Upconverting nanoparticles (UCNPs) are nanoscale particles (diameter 1–100 nm) that exhibit photon upconversion. In photon upconversion, two or more incident photons of relatively low energy are absorbed and converted into one emitted photon with higher energy. Generally, absorption occurs in the infrared, while emission occurs in the visible or ultraviolet regions of the electromagnetic spectrum. UCNPs are usually composed of rare-earth based lanthanide- or actinide-doped transition metals and are of particular interest for their applications in in vivo bio-imaging, bio-sensing, and nanomedicine because of their highly efficient cellular uptake and high optical penetrating power with little background noise in the deep tissue level. They also have potential applications in photovoltaics and security, such as infrared detection of hazardous materials.
Before 1959, the anti-Stokes shift was believed to describe all situations in which emitted photons have higher energies than the corresponding incident photons. An anti-Stokes shift occurs when a thermally excited ground state is electronically excited, leading to a shift of only a few kBT, where kB is the Boltzmann constant, and T is temperature. At room temperature, kBT is 25.7 meV. In 1959, Nicolaas Bloembergen proposed an energy diagram for crystals containing ionic impurities. Bloembergen described the system as having excited-state emissions with energy differences much greater than kBT, in contrast to the anti-Stokes shift.
Advances in laser technology in the 1960s allowed the observation of non-linear optical effects such as upconversion. This led to the experimental discovery of photon upconversion in 1966 by François Auzel. Auzel showed that a photon of infrared light could be upconverted into a photon of visible light in ytterbium–erbium and ytterbium–thulium systems. In a transition-metal lattice doped with rare-earth metals, an excited-state charge transfer exists between two excited ions. Auzel observed that this charge transfer allows an emission of photon with much higher energy than the corresponding absorbed photon. Thus, upconversion can occur through a stable and real excited state, supporting Bloembergen's earlier work. This result catapulted upconversion research in lattices doped with rare-earth metals. One of the first examples of efficient lanthanide doping, the Yb/Er-doped fluoride lattice, was achieved in 1972 by Menyuk et al.
Physics
Photon upconversion belongs to a larger class of processes by which light incident on a material induces anti-Stokes emission. Multiple quanta of energy such as photons or phonons are absorbed, and a single photon with the summed energy is emitted. It is important to make the distinction between photon upconversion, where real metastable excited states allow for sequential absorption, and other nonlinear processes like second-harmonic generation or two-photon excited fluorescence which involve virtual intermediate states such as the "simultaneous" absorption of two or more photons. It is also distinct from more weakly anti-Stokes processes like thermoluminescence or anti-Stokes Raman emission, which are due to initial thermal population of low-lying excited states and consequently show emission energies only a few kBT above the excitation. Photon upconversion is distinctly characterized by emission-excitation differences of 10–100 kBT. and an observable fluorescence lifetime after the excitation source has been switched off.
Photon upconversion relies on metastable states to facilitate sequential energy absorption. Therefore, a necessary condition for upconverting systems is the existence of optically active long-lived excited states. This role is traditionally filled by lanthanide metal ions embedded in an insulating host lattice. Generally in the +3 oxidation state, these ions have 4fn electronic configurations and typically exhibit f-f transitions. These 4f orbitals allow for complex electronic structures and a large number of possible electronic excited states with similar energies. When embedded in bulk crystals or nanostructures, the energies of these excited states will further split under crystal field, generating a series of states with many closely spaced energies. The 4f shell is localized near the core of the ion and is therefore non-bonding, while the 5s and 5p shells provide further shielding from the exterior crystal field. Thus, the coupling of electronic excited states to the surrounding lattice is weak, leading to long excited state lifetimes and sharp optical lineshapes.
The physical processes responsible for upconversion in nanoparticles are the same as those in bulk crystals on the microscopic level, although total efficiency and other ensemble effects will have unique considerations in the nanoparticle case. The processes contributing to upconversion may be grouped according to the number of ions involved. The two most common processes by which upconversion can occur in lanthanide-doped nanoscale materials are excited state absorption (ESA) and energy transfer upconversion (ETU).
A single ion in the lattice sequentially absorbs two photons and emits a photon of higher energy as it returns to the ground state. ESA is most common when dopant concentrations are low and energy-transfer is not probable. Since ESA is a process where two photons must be absorbed at a single lattice site, coherent pumping and high intensity are much more important (but not necessarily required) than for ETU. Because of its single-ion nature, ESA does not depend on the lanthanide ion concentration.
Two-ion processes are usually dominated by ETU. This is characterized by the successive transfer of energy from singly excited ions (sensitizers/donors), to the ion which eventually emits (activators/acceptors). This process is commonly portrayed as the optical excitation of the activator followed by further excitation to the final fluorescing state due to energy transfer from a sensitizer. While this depiction is valid, the more strongly contributing process is the sequential excitation of the activator by two or more different sensitizer ions.
The upconversion process is said to be cooperative when there are one or more elementary steps (sensitization or luminescence) in the process which involve multiple lanthanide ions. In cooperative sensitization process, two ions in their excited state simultaneously decay to their ground states, generating a higher energy photon. Similarly, in cooperative luminescence, two excited state ions transfer their energy to a neighboring ion in one elementary step.
Energy migration-mediated upconversion (EMU) involves four types of luminescent ion centers with different roles. They are located in separate layers of a core-shell structure of the nanomaterial to inhibit relaxation processes between ions. In this case, low-energy photons are excited in an ETU process that populates an excited state of another ion. Energy from this state can transfer to an adjacent ion through a core-shell interface and is then emitted.
Recently, moving forward in the challenge of designing particles with tunable emissions, important progress in synthesis of high-quality nano-structured crystals has enabled new pathways for photon upconversion. This includes the possibility of creating particles with core/shell structures, allowing upconversion through interfacial energy transfer (IET), upon which the interactions between typical lanthanide donor-acceptor pairs including Yb-Er, Yb-Tm, Yb-Ho, Gd-Tb, Gd-Eu and Nd-Yb can be precisely controlled on the nanoscale.
Photon avalanche occurs under conditions of a high ESA/GSA ratio and efficient cross-relaxation (CR). Ions, typically Tm3+, Pr3+, Nd3+, initially in the ground state weakly absorb energy from the excitation source. This ground-state absorption (GSA) raises the ions to intermediate excited states. The excited-state ions absorb the energy more strongly than the ground-state ions. Highly excited ions undergo CR with neighboring ground-state ions, producing two ions in intermediate excited states. Further cycles of ESA and CR exponentially increase the number of intermediate excited ions. They eventually relax back to the ground state, emitting a large number of photons. Such systems with high nonlinearity exhibit a steep increase in emission intensity with increasing excitation power.
The mechanism for photon upconversion in lanthanide-doped nanoparticles is essentially the same as in bulk material, but some surface and size-related effects have been shown to have important consequences. While quantum confinement is not expected to have an effect on the energy levels in lanthanide ions since the 4f electrons are sufficiently localized, other effects have been shown to have important consequences on emission spectra and efficiency of UCNPs. Radiative relaxation is in competition with non-radiative relaxation, so phonon density of states becomes an important factor. In addition, phonon-assisted processes are important in bringing the energy states of the f orbitals in range so that energy transfer can occur. In nanocrystals, low frequency phonons do not occur in the spectrum, so the phonon band becomes a discrete set of states. With non-radiative relaxation decreasing the lifetimes of excited states and phonon-assistance increasing the probability of energy transfer, the effects of size are complicated because these effects compete with one another. Surface-related effects can also have a large influence on luminescence color and efficiency. Surface ligands on nanocrystals can have large vibrational energy levels, which can significantly contribute to phonon-assisted effects.
Lanthanide-doped upconversion nanocrystals have found broad applications across various fields. However, their low conversion efficiency remains a significant challenge. In recent decades, researchers have developed innovative solutions to synthesize upconversion nanocrystals with greatly improved efficiency. One frequently used approach is surface passivation, which aims to reduce quenching from surface impurities, ligands, and solvent molecules through multiphonon relaxation. Additionally, techniques such as organic dye sensitization, surface plasmon coupling, dielectric superlensing modulation, and resonant dielectric metasurface modulation have been widely employed for nanophotonic control. These techniques play a crucial role in enhancing upconversion luminescence, contributing to the overall improvement in efficiency.
Chemistry
The chemical composition of upconverting nanoparticles, UCNPs, directly influences their conversion efficiency and spectral characteristics. Primarily, three compositional parameters influence the particles' performance: the host lattice, activator ions, and sensitizer ions.
The host lattice provides structure for both the activator and sensitizer ions and acts as a medium that conducts energy transfer. This host lattice has to satisfy three requirements: low lattice phonon energies, high chemical stability, and low symmetry of the lattice. The major mechanism responsible for reduced upconversion is nonradiative phonon relaxation. Generally, if large numbers of phonons are needed to convert excitation energy into phonon energy, the efficiency of the nonradiative process is lowered. Low phonon energies in the host lattice prevent this loss, improving the conversion efficiency of incorporated activator ions. The lattice must also be stable under chemical and photochemical conditions, as these are the environments the conversion will take place within. Finally, this host lattice ought to have low symmetry, allowing for a slight relaxation of the Laporte selection rules. The normally forbidden transitions lead to an increase in the f-f intermixing and thus enhancement of the upconversion efficiency.