Deposition Michael Getz a , Per-Anders Hansen a , Mohammed A. K. Ahmed a , Helmer Fjellvåg a , and Ola Nilsen* a Received (in XXX, XXX) Xth XXXXXXXXX 20XX, Accepted Xth XXXXXXXXX 20XX DOI: 10.1039/b000000x UV to visible and near-infrared converting thin films of YbVO 4 have been deposited by atomic layer deposition, 5 using the precursor combinations Yb(thd) 3 (thd = 2,2,6,6-tetramethyl-3,5-heptanedione) and O 3, and VO(thd) 2 and O 3 at a deposition temperature of 240 °C, followed by post deposition annealing at 400-1000 °C. The UV absorption and the visible and near-infrared emission have been investigated in detail. The structure, thickness and composition of the deposited films have been studied by X-ray diffraction, ellipsometry, and X-ray fluorescence, respectively. The optimal pulse ratio of Yb(thd) 3 and VO(thd) 2 with respect to near-infrared emission was found to 10 be 1:3, which also yielded the most crystalline sample after annealing. Crystallization of the films is accelerated when an excess of V 2 O 5 is present, enabling crystallization at temperatures as low as 500 °C, probably through a flux aided process. Introduction With the ever growing need for renewable energy, it has become 15 clear that the sun will be one of the key energy sources in the years to come. Currently, the competition within the solar industries is fierce, leading to limited margins and a constant drive for increased efficiency. The silicon solar cells are, however, already approaching their theoretical limit of 30%, 1 20 with Panasonic currently holding the record of industrial scale panels that operate at an efficiency of 25.6%. 2 It is difficult to go beyond this due to the mismatch between the solar spectrum and the bandgap of silicon. With a bandgap of 1.11 eV, more than half of the energy in each photon in the UV part of the spectrum 25 is wasted on thermalization, which reduces the efficiencies even further if the cell is allowed to increase in temperature under operation. In addition, most solar cells have a heavily doped emitter region in the top part of the solar cell, i.e. where the UV- light is absorbed. As this region is filled with defects, the lifetime 30 of the excited carriers is small and consequently many of them never reach the contacts. In order to utilize the UV region better and to go beyond the Shockley-Queisser efficiency limit of 30%, down conversion of each high energy UV photon to two lower energy photons would 35 be beneficial. The process was first described by Dexter in 1957 3 and was demonstrated in YF 3 :Pr 3+ phosphor under the excitation of 185 nm with a quantum efficiency of around 140% in 1974. 4 Currently no one has been able to demonstrate an external quantum efficiency of more than 100% with photons from the 40 solar spectrum experimentally, even though down conversion has been reported numerous times. 5-12 The Tb 3+ /Yb 3+ couple has been proven to display second order cooperative energy transfer (CET), 5 but due to the intraconfigurational 4f transitions being parity forbidden, the absorption cross section of Tb 3+ and other 45 suitable lanthanides like Tm 3+ and Pr 3+ is extremely poor. Thus, they are not very useful as conversion materials for solar cells on their own. While YbVO 4 is not a common luminescent material, it closely resembles YVO 4 , which is well-known for its excellent 50 luminescent properties when doped with various lanthanides. 12-16 YVO 4 :Yb 3+ has previously been suggested as a quantum cutting material due to strong charge transfer absorption of (VO 4 ) 3- , and blue emission at twice the energy of Yb 3+ emission 8,12 . This should imply that the host can display resonant energy transfer to 55 Yb 3+ . In a study by Cheng et.al., 12 it was indeed determined that a sample with 1 mol% Yb 3+ should have a theoretical quantum efficiency of 148.7%, under the assumption that the energy not transferred to Yb 3+ results in radiative decay and that there is 100% emission efficiency from Yb 3+ . The actual quantum 60 efficiency of this system has not been reported, most likely due to the challenge of quantifying the UV excitation as well as the visible and NIR emission in such a large range. It is expected that energy transfer from both the (VO 4 ) 3- complex and Yb 3+ to quenching centres is significant. The amount and type of 65 quenching centres should depend on the synthesis routes. It is desirable with a method that produces high quality crystals which can be applied as thin films on top of a solar cell. Atomic layer deposition (ALD) is a technique in which the material is grown layer by layer, providing very fine control of 70 the atomic distribution. Yb 3+ ions can thus be evenly distributed during deposition, providing excellent samples for studies of properties during annealing towards high quality crystals. Vanadium has mostly been deposited as its binary oxide by ALD. We here expand on this process by converting it to its polyanion 75 in the YbVO 4 structure. Insight into the deposition of YbVO 4 is expected to be useful for deposition of more complex systems like YVO 4 :Yb 3+ in the future, and as YbVO 4 can be considered to be an YVO 4 :Yb 3+ system in which 100% of the lanthanide positions are occupied by Yb 3+ , it is also interesting to investigate 80