Magnetic and upconverted luminescent properties of multifunctional lanthanide doped cubic KGdF₄ nanocrystals

Abstract

Lanthanide (Ln³⁺) doped KGdF₄ (Ln = Yb³⁺, Er³⁺, Ho³⁺, Tm³⁺) nanocrystals with a mean diameter of approximately 12 nm were synthesized by a hydrothermal method using oleic acid as a stabilizing agent at 180 °C. The nanocrystals crystallize in the cubic phase as α-NaGdF₄. When excited by a 980 nm laser, these Ln³⁺ doped nanocrystals exhibit multicolor up-conversion (UC) emissions in red, yellow, blue and white. The calculated color coordinates demonstrate that white UC emission (CIE-X = 0.352, CIE-Y = 0.347) can be obtained by varying the dopant concentrations in the Yb³⁺/Ho³⁺/Tm³⁺ triply-doped nanocrystals to yield different RGB emission intensities. The measured field dependence of magnetization (M-H curves) of the KGdF₄ nanocrystals shows their paramagnetic characteristics that can be ascribed to the non-interacting localized nature of the magnetic moment of Gd³⁺ ions. Moreover, low temperature thermal treatment can enhance UC properties, magnetization and magnetic mass susceptibility of Ln³⁺ doped KGdF₄ nanocrystals. The multifunctional Ln³⁺ doped KGdF₄ nanocrystals have potential applications in color displays, bioseparation, and optical–magnetic dual modal nanoprobes in biomedical imaging.

---

1. Introduction

Interests in lanthanide (Ln³⁺) doped nanocrystals with up-conversion (UC) emission under the excitation of an infrared laser are increasing due to its potential applications in solid-state lasers,¹ multi-color three-dimensional displays,² optical processing sensors³ and solar cells.⁴ In particular, UC nanocrystals are envisioned to be the most appropriate candidate for biological fluorescent labels and hybrid magnetic–optical bimodal nanoprobes in biomedicine imaging due to advantages such as the deep penetration depth of infrared excitation, excellent photo-stability and absence of autofluorescence in contrast to traditional fluorescent dyes and semiconductor quantum dots.^5–7 The UC processes involving Ln³⁺ have heretofore been observed from different matrices such as oxides,⁸ phosphates,⁹ vanadates,¹⁰ sulfides¹¹ and fluorides.^12–20 Among them, fluorides have been extensively investigated due to the strong visible UC emission. By controlling the nanocrystal size, dopant–host combination, and dopant concentration, multicolor UC emissions spanning infrared to deep ultraviolet including white light have been observed from fluoride nanocrystals.^15–22 However, studies on the corresponding dispersible nanocrystals have mostly been restricted to AReF₄ (A = sodium, Re = Ln³⁺) fluoride compounds and there have been few investigations on other important fluorides such as CaF₂, nMF–mGdF₃ (M = Li, K, and Cs), and xAF₂–yGdF₃ (A being a group-II element). Very recently, Wang et al. reported that monodispersed CaF₂ : Yb³⁺/Er³⁺ nanocrystals showed more efficient UC emission than NaYF₄ when the size of the nanocrystals was less than 10 nm.²³ Hence, there may be other fluoride nanocrystals with even better properties.

The Ln³⁺ doped nKF–mGdF₃ system is interesting due to both its structural and optical properties. Firstly, it constitutes a good host matrix for luminescent Ln³⁺. Visible quantum-cuttings based on down-conversion under the excitation of vacuum ultraviolet radiation in KGdF₄ : Eu³⁺, K₂GdF₅ : Tb³⁺ and KGd₃F₁₀ : Eu³⁺ have been reported.^24–26 Secondly, Gd³⁺ is an ideal paramagnetic relaxation agent used in magnetic resonance imaging because of its large magnetic moment and nanosecond time scale electronic relaxation time.^6,7 This makes the nKF–mGdF₃ system a very attractive choice to dope with different Ln³⁺ ions to produce single-phase multi-functional nanocrystals (i.e., paramagnetism and multicolor up- or down-emission). Despite previous research on the Ln³⁺ doped nKF–mGdF₃ system, there have been few studies on the corresponding nanocrystals.^27,28 In this paper, we report the realization of Ln³⁺ co-doped cubic KGdF₄ nanocrystals with intrinsic paramagnetism and multicolor UC emissions.

2. Experimental details

Colloidal KGdF₄ nanocrystals with different Ln³⁺ concentrations were synthesized by a hydrothermal method at 180 °C for 20 h using oleic acid as a stabilizing agent under basic conditions. The detailed experimental processes can be found elsewhere.¹⁸ Succinctly speaking, 0.982 g (17.5 mmol) of KOH, 7.1 g (22.6 mmol) of oleic acid (90 wt%), and 12.67 ml of ethanol were mixed to get a semitransparent solution at room temperature. Then, 6 ml (6 mmol) of a 1M KF solution was added under vigorous stirring until a translucent solution was obtained. Subsequently, 2.24 ml (1.12 mmol) of 0.5M of Gd(NO₃)₃ with a pre-determined Ln³⁺ doping content was poured into the above solution under vigorous stirring. Before transferring to a Teflon-lined autoclave with an internal volume of 25 mL, the solution was aged for 20 min at room temperature. The hydrothermal reactions were conducted in an oven at 180 °C for 20 h. After the reactions, the white products were harvested by centrifugation, thorough washing with ethanol and deionized water sequentially three or four times and then drying at 60 °C for 48 h.

The crystal structures of the synthesized samples were determined by X-ray diffraction using a copper K_α radiation source. The morphology and microstructure were characterized by transmission electron microscopy (JEOL 2100) equipped with selected area electron diffraction (SAED). Differential scanning calorimetric (DSC) and thermogravimetric analyses were conducted on the thermal analysis systems DSC-7 and TGA Q50. The UC spectra were recorded on a spectrophotometer (R-500) under excitation by a 980 nm laser diode (LD) after the powders were compressed into smooth pellets. To eliminate the influence from adsorbates such as carboxyl, H₂O, and OH⁻, the samples were heated to 400 °C in air for 2 h before optical measurements. The fluorescence spot excited by the parallel laser beam on the sample had a diameter of about 0.4 cm and the measurements were performed at room temperature. Magnetization as a function of applied magnetic field was measured using a Lakeshore vibrating sample magnetometer at room temperature (RM) and 79 K.

3. Results and discussion

Fig. 1 shows the XRD patterns of the as prepared Ln³⁺ co-doped nanocrystals after annealing at different temperatures. Compared to the standard PDF data, the XRD results are not in accordance with that of orthorhombic KYF₄ (JCPDS NO.33-1007) or the diffraction data of cubic KYb₃F₁₀ (JCPDS NO.74-2204). Instead, they are similar to those of cubic NaGdF₄ (a = 5.52 Å, space group Fm3m, JCPDS NO.27-0697). Indexing the diffraction profile indicates a cubic unit cell with a = 5.73 Å. Based on this value, we simulated the diffraction peaks based on the crystallographic data of cubic NaGdF4 (α-NaGdF₄, space group Fm-3m, Na⁺ replaced by K⁺, and a = 5.52 Å replaced by a = 5.73 Å) by employing the MDI JADE 5.0 software. The positions of calculated diffraction peaks (red triangle in Fig. 1) are consistent with the experimental results. It can thus be inferred that our nanocrystalline samples crystallize in the fluoride structure known for α-NaGdF₄ and the enlarged cell constant results from the larger ionic radius of K⁺. In addition, according to the XRD results, one can infer that the size of the nanocrystals only increases slowly during thermal treatment at 300 and 400 °C since no obvious decrease in the full-width at half-maximum (FWHM) can be observed from the XRD peaks.

Fig. 1 XRD patterns of the Ln³⁺ co-doped KGdF₄ nanocrystals annealing at different temperature. The diffraction peaks labeled by red triangles are those from the calculation via MDI JADE 5.0 software based on the crystallographic data of cubic NaGdF₄ (space group Fm-3m, Na⁺ replaced by K⁺, a = 5.52 Å replaced by a = 5.73 Å). The impurity peaks at 27.8 °C, 31.9 °C, 46.2 °C and 54.7 °C in the sample annealing at 600 °C can be indexed to (111), (200), (202), (311) lattice planes of tetragonal GdOF (JCPDS NO. 26-0658), respectively.

Fig. 2(a) depicts the typical TEM image of the synthesized Ln³⁺ doped KGdF₄ nanocrystals revealing that the nanocrystals are nearly spherical in shape. The average size of the nanocrystals is determined to be 11.6 nm from the statistical histogram (see the inset) which also shows the narrow size distribution and that the size of the majority of the nanocrystals is 7–15 nm. Fig. 2(c) displays the typical high resolution TEM image of the nanocrystals. The lattice arrangement in the nanocrystals is clearly visible suggesting a highly crystalline nature. The distance between the fringes (d-spacing) is ∼2.73 Å, which corresponds to the distance of the (200) plane of the standard cubic KGdF₄ and is consistent with the SAED data [inset in Fig. 2(b)]. Further energy-dispersive X-ray analysis (EDS) confirms that the major constituents are K, Gd, F, and Yb.

Fig. 2 (a) TEM images and (b) the SEAD pattern acquired from as-synthesized Ln³⁺ co-doped KGdF₄ nanocrystals. The inset in (a) shows the statistical histogram of the nanocrystal size distribution. (c) HRTEM image of a single nanocrystal. (d) EDS spectrum of the Ln³⁺ co-doped KGdF₄ nanocrystals.

When excited by a 980 nm LD, the Yb³⁺/Er³⁺, Yb³⁺/Ho³⁺ and Yb³⁺/Tm³⁺ co-doped KGdF₄ nanocrystals exhibit eye-visible red, yellowish-green, and blue luminescence, respectively (see the inset fluorescence photographs in Fig. 3). Fig. 3(a), (b), and (c) depict the corresponding UC emission spectra from Yb³⁺/Er³⁺ (10 : 0.5 mol%), Yb³⁺/Ho³⁺ (10 : 0.5 mol%), Yb³⁺/Tm³⁺(20 : 0.5 mol%) co-doped KGdF₄ nanocrystals. In panel (a), the strong green emission bands centered at 546 nm and 523 nm and red UC band at 660nm can be attributed to the ²H_11/2/⁴S_3/2 → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions of Er³⁺, respectively, as shown in Fig. 4. In panel (b), the strong green and red UC emission bands centered at 650 nm and 540 nm can be assigned to the ⁵F₅ → ⁵I₈ and ⁴S₂/⁵F₄ → ⁵I₈ transitions of Ho³⁺, respectively. In panel (c), the blue emission bands at 450 and 475 nm correspond to the ¹D₂→³F₄ and the ¹G₄→³H₆ transitions. The red emission bands at 647 arises from the ¹G₄→³F₄ transition and the intense near-infrared emission at 803 nm stems from the ³H₄→³H₆ transitions of Tm³⁺ ions. To clarify the mechanism, the power dependent UC behavior is investigated. Fig. 3(d), (e), and (f) show the double-logarithmic plots of the UC emission intensity (I_UP) versus infrared pump-power (I_IR) for the Yb³⁺/Er³⁺, Yb³⁺/Ho³⁺ and Yb³⁺/Tm³⁺ co-doped KGdF₄ nanocrystals, respectively. The slopes of the linear fits of log(I_UP) versus log(I_IR) are 2.23, 2.22 and 1.79 for the green and red bands at 523, 546 and 660 nm from the Yb³⁺/Er³⁺ co-doped sample, 2.2 and 2.18 for the green and red bands at 540 and 650 nm from the Yb³⁺/Ho³⁺ co-doped sample, and 2.51 and 1.74 for the blue and infrared bands at 475 and 803 nm from the Yb³⁺/Tm³⁺ codoped sample, respectively. The results reveal that two pump photons from the excited Yb³⁺ ions via successive energy transfer (ET) are necessary to produce the red and green bands from these co-doped samples, whereas three pump photons are needed to emit the blue bands. They are consistent with previous results obtained from NaYF₄, Y₂O₃ and BaYF₅ nanocrystals.^29–32

Fig. 3 Up-conversion emission spectra and corresponding double-logarithmic plot of the up-conversion emission intensity versus the pump power density of the KGdF₄ nanocrystals doped with 10 mol% Yb³⁺, 0.5 mol% Er³⁺ (a, d); doped with 10 mol% Yb³⁺, 0.5 mol% Ho³⁺ (b, e); and doped with 20 mol% Yb³⁺, 0.5 mol% Tm³⁺ (c, f), respectively. The insets are the digital photographs acquired from the corresponding samples at the power density of 18.09 W cm⁻².

Fig. 4 The energy diagram of the Yb³⁺, Er³⁺, Ho³⁺ and Tm³⁺ dopant ions and the possible UC mechanisms under the excitation of a 980 nm LD. The full, dotted and dashed arrows represent emission, energy transfer and multi-phonon relaxation processes respectively.

The low temperature thermal treatment can notably enhance the UC intensity of the samples. Fig. 5(a) shows the UC spectra from the Yb³⁺–Tm³⁺ co-doped KGdF₄ nanocrystals annealed at different temperatures demonstrating enhanced UC emissions at both 475 nm and 803 nm of Tm³⁺. Generally, with miniaturization down to the nanometre scale, some surface and interface states have great influence on the optical properties of Ln³⁺ doped nanocrystals.^33,34 These surface states are closely related to some anionic groups such as OH⁻ and carboxyl groups which originate from the precursors in the synthesis and cannot be removed by simply drying at 60 °C. Here, the nanocrystals are produced by a hydrothermal method using oleic acid as the stabilizing agent. It is reasonable that certain amounts of OH⁻ and carboxyl groups adsorb on the surface of nanocrystals. The OH⁻ and carboxyl groups yield high-energy vibrational modes (2800–3600 cm⁻¹), which may quench the intermediate or excited states of Tm³⁺ by multi-phonon relaxation to lower UC efficiency. In the Tm³⁺ ions, the ⁴H₄ states are both the intermediate levels responsible for the emission at 475 nm and the excited state for emission at 803 nm. There is an energy gap of ∼4100 cm⁻¹ between the ⁴H₄ state and its nearest lower ³H₅ state. According to the Miyakawa–Dexter theory,³⁵ the high-energy vibrations of the OH⁻ and carboxyl groups render multi-phonon relaxation (⁴H₄ → ³H₅) much more probable than that of the intrinsic phonons in KGdF₄ nanocrystals, and only two vibration phonons of OH⁻ are required to bridge the gap. Therefore, the presence of OH⁻ and carboxyl groups on the nanocrystal surface reduces the UC intensity if they are coupled to the Tm³⁺ ions. Obviously, the presence of OH⁻ and carboxyl groups imposes a similar quenching effect on the UC emission from Er³⁺ and Ho³⁺. Fig. 5(b) shows the FTIR spectra from the Yb³⁺-Tm³⁺ co-doped KGdF₄ nanocrystals annealed at different temperatures. Four peaks and one broad band can be observed from as prepared KGdF₄ nanocrystals. The broad band at 3446 cm⁻¹ characterizing mode is appreciable for the O–H stretching vibration.³⁶ The peaks at 2916 and 2838 cm⁻¹ are assigned to the asymmetric and symmetric stretching vibration of methylene (CH₂) in the long alkyl chain of the oleic acid molecule, respectively. The peaks at 1566 and 1438 cm⁻¹ can also be assigned to the asymmetric and symmetric stretching vibration of the carboxylic group (–COOH), respectively. After the low temperature thermal treatment, four peaks at 2916, 2838, 1566 and 1438 cm⁻¹ completely disappear and the intensity of the broad band at 3446 cm⁻¹ characterizing mode is notably discreased. The inset in Fig. 5 shows the DSC-TGA results of the Yb³⁺–Tm³⁺ co-doped KGdF₄ nanocrystals. Two endothermic peaks at around 104 °C and 303 °C can be observed. The former weak endothermic peak is due to the desorption of absorbed water and organic solvent, whereas the latter broad one is attributed to the desorption of adsorbates such as carboxyl and hydroxyl groups.³⁷ The FTIR and DSC-TGA results confirm that careful elimination of such adsorbates decreases the rate of non-radiative multi-phonon relaxation consequently improving the UC efficiency.

Fig. 5 (a) Up-conversion emission spectra obtained from the 10 mol% Yb³⁺ and 0.5 mol% Tm³⁺ codoped KGdF₄ nanocrystals annealed at different temperature. The inset shows the DSC-TGA curves of the Yb³⁺/Tm³⁺ codoped KGdF₄ nanocrystals. (b) FTIR spectra from the 10 mol% Yb³⁺ and 0.5 mol% Tm³⁺ co-doped KGdF₄ nanocrystals annealed at different temperature.

Based on the generation of RGB bands from different co-doped nanocrystals, it is possible to produce white emission by introducing three dopants, namely Yb³⁺, Tm³⁺, and Er³⁺ or Ho³⁺ into the KGdF₄ nanocrystals. Fig. 6(a) shows the UC emission spectra and corresponding digital photographs (see inset) obtained from KGdF₄ nanocrystals doped with different molar ratios of Yb³⁺, Ho³⁺ and Tm³⁺ under excitation by a 980 nm LD with a pump power density of 16.2 W cm⁻². Referring to Fig. 3(b) and (c), it is easy to assign the origins of the observed emission bands. The intense blue emission at 475 nm and weak one at 450 nm can be assigned to the intra-4f electronic transitions of Tm³⁺ and the green emission around 540 nm arises from Ho³⁺. The red emission includes the contribution from both Tm³⁺ and Ho³⁺ according to the change of fine spectral shape in the red band. As the ratio of Ho³⁺ and Tm³⁺ increases, the intensity of the green and red emissions increase faster than that of the blue emission from Tm³⁺, resulting in a color shift from blue-white to red-white. Therefore, the contribution to the red emission from Ho³⁺ is more than that from Tm³⁺ in the triple-doped nanocrystals. Fig. 6(b) shows the CIE 1931 chromaticity diagram together with the calculated color coordinates of Fig. 6(a). The calculated color coordinates of (0.333, 0.261), (0.312, 0.283), (0.352, 0.347) and (0.358, 0.361) correspond to the different molar ratios of (25 : 0.75 : 1), (25 : 1 : 1), (25 : 1.25 : 1), (25 : 1.5 : 1), respectively. The chromaticity coordinate of the Yb³⁺/Ho³⁺/Tm³⁺ (25 : 1.25 : 1) doped KGdF₄ nanocrystals falls within the white region of the CIE 1931 chromaticity diagram and is very close to that of standard white light (X = 0.33, Y = 0.33), suggesting potential use as a white light source.

Fig. 6 (a) Up-conversion emission spectra and digital photographs (inset) of KGdF₄ nanocrystals doped with different molar ratios of Yb³⁺, Ho³⁺ and Tm³⁺ under excitation by a 980 nm LD. The photograph and spectra were obtained using a fixed pump power density of 16.2 W cm⁻². (b) Corresponding calculated CIE(X,Y) coordinate diagram showing the chromaticity points of the up-conversion luminescence: point 1–25 : 0.75 : 1; point 2–25 : 1 : 1; point 3–25 : 1.25 : 1; point 4–25 : 1.5 : 1.

Fig. 7 shows the magnetization as a function of applied magnetic field of the Yb³⁺–Er³⁺ co-doped KGdF₄ nanocrystals with and without annealing.The nanocrystals at both RM and 79 K (applied field ranging from −17 to 17 kOe) show paramagnetism, unlike the behavior of Gd atoms which exhibit a ferromagnetic behavior below 289 K. In general, the magnetic properties of Gd³⁺ arise from seven unpaired inner 4f electrons which are closely bound to the nucleus and effectively shielded by the outer closed shell electrons 5s² 5p⁶ from the crystal field.^38,39 If the separation between the Gd³⁺ ions in the matrix are too far, it may inhibit sufficient overlap of the orbitals associated with the partially filled 4f electrons shells of the Gd³⁺ ions necessary for ferromagnetism. Therefore, the paramagnetism which is different from that of the Gd atom even at 79 K in the KGdF₄ nanocrystals can be ascribed to the fact that the magnetic moments associated with Gd³⁺ are all localized and noninteracting.^38,39 The magnetic mass susceptibility of the as prepared and annealed KGdF₄ nanocrystals at RM are found to be 1.17 × 10⁻⁵ and 1.64 × 10⁻⁵ emu g⁻¹ Oe⁻¹, respectively. According to the aforementioned XRD, FTIR and DSC-TGA results, the ehancement of magnetic mass susceptibility likely originates from the desorption of adsorbates such as carboxyl and hydroxyl. At low temperature of 79 K, the magnetic susceptibility for the as prepared and annealed KGdF₄ nanocrystals are 3.89 x10⁻⁵ and 5.27 x10⁻⁵ emu g⁻¹ Oe⁻¹, respectively. The increase in the magnetic susceptibility with decreasing the temperature is due to the reduction in thermal fluctuation, which is a typical behavior in paramagnetic materials described by the Curie's law. Moreover, the RM magnetization of the annealed KGdF₄ nanocrystals at 15 kOe is around 0.29 emu g⁻¹ and it is close to the reported value of the GdF₃ : Eu³⁺ and Gd₂O₃ : Eu³⁺ nanocrystals used in common bio-separation.^38–40

Fig. 7 Magnetization vs. magnetic field of the Yb³⁺–Er³⁺ codoped KGdF₄ nanocrystals with and without annealing measured at room temperature (RM) and 79 K.

4. Conclusion

Multifunctional Ln³⁺ doped KGdF₄ (Ln = Er³⁺, Ho³⁺, Tm³⁺, Yb³⁺) nanocrystals with paramagnetism and multicolor UC emissions including red, yellow, blue and white under the excitation of a 980 nm diode laser (LD) have been synthesized by a hydrothermal method. TEM and XRD indicate that the average size of the nanocrystals in the cubic structure is approximately 12 nm. The power dependence studies indicate that a two-photon successive ET process from excited Yb³⁺ is responsible for the red and yellow emissions, whereas a three-photon successive ET process gives rise to the blue emission. The calculated color coordinates demonstrate that white UC emission (CIE-X = 0.352, CIE-Y = 0.347) can be obtained by varying the dopant concentrations in the Yb³⁺/Ho³⁺/Tm³⁺ triply-doped nanocrystals. The nanocrystals at both RM and 79 K exhibit paramagnetic properties which can be ascribed to the non-interacting localized nature of the magnetic moment of Gd³⁺. Furthermore low temperature thermal treatment can enhance UC properties, magnetization and magnetic mass susceptibility of Ln³⁺ doped KGdF₄ nanocrystals. The Ln³⁺ doped KGdF₄ nanocrystals have potential applications in color displays, bio-separation, and optical–magnetic dual modal nanoprobes in biomedicine imaging.^41–43

Acknowledgements

This work was supported by the Grants from National Natural Science Foundation of China (No.10802071 and No.10772157) and Technical Innovation Project (708068), Ministry of Education of China as well as City University of Hong Kong Strategic Research Grant (SRG) No. 7008009.

References

1. F. Auzel, Chem. Rev., 2004, 104, 139–174.
2. E. Downing, L. Hesselink, J. Ralston and R. Macfarlane, Science, 1996, 273, 1185.
3. C. Jacinto, M. Vermelho, E. Gouveia, M. de Araujo, P. Udo, N. Astrath and M. Baesso, Appl. Phys. Lett., 2007, 91, 071102–071102.
4. B. van der Ende, L. Aartsa and A. Meijerink, Phys. Chem. Chem. Phys., 2009, 11, 11081–11095.
5. F. Wang and X. Liu, Chem. Soc. Rev., 2009, 38, 976–989.
6. R. Kumar, M. Nyk, T. Ohulchanskyy, C. Flask and P. Prasad, Adv. Funct. Mater., 2009, 19, 853–859.
7. Y. Park, J. Kim, K. Lee, K. Jeon, H. Na, J. Yu, H. Kim, N. Lee, S. Choi, S. Baik, H. Kim, S. ParK, B. Park, Y. Kim, S. Lee, S. Yoon, I. Song, W. Moom, Y. Suh and T. Hyeon, Adv. Mater., 2009, 21, 4467–4471.
8. G. Glaspell, J. Anderson, J. Wilkins and M. El-Shall, J. Phys. Chem. C, 2008, 112, 11527–11531.
9. K. Koempe, H. Borchert, J. Storz, A. Lobo, S. Adam, T. Moeller and M. Haase, Angew. Chem., Int. Ed., 2003, 42, 5513–5516.
10. Y. Sun, H. Liu, X. Wang, X. Kong and H. Zhang, Chem. Mater., 2006, 18, 2726–2732.
11. T. Mirkovic, M. Hines, P. Nair and G. Scholes, Chem. Mater., 2005, 17, 3451–3456.
12. J. Boyer, L. Cuccia and J. Capobianco, Nano Lett., 2007, 7, 847–852.
13. S. Heer, K. Kömpe, H. Güdel and M. Haase, Adv. Mater., 2004, 16, 2102–2105.
14. J. Stouwdam and F. van Veggel, Nano Lett., 2002, 2, 733–737.
15. F. Wang and X. Liu, J. Am. Chem. Soc., 2008, 130, 5642–5643.
16. J. Zeng, J. Su, Z. Li, R. Yan and Y. Li, Adv. Mater., 2005, 17, 2119–2123.
17. F. Zhang, Y. Wan, T. Yu, Y. Shi, S. Xie, Y. Li, L. Xu, B. Tu and D. Zhao, Angew. Chem., 2007, 119, 8122–8125.
18. L. Yang, H. Han, Y. Zhang and J. Zhong, J. Phys. Chem. C, 2009, 113, 18995.
19. V. Mahalingam, F. Vetrone, R. Naccache, A. Speghini and J. Capobianco, Adv. Mater., 2009, 21, 4025.
20. F. Wang, Y. Han, C. Lim, Y. Lu, J. Wang, J. Xu, H. Chen, C. Zhang, M. Hong and X. Liu, Nature, 2010, 463, 1061–1065.
21. H. Mai, Y. Zhang, L. Sun and C. Yan, J. Phys. Chem. C, 2007, 111, 13721–13729.
22. Z. Yan and C. Yan, J. Mater. Chem., 2008, 18, 5046–5059.
23. G. Wang, Q. Peng and Y. Li, J. Am. Chem. Soc., 2009, 131, 14200–14201.
24. N. Kodama and Y. Watanabe, Appl. Phys. Lett., 2004, 84, 4141.
25. T. Lee, L. Luo, W. Diau, T. Chen, B. Cheng and C. Tung, Appl. Phys. Lett., 2006, 89, 131121.
26. R. Wegh, H. Donker, K. Oskam and A. Meijerink, Science, 1999, 283, 663.
27. M. Karbowiak, A. Mech, A. Bednarkiewicz and W. Strk, J. Alloys Compd., 2004, 380, 321–326.
28. Y. Du, Y. Zhang, L. Sun and C. Yan, Dalton Trans., 2009,(40), 8574–8581.
29. G. Chen, Y. Liu, Y. Zhang, G. Somesfalean, Z. Zhang, Q. Sun and F. Wang, Appl. Phys. Lett., 2007, 91, 133103.
30. F. Vetrone, V. Mahalingam and J. Capobianco, Chem. Mater., 2009, 21, 1847–1851.
31. J. Yang, C. Zhang, C. Peng, C. Li, L. Wang, R. Chai and J. Lin, Chem.–Eur. J., 2009, 15, 4649–4655.
32. A. Yin, Y. Zhang, L. Sun and C. Yan, Nanoscale, 2010, 2, 953.
33. G. De, W. Qin, J. Zhang, J. Zhang, Y. Wang, C. Cao and Y. Cui, Solid State Commun., 2006, 137, 483–487.
34. F. Vetrone, J. Boyer, J. Capobianco, A. Speghini and M. Bettinelli, J. Phys. Chem. B, 2002, 106, 5622–5628.
35. T. Miyakawa and D. L. Dexter, Phys. Rev. B: Solid State, 1970, 1, 2961.
36. L. Wang and Y. Li, Nano Lett., 2006, 6, 1645–1649.
37. S. Lu, B. Lee, Z. Wang and W. Samuels, J. Cryst. Growth, 2000, 219, 269–276.
38. H. Wong, H. Chan and J. Hao, Opt. Express, 2010, 18, 6123–6130.
39. H. Wong, H. Chan and J. Hao, Appl. Phys. Lett., 2009, 95, 022512.
40. H. Yang, S. Zhang, X. Chen, Z. Zhuang, J. Xu and X. Wang, Anal. Chem., 2004, 76, 1316.
41. L. Xiong, Z. Chen, Q. Tian, T. Cao, C. Xu and F. Li, Anal. Chem., 2009, 81, 8687–8694.
42. M. Yu, F. Li, Z. Chen, H. Hu, C. Zhan, H. Yang and C. Huang, Anal. Chem., 2009, 81, 930–935.
43. J. Zhou, Y. Sun, X. Du, L. Xiong, H. Hu and F. Li, Biomaterials, 2010, 31, 3287–3295.

---