Researchers: L. Perondi, Philippe Binder*, Timothy Gosnell+
and K. Kaski
*Universidad de los Andes, Colombia
+Los Alamos National Laboratory, USA
The advent of erbium-doped devices has had a significant impact on the area of optical transmission and amplification. Erbium-doped glass fibres are already the standard data transmission medium for wide area networks while highly doped erbium silica-based plane waveguides hold the promise of becoming one of the key elements of highly integrated optical devices. Some important technological and scientific questions, however, have to be addressed before highly compact and efficient amplifiers become a reality. It is well known that the efficiency of a planar Er doped waveguide amplifier decreases when the concentration of erbium atoms exceeds a threshold value. The loss of efficiency is explained by a cooperative effect, generally termed upconversion, involving two or more excited ions, in which one ion looses its excitation energy to phonons of the lattice. With increasing ion concentration, the average distance between ions decreases, favouring the existence of larger clusters, with a markedly increase in the upconversion rate. Migration of excitations is also believed to play a major role in this regime. The net result is a fast mechanism for depopulation of the first excited state of Er+, which is used for amplification of light in the 1550-nm wavelength band.
The purpose of the study is to model the migration and interaction of erbium excitations in a model system through Monte Carlo simulations. The MC simulations for Er doped silicate glasses will include excited states beyond the first one and the possibility of energy transfer ('excitation hops') beyond the first neighbours. Radiative decay will also be considered when studying the concentration region in which the upconversion 'takes over' as the dominant depopulation mechanism. Different lattice geometries, representing the possibly different spatial arrangements of the Er ions in the matrix, will also be considered. In a second stage, the model will be tested against experimental data. At this stage, the effect of different fabrication techniques on the upconversion rate will also be studied. To this effect, the simulation model will be upgraded in order to allow for the existence of impurities other than Er ions. This feature will enable a detailed study of the effect of different oxides in the glass composition on the upconversion rate. Among the expected results we cite the quantitative estimate of the de-excitation rate as a function of erbium concentration for different lattice topologies, degrees of disorder and impurities.