Upconversion Nanoparticles: Synthesis, Surface Modification, and Biological Applications

Abstract

New generation fluorophores, also termed upconversion nanoparticles (UCNPs), have the ability to convert near infrared radiations with lower energy into visible radiations with higher energy via a non-linear optical process. Recently, these UCNPs have evolved as alternative fluorescent labels to traditional fluorophores, showing great potential for imaging and biodetection assays in both in vitro and in vivo applications. UCNPs exhibit unique luminescent properties, including high penetration depth into tissues, low background signals, large Stokes shifts, sharp emission bands, and high resistance to photo-bleaching, making UCNPs an attractive alternative source for overcoming current limitations in traditional fluorescent probes. In this review, we discuss the recent progress in the synthesis and surface modification of rare earth doped UCNPs with a specific focus on their biological applications.

1. Introduction

Recent advancements in science have allowed a greater availability of enhanced sensitive analytical techniques, in particular, advanced tools for fluorescence imaging.¹ The last decade has provided a tremendous awareness regarding the use of fluorescent labeled molecules for cell and tissue labeling.² Despite the remarkable applicability of fluorescent dyes in imaging, their current use in visualizing mammalian cells is still greatly limited by the autoflourescence resulting from the excitation of fluorescent dyes. Thus, In order to meet the demands of modern technology, the development of next generation fluorescence probes is highly essential.³

Recent progress in nano-science has allowed scientists to develop new fluorescent nanoparticles (NPs) for biolabeling. These fluorescent labels are conjugated with biomolecules to generate detectable fluorescent signals used for investigating and understanding the complexity and dynamics of biological interactions at the molecular level.⁴ In general, an ideal fluorescent probe should be ultra sensitive, resistant to photo-bleaching, biocompatible, nontoxic and possess high fluorescent efficiency and superior chemical and physical stability.^{3, 5}

In the last decade, organic dyes and fluorescent proteins have been the two most commonly used fluorescent probes for biological detections. Organic fluorescent dyes have several advantages over other fluorescent molecules as a result of their small size, good biocompatibity, easy surface modifications for covalent conjugations, and relatively high fluorescent intensity.³ However; they are limited by short detection times due to high photo-bleaching efficiency and chemical degradation.⁶ In addition, organic dyes often have narrow absorption and broad emission spectra with long tailing which limits their detection.^{3, 6}

With the evolution of nanotechnology, semiconductor quantum dots (QDs) which possess size tunable emission, bright photoluminescence, good photo stability, broad ultraviolet (UV) excitation and narrow emission have emerged as alternative and promising fluorescent labels to conventional organic dyes.⁷ These QDs have proven to have broad applications in areas such as fluorescence resonance energy transfer (FRET) analysis, gene delivery, cell labeling, and tissue imaging.⁸ However, increasing concerns for QDs such as inherent cytotoxicity and chemical instability ⁹ have limited the employment of QDs to long term usage. Moreover, the excitation of traditional biolabels (organic dyes, fluorescent proteins and QDs) usually requires the use of UV or short wavelength radiation which results in serious limitations such as (i) low light-penetration depth inherent to the short wavelength excitation light, (ii) possible damage or even death of biomolecules caused by long-term irradiation, and most importantly, (iii) low signal-to-noise ratio due to significant autofluorescence (background) from biological samples in the UV short wavelength regions.

Therefore, it is highly essential to develop more efficient biolabels to overcome the limitations of traditionally used biolabels. Recently, the conversion of near infrared radiations into visible light via non-linear optical processes, termed upconversion (UC), has generated great interest from researchers. Currently, applications of upconversion nanoparticles (UCNPs) have become very prominent in the medical field. In particular, Rare earth (RE) doped near infrared (NIR)-to-visible UCNPs serve as an alternative and excellent substitute for traditional fluorescent labels.

RE doped UCNPs, which can convert long wavelength radiation (e.g., NIR light) into short wavelength fluorescence (e.g., visible light) via a two-photon or multi-photon mechanism,¹⁰ is emerging as a new class of fluorescent biolabels.^11–15 Distinguished from traditional fluorescent biolabels, RE doped UCNPs can be excited by NIR radiation and possesses several advantages which include (i) an excellent signal-to-noise ratio and improved detection sensitivity owing to the absence of autofluorescence, (ii) deeper NIR light penetration into biological tissue causing less photo damage to biological samples, and (iii) excitation via a low power NIR laser which is compact and inexpensive.^11–15 Additional advantages of UCNPs include narrow emission peaks, large Stokes shifts, good chemical and physical stability, and low toxicity.⁵ Therefore, RE doped UCNPs are very promising alternatives to traditional fluorescent biolabels for potential medical applications.

The main purpose of this review is to provide insight into the recent progress of the synthesis, surface modification, and biological applications of RE doped UCNPs. The challenges and perspectives of these UCNPs are also discussed at the end of this review.

2. Properties of upconversion nanoparticles (UCNPs) and rare earth (RE) doped UCNPs

2.1. Principle and types of Upconversion (UC) processes

The process of UC has been extensively studied in the past few years and has proven to be a successful method for generating visible light from NIR radiation. UC is a non-linear optical process by which excitation of lower electronic levels with low energy radiation (NIR light) results in higher energy emission (visible or UV light) at higher electronic levels, and can therefore be ascribed as an anti-Stokes mechanism. This process requires the absorption of two or more photons to provide the sufficient energy for the UC emission to occur. There are three different classes of UC processing mechanisms which can lead to the two or more photon absorption: excited state absorption, energy transfer, and photon avalanche.¹⁶

2.1.1. Excited state absorption (ESA)

Excited state absorption (ESA), also known as sequential two-photon absorption, is one of the most accepted models of UC, proposed by Bloembergen in 1959.¹⁷ ESA involves the absorption of photons via a single ion, and is the only UC process which occurs in materials with a low dopant concentration. The general energy scheme of ESA is shown in figure 1 and involves the successive absorption of two photons. The first photon causes an ion from the ground state (state 1) to enter a long-lived intermediate excited state (state 2), known as the ground state absorption (GSA). A second photon then promotes this ion from state 2 to the higher excited state (state 3) in optical transition and results in UC emission.

Figure 1.

General energy schemes related to the ESA process

2.1.2. Energy transfer

UC via energy transfer (ET) between excited ions was studied extensively in the mid 1960’s. The pioneering contributions of Auzel resulted in the observation of the ATPE (addition de photons par transfert d’energie) effect,¹⁸ which was later termed energy transfer (ET) UC. Distinguished from the ESA process, the ET process involves a sequential absorption of two photons that transfer energy from an excited ion (sensitizer or donor) to another neighboring ion (activator or acceptor), and is observed in materials with high dopant ion concentrations. There are different types of well known ET mechanisms including ET followed by ESA (EFE), successive energy transfer (SET), cross relaxation (CR), cooperative sensitization (CS), and cooperative luminescence (CL) which are listed in Table 1.

Table 1.

Several types of ET processes

Types	Schematic of strategies	Remarks
EFE		Energy is transferred from the sensitizing ion from an excited state to the activating ion in its state 1 by an ET, promoting the activating ion to its state 2. Next, the activating ion is promoted to its state 3 through an ESA.
SET		The activating ion in state 1 is promoted to its state 2 by an ET. Next, an activating ion is promoted again to its state 3 via a second ET. Only the sensitizing ion can absorb photos from the incident light.
CR		The sensitizing ion and the activating ion are identical ions. Photons from the incident light are absorbed by both ions, promoting these ions into state 2. An ET then promotes the activating ion to its state 3 while the sensitizing ion goes into its lower energy state.
CS		The energy accumulated from two sensitizing ions in their respective excited states is transferred to a single activating ion, promoting the activating ion to its higher excited state.
CL		The emission comprises a single process of one photon from two excited interacting ions which act as both the sensitizing ion and the activating ion.

2.1.3. Photon avalanche

A photon avalanche (PA), also known as an absorption avalanche, is among the most efficient types of UC which was first discovered by Chivian in 1979.¹⁹ The PA process is the least observed among all the UC schemes involving an ESA of incident light as well as intrinsic CR. Figure 2 shows the most simple energy scheme for a PA process.

Figure 2.

General energy schemes related to the PA process

Initially, the sensitizing ion (ion 1) in state 1 is promoted to its state 2 by ground state absorption (GSA). Next, an incident photo promotes it to its state 3 via an ESA. The fundamental nature of the PA process is that one sensitizing ion (ion 1) in state 3 can interact with a neighboring ion (ion 2) in the ground state to produce two ions (ion 1 and 2) in state 2 as a result of CR. The two resultant ions act as sensitizing ions which can produce an additional four ions, which in turn can produce another eight etc.. Eventually the intermediate excited state (state 2) acts as a storage reservoir for energy so that an avalanche of the ion population in state 2 can be established.

2.2. RE doped UCNPs

Recently, RE doped UCNPs have gained considerable attention because of their unique ability to emit light with higher energy than the excitation wavelength. In most cases, RE doped UCNPs are composed of three components: a host matrix, sensitizer, and activator.¹¹ Host matrix is one of the most important components of UCNPs as they provide essential and unique UC optical properties including UC efficiency and emission profile. Sensitizer can be excited effectively from the energy of the incident light source which transfers this energy to the activator where radiation can be emitted. Therefore, activator is the luminescence center in UCNPs while the sensitizer enhances the UC luminescence efficiency. The dopants, sensitizer and activator are added to the host lattice in relatively low concentrations (usually ~20 mol% for the sensitizer and <2 mol% for the activator).

In general, all trivalent RE ions exhibit similar ionic sizes and chemical properties. In addition, their inorganic compounds provide an ideal host matrix for UCNPs. Host matrixes with low lattice photon energies, however, are required to minimize the non-radiative loss and maximize the radiative UC emission. The trivalent Yb³⁺ ion, having an extremely simple energy level scheme, is suitable for use as a UC sensitizer. Most RE³⁺ ions (except La³⁺, Ce³⁺, Yb³⁺, and Lu³⁺ ions) have more than one excited level and are suitable for acting as UC activators. Er³⁺, Tm³⁺, and Ho³⁺ ions have energy levels in the form of a ladder and hence are most frequently used as UC activators. Table 2 shows several types of known UCNPs as well as their luminescent properties (under 970–980 nm NIR excitation).

Table 2.

Different UCNPs and their luminescent properties

Host matrix	Sensitizer/activator	Emissions (nm)	Ref.
Oxide
Y2O3	Yb/Er	660	17
	Yb/Tm	450, 480	19
	Yb/Ho	549, 666	21
Lu2O3	Yb/Er	662	22
	Yb/Tm	477, 490	22
La2O3	Yb/Er	530, 549, 659, 672	24
Gd2O3	Yb/Er	520–580, 650–700	26
Oxysulfide
Y2O2S	Yb/Er	520–560, 650–680	28
	Yb/Tm	450–500, 650, 690	28
Gd2O2S	Yb/Er	520–580, 650–700	31
La2O2S	Yb/Pr	500, 508, 830	33
Oxyhalide
GdOF	Yb/Er	521, 545, 659	36
YOF	Yb/Er	525, 545, 656	38
Phosphate
LaPO4	Yb/Er	535–556	40
LuPO4	Yb/Tm	476	42
Molybdate
La2(MoO4)3	Yb/Er	519, 541	45
	Yb/Tm	472, 647	47
tungstate
NaY(WO4)2	Yb/Er	526, 553, 660	18
	Yb/Tm	476, 647	20
gallate
Gd3Ga5O12	Yb/Tm	454,484, 640–680	23
vanadate
YVO4	Yb/Er	547, 554, 660–670	25
fluoride
LaF3	Yb/Er	521, 545, 659	27
	Yb/Tm	475, 698, 800	29
	Yb/Ho	541, 643	30
YF3	Yb/Er	411, 526, 552, 664	32
	Yb/Tm	347, 363, 454, 477	34
LuF3	Yb/Tm	481	35
NaYF4	Yb/Er	525, 547, 660	37
	Yb/Tm	450, 476	39
	Yb/Ho	541	39
LiYF4	Yb/Tm	361, 450, 479, 647	41
NaGdF4	Yb/Ho	541, 647, 751	43
KY3F10	Yb/Er	522, 545, 656	44
KGd2F7	Yb/Er	525, 552, 666	46
BaYF5	Yb/Tm	475, 650, 800	48

2.3. RE doped NaYF₄ UCNPs

Among the available different types of UC host materials, fluorides have proven to be ideal host candidates for UC because of their very low phonon energies and high chemical stability. In particular, Yb³⁺/Er³⁺ and Yb³⁺/Tm³⁺ doped NaYF₄ have been previously reported as the most efficient UC materials for giving green and blue emissions.^{51, 52} Recently, Liu’s group along with other have focused on the synthesis, surface modification, and biological applications of RE doped NaYF₄ UCNPs.^{11–15, 53}

There are two types of crystalloid phases that exist in NaYF₄ at ambient pressure: the cubic phase (α-phase) and the hexagonal phase (β-phase).⁵⁴ It has been reported that the more thermodynamically stable, β-NaYF₄, can be transformed from the α-NaYF₄. Interestingly, the NaYF₄ phase we often obtained was the thermodynamically meta stable α-phase rather than the stable β-phase. That is due to the fact that the free-energy barrier for the conversion from α-NaYF₄ to β-NaYF₄ is very high, requiring more energy for the transition from the cubic-to-hexagonal phase.⁵⁵ In most cases, β-NaYF₄ can be obtained via a heating treatment such as annealing, hydrothermal or solvothermal treatments, often requiring harsh conditions with high reaction temperatures and long reaction times.

The UC intensity of RE doped NaYF₄ is closely related to its phase where the UC efficiency of β-NaYF₄ is much higher than that of α-NaYF₄. For example, the green emission of β-NaYF₄:Yb,Er is 10 times stronger than that of α-NaYF₄:Yb,Er and the overall (green plus red) emission of the former is 4.4 times higher than that of the latter. Thus, in order to overcome the current limitations of traditional fluorescent probes in biological applications, RE doped β-NaYF₄ UCNPs are appropriate substituents due to their superior UC properties.

3. Synthesis and formation mechanism of RE doped NaYF₄ UCNPs

3.1. Coprecipitation method

The coprecipitation method is one of the easiest and most convenient approaches for synthesizing RE doped NaYF₄ UCNPs because of their mild reaction conditions, low costs for required equipment, simple protocols, and short reaction times. Yi et al.⁵⁶ firstly reported the coprecipitation method in synthesizing NaYF₄:Yb,Er UCNPs with the aid of ethylenediamine tetraacetic acid (EDTA). In this procedure, Yi et al. quickly injected an RE-EDTA complex into a NaF solution while stirring vigorously. This resulted in the formation of α-NaYF₄:Yb,Er UCNPs through a homogeneous nucleation process. The sizes of the resulting UCNPs can be controlled from 37 to 166 nm by varying the molar ratio of EDTA to the total RE³⁺ ions. Because the resulting UC fluorescent intensity of the α-NaYF₄:Yb,Er UCNPs was too weak to be applied to biological labeling, an annealing treatment was used to enhance the UC fluorescent intensity. This resulted in the phase transition from a cubic-to-hexagonal phase, thereby increasing the fluorescent intensity of the UCNPs by up to 40-fold.

However, after annealing, the UCNPs tended to aggregate into larger sizes at higher temperatures (above 400 °C), which limited their potential applications. Later, Wei et al.⁵⁷ reported a similar method to synthesize NaYF₄:Yb,Tm UCNPs in the presence of EDTA. Their studies showed that changes in pH greatly affected the morphology of the resulting NPs. For example, the particles tended to aggregate when the pH value was 10.0, and began to form fibroid like structures when the pH value was increased to 12.0. The differential scanning calorimetric (DSC) experiments and analysis showed that EDTA molecules capped on the surface of the NPs could suppress the cubic-to-hexagonal phase transition, prohibiting transformation from the α-phase to β-phase of the NPs even when annealed at 600 °C for 5 h.

For the synthesis of RE doped NaYF₄ UCNPs by the coprecipitation method, post-heat treatment (annealing) is typically required to promote UC fluorescent intensity which in turn allows the NPs to aggregate and become larger. Since annealing with capping reagents such as EDTA can become carbonized after the annealing process, reducing the hydrophilicity of the NPs, further surface modifications such as surface silica coating is often required to improve the hydrophilicity of the NPs to allow further increasing of NP size. Therefore, the application of RE doped NaYF₄ UCNPs synthesized via the coprecipitation method in biological field is limited.

3.2. Thermal decomposition method

Since metallic trifluoroacetates can thermally decompose to their corresponding metal fluorides, Zhang et al.⁵⁸ first reported a novel synthesis of single-crystalline and monodisperse LaF₃ triangular nanoplates via the thermal decomposition of lanthanum trifluoroacetates (La(CF₃COO)₃). Since its discovery, this method has become a common route for synthesizing high quality RE doped NaYF₄ UCNPs.

Mai et al.⁵⁵ reported a general synthesis of high-quality (monodisperse, single-crystalline, well-shaped and phase-pure) NaREF₄ (RE = Pr to Lu, Y) NPs using Na(CF₃COO) and RE(CF₃COO)₃ as precursors. The combined solvents were composed of both the coordinating solvent and non-coordinating solvent. 1-octadecene (ODE) with a high boiling point (315 °C) was used as the non-coordinating solvent to provide a high temperature environment. Oleic acid (OA) and oleyamine (OM), having good coordinate abilities, acted as the coordinating solvent which aids in capping the surface of NPs to prevent agglomeration. From a series of experiments, they demonstrated that pure β-NaYF₄ could be prepared from the OA-ODE system under harsh conditions (high Na/RE ratio, high temperature, and long reaction time), while pure α-NaYF₄ could be obtained from the OA-OM-ODE system under relatively mild conditions (low Na/RE ratio, low temperature, and short reaction time). These methods allowed the successful synthesis of high quality α- and β-phase NaYF₄:Yb,Er/Tm UCNPs.

Boyer et al.⁵⁹ also reported a similar procedure to synthesize α-NaYF₄:Yb,Er/Tm UCNPs using Na(CF₃COO) and RE(CF₃COO)₃ as precursors, however, the particle size distribution resulted in a broad range (10–60 nm). They slowly added the dissolved precursors to the reaction solution which resulted in monodisperse α-NaYF₄:Yb,Er/Tm UCNPs with narrow size distributions (22–32 nm).⁶⁰ They showed that the rate of decomposition and particle formation could be controlled via the addition of precursors to the reaction solution over a long period of time. Thus, the nucleation and growth stages of the NPs could be separated resulting in monodisperse NPs with narrow distributions. However, it is still necessary to understand the detailed mechanism of size and phase controlled synthesis of UCNPs.

Mai et al.⁶¹ found that the UC emission (the intensity and ratio of green to red emission) was sensitive to the growth process (nucleation and phase transition process) of NaYF₄:Yb,Er UCNPs and provided a powerful tool for investigating the growth and phase transition stages of NPs. In their studies, they demonstrated that NaYF₄:Yb,Er UCNPs synthesized by the thermal decomposition of precursors, were formed via a unique delayed nucleation pathway in a solution phase. In addition, monodisperse β-NaYF₄:Yb,Er UCNPs with tunable sizes could be obtained from α-NaYF₄:Yb,Er monomers by restricting or enhancing the Ostwald-ripening process in which the cubic-to-hexagonal phase transition process happened in a delayed time. This method was later developed as a common route to synthesize other kinds of NPs.

Ehlert et al.⁶² have developed a series of RE doped NaYbF₄ UCNPs via a typical thermal decomposition method. More interestingly, the Er³⁺, Tm³⁺ and Ho³⁺ ion doped NaYbF₄ UCNPs could produce red, blue and green light under a 980 nm radiation, respectively, showing their potential use in multiplex analysis. In another example, Du et al.⁶³ reported the synthesis of NaMF₃ (M = Mn, Co, Ni, Mg) and LiMAlF₆ (M = Ca, Sr) nanocrystals using a similar approach. It should be noted that NaMgF₃:Yb,Er nanorods with the size of (337.2 ± 66.3) × (10.3 ± 1.6) nm² synthesized by Du et al. was the first report to show a strong red emission under 980 nm radiation.

More recently, Shan et al.⁶⁴ reported a new thermal decomposition method for the synthesis of monodisperse β-NaYF₄:Yb,Er/Tm/Ho UCNPs using a single solvent (trioctylphosphine oxide, TOPO) rather than the previously mentioned component solvent (OA/ODE or OA/OM/ODE). Here, TOPO was used as a boiling solvent and capping reagent to control crystalline growth by providing a broad temperature window for the synthesis of β-NaYF₄:Yb,Er/Tm/Ho UCNPs. The as-prepared UCNPs were small in particle size (~10 nm) with a high UC efficiency and narrow size distribution. They demonstrated that the energy barrier of the cubic-to-hexagonal phase transition process was significantly reduced using TOPO allowing more efficient formation of β-phase UCNPs. Later, they reported a similar synthesis of β-NaYF₄:Yb,Er UCNPs using ODE and the combined capping reagents of OA and trioctylphosphine (TOP). The kinetic mechanisms for the particle phase transition and growth were discussed in detail.⁶⁵ From a series of experimental results, they proposed that the formation of β-phase UCNPs could be divided into two stages: firstly, a kinetic controlled precipitation stage for the formation of α-phase UCNPs, and secondly, the diffusion controlled growth and size focusing stage for the formation of β-phase UCNPs. They also found that the UC intensity was proportional to the size of the β-phase UCNPs.

Chen et al.⁶⁶ reported a novel thermal decomposition method for synthesizing α- and β-phase NaYF₄:Yb,Er UCNPs using RE-oleate complexes as the precursors and ODE as the reaction solvent. This could also be attributed to a liquid-solid two-phase approach. By adjusting the reaction temperature, α-NaYF₄:Yb,Er UCNPs could be obtained at 210 °C for 6 h, and β-phase UCNPs could be prepared at 260 °C for 6 h which is relatively milder than other thermal decomposition methods mentioned above.

Although the thermal decomposition method has proven to be an effective approach for fabricating monodisperse, single-crystalline, well-defined, and phase-pure UCNPs, there remains some disadvantages such as required rigorous (anhydrous and oxygen-free) and harsh (long reaction time and high reaction temperature) synthesis conditions. In addition, the thermal decomposition of the metal trifluoroacetates produces fluorinated and oxy-fluorinated carbon species which are considered to be toxic. More importantly, the capping ligands such as OA, OM, and TOPO bind to the surface of NPs via outward hydrophobic alkyl chains, rendering the NP surface hydrophobic. This limits the use of these UCNPs for biological applications in which further surface modifications are essentially required.

3.3. Hydrothermal/solvothermal method

The hydrothermal/solvothermal method refers to a chemical synthesis procedure within a sealed environment under high pressure and temperature (often above the critical point of the solvent). In a typical hydrothermal/solvothermal synthesis, specialized reaction vessels known as autoclaves are often used to provide the sealed reaction environment.

Sun et al.⁶⁷ reported a hydrothermal synthesis of α- and β-phase NaYF₄:Yb,Er UCNPs using RE-EDTA or RE-citrate complexes as precursors. EDTA and citrate were used as capping ligands to control the size and morphology of the NPs. They found that the particle size was dependent on the nucleation rate which could be controlled by the concentration of the reactants, molar ratio of RE, capping ligands to NaF, and choice of capping ligands. There are several examples for the hydrothermal synthesis of high-quality UC micro/nanoparticles using EDTA or citrate as capping reagents.^68–74

Liang et al.⁷⁵ also reported a solvothermal synthesis of branched NaYF₄:Yb,Er UCNPs in the presence of cetyltrimethylammonuim bromide (CTAB) in a methanol-water system. Here, CTAB was used as a regulating reagent to facilitate the branched growth of NPs. Due to its unique structure, these branched NaYF₄:Yb,Er UCNPs could be introduced into polystyrene to form a novel multifunctional polymer which showed green excitation emissions under 980 nm. Zeng et al.⁷⁶ reported a solvothermal synthesis of β-NaYF₄:Yb,Er UCNPs in selected solvents (water, acetic acid or ethanol) with the aid of EDTA and CTAB. In their studies, the size and morphology of the NPs were controlled via EDTA and CTAB, respectively. EDTA possesses excellent chelating abilities that can reduce particle size and CTAB acts as a surfactant for tuning the morphology of nanocrystals from nanoparticles to nanorods. Later, EDTA-CTAB was used to aid the solvothermal method developed by Zeng et al.⁷⁷ as a general synthesis route for preparing complex RE fluoride NPs such as NaYF₄:Yb,Er/Tm, NaGdF₄:Eu, and NaCeF₄ NPs.

Wang et al.⁷⁸ developed a one-step synthesis of polyethylenimine (PEI) coated NaYF₄:Yb,Er/Tm UCNPs via a solvothermal approach. This was the first report made on the one-step synthesis of water-soluble and biocompatible NaYF₄:Yb,Er/Tm UCNPs. PEI, an organic polymer surfactant, was used in this study to control the particle size and prevent particle aggregation. More importantly, the free amine groups on the surface of the UCNPs could bind with biomolecules (such as antibodies), facilitating the use of UCNPs in biological applications. They also demonstrated that the PEI coated UCNPs were biocompatible with mammalian cells. Later, they reported a similar approach for synthesizing PEI-coated NaYF₄:Yb,Er/Tm UCNPs using ethylene glycol as the reaction solvent.⁷⁹ Interestingly, they developed three-component (Yb, Er, Tm) doped NaYF₄ UCNPs. After fine-tuning by adjusting the amount of dopants, these UCNPs could emit multi-colors ranging from blue to white which has great potential uses for multiplexed labeling.

Wang et al.⁸⁰ developed a novel liquid-solid-solution (LSS) strategy for the synthesis of NPs which exhibited semiconducting, fluorescent, magnetic, and dielectric properties. This LSS strategy involved two main processes called phase transfer and phase separation and was further developed as an effective way to synthesize RE fluoride NPs^{78, 81} including RE fluoride UCNPs.⁸² Later, Wang et al.⁸³ reported a solvothermal synthesis of monodisperse NaYF₄:Yb,Er single-crystal nanorods, hexagonal nanoplates, and nanoparticles using the LSS synthetic strategy. From the experimental results, they indicated that high NaF content, low RE³⁺ content, appropriate temperatures, and long reaction times could facilitate the epitaxial growth of NaYF₄:Yb,Er NPs.

Our group, Wang et al.⁴² reported a two-phase solvothermal synthesis of NaYF₄:Yb,Er/Tm/Ho UCNPs in a water-ethanol-oleic acid system (Figure 3). In their studies, RE stearate was first used as the precursor. Later, they developed a one-step synthesis of polymer coated NaYF₄:Yb,Er UCNPs by employing a series of polymers such as polyvinylpyrrolidone (PVP), polyethylene glycol (PEG), polyacrylic acid (PAA), and polyethylenimine (PEI) rather than oleic acid.⁸⁴ These polymers which were capped on the surface of the UCNPs, proved highly important as they made the NPs hydrophilic and prevented them from aggregating.

Figure 3.

Mechanism for the synthesis of NaREF₄ UCNPs by solvothermal method⁴²

The advantages for adopting a hydrothermal/solvothermal method for synthesizing high quality UCNPs include (i) a high purity of product, (ii) easy control of size, structure, and morphology of the NPs, (iii) relatively lower reaction temperatures (in general below 200 °C),⁸⁵ and (iv) simple equipment and overall process.⁸⁶ However, in most cases, the hydrophilicity of UCNPs prepared using the hydrothermal/solvothermal method was not sufficient due to the presence of hydrophobic organic ligands (such as oleic acid) on the surface of the NPs. Therefore, to improve the water-solubility and biocompatibility of the UCNPs, surface modifications are a high priority.

3.4. Other methods

Polyols which have high boiling points and good water-solubility, have widely been used as solvents for preparing different varieties of NPs including metals,⁸⁷ oxides,⁸⁸ and phosphates.⁸⁹ Wei et al.⁹⁰ reported a polyol method for preparing NaYF₄:Yb,Er/Tm UCNPs. In their studies, three kinds of polyols which included ethylene glycol, diethylene glycol, and glycerol were used not only as solvents but also as capping ligands for limiting the growth of NPs and stabilizing them from aggregation. After the solvothermal procedure, the phase of the UCNPs was converted from the α-phase to β-phase resulting in an enhancement of the fluorescent intensity. Since polyols are water soluble, the as-prepared UCNPs could be well dispersed in water. Other kinds of fluorides such as NH₄Y₃F₁₀ and YF₃ were also synthesized using this polyol-mediated method.⁹¹

Li et al.⁹² reported a simple and user-friendly method for the synthesis of β-NaYF₄:Yb,Er/Tm UCNPs with enhanced fluorescence. This system has two major steps: the first, nucleation at room temperature and the second, particle growth at elevated temperatures (300 °C). Similar to typical thermal decomposition methods, OA and ODE were used in this method to provide a high-temperature environment for the cubic-to-hexagonal phase transition. Here, the use of metal trifluoroacetates was avoided to limit the toxicity of fluorine species. However, use of rigorous synthetic conditions involving anhydrous and oxygen free environments were still required.

Heer et al.⁹³ reported a three-step synthesis of NaYF₄:Yb,Er/Tm UCNPs using N-(2-hydroxyethyl)-ethylenediamine (HEEDA), a high boiling organic solvent. This method involved the preparation and reaction of two educt-solutions (containing Na⁺ and RE³⁺ cations, and F⁻ anions, respectively) in HEEDA. The resulting UCNPs were dispersed in dimethyl sulfoxide (DMSO) to form a transparent colloidal solution. This method is the first reported synthesis of NaYF₄:Yb,Er/Tm UCNPs.

4. Modification of RE doped NaYF₄ UCNPs

In general, RE doped NaYF₄ UCNPs synthesized using the above mentioned methods are not water soluble due to the presence of hydrophobic organic ligands such as oleic acid on the surface of the NPs. In most cases, even if the NPs are water dispersible, there are no appropriate functional groups (such as carboxyl groups or amino groups) on the surface of the NPs. Therefore, the conjugation between the UCNPs and biomolecules becomes difficult for practical biomedical applications. To overcome this problem, further surface modifications are necessary. The surface modification of the UCNPs can be achieved in two ways: (i) surface modification with an inorganic shell layer, and (ii) surface modification with organic capping ligands.

4.1. Surface modification with an inorganic shell layer

In most cases, surface modification with an inorganic shell layer is achieved via surface silanization. In this procedure, the surfaces of synthesized NPs are further coated with amorphous silica. A well-known Stober procedure which involves the hydrolysis and condensation of siloxane precursors such as tetraethoxysilane (TEOS) in the presence of ethanol and ammonia, has been widely used for the surface silanization of UCNPs.⁹⁴

A good example of coating the surface with silica was reported by Li et al., who coated polyvinylpyrrolidone (PVP) stabilized NaYF₄:Yb,Er/Tm UCNPs with a layer of silica via a typical Stober method as shown in Figure 4.⁹⁵ In their studies, the thickness of the silica shell was approximately 10 nm and could be controlled within 1–3 nm by adjusting the initial amount of TEOS taken. After coating with silica, the UCNPs were quite stable in water and still emitted strong UC fluorescence. The use of Stober method to coat silica is not only limited to UCNPs synthesized via the hydrothermal/solvothermal method but is also applicable to UCNPs prepared via the thermal decomposition method. Shan et al.⁹⁶ modified the NaYF₄:Yb,Er UCNPs synthesized in hydrophobic organic solvents (OA-ODE and OA-TOP-ODE) with a layer of silica using a Stober-based microemulsion reaction. However, the silica-coated NPs tend to be aggregate when removed from the microemulsion. Later, Johnson et al.⁹⁷ reported a facile method to solve this problem. They changed the OA-coated hydrophobic NaYF₄:Yb,Er UCNPs to hydrophilic, which involved a ligand exchange between OA and PVP, and then modified the PVP-stabilized UCNPs with a layer of silica. The results show that the resultant silica-coated NPs have a higher colloidal stability and a better monodispersity, compared with the NPs prepared in reverse microemulsion.

Figure 4.

Schematic illustration of the surface silica-coating of NaYF₄:Yb,Er@PVP UCNPs

Recently, Li et al.⁹⁸ reported a novel surface modification combined with the fluorescence resonance energy transfer (FRET) mechanism to generate multicolor silica-coated UCNPs. In their work, some organic dyes or QDs were first encapsulated together with NaYF₄:Yb,Er or NaYF₄:Yb,Tm UCNPs, respectively, using a Stober-based microemulsion method. When excited by a 980 nm radiation, the UC fluorescence emitted from the UCNPs (core) was transferred to the organic dyes or QDs (shell) via a FRET process, thus, different colors could be generated using different organic dyes or QDs (Figure 5). These multicolor NPs have great potential in multiplexed bioassays.

Figure 5.

Schematic illustration of silica coated multicolor UCNPs based on the FRET mechanism

Even though coating the surface of NPs with silica allows them to be water dispersible, it may not provide the required functional groups for bioconjugation. To solve this problem, the hydrolysis of amino siloxanes such as (3-aminopropyl)triethoxysilane (APS) are often performed to generate functional amino groups on the surface of the silica coated NPs.

Recently, synthesis of multi-functional NPs have drawn a great deal of attention.⁹⁹ Lu et al.¹⁰⁰ reported the synthesis of NaYF₄:Yb,Er core-shell UCNPs with multi functional capabilities including magnetic, fluorescent, and bioaffinity properties. In this procedure, they used iron oxide NPs as a magnetic core covered by a layer of NaYF₄:Yb,Er via a coprecipitation method. After annealing, the fluorescence intensity of the NPs was greatly enhanced. The resulting NPs were then coated with silica and amino-functionalized by the hydrolysis of TEOS and APS which rendered the UCNPs water-soluble and biocompatible (Figure 6).

Figure 6.

Schematic illustration of the modification of magnetic NaYF₄:Yb,Er UCNPs with silica coating

Liu et al.¹⁰¹ also reported a synthesis of superparamagnetic NaYF₄:Yb,Er core-shell UCNPs. In their synthesis, NaYF₄:Yb,Er UCNPs together with Fe₃O₄ superparamagnetic NPs acted as a core and were encapsulated in a silica shell via the hydrolysis of TEOS and APS in a reverse microemulsion system. These biocompatible NPs exhibited both UC fluorescence and magnetic properties and are emerging as promising materials for bioimaging, drug targeting, and bioseparation.

Likewise, Mi et al.¹⁰² have recently reported that active functional groups on the surface of Fe₃O₄/NaYF4:Yb,Er magnetic/luminescent nanocomposites can be used to be successfully conjugated to antibodies to recognize mammalian cells. In their recent paper, Mi et al. have successfully conjugated trasferrin protein on to luminescent nanocoposites and were able to specifically recognize the transferrin receptors which were over expressed on HeLa cells.

4.2. Surface modification with organic capping ligands

Schafer et al.¹⁰³ reported the first synthesis of NaYF₄:Yb,Er/Tm UCNPs in HEEDA, and further modified the surface with an organic ligand, namely, 1-hydroxyethane-1,1-diphosphonic acid (HEDP). After surface modification, the resulting NaYF₄:Yb,Er/Tm UCNPs were water soluble, These UCNPs were further purified by the removal of HEEDA which resulted in an increased UC efficiency. This method, however, required high temperatures (180 or 320 °C) and a vacuum (p < 0.1 mbar).

Yi et al.¹⁰⁴ reported the synthesis of oleyl amine (OM) stabilized β-NaYF₄:Yb,Tm UCNPs via a typical thermal decomposition method. The hydrophilic amino groups of OM could bind to the surface of the UCNPs while the hydrophobic alkyl chains extended outward resulting in a hydrophobic NP surface. Later, they changed the hydrophobic surface to hydrophilic using a bipolar surfactant, polyethylene glycol 600 diacid (PEG600 diacid), which involved a ligand exchange between OM and PEG600 diacid. In the ligand exchange process, the carboxyl group on one end of PEG600 diacid was capped onto the surface of the UCNPs while the carboxyl group on the other end extended outward, making the UCNPs more hydrophilic (Figure 7a). After the ligand exchange, the NPs could be dispersed in water to form a clear and stable colloid solution for at least two weeks. The free carboxyl groups on the surface of the UCNPs are of great importance for further conjugation with biomolecules. Zhang et al.¹⁰⁵ also reported a similar ligand exchange approach for converting the OM modified NaYF₄:Yb,Er/Tm UCNPs from hydrophobic to hydrophilic using hexane diacid as the modifier. Boyer et al.¹⁰⁶ changed the hydrophobic OA-coated NaYF₄ NPs to hydrophilic by the aid of PEG-phosphate ligands via a ligand exchange process, and then coated the resultant NPs with a silica shell to avid the luminescent quenching effect of the NPs in aqueous environments.

Figure 7.

The mechanisms for organic ligand modification of UCNPs: (a) ligand exchange, (b) ligand attraction, (c) ligand oxidation, and (d) ligand assembly

Later, Yi et al.¹⁰⁷ reported another method for modifying the surface of OM stabilized β-NaYF₄:Yb,Er/Tm NaYF₄ core-shell UCNPs via a ligand attraction process. In their studies, the β-NaYF₄:Yb,Er/Tm core was first coated with a NaYF₄ shell to enhance the UC fluorescence intensity of the NPs. Then, an amphiphilic layer, 25% octylamine and 40% isopropylamine modified polyacrylic acid (PAA), was used as the modifier to make the UCNPs hydrophilic. This process involved hydrophobic interactions between the alkyl chains, the octyl on the surfaces of the NPs and the octyl and isopropyl groups in the modified PAA molecules (Figure 7b). After achieving surface modification, the hydrophilic carboxyl groups of the modified PAA extended outwards which made the NPs water-soluble, facilitating their attachment to biomolecules. This method is suitable for UCNPs with hydrophobic functional groups on the surface.

Based on the fact that a carbon-carbon double bond can be oxidized directly to form two carboxyl groups with the aid of a Lemieux-von Rudloff reagent (KMnO₄ and NaIO₄ aqueous solution), Chen et al.¹⁰⁸ developed a novel ligand oxidation method for modifying OA stabilized NaYF₄:Yb,Er UCNPs. In this ligand oxidation process, the oleic acid capped on the surface of UCNPs was oxidized into azelaic acid, resulting in an extension of the newly generated free carboxylic groups (Figure 7c). They demonstrated that the oxidation did not affect the morphology, phase transition, composition, or UC fluorescence of the UCNPs. These modified UCNPs could be easily dispersed in water and further conjugated with biomolecules such as streptavidin due to the free carboxylic groups on the surface of the UCNPs. However, this method is only applicable to a specific class of UCNPs such as those capped with a ligand containing an unsaturated carbon-carbon bond in its long alkyl chain.

Wang et al.¹⁰⁹ used a more applicable method, layer-by-layer (LbL) assembly to convert the hydrophobic UCNPs into hydrophilic. This method involved electrostatic attractions between the oppositely charged species deposited. At weakly basic conditions (pH 8.5), the UCNPs were negatively charged and fabricated using positively charged polyallylamine hydrochloride (PAH), giving positively charged NPs. Subsequently, negatively charged polystyrene sulfonate (PSS) was used to assemble the resultant NPs which was then assembled by PAH again (Figure 7d). After the LbL assembly process, amino groups from the amino-rich PAH were introduced onto the surface of the UCNPs, allowing the modified NPs to be attached to biomolecules such as biotin. The advantage of the LbL assembly method is that it permits the preparation of coated colloids with varying shapes and sizes as well as precise control of the thickness layer of the polymer. However, before modification, the NPs are usually required to be hydrophilic.

5. Biological applications of RE doped NaYF₄ UCNPs

5.1. In vitro cellular imaging and in vivo tissue imaging

The excitation of RE doped UCNPs often requires IR radiation. The use of IR radiation has great advantages in cell and tissue imaging because they provide (i) a high signal-to-noise ratio, (ii) strong penetration ability (can penetrate tissues up to several inches) inherent to the long wavelength excitation light, and (iii) less photo damage to the cell/tissue under long-term irradiation.¹¹⁰ Therefore, RE doped UCNPs are promising alternatives to traditional fluorescent biolabels (such as organic dyes and QDs) for in vitro cell and in vivo tissue imaging.

5.1.1. In vitro cell labeling and imaging

Three years ago, Chatterjee et al.¹¹¹ reported the first use of NaYF₄:Yb,Er UCNPs for cellular imaging in which the NaYF₄:Yb,Er UCNPs were first functionalized with PEI and later covalently conjugated with folic acid to form folic acid modified NaYF₄:Yb,Er UCNPs. The folic acid modified NaYF₄:Yb,Er UCNPs were then incubated with human HT29 adenocarcinoma cells and human OVCAR3 ovarian carcinoma cells under physiological conditions for 24 h. As a result of the abnormally high levels of folate receptors expressed on the surface of both cells, the folic acid modified UCNPs were able to specifically target the cells. As shown in Figure 8, when attached, the UCNPs could emit green UC fluorescence under a confocal microscope equipped with a 980 nm laser. Their control experiment demonstrated that the binding between the folic acid modified UCNPs and the folic acid receptors expressed on the cell surface was target specific.

Figure 8.

Fluorescent images of live human ovarian carcinoma cells (OVCAR3, top row) and human colonic adenocarcinoma cells (HT29, bottom row) after incubation with folic acid modified PEI/NaYF₄ UCNPs. The left rows are images in bright field, the middle rows are fluorescent images in dark field, and the right rows are overlays of the left and middle rows¹¹¹

In the same year, Jalil et al.¹¹² reported the in vitro cytotoxic effect of silica coated NaYF₄:Yb,Er UCNPs on two types of cells: rat skeletal myoblasts and bone marrow-derived stem cells (BMSCs). In their studies, the silica coated NaYF₄:Yb,Er UCNPs excluding any functionalization, were incubated under physiological conditions with the two cell types for 24 h. As shown in Figure 9, observed using a confocal microscope under 980 nm radiation, the silica coated UCNPs were mainly located in the cytoplasmic and perinuclear regions shown by bright green fluorescence. They also demonstrated that both the rat skeletal myoblasts and BMSCs survived reasonably well even at relatively high concentrations of UCNPs (100 μg/ml in this study) indicating that the silica coated NaYF₄:Yb,Er UCNPs had good in vitro biocompatibility for cellular imaging.

Figure 9.

Bright field (left) and fluorescent (right) images of bone marrow-derived mesenchymal stem cells (A, B) and skeletal myoblasts (C, D) treated with silica coated NaYF₄:Yb,Er UCNPs¹¹²

Nyk et al.¹¹³ reported high contrast imaging of human pancreatic cancer cells (Panc 1) using NaYF₄:Yb,Tm UCNPs. In their studies, OA coated NaYF₄:Yb,Tm UCNPs were first modified with 3-mercaptopropionic acid (MPA) to make them water-soluble and then incubated with Panc 1 cells at 37 °C for 2 h. Under a 980 nm excitation, the NPs on the surface of the cells could emit IR light at approximately 800 nm (Figure 10), when viewed under a Nuance GNIR CCD camera. It should be noted that the photoluminescence imaging showed high contrast, inherent three-dimensional localization features and a complete absence of autofluorescence. In addition, no overt toxicity of these UCNPs was found during the cell viability assay.

Figure 10.

Bright field (a) and fluorescence (b) images of Panc 1 cells treated with NaYF₄:Yb,Tm UCNPs¹¹³

Our group, Wang et al.^{114, 115} have recently reported an effective and time efficient method for immunolabelling and imaging of HeLa cells via silica coated NaYF₄:Yb,Er and NaYbF₄:Er/Tm/Ho UCNPs. In this study, the silica coated amino-modified UCNPs were first linked to the rabbit anti-CEA8 antibody to form antibody modified UCNPs and were then incubated with HeLa cells in physiological conditions for 1 h. During the incubation period, the immunoreaction between the rabbit anti-CEA8 antibody modified UCNPs and carcinoembryonic antigens (CEA) expressed on HeLa cell membranes occurred resulting in an attachment of the UCNPs to the surface of the cells (Figure 11a). As a result, HeLa cells were labeled by UCNPs. From the control experiments, we demonstrated that this immunolabelling exhibited high specificity (Figure 11b–d).

Figure 11.

Mechanism for the immunolabeling of HeLa cells using rabbit anti-CEA8 antibody conjugated NaYF₄:Yb,Er NPs (a), and fluorescent images of HeLa cells after incubation with rabbit anti-CEA8 antibody conjugated NaYF₄:Yb,Er NPs (b–d). The left row (b) is an image in bright field, the middle row (c) is an image in dark field, and right row (d) are the overlays of the left and middle rows¹¹⁴

Jiang et al.¹¹⁶ also used folic acid and anti-Her2 antibody conjugated NaYF₄:Yb,Er UCNPs for labeling and fluorescent imaging of HT-29 cells and SK-BR-3 cells, respectively. More interestingly, NaYF₄:Yb,Er UCNPs were used in this work for the delivery and tracking of siRNA (small interference RNA). The siRNA was first attached to anti-Her2 antibody conjugated UCNPs and then attached to SK-BR-3 cells via an immunoreaction between the anti-HER2 antibody and HER2 receptors of SK-BR-3 cells. This led to the delivery of siRNA to the SK-BR-3 cells. The intracellular delivery was tracked using a confocal laser scanning microscopy equipped with an NIR laser and the effect of siRNA on gene silencing was studied using a luciferase assay.

Shan et al.¹¹⁷ reported the use of carboxyl functionalized NaYF₄:Yb,Er UCNPs for the imaging of AB12 mouse mesothelioma cells. From the TEM and fluorescent images, they concluded that these NaYF₄:Yb,Er UCNPs could enter into the cells indicating a high biocompatibility of the carboxyl functionalized NaYF₄:Yb,Er UCNPs. In addition, other kinds of UCNPs such as LaF₃:Yb,Ho NPs were also used as biological labels for cell imaging.¹¹⁸

Cytotoxicity tests can be used as an effective and sensitive way to assess the harmful components contained in UCNPs. The cytotoxicity of UCNPs to cells is often measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays ^111–113 or 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazoli um, innersalt (MTS) assays ^116–119. Typically, the UCNPs in various concentrations are incubated with cells for a period of incubation times, and then cell viability is scored under both control and exposed conditions to characterize the cytotoxic effect. Table 3 lists the data of cytotoxicity tests on UCNPs to different cells, which can prove that RE doped UCNPs exhibit low cytotoxicity and are even non-cytotoxic to a broad range of cells.

Table 3.

Cytotoxicity data on UCNPs to different cells

cells	nanoparticles (NPs)	NPs concentration/incubation time	cell viability	Ref.
BMS cells 1	NaYF4:Yb,Er@PEI	1 μg/ml for 24–48 h	~100%	111
		25 μg/ml for 24–48 h	>90%	111
BMS cells	NaYF4:Yb,Er@SiO2	25 μg/ml for 24 h	91.1%	112
skeletal myoblasts	NaYF4:Yb,Er@SiO2	50 μg/ml for 24 h	93.3%	112
Panc 1 cells 2	NaYF4:Yb,Tm	2 mg/ml for 10 min	>90%	113
SK-BR-3 cells	NaYF4@Ab-siRNA3	50 μg/ml for 24 h	98.6%	116
		80 μg/ml for 24 h	92.5%	116
human osteosarcoma cells	NaYF4:Yb,Er@SiO2	1 mg/ml for 9 days	96.2%	117
	NaYF4:Yb,Er@PAA	1 mg/ml for 9 days	92.8%	117
KB cells 4	LaF3:Yb,Ho@mPEG5	125 μg/ml for 12 h	>85%	118
		250 μg/ml for 12 h	>80%	118
KB cells	NaYF4:Yb,Tm@PAA	6–480 μg/ml for 24 h	>94%	119
		480 μg/ml for 48 h	>80%	119

¹bone marrow-derived stem cells

²human pancreatic cancer cells

³anti-Her2 antibody conjugated NaYF₄ UCNPs with siRNA attached

⁴human nasopharyngeal epidermal carcinoma cell line

⁵mPEG = polyethylene glycol monomethyl ether

5.1.2. In vivo tissue imaging

Chatterjee et al.¹¹¹ first reported the in vivo imaging of deep tissues in the Wistar rat using UCNPs. In their studies, PEI (5 wt%) was used for the coating of NaYF₄:Yb,Er UCNPs, and then 100 μl of resultant UCNPs (4.4 mg/ml) were injected subcutaneously into the groin and upper leg regions of the rat with a depth up to 10 mm. The rat was then excited using a 980 nm laser. As a result, the UCNPs injected below the abdominal skin, thigh muscles, and skin showed visible fluorescence under a 980 nm excitation. However, when the QDs (used as a control) were injected into the thicker skin or abdomen, they did not show any fluorescence under a UV excitation. Only QDs injected into the translucent skin of the foot could emit fluorescence (Figure 12a–f). Therefore, the NIR radiation proved to have better penetration ability than the UV light. UCNPs excited with NIR radiation therefore have great potential for in vivo imaging. Later, Nyk et al.¹¹³ reported high contrast whole-body in vivo imaging of the Balb-c mouse injected with NaYF₄:Yb,Tm UCNPs (Figure 12g–h). They demonstrated that the UCNPs showed no significant toxicity in the mice after 48 h post injection. Their study provided a foundation for the development of 3-D imaging systems for advanced whole body optical imaging. However, from available literature, the use of UCNPs for in vivo imaging is still in its preliminary stages.^{120, 121}

Figure 12.

In vivo imaging of rats: QDs injected into the translucent skin on the foot (a) shows fluorescence, but not through the thicker skin of the back (b) or abdomen (c); PEI/NaYF₄:Yb,Er UCNPs injected below the abdominal skin (d), thigh muscles (e), or below the skin of the back (f) show luminescence.¹¹¹ Whole body images of a rat iv injected with NaYF₄:Yb,Tm UCNPs; intact rat (g), same rat after dissection (h)¹¹³

5.2. Biological detection and analysis

5.2.1. Detection based on fluorescence resonance energy transfer

Fluorescence resonance energy transfer (FRET) is a nonradiative process where the electron excitation energy of a donor chromophore is transferred to a nearby acceptor molecule via long-range dipole-dipole interactions. In recent years, FRET-based analytical methods have gained considerable attention as powerful tools for biological detections because of their easy handling and high sensitivity.

To the best of our knowledge, the first demonstration of a FRET-based biological detection using NaYF₄:Yb,Er UCNPs was reported by Wang et al..¹⁰⁹ In the FRET system, the amino-functionalized NaYF₄: Yb,Er UCNPs were conjugated with biotin to serve as the donor and the biotin modified Au NPs acted as the acceptor. When avidin was added into the system, a reaction between the biotin and avidin occurred where the distance between the donor and the acceptor was shortened and energy emitted from the donor was transferred to the acceptor effectively. The more avidin that was added, the more energy could be transferred to the Au NPs. Thus, a linear relationship between the relative fluorescence intensity and the concentration of avidin in the FRET system was established (Figure 13a). Using this method, trace amounts of avidin in the concentration range from 0.5 to 370 nM could be detected.

Figure 13.

FRET-based detection of (a) avidin¹⁰⁹ and (b) DNA¹⁰⁸ using the NaYF₄:Yb,Er UCNPs

Chen et al. utilized NaYF₄:Yb,Er UCNPs to demonstrate their possible applications in the FRET system.¹⁰⁸ In their studies, carboxyl functionalized NaYF₄:Yb,Er UCNPs were covalently linked with streptavidin and then conjugated with biotin labeled DNA (capture-DNA). The resulting UCNPs were modified with capture-DNA which served as the donor, and TAMRA labeled reporter-DNA which could emit red light at the excitation of green light which acted as the acceptor (Figure 13b). When target-DNA was added into the system, a sandwich-type hybridization occurred where two shorter DNA (capture-DNA and reporter-DNA) were captured by a longer DNA (target-DNA). The energy could then be effectively transferred from the UCNPs to TAMRA, enabling TAMRA to emit red light due to the shortened distance between the UCNPs and TAMRA. With an increasing concentration of target-DNA added into the system, the green emission at 540 nm measured from the UCNPs gradually decreased while the yellow emission at 580 nm from TAMRA was enhanced. In addition, the fluorescent intensity ratio (I₅₈₀/I₅₄₀) varied linearly with the concentration of target-DNA in the range of 10–50 nM.

Our group, Wang et al.¹²² have also recently developed a luminescence resonance energy transfer (LRET)-based immunoassay. In this approach, amino-functionalized NaYF₄:Yb,Er UCNPs acted as a source for energy donor efficiency while gold NPs acted as an energy acceptor by absorbing the energy released by the donor molecules (Figure 14). This new sandwich based immunoassay system can be efficiently used for the detection of trace amounts of goat anti-human antibodies in the mg/ml range.

Figure 14.

Mechanism for the LRET process between NaYF₄:Yb,Er UCNPs and gold NPs¹²¹

The above examples are significant contributions as they provide good models for FRET-based biological detections using RE doped UCNPs and can be used in the detection of other kinds of biomolecules such as antibodies.

5.2.2. Detection based on magnetic separation

Wang et al.¹²³ applied NaYF₄:Yb,Er UCNPs for the sensitive detection of trace amounts of DNA with the aid of magnetic bioseparation and concentration technology. First, the Fe₃O₄ magnetic NPs were covalently linked to capture-DNA and NaYF₄:Yb,Er UCNPs were conjugated with probe-DNA. As in a typical sandwich-type assay, the short capture-DNA on the surface of magnetic NPs was first hybridized with long target-DNA. Then the overhanging region of target-DNA was matched to UCNPs labeled with probe-DNA. As a result, the UCNPs were captured by the magnetic NPs through hybridization of the oligonucleotides and then purified by magnetic separation (Figure 15). The UC fluorescence intensity of the captured UCNPs was found to have a linear relationship with the concentration of target DNA in the range of 7.8–78.0 nM. This method detects trace amounts of DNA and is a good alternative to PCR. This magnetic separation and concentration technology may serve as a useful model for the development of bioanalytical devices for biodetections in the near future.

Figure 15.

Magnetic separation-based detection of DNA using NaYF₄:Yb,Er UCNPs¹²²

6. Summary and Future Outlook

This review is focused on the synthesis of RE doped NaYF₄ using various methodologies and summarizes the general synthetic strategies, surface modifications and biological applications of emerging next generation luminescent UCNPs. Even though, the UC luminescence phenomenon of RE doped nanomaterials was observed several years ago, only recently are its practical applications greatly realized. Owing to their unique luminescent properties, UCNPs are currently evolved as an alternative source to traditional fluorescent probes by overcoming their current limitations. These UCNPs offer a wide range of biomedical applications including in vitro labeling, imaging, and analysis including the detection based on FRET and magnetic separation. In the future these UCNPs applications are expected to find more applications in in vivo biomedical studies. Despite the significant advancement in developing different strategies to synthesize and modify UC nanomaterials, till date it lacks a generalized protocol for synthesizing ultra-small NPs with strong UC luminescence intensity for biomedical applications. Current studies confirm that UCNPs can provide an excellent platform for in vitro based optical imaging; however, more detailed studies are necessary in the future to prove its wide applicability in animal models. There is also a greater scope to develop biocompatible multifunctional UC nanomaterials with high photoluminescence ability, strong superparamagnetism and easy bioconjugation for magnetic separation purposes in the coming years. Likewise, more focus is needed to develop new strategies for synthesizing UCNPs that can be extended to a variety of substrates for generating multiplexed detection devices.