The most commonly proposed mechanisms for NIR-to-red upconversion in the well-studied material β-NaYF4:Er(3+),Yb(3+) are evaluated in order to resolve inconsistencies that persist in the literature. Each of four possible mechanisms is evaluated in terms of the direct analysis of spectroscopic data. It is shown that there are no important mechanisms that involve the first excited state of Er(3+), (4)I13/2, as an intermediate state. A large body of evidence overwhelmingly supports the proposed mechanism of Anderson et al., which suggests an intimate connection between NIR-to-red and NIR-to-blue upconversion. Namely, both red and blue upconversion are produced primarily by a three-photon excitation process that proceeds through the green emitting state to a dense manifold of states, (4)G/(2)K, above the blue emitting state, (2)H9/2. Competing relaxation mechanisms out of (4)G/(2)K determine the relative amounts of blue and red upconversion produced. Multiphonon relaxation from (4)G/(2)K results in blue upconversion, whereas back energy transfer from Er(3+)((4)G/(2)K) to Yb(3+)((2)F7/2) results in red emission.

Here, we show that the long-accepted mechanism for the production of red and blue emission through upconversion (UC) of 1 μm excitation in Yb(3+)/Er(3+)-doped materials does not apply in the popular β-NaYF4 host. We propose a new mechanism involving Yb(3+)-to-Er(3+) energy-transfer UC out of the green-emitting (2)H11/2,(4)S3/2 states that quantitatively accounts for all of the observed optical behavior. Rate constants for the relevant radiative and nonradiative processes are reported along with a prediction of the power dependence of the pulsed and continuous-wave UC quantum efficiency.

The temperature dependence of the Eu3+ 5D0 relaxation rate in solid europium tris(2,2,6,6-tetramethyl-3,5-heptanedionato) is presented. Eu3+ emission is strongly quenched in the region of room temperature. It is proposed that thermal quenching of 5D0 emission is via crossover from the 5D0 to a ligand-to-lanthanide charge transfer state. Activation energies for the crossover process were determined using the models of Mott and of Englman and Jortner, yielding values of 4120 ± 20 and 5700 ± 300 cm-1, respectively. The crossover process is less efficient in media in which Eu(thd)3 exists as a monomer.

Nanocrystals of β-NaYF4:Yb3+, Er3+ generally have lower NIR-to-visible upconversion (UC) internal quantum efficiency, IQE, compared to high-quality bulk materials, and exhibit more rapid UC dynamics, typical of quenching, when excited with a pulsed source near 980 nm. The addition of a protective shell increases the IQE of the nanocrystals and slows the overall excited-state dynamics. Here, we show that an extension of a recently developed model for UC in powders of micron-sized β-NaYF4:18%Yb3+, 2%Er3+ crystals correctly predicts the time-resolved luminescence curve shapes, relative intensities, and observed drop in IQE of the various emission lines for core and core–shell nanoparticles following pulsed excitation. The model clearly shows that the nanoscale effect on visible upconversion luminescence in these materials, with typical high-Yb3+ and low-Er3+ doping, is largely due to rapid energy migration among Yb3+(2F5/2) and Er3+(4I11/2) ions at the 1 μm energy level, such that an equilibrium is achieved between interior sites and rapidly relaxing surface sites. The faster kinetics observed in visible emission following pulsed NIR excitation is mainly a propagation of the effect of surface quenching of the 1 μm reservoir states and is not due to direct quenching of the visible emitting states themselves. For Er3+ ions contributing to UC emission, the relaxation rate constants for the blue (2H9/2), green (2H11/2, 4S3/2), and red (4F9/2) emitting states are essentially unchanged from their bulk values, indicating that Er3+ ions close to the nanoparticle surface are nearly silent with regard to UC. The addition of a passive β-NaYF4 shell retards the drain of the 1 μm excitation reservoir and recovers the participation of outer Er3+ sites in UC. The dependence of IQE on shell thickness is well explained in terms of a Förster-type model describing an energy donor (Er3+, Yb3+) interacting with a thin plane layer of acceptors (oleate). The UC behavior of both the core and the core–shell nanocrystals can be modeled, almost quantitatively, solely on the basis of quenching at the 1 μm level, without separate consideration of a near-surface Er3+ population. However, a two-layer model for the core nanoparticles is revealing with regard to the modest extent to which near-surface ions do participate in UC and gives a better representation of the detailed dynamics of the NIR emitting states. A method is presented for allowing investigators to estimate the IQE for any nanosample (with 18% Yb3+, 2%Er3+ doping) as a function of excitation power density (cw) or pulse-energy density based on the low pulse energy measurement of the decay constant for the 1 μm emission.

A method is described for producing highly luminescent composite NIR-to-visible upconversion thin films, made from β-NaYF4:3%Er,17%Yb nanocrystals in a polymethyl methacrylate (PMMA) matrix, which require no postdeposition heat treatment. Nanocrystals are synthesized via a single-phase, high-boiling-point solvent method, which requires neither metal-trifluoroacetate precursors nor the use of autoclaves. Highly luminescent films are produced that can be varied in thickness down to dimensions approaching those of the nanocrystals themselves. The physical properties of the films are characterized by AFM and TEM, whereas the spectroscopic properties are characterized by NIR-to-visible confocal microscopy and by the time-dependence of upconversion luminescence following pulsed NIR excitation. It is shown that dispersal of β-NaYF4:3%Er,17%Yb nanocrystals in PMMA has no adverse effect on the intrinsic quantum efficiency of upconversion. By focusing the NIR pump beam (980 nm, cw) in the film, linear intensity response and constant color balance are achieved at pump powers down to 40 μW. It is also demonstrated that the thin-film method can be modified to produce large NIR-to-visible upconversion monoliths of high optical quality. This study supports an earlier assertion that the upconversion properties of β-NaYF4:Er,Yb nanocrystals approach those of the bulk material when nanocrystal size is greater than ∼70 nm.