Applications of Upconversion Nanoparticles in Imaging, Detection and Therapy

Lei Yin Ang; Meng Earn Lim; Li Ching Ong; Yong Zhang


Nanomedicine. 2011;6(7):1273-1288. 

In This Article

Upconversion Phenomenon

Upconversion (UC) is a type of nonlinear process, capable of converting a lower energy excitation light into a higher energy emission.[1] To understand this phenomenon, we will first look at the differences between a nonlinear process and the typical single photon excitation fluorescence and how UC compares with the other common nonlinear processes. As illustrated in Figure 1A, typical fluorophores exhibit the phenomenon of Stokes shift, in which the emission light is of a longer wavelength (λ) than the excitation source (i.e., λex < λem). In terms of energy, this implies that emission energy (Eem) is less than that of excitation (Eex).[2] Additionally, since only one photon is involved in the excitation, the process can be termed as single photon (linear) excitation fluorescence. Hence, an obvious difference between this and the nonlinear processes is the number of photons involved. Nonlinear processes will involve more than one photon (illustrated as two photons in Figure 1B, but can actually involve more than two). Then, as a consequence of this multiphoton excitation, Eex for the nonlinear processes is correspondingly less than that of the emission. Correspondingly, the excitation light has a longer wavelength than the emission light.[3]

Figure 1.

The main differences between (A) single photon excitation fluorescence (linear) and (B) nonlinear processes.
λem: Emission wavelength; λ ex: Excitation wavelength; Eem: Emission energy; Eex: Excitation energy.

Common nonlinear processes (Figure 2) include simultaneous two photon absorption (STPA), second harmonic generation (SHG) and UC. In contrast to STPA in which the excitation photons have to be absorbed almost simultaneously (within an attosecond or 10–18 s), UC's photon absorption process is sequential.[1,4] Furthermore, this sequential absorption is made possible by the existence of real metastable states of the material.[1] This is indeed distinct from STPA, in which such states do not exist between the ground and excited states. In addition, when comparing with SHG, based on hyper Rayleigh scattering and not photon adsorption, the excitation photons are absorbed for UC.[1,4]

Figure 2.

Simplified energy level diagrams illustrating simultaneous two photon absorption, second harmonic generation and upconversion.
SHG: Second harmonic generation; STPA: Simultaneous two photon absorption; UC: Upconversion.

Basic Structure of Upconversion Nanoparticles

The basic structure of upconversion nanoparticles (UCNs) consists of transition metal (3d, 4d, 5d), lanthanide (4f) or actinide (5f) dopant ions embedded in the lattice of an inorganic crystalline host. Of these dopant ions, the trivalent lanthanides are predominantly used because most of them (except lanthanum, cerium, ytterbium and lutetium) have multiple metastable states, making them well-suited for UC. Although a single lanthanide ion is sufficient to produce the UC effect, co-doping is usually favored as most lanthanide ions have low absorption cross-sections leading to weak emission. To enhance UC efficiency, co-doping between two different lanthanide ions, the first serving as an absorber and other acts as an emitter, is generally performed to exploit the energy transfer UC process. Ytterbium (Yb3+) ion is often used as the absorber ion due to its larger absorption cross-section in the near-infrared (NIR) region. Conversely, erbium (Er3+), thulium (Tm3+) and holmium (Ho3+) ions are frequently used emitter ions due to their equally spaced energy levels that facilitate photon absorption and energy transfer in UC processes.[5]

Another important component of UCNs is the host materials, which determines the optical properties and emission efficiency. The desired host materials should have close lattice matches with the dopant ions and low lattice phonon energies to minimize energy losses and maximize radiative emissions. Halide-based compounds are mostly used due to their low phonon energies but the hygroscopic nature of the heavier halides makes the fluorides a more popular choice.[5] To date, many host materials, as well as lanthanide dopant ions, have been used to produce UCNs with different emissions by varying host-dopant combinations (Table 1). Among these lanthanide-doped UCNs, NaYF4 co-doped with Yb3+/Er3+ or Yb3+/Tm3+ nanoparticles have been reported as the materials with the highest UC efficiency.[6]

UCNs Properties

When using UCNs for biological applications, it is important to obtain biocompatible particles with uniform size, homogeneous shape, excellent water dispersity and functional groups for bioconjugation. UCNs with controllable size and shape are typically achieved at the point of particle synthesis. Generally speaking, size and morphology control can be attained by adjusting reaction parameters such as reactant concentration, reaction time and temperature. Different synthesis methods such as co-precipitation, hydrothermal synthesis and thermal decomposition have been applied successfully in this aspect. The various synthesis strategies have been reviewed extensively by Liu's[5] and Xu's[7] groups. Conversely, conferring water dispersity and functional groups to UCNs can be done either during synthesis or post-synthesis. For direct synthesis of water-soluble UCNs, hydrophilic molecules such as oleate anion,[8] polyvinylpyrrolidone (PVP),[9,10] polyacrylic acid (PAA),[10,11] polyethylenimine (PEI),[10,12,13] aminohexanoic acid (AA)[14,15] and polyols[10,16,17] are used as ligands to control particles growth. The majority of these ligands also confer functional groups that are useful for bioconjugation. More specifically, PAA and oleate anions confer the carboxyl groups while UCNs capped with PEI and AA ligands are amine-functionalized. In addition, a synthetic strategy involving an ionic liquid, 1-butyl-3-methylimidazolium tetrafluoroborate [Bmim][BF4], as solvent, has been employed.[18,19] Utilizing the hydrophilic ionic nature of the solvent, the synthesized UCNs are water-dispersible and exhibit strong positive charge. Additionally, when incorporating PEI into the ionic liquid solvent during particle synthesis, the prepared UCNs are amine-functionalized and allow further bioconjugation with recognition elements such as antibodies.[18]

For the hydrophobic UCNs, further modifications to achieve surface hydrophilicity and functional groups are required. These approaches can be broadly classified into two main categories. The first approach involves modifying the hydrophobic capping ligands in one of the following methods: ligand removal via acid-base titration or repeated washing, chemical conversion of ligands into hydrophilic ones or ligand exchange. Ligand removal treatment with hydrochloric acid protonates the carboxyl groups on capping oleate ligand, disrupting the ligand-particles interactions and effectively removing the ligand from the particles' surface.[20] Capping oleic acid may also be removed by washing the particles repeatedly in ethanol under ultrasonication.[21] Though these ligand-free UCNs are hydrophilic and are well-dispersed in aqueous solutions, they do not contain any useful functional groups and may require additional modifications. The oleic acid ligands can also be oxidized by using Lemieux-von Rudloff reagent[22–26] or ozone,[27] resulting in carboxyl-functionalized UCNs. Alternatively, exchange of hydrophobic ligands with hydrophilic ones has been frequently adopted to yield water-soluble UCNs. Some examples of these hydrophilic ligands include hexanedioic acid (HDA),[28,29] polyethylene glycol (PEG) derivatives,[30,31] PAA and its derivatives,[26,32–34] thioglycolic acid (TGA)[35] and citrate.[36,37] Silica coating of UCNs in a reverse microemulsion system is another approach used to produce water-dispersible particles.[34,38–43] For this, the hydrophobic UCNs would be dispersed in an oil phase with surfactants and ammonia which act as the water phase to form a water-in-oil microemulsion. The growth of UCNs in the water pool is restrained by the surfactants and thus created a stable environment for silica-coating. A thin silica layer is then formed surrounding the UCNs, producing core-shell structured particles after hydrolysis of organosilane, usually TEOS. Other than performing silica coating, carbonized glucose coating of UCNs via a similar microemulsion system also produces colloidal particles that possess useful functional groups.[44]

After the UCNs have been appropriately modified, their biocompatibility must be addressed to establish their suitability for biological applications. This evaluation is often based on in vitro cytotoxicity and in vivo biodistribution studies. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) and MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) assays are commonly used to determine in vitro cytotoxicity via quantification of viable cells. Studies involving these assays suggest that UCNs display excellent biocompatibility, exhibiting no significant to low in vitro cytotoxicity to a broad range of cell lines (Table 2). Most of the in vivo biodistribution studies involve injecting the nanoparticles into an animal model and quantifying amount of UCNs retained in different organs after a certain period of time. The animals' weights and behavior are also recorded to assess their health status. Recent work by Yu et al.[45] showed that mice injected intravenously with peptide-modified NaYF4:Yb,Er/Ce UCNs remained healthy and behaved normally 7 days post-injection. Their weight increased continuously over this period of time. Abdul Jalil et al.[43] and Chatterjee et al.[46] also carried out similar studies over an observation period of 7 days. The amount of silica-coated and PEI NaYF4: Yb,Er UCNs were significantly reduced 24 h postinjection. Recently, Zhou et al.[22] have studied the biodistribution of azelaic acid-coated NaGdF4:Tm,Er,Yb in Kunming mice and obtained images of tissues and whole mice 40 min post-intravenous injection. Vital organs were then isolated to quantify the gadolinium concentration. Inductively coupled plasma-mass spectrometry indicated that the uptake of UCNs peaks at spleen and liver, and lower in other organs. The same group later went on to investigate long-term in vivo biodistribution and toxicity of PAA-coated NaYF4:Yb,Tm nanoparticles, accessing their tissue uptake and elimination from Athymic nude mice over a period of 115 days. Histological, hematological and biochemical analysis were done and reported no obvious toxic effects at such long exposure duration.[33] All these data highlight UCNs' biocompatiblity and suitability for use in biological applications.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.