Preparation and luminescence properties of monodisperse silica/aminosilane-coated Y 2 O 3 :Yb,Ho upconversion nanoparticles

Monodisperse silica/aminosilane-coated Y 2 O 3 :Yb,Ho nanoparticles are prepared via homogenous precipitation combined with a polyvinylpyrrolidone-assisted ammoniation method. The factors that contribute to the success of the coating are examined, and the procedure is optimized. Compared with uncoated nanoparticles, coated nanoparticles exhibit an increased ratio of green to red emission intensity, which can mainly be attributed to the decreased number of surface defects induced by the surface coating. silica coating, yttrium oxides, nanoparticles, upconversion luminescence

Upconversion luminescence is a physical process where low frequency light is converted to high frequencies via multiphoton absorption [1]. Due to potential applications in solid state lasers, three dimensional displays and anti-forgery, considerable effort has been devoted to the study of upconversion luminescence since the concept was first formulated by Auzel in the mid-1960s [1][2][3][4][5]. Recently, fluorescent bioprobes based on upconversion nanocrystals have attracted much attention because they not only overcome the inherent disadvantages of conventional fluorescence markers such as photobleaching, optical damage and auto-fluorescence, but also allow in vivo imaging of deep tissues [6][7][8][9][10][11].
Yttria-based upconversion nanocrystals (YUNs) are well known as an excellent candidates for upconversion fluorescent bioprobes because of their favorable chemical and thermal stability, nontoxicity and high emissive intensity [12][13][14][15][16][17][18][19][20][21]. However, unmodified YUNs tend to agglomerate in water and lack surface groups capable of coupling biomolecules, which limits their application in bioassays. Various methods have been developed to solve these problems *Corresponding author (email: whcao@online.ln.cn) [17][18][19][20][21], among which surface coating the YUNs with silica and/or siloxane layers is the simplest and effective method. To date, the silica or silica/siloxane layers coated on the aggregated YUNs are relatively thick (more than 20 nm), and the particles coated are larger than 100 nm. However, it has been reported that ideal fluorescence biolabeling materials should be well-dispersed nanoparticles coated with thin shells [22].
In this paper, silica/aminosilane-coated Y 2 O 3 :Yb,Ho (Y-Y-H) nanoparticles with cores that are about 40 nm in diameter and shells of about 5 nm in thickness were prepared via a modified homogenous precipitation method combined with a polyvinylpyrrolidone (PVP)-assisted ammoniation process. The factors playing key roles in achieving a thin coating, and the effects of the coating on the upconversion luminescence of the Y-Y-H nanoparticles are investigated.

Particle Characterization
Transmission electron microscopy (TEM) was carried out on a Tecnai G2 20 microscope. Fourier transform infrared (FTIR) spectra over the range 400-4000 cm -1 were obtained on a Magna-IR 550 spectrometer using KBr pellets (the weight ratio of sample to KBr was 1%). Upconversion spectra were recorded on a Hitachi F-4500 fluorescence spectrometer equipped with a 980 nm continuous wave laser diode (LD).

Synthesis of Y-Y-H nanoparticles
A procedure from [23] was modified to prepare Y 2 O 3 nanoparticles doped with 6% Yb and 1.5% Ho. An aqueous solution containing rare-earth nitrates (0.01 mol/L) was made by mixing stoichiometric amounts of Y(NO 3 ) 3 , Yb(NO 3 ) 3 and Ho(NO 3 ) 3 stock solutions, followed by dilution with water. Similarly, an identical volume of an aqueous solution of urea (6 mol/L) was also made. The two solutions were mixed and then aged at 83°C for 9 min. The resulting white precipitate was collected by filtration, washed three times with water and ethanol, and then dried at 40°C for 12 h. Finally, the Y-Y-H nanoparticles were obtained by firing the white precipitate at 800°C for 1 h.

Synthesis of individually silica/aminosilane-coated Y-Y-H nanoparticles
Silica/aminosilane-coated Y-Y-H nanoparticles were prepared by PVP-assisted ammoniation. In a typical procedure, Y-Y-H nanoparticles (0.06 g) were dispersed in a solution of ethanol (50 mL) containing PVP (0.3 g), water (5 mL) and ammonia (4 mL) using an ultrasonic bath. The suspension was then stirred magnetically at room temperature and a solution of TEOS (0.04 mL) in ethanol (20 mL) was added. After 3 h, a solution of APES (0.04 mL) in ethanol (20 mL) was added and the reaction was stirred for an additional 1 h. The resulting product (Y-Y-H/S-N2) was precipitated using a centrifuge, washed three times with ethanol and then redispersed in ethanol. For comparison, and Y-Y-H/S-N1 samples were also prepared. The details of the conditions used are presented in Table 1. Figure 1. TEM observation showed that the Y-Y-H nanoparticles are well dispersed and have diameters of about 40 nm. In contrast, the Y-Y-H/S1 nanoparticles form a micrometer-sized aggregate of crystallites with diameters of 40 nm. In fact, we concluded that the Y-Y-H/S1 nanoparticles aggregate without the need for TEM observation because they settle out of the reaction solution after only 10 min. According to [24], Y-Y-H nanoparticles are very unstable under the coating conditions, resulting in a tendency to aggregate. Thus the most important requirement for preparing monodisperse silica-coated Y-Y-H nanoparticles is to prevent the aggregation of Y-Y-H during silica coating. Adsorbing PVP onto Y-Y-H prior to silica coating appears to be a good method for preventing aggregation because (1) PVP can adsorb onto a broad range of different materials, (2) PVP has a larger hydrodynamic radius than Y-Y-H, which can offer sufficient steric hindrance to prevent the flocculation of Y-Y-H, and (3) the adsorption of PVP onto Y-Y-H aids the deposition of silica onto Y-Y-H [22,[25][26][27][28][29]. As expected, monodisperse Y-Y-H/S2 was obtained by prior addition of PVP. The silica shells are difficult to observe because Y-Y-H/S2 is an unwashed sample that includes large amounts of core-free silica particles. In the studies of other silica-coated materials (e.g. magnetic materials), researchers proved that the secondary nucleation of silica can be avoided by carefully controlling the process parameters (including reaction temperature, reaction time, ratio of TEOS to core particles, and the content of water and/or ammonia). However, the optimum parameters are not universal because the nanoparticles vary in size and morphology as well as in their physical and chemical properties. In other words, it is necessary to optimize the procedure for each new material, which inevitably increases the testing cost and prolongs the experimental period. There exists a need to develop a simple but effective method that will allow the secondary nucleation of silica particles to be avoided when applying silica coatings  to nanoparticles. Considering the large difference in size between the Y-Y-H/S2 and silica particles, centrifugation was used to remove the silica particles. However, centrifugation results in the aggregation of Y-Y-H/S2 (Figure 1(d)) and redispersion did not occur even after a long period in an ultrasonic bath. This indicates that the silica-coated particles have been linked by the condensation of hydroxyl groups. Addition of ammonia is a feasible solution to shield the hydroxyl groups from condensation because it bypasses the need for removing silica-coated particles and grafts -NH 2 groups onto the silica-coated particles. Figure 1 Because thick shells will prevent upconversion nanocrystals from behaving as bioprobes [22], only the effect of thin shells on the luminescence properties of Y-Y-H has been investigated. As shown in Figure 3   the 5 F 4 / 5 S 2 and 5 F 5 states, respectively, and thus the change of I g /I r suggests that the coating affects the population of the 5 F 4 / 5 S 2 and 5 F 5 levels. It is well known that at low concentrations of Ho 3+ ions, a two-step energy transfer process from Yb 3+ to Ho 3+ ions is dominant for the Yb 3+ /Ho 3+ codoped systems [17]. From the energy level diagram shown in Figure 4, the population of the 5 F 5 level is mainly dependent on the nonradiative relaxation processes of 5 I 6 → 5 I 7 and 5 F 4 / 5 S 2 → 5 F 5 . The smaller the nonradiative relaxation rate, the smaller the population of 5 F 5 and the larger I g /I r will be. Therefore, the increasing I g /I r indicates that the silica/aminosilane coating decreases the nonradiative relaxation rates of 5 I 6 → 5 I 7 and 5 F 4 / 5 S 2 → 5 F 5 . This can be explained by a decrease in the number of surface defects (i.e. nonradiative relaxation centers) after coating [30,31]. Recently, Lü et al. [20,21] reported that in addition to the decrease of surface defects, a coating could also change the crystal field around the dopant ions at the surface of the nanoparticles. However, it is thought that the silica/aminosilane coating used herein has little influence on the crystal field around the dopant ions at the surface because the position and shape of the emission bands do not change upon addition of the coating (see Figure 3) [32,33]. The reason for this may be that the introduction of PVP weakens the effect of the silica/aminosilane coating on the crystal field around the surface dopant ions.

Conclusion
A simple method to prepare monodisperse silica/aminosilanecoated Y-Y-H nanoparticles was reported, and the factors playing key roles in the success of the coating were discussed. The coated Y-Y-H nanoparticles show enhanced I g /I r compared with the uncoated counterpart, which was attributed to a decrease in the number of surface defects upon coating.