Synthesis of Nd³⁺/Yb³⁺ sensitized upconversion core–shell nanocrystals with optimized hosts and doping concentrations

Abstract

Development of upconversion nanocrystals (UCNs) under 808 nm excitation rather than 980 nm is very important to biological applications for avoiding tissue over-heating. Nd³⁺ and Yb³⁺ dual-sensitized UCNs are proven to be promising candidates but how to select host materials to construct core–shell UCNs with strong UC emissions still remains unexplored. Herein, we prepare a series of homogeneous and heterogeneous core–shell UCNs using NaYF₄ and NaGdF₄ as the core–host and/or as the shell–host, respectively, through the seed-mediated synthetic approach. Our results show that selecting the core–host in the core–shell UCNs plays the key role in determining their final UC intensities. Furthermore, homogeneous core–shell UCNs can give stronger UC fluorescence than the heterogeneous ones due to the low crystal defects at the core–shell interface. Moreover, the concentration effect of activator ions and sensitizer ions in these core–shell UCNs on their UC emissions is analyzed, and optimal doping under different NIR excitation (808 nm, 980 nm or 808 nm/980 nm) is achieved.

---

Introduction

Lanthanide nanocrystals (NCs) with upconversion (UC) fluorescence have attracted extensive research interest in the past decade.^1–10 Owing to their unique upconverting properties and other features such as narrow emissions, long luminescence lifetimes, and high photostability, upconversion nanocrystals (UCNs) have shown potential applications in a variety of fields including imaging, sensing, drug delivery, and photodynamic therapy.^11–16 Typically, a UCN is constituted by the host nanomatrix which is simultaneously doped with sensitizer ions (usually Yb³⁺) and activator ions (Er³⁺, Tm³⁺, or Ho³⁺ etc.).^17–19 The sensitizer ions continuously absorb near-infrared (NIR) photons and then transfers energy to the activator ions for luminescence. Although great advances have been made in the past years, recent progress on UCNs has been largely hindered because most of the developed UCNs are sensitized by Yb³⁺ ions which only respond to a narrow-band NIR excitation centered at 980 nm. The absorption of Yb³⁺ ions overlaps the maximal absorption of water molecules that are dominant in biological samples. Over-exposure of bio-species under 980 nm diode laser may induce potential thermal damages to cells and tissues due to over-heating, limiting these UCNs for deep tissue imaging. To address this issue, exploration of new sensitized UCNs which can absorb other NIR bands in addition to 980 nm is an elegant solution. In searching of this type of sensitizer ions, Nd³⁺ has been proven to be a good choice because Nd³⁺ doped UCNs exhibits intense absorption around 808 nm at which water absorption is minimal. Furthermore, Nd³⁺ has a larger absorption cross-section in the NIR region than that of Yb³⁺.²⁰ Particularly, it has been revealed that the energy transfer of Nd³⁺ → Yb³⁺ has a high efficiency in many types of host materials,^21–26 suggesting that co-doping of Nd³⁺ and Yb³⁺ is a practical approach to develop new types of UCNs under the excitation of 808 nm.

Recently, several attempts have been made to the synthesis of Nd³⁺ sensitized UCNs which show impressive UC emission under 808 nm excitation. For examples, Han et al. pioneered the synthesis of Yb/Er/Nd triply doped NaYF₄ NCs which can be excited under 808 nm.²² However, the deleterious cross-relaxation between activator ions and Nd³⁺ ions requires low Nd³⁺ doping concentration (typically <2%), which leads to weak absorption at 808 nm and low UC emissions. To increase the Nd³⁺ doping, core–shell structured NCs (such as NaGdF₄@NaGdF₄ and NaYF₄@NaYF₄) were developed by Yan's group and Liu's group, respectively.^20,23 In such configuration, the activators ions and Nd³⁺ ions can be spatially separated in a single particle and thus serous cross-relaxation are suppressed. Very recently, Wang et al. also reported a four-layered core–shell NaGdF₄ NCs which can emit both down- and up-conversion emissions at single 808 nm excitation.²⁴ All above works have shown important insights into the design and synthesis of Nd³⁺ sensitized UCNs under the NIR band at 808 nm.

Despite above progress, there are still a few questions remains unexplored and deserve further investigation. First, either NaYF₄ or NaGdF₄ was monotonously used as host materials in these developed core–shell UCNs.^20–24 Considering that both of them are excellent host matrixes for UC luminescence upon lanthanide doping, which kind of core–shell configuration is the best if homogeneous core–shell hosts (NaLn(1)F₄@NaLn(1)F₄) and heterogeneous core–shell hosts (NaLn(1)F₄@NaLn(2)F₄) are compared (Ln(1) and Ln(2) mean different lanthanide elements)? Second, when referring to the synthesis of dual Nd³⁺/Yb³⁺ sensitized UCN aiming to be excited at both 808 nm and 980 nm, how to control the doping concentrations of sensitizer ions in the shell and activator ions in the core?

To clarify above issues, herein we employ a seed-mediated synthetic approach to prepare both homogenous core–shell NCs (e.g., NaYF₄@NaYF₄ and NaGdF₄@NaGdF₄) and heterogeneous core–shell NCs (e.g., NaYF₄@NaGdF₄ and NaGdF₄@NaYF₄). After doping different amounts of activator and/or sensitizer ions, we have systematically investigated the UC emissions from these core–shell UCNs under 808 nm and 980 nm excitation, respectively. Our results show that selecting nanohosts for the core play a key role in determining the final UC emissions of these core–shell UCNs. Furthermore, it is found that homogenous core–shell structure is better than the heterogonous one due to low crystal defects at the interface between the core and the shell. Moreover, doping effects of the activator ions in the core and sensitizer ions in the shell have been tested, and an optimal doping concentration is achieved in the core–shell UCNs. This work may shed some new lights on the design and synthesis of dual Nd³⁺/Yb³⁺ sensitized UCNs with a core–shell configuration.

Experimental section

Synthesis of the core–shell structured UCNs

High-quality NaYF₄:Yb (20%),Tm (0.5%) NCs were synthesized using a user-friendly method we have previously developed.²⁷ To prepare core–shell structured NaYF₄:Yb,Tm@NaYF₄:Yb,Nd (termed as Y@Y) NCs, the NaYF₄:Yb (20%),Tm (0.5%) NCs were used as seeds to epitaxial grow a NaYF₄:Yb (10%), Nd (10%) layer on their surface.²⁸ In a typical synthesis, 0.8 mmol YCl₃, 0.1 mmol YbCl₃ and 0.1 mmol NdCl₃ were mixed with 6 mL oleic acid and 15 mL 1-octadecene in a 50 mL flask. The solution was heated to 160 °C to form a homogeneous solution, and then cooled down to 70 °C. 1 mmol NaYF₄:Yb (20%),Tm (0.5%) NCs in 10 mL cyclohexane were then added into the solution and the cyclohexane solution was gradually removed upon evaporation. After that, 10 mL methanol solution containing 4 mmol NH₄F and 2.5 mmol NaOH was added and stirred for 30 min. Subsequently, the solution was slowly heated to remove methanol, degassed at 100 °C for 10 min, and then heated to 300 °C and maintained for 1 h under Ar protection. After the solution was cooled down to room temperature naturally, the products were precipitated from the solution with ethanol, and washed with ethanol/cyclohexane (1 : 1 v/v) for three times. Finally, the prepared core–shell NCs was dispersed in 10 mL cyclohexane for further use.

The NaGdF₄:Yb,Tm NCs were prepared with the similar protocol to the synthesis of NaYF₄:Yb,Tm NCs. Other core–shell NCs such as NaYF₄:Yb,Tm@NaGdF₄:Yb,Nd, NaGdF₄:Yb,Tm@NaGdF₄:Yb,Nd and NaGdF₄:Yb,Tm@ NaYF₄:Yb,Nd (termed as Y@Gd, Gd@Gd and Gd@Y), were also prepared according the seed-mediated approach as the synthesis of Y@Y.

Characterizations

The morphologies of samples were observed by transmission electron microscopy (TEM) which was taken on a JEOL 2010F TEM operating at 200 kV. The TEM samples were prepared by dropping a suspension of NCs on a carbon-film coated copper grid. X-ray powder diffraction (XRD) was carried out on a Japan Rigaku D/max rA X-ray diffractometer equipped with a Cu Kα radiation. X-ray photoelectron spectra (XPS) were collected on an ESCAlab MKII X-ray photoelectron spectrometer. Fluorescence spectra were acquired on a Hitachi F-7000 spectrometer equipped with commercial 980 nm and 808 nm NIR laser, respectively. Size distribution of the NCs was obtained using a Malvern Zetasizer Nano (ZEN3690).

Results and discussion

TEM characterizations of the core–shell NPs

TEM image of the prepared NaYF₄:Yb,Tm NCs are as shown in Fig. 1A. These NCs are uniform in size with a diameter of 32 nm. When these NCs were used as seeds, an layer of NaYF₄:Yb,Nd can be epitaxial grown on the seeded NCs. As a result, core–shell Y@Y with different dopants in the core and the shell are produced. From the TEM image (Fig. 1B), one can see that the size of the Y@Y NCs has increased to 40 nm and the structural uniformity of the core–shell NCs is reserved as the core NCs. Using the similar protocol, an NaGdF₄:Yb,Nd shell can also be produced on the seeded NaYF₄:Yb,Tm NCs (Fig. 1C), because NaGdF₄ and NaYF₄ crystals have the same crystal structure (hexagonal phase) and growing habits during the synthesis. This kind of seed-mediated approach has been well established in the synthesis of core–shell structured lanthanide fluoride NCs.²⁹ Similarly, we have also prepared NaGdF₄:Yb,Tm NCs and epitaxial grown a NaYF₄:Yb,Nd shell or a NaGdF₄:Yb,Nd shell on them, respectively (see Fig. 1D–F). The obtained core–shell Gd@Y and Gd@Gd NCs also exhibit uniform structures as the seeded NaGdF₄:Yb,Tm NCs except the incensement in particle size. Size distributions of these NCs before and after the shell growth have also been measured (see Fig. S1†). This result also show that both the core NCs and corresponding core–shell NCs have a narrow size distribution, confirming that a uniform shell has been produced on these core NCs. XRD patterns and XPS analyses of NaYF₄:Yb,Tm NCs before and after growing a NaGdF₄:Yb,Nd shell were also investigated. XRD data confirm that the core and the shell are much similar in phase and lattice (see Fig. S2†). XPS data reveal that the surface of Y@Gd NCs is dominated by element Gd (see Table S1†), implying a NaGdF₄ was covered on NaYF₄:Yb,Tm NCs.

Fig. 1 TEM images of the seeded NCs and corresponding core–shell NCs: (A) NaYF₄; (B) NaYF₄@NaYF₄; (C) NaYF₄@NaGdF₄; (D) NaGdF₄; (E) NaGdF₄@NaYF₄; (F) NaGdF₄@NaGdF₄ (scale bars: 100 nm).

UC emissions from different core–shell samples

For the Yb³⁺ sensitized UCNs (980 nm), Yb³⁺ ions can effectively transfer energy to activator ions (e.g., Tm³⁺) and avoid deleterious cross-relaxation.^30,31 Thus, Yb³⁺ and Tm³⁺ can be homogenously doped together in a single NC for UC emissions. For the Nd³⁺ sensitized UCNs (808 nm), however, serous deleterious cross-relaxation appears between Nd³⁺ and Tm³⁺ ions when they are homogenously doped in one single NC.²² To suppress the deleterious cross-relaxation, core–shell structure is required to spatially separate the Nd³⁺ ions from Tm³⁺ ions by doping them in the shell and in the core, respectively. For above consideration, core–shell structure is thus preferred for constructing dual Nd³⁺/Yb³⁺ sensitized UCNs in our case. On the other hand, since the energy transfer (ET) of Nd³⁺ → Yb³⁺ → Tm³⁺ is an efficient channel to give UC emissions under 808 nm excitation, co-doping of Yb³⁺ ions in both the shell and the core is favorable, for the ET and sensitization purpose, respectively. UC mechanism in the dual sensitized core–shell NCs can be illustrated in Fig. 2.

Fig. 2 Proposed energy transfer upconversion mechanism in the core–shell structured NCs.

To investigate the host effect on the core and the shell, individually, we have prepared core–shell Y@Y, Y@Gd, Gd@Gd and Gd@Y NCs which doped with the same amount of lanthanide ions in the core (20%Yb and 0.5%Tm) and in the shell (10%Yb and 10%Nd). UC spectra of these core–shell NCs were measured and shown in Fig. 3A and B, respectively. Under either 980 nm or 808 nm excitation, strong emissions at 360 nm, 450 nm and 475 nm are observed. The ultra-violet emission at 360 nm and blue emission at 450 nm are from four-photon processes, owing to the transitions of ¹D₂ → ³H₆ and ¹D₂ → ³H₄ in Tm³⁺ ions, respectively. The blue emission at 475 nm is from three-photon processes due to the transition of ¹G₄ → ³H₆ in Tm³⁺ ions.³² In both UC spectra, it is observed that Y@Y NCs display the highest intensity under the same concentration and measuring condition, and the Y@Gd NCs possess the second position. In contrast, Gd@Gd and Gd@Y NCs exhibit relatively low intensity and take the third and fourth position, respectively. Note that UC intensities from these core–shell hosts are in the same sequence under either 980 nm or 808 nm excitation, namely, Y@Y ＞ Y@Gd ＞ Gd@Gd ＞ Gd@Y.

Fig. 3 UC emission spectra of four core–shell UCNs constituted by different core–host and/or shell–host with the same doping concentrations under different NIR excitation: (A) 808 nm NIR laser; (B) 980 nm NIR laser.

Host effect in the core–shell UCNs

Judging from the UC intensities from above four core–shell samples, one can see that two samples prepared with NaYF₄:Yb,Tm cores shows stronger emissions than the other two samples prepared with NaGdF₄:Yb,Tm cores. This result indicates that selecting the core–host play the key role in determining their final UC intensity. In previous research, it has been revealed that both NaYF₄ and NaGdF₄ are efficient hosts for UC luminescence among the fluoride NCs when doped with Yb³⁺ and Tm³⁺ (or Er³⁺).^33–36 Since the UC emissions are resulted from electron transitions of the activator ions, a better host for the activator ions is thus crucial for these core–shell UCNs. In our case, the seeded NaYF₄:Yb(20%),Tm(0.5%) NCs show stronger UC fluorescence than the seeded NaGdF₄:Yb(20%),Tm(0.5%) (see Fig. S3†). This is probably due to the fact that Yb³⁺ and Tm³⁺ doped in NaYF₄ may produce less crystals defects and lattice stress than those doped in NaGdF₄, since the ionic radius of Y³⁺ is closer to those of Yb³⁺ and Tm³⁺ than that of Gd³⁺.^37,38 Therefore, Y@Y and Y@Gd display higher intensities than Gd@Gd and Gd@Y with the same doping concentrations.

Once the host material for the core is determined, the final UC emission intensity of sample will depend on the selection of the shell materials. Among our four core–shell samples, it is also observed that homogeneous core–shell hosts are better than the heterogeneous core–shell hosts (i.e., Y@Y > Y@Gd and Gd@Gd > Gd@Y). It means that an ideal host material for the shell does not rely on what kind of material it is but depends on what core material has already employed as the seeds in these core–shell UCNs. Selecting the same material for the shell as the core will favor better UC fluorescence under the same doping condition. It is known that crystal defects are fluorescence quenchers to lanthanide NCs. The difference between heterogeneous core–shell NC and homogenous core–shell NC is that there is an obvious interface between the core and the shell (see Fig. 4). Obviously, more crystal defects will appear at the interface of the heterogeneous core–shell NCs because they are made by different materials despite that they have similar crystal structure (hexagonal phase). Therefore, homogeneous core–shell hosts will achieve higher intensity than heterogeneous core–shell hosts. As a result, the sequence in UC emissions is Y@Y ＞ Y@Gd ＞ Gd@Gd ＞ Gd@Y. In short, the selection rule for the core–shell hosts is as follows: (1) selecting an ideal core–host for doping the activator ions; (2) selecting the same material for the shell–host as the core–host to avoid interfacial crystal defects.

Fig. 4 Schematic illustration of the core–shell structure of a single NaYF₄@NaYF₄ and NaYF₄@NaGdF₄ NC. More crystal defects may appear at the core/shell interface of NaYF₄@NaGdF₄ NC due to the different ionic radii of Y³⁺ and Gd³⁺.

Optimizing doping concentrations in the core–shell NCs

In order to reach the best UC performance, we have investigated the effect of doping concentrations of different ions on their final UC fluorescence, using the Y@Y NCs as an example. Firstly, we doped different amount of Yb³⁺ ions in the core and fixed the doping concentration of other ions. When the concentration of Yb³⁺ ions in the core increases from 10% to 15% and 20%, the total fluorescence of NCs has distinctly enhanced under the 808 nm excitation (Fig. 5A). This enhancement indicates that more Yb³⁺ ions can enhance the ET process from the shell to the core since Yb³⁺ does not directly absorb 808 nm light. However, obvious decrease in intensity appears once the doping concentration is over 25%, suggesting that obvious cross-relaxation between Yb³⁺ ions will arise at high Yb³⁺ doping. Therefore, optimal doping concentration of Yb³⁺ is around 20% in the core for the Y@Y NCs. In the similar way, we have tried different amount of Tm³⁺ ions in the core from 0.1% to 0.8% and fixed the Yb³⁺ doping at 20% (Fig. 5B). The result implies that 0.5% Tm³⁺ is the best doping level for UC luminescence and over-doping of Tm³⁺ will induce obvious cross-relaxation between the activator ions. Note that the best doping concentration of Yb³⁺ and Tm³⁺ in the core–shell NCs are similar to the optimal doping of Yb³⁺ and Tm³⁺ in the NaYF₄ seeds,³⁹ confirming that the key to strong UC fluorescence mainly relies on the performance of the core NCs. The requirement of high Yb³⁺ concentration and relatively low Tm³⁺ concentration is because the three- and four-photon UC process of Tm³⁺ ions which require sufficient energy from the surrounded sensitizer ions.

Fig. 5 Comparison of UC emission intensities of NaYF₄:Yb,Tm@NaYF₄:Yb,Nd UCNs under 808 nm excitation by changing the doping concentration of different ions: (A) Yb³⁺ in the core; (B) Tm³⁺ in the core; (C) Yb³⁺ in the shell; (D) Nd³⁺ in the shell.

Similarly, we also optimized the doping concentration of Yb³⁺ and Nd³⁺ ions in the shell by fixing the doping ions in the core. Under 808 nm excitation, the Yb³⁺ in the shell serve as a bridge to transfer energy from the Nd³⁺ ions to Tm³⁺ ions in the core. As such, too low Yb³⁺ concentration does not favor the ET process while too much Yb³⁺ ions will prolong the ET pathway and reduce the ET efficiency.²⁰ Therefore, there is an optimal doping concentration of Yb³⁺ in the shell host, too. The optimal level of Yb³⁺ ions is found to be 10% when the Nd³⁺ doping is fixed at 10% (a pretty high concentration for Nd³⁺) (see Fig. 5C). On the other hand, the Nd³⁺ ions serve as sensitizers to directly absorb 808 nm light and then transfer energy to Yb³⁺ ions. Although core–shell configuration can suppress deleterious cross-relaxation between Tm³⁺ and Nd³⁺ to some extent, over-doping of Nd³⁺ in the shell will also enhance such cross-relaxation due to more Nd³⁺ ions will appear at the interface.²⁴ In the Y@Y NCs, the optimal Nd³⁺ doping for strong UC fluorescence is around 10% (see Fig. 5D).

We also tested the UC fluorescence of above samples under 980 nm excitation (see Fig. S4†). It was found that the optimal doping for Yb³⁺ and Tm³⁺ in the core is similar to the samples excited under 808 nm. The difference lies in the doping concentration of Yb³⁺ ions in the shell. Under 980 nm excitation, the UC intensities can be greatly improved along with the increase of Yb³⁺ concentration and the optimal doping has increases to 20%. This is because that Yb³⁺ in the shell can help to absorb 980 nm light along with the Yb³⁺ doped in the core, serving as an ‘active shell’ in the core–shell NCs.⁴⁰ The concentration of Nd³⁺ in the shell (below 10%) has little effect on the UC emissions since they did not absorb the 980 nm light.

The UC fluorescence of samples doped with various amount of ions were also evaluated under simultaneous NIR excitation of 808 nm/980 nm (two NIR lasers were simultaneously used with the same power) (see Fig. 6). Interestingly, it is found that the optimal doping of Yb³⁺ and Tm³⁺ in the core and Nd³⁺ in the shell is similar to the samples excited under single 808 nm excitation. This is because these doping ions have similar functions under either 808 nm or 980 nm excitation. However, a relatively higher Yb³⁺ doping in the shell (25%) is required for achieving the best UC fluorescence. This is attributed to the fact that the Yb³⁺ ions in the shell have two functions when the samples were simultaneously excited under 808 nm and 980 nm. They not only serve as sensitizer ions for the absorption of 980 nm light but also work as energy migrators in the Nd³⁺ → Yb³⁺ → Tm³⁺ channel at 808 nm excitation. As such, the optimal Yb³⁺ doping should more than 20% for both the sensitization and ET purpose. Core–shell UCNs with excitation at two NIR bands may also have potential in the hybrid UC/semiconductor photocatalysts in addition to the biological applications, since they can show strong UC emissions under a wider NIR region.

Fig. 6 Comparison of UC emission intensities of NaYF₄:Yb,Tm@NaYF₄:Yb,Nd UCNs under simultaneous 808 nm/980 nm excitation by changing the doping concentration of different ions: (A) Yb³⁺ in the core; (B) Tm³⁺ in the core; (C) Yb³⁺ in the shell; (D) Nd³⁺ in the shell.

Conclusions

In summary, we employed the seed-mediated approach and prepared a series of homogenous and heterogeneous core–shell UCNs using NaYF₄ and NaGdF₄ as core–host and/or shell–host, respectively. After lanthanide doping, the UC emissions of these samples were mechanically investigated. It is found that selecting the core–host plays the key role in determining their final UC intensities. At the same time, homogenous core–shell UCNs can give stronger UC fluorescence than the heterogeneous one due to the low crystal defects at the core–shell interface. Through doping different activator and/or sensitizer ions in the core and in the shell, we also evaluated the concentration effect of various lanthanide ions on their UC fluorescence. Optimal doping concentration under different NIR excitation (808 nm, 980 nm and both 808 nm/980 nm) is analyzed. This work may shed some new lights on the design and synthesis of core–shell lanthanide NCs with dual sensitized features.

Acknowledgements

The authors acknowledge financial support from National Nature Science Foundation of China (No. 21273203 and 21201151) and Zhejiang Provincial Natural Science Foundation (No. LR15B0100 01 and LR12B040001).

References

1. T. Kano, H. Yamamoto and Y. Otomo, J. Electrochem. Soc., 1972, 119, 1561–1564.
2. N. Menyuk, K. Dwight and J. W. Pierce, Appl. Phys. Lett., 1972, 21, 159–161.
3. J. L. Sommerdijk, J. Lumin., 1973, 6, 61–67.
4. R. Kapoor, C. S. Friend, A. Biswas and P. N. Prasad, Opt. Lett., 2000, 25, 338–340.
5. S. Heer, O. Lehmann, M. Haase and H. U. Gudel, Angew. Chem., Int. Ed., 2003, 42, 3179–3182.
6. F. Auzel, Chem. Rev., 2004, 104, 139–173.
7. F. Vetrone, J. C. Boyer and J. A. Capobianco, J. Appl. Phys., 2004, 96, 661–667.
8. T. Hirai and T. Orikoshi, J. Colloid Interface Sci., 2004, 269, 103–108.
9. F. Wang and X. G. Liu, Chem. Soc. Rev., 2009, 38, 976–989.
10. N. Liu, W. P. Qin, G. S. Qin, T. Jiang and D. Zhao, Chem. Commun., 2011, 47, 7671–7673.
11. Q. Liu, Y. Sun, T. S. Yang, W. Feng, C. G. Li and F. Y. Li, J. Am. Chem. Soc., 2011, 133, 17122–17125.
12. Y. H. Wang, L. Bao, Z. H. Liu and D. W. Pang, Anal. Chem., 2011, 83, 8130–8137.
13. C. Wang, L. Cheng and Z. Liu, Biomaterials, 2011, 32, 1110–1120.
14. S. S. Cui, D. Y. Yin, Y. Q. Chen, Y. F. Di, H. Y. Chen, Y. X. Ma, S. Achilefu and Y. Q. Gu, ACS Nano, 2013, 7, 676–688.
15. Y. S. Liu, D. T. Tu, H. M. Zhu and X. Y. Chen, Chem. Soc. Rev., 2013, 42, 6924–6958.
16. S. L. Gai, C. X. Li, P. P. Yang and J. Lin, Chem. Rev., 2014, 114, 2343–2389.
17. J. B. Zhao, Z. D. Lu, Y. D. Yin, C. McRae, J. A. Piper, J. M. Dawes, D. Y. Jin and E. M. Goldys, Nanoscale, 2013, 5, 944–952.
18. H. Na, K. Woo, K. Lim and H. S. Jang, Nanoscale, 2013, 5, 4242–4251.
19. N. Niu, F. He, S. L. Gai, C. X. Li, X. Zhang, S. H. Huang and P. P. Yang, J. Mater. Chem., 2012, 22, 21613–21623.
20. Y. F. Wang, G. Y. Liu, L. D. Sun, J. W. Xiao, J. C. Zhou and C. H. Yan, ACS Nano, 2013, 7, 7200–7206.
21. G. Y. Chen, T. Y. Ohulchanskyy, S. Liu, W. C. Law, F. Wu, M. T. Swihart, H. Agren and P. N. Prasad, ACS Nano, 2012, 6, 2969–2977.
22. J. Shen, G. Y. Chen, A. M. Vu, W. Fan, O. S. Bilsel, C. C. Chang and G. Han, Adv. Opt. Mater., 2013, 1, 644–650.
23. X. J. Xie, N. Y. Gao, R. R. Deng, Q. Sun, Q. H. Xu and X. G. Liu, J. Am. Chem. Soc., 2013, 135, 12608–12611.
24. H. L. Wen, H. Zhu, X. Chen, T. F. Hung, B. L. Wang, G. Y. Zhu, S. F. Yu and F. Wang, Angew. Chem., Int. Ed., 2013, 125, 1–6.
25. X. M. Li, R. Wang, F. Zhang, L. Zhou, D. K. Shen, C. Yao and D. Y. Zhao, Sci. Rep., 2013, 3, 3536.
26. M. K. G. Jayakumar, N. M. Idris, K. Huang and Y. Zhang, Nanoscale, 2014, 6, 8441–8443.
27. Z. Q. Li, Y. Zhang and S. Jiang, Adv. Mater., 2008, 20, 4765–4769.
28. H. Guo, Z. Q. Li, H. S. Qian, Y. Hu and I. N. Muhammad, Nanotechnology, 2010, 21, 125602.
29. H. S. Qian and Y. Zhang, Langmuir, 2008, 24, 12123–12125.
30. H. Zhang, Y. J. Li, Y. C. Lin, Y. Huang and X. F. Duan, Nanoscale, 2011, 3, 963–966.
31. C. Z. Zhao, X. G. Kong, X. M. Liu, L. P. Tu, F. Wu, Y. L. Zhang, K. Liu, Q. H. Zeng and H. Zhang, Nanoscale, 2013, 5, 8084–8089.
32. A. X. Yin, Y. W. Zhang, L. D. Sun and C. H. Yan, Nanoscale, 2010, 2, 953–959.
33. S. Heer, K. Kömpe, H. U. Güdel and M. Haase, Adv. Mater., 2004, 16, 2102–2105.
34. J. H. Zeng, J. Su, Z. H. Li, R. X. Yan and Y. D. Li, Adv. Mater., 2005, 17, 2119–2123.
35. F. Vetrone, R. Naccache, V. Mahalingam, C. G. Morgan and J. A. Capobianco, Adv. Funct. Mater., 2009, 19, 2924–2929.
36. R. Naccache, F. Vetrone, V. Mahalingam, L. A. Cuccia and J. A. Capobianco, Chem. Mater., 2009, 21, 717–723.
37. M. Haase and H. Schafer, Angew. Chem., Int. Ed., 2011, 50, 5808–5829.
38. A. Aebischer, M. Hostettler, J. Hauser, K. Kramer, T. Weber, H. U. Gudel and H. B. Burgi, Angew. Chem., Int. Ed., 2006, 45, 2802–2806.
39. F. Shi, J. S. Wang, D. S. Zhang, G. S. Qin and W. P. Qin, J. Mater. Chem., 2011, 21, 13413–13421.
40. B. Zhou, L. L. Tao, Y. H. Tsang and W. Jin, J. Mater. Chem. C, 2013, 1, 4313–4318.

---

Footnote

† Electronic supplementary information (ESI) available: Size distributions and UC emissions of the seeds and core–shell UCNs. See DOI: 10.1039/c5ra09873d

---