• anti-Stokes
• cooling
• emission
• fluorescence
• lanthanides
• laser
• rare earths
• ytterbium

The fifteen lanthanide elements, along with Scandium and Yttrium, are collectively known as the rare earth elements. Since their discovery, these atoms have been exten- sively studied for many applications, such as permanent magnets and solid-state lasers. A less known use of these elements is the optical refrigeration of solids. Although light is intuitively associated with heat, such as the sunlight or laser cutting machines, it is indeed possible to cool down a solid by pumping it with a laser. This effect exploits the emission of light with wavelengths shorter than the incoming one, called “anti-Stokes fluorescence”, to subtract thermal energy from a suitable system, hence reducing its temperature. Ideally, an object should cool down when pumped with a wavelength longer than its mean emission. However, more realistic models consider losses due to incomplete absorption and due to escape probability of the fluorescence. These factors are very limited in the cooling of solids. For example, the probability that an excited ion will emit a photon that exits the material must be higher than 97%. These tight constraints distinguish optical refrigerators from other devices that produce light with energies higher than the pump, such as up-conversion based systems, which have much lower quantum efficiencies.
Recently, the researches on this field have been moved to Ytterbium-doped fluoride crystals. Among the rare earths, Ytterbium exhibits a simpler level structure, with only two manifolds. The absence of upper levels inhibits quenching processes between the doping ions, making the quantum efficiency of these systems close to unity. Fluoride materials have been shown to offer several advantages over the other hosts, such as longer upper-state lifetimes. In particular, the Yttrium Lithium fluoride (LiYF4) or YLF, is emerging as solid-state laser medium.
This thesis work is dedicated to the study of Ytterbium-doped YLF crystals to measure their optical refrigeration performances. With this goal in mind, several YLF crystals, doped with 5%, 7.5% and 10% of Ytterbium, were grown in a custom-designed furnace at the New Materials for Laser Applications laboratories, in Pisa. The formed rods were examined to check the absence of fractures or other structural defects, then oriented to determine the crystallographic axes. After the orientation, the rods were cut into smaller samples, made to perform the necessary experiments.
Optical properties were measured for all the samples, such as absorption coefficient of the Ytterbium ions, fluorescence spectrum and decay time of the upper manifold. The spectroscopical measurements were carried out at several temperatures, in a range
between 10 K and 300 K, to verify how the temperature modified the spectra of the crystals. These analyses ensured the high quality of the samples and were useful to compare the various doping concentrations. Moreover, from the acquired data, it was possible to estimate the potential cool down capability of all the samples, when pumped with an appropriate source.
The refrigeration properties of the crystals were tested in a homemade, custom- built, vacuum chamber, developed to insulate the cooling system and minimize heat transfer from the environment. The chamber was equipped with four windows, designed to pump and observe the sample at the same time. Two contact-less techniques were employed to detect the temperature of the crystal, without altering its heat capacity.
Only a small subset of the samples was capable to cool down, even though all of them had similar optical characteristics, thus indicating that the Ytterbium spec- troscopy was not enough to fully characterize the refrigeration behavior. Indeed, the growth of crystals for optical refrigeration appears to be a more complex task than the growth for other experiments, such as laser applications. The optical refrigeration efficiency was estimated for the three samples that effectively cooled down, two doped with 5% of Ytterbium, and one with 7.5% concentration.
A lower efficiency than expected was measured in one of the two samples with 5% of Ytterbium, although they had the same spectral characteristics. This difference was attributed to the amount of impurities, in particular other rare earth ions, enclosed in- side the crystals. Several spectroscopical measurements were performed on the samples, at various wavelength regions, with the purpose of identifying the contaminants from their fluorescences. These investigations, along with an independent chemical analysis, discovered the presence of Erbium, Holmium and Thulium inside the crystals.
During the cooling measurements, these impurities were not excited by the pump laser, but instead they gained energy, both directly and indirectly, from the Ytterbium ions. The amount of contaminants directly activated from the Ytterbium decreased the number of ions involved in the refrigeration process. The results of the analyses were summarized in a descriptive model of the energy transfer processes between the ions in a crystal. This model identifies which contaminants were directly pumped by the Ytterbium, hence, which elements were more likely to cause the reduction of the cooling performances in one of the samples.