Special Tm³⁺ transition and white upconversion luminescence in the Yb³⁺/Tm³⁺ co-doped KMnF₃

Abstract

Special Tm³⁺ transition and a bright green light emission is first observed in the Yb³⁺/Tm³⁺ co-doped KMnF₃, which are synthesized by the molten salt synthesis method. The RGB light emissions are simultaneously obtained with a single Tm³⁺ activator, and white upconversion luminescence is observed only by changing the reaction time.

---

1. Introduction

Rare-earth (RE) ion doped materials with frequency upconversion (UC) luminescence have been the focus of the related scientific research owing to their applications in solid-state blue-green lasers, three-dimensional color displays, solar cells, and fluorescent bio-labels.^1–4 In particular, RE³⁺-doped materials capable of white UC emission have attracted enormous attention.^5–10 In the field of color-tunable UC luminescence, more and more efforts are being devoted for the development of materials emitting white UC light. Generally, white light is obtained by mixing three monochromatic light sources RGB (red, green, and blue). The common method of inducing RGB light emissions is to add a variety of rare earth ions with colorful UC emissions.^11–13

For example, white UC luminescence is observed in the tri-doped Yb³⁺, Tm³⁺ and Er³⁺ system upon 980 nm diode laser excitation, which consisted of blue (¹D₂ → ³F₄, ¹G₄ → ³H₆ of Tm³⁺), green (²H_11/2/⁴S_3/2 → ⁴I_15/2 of Er³⁺), and red (⁴F_9/2 → ⁴I_15/2 of Er³⁺) UC emissions.¹⁴ Even though green (¹D₂ → ³H₅, 550 nm) and red (¹G₄ → ³F₄, 650 nm) emissions can also be produced by the energy level transition of Tm³⁺, they are generally weak or even cannot be detected due to their low transition probabilities.¹⁵ Consequently, if the transition probabilities of the ¹D₂ and ¹G₄ levels of Tm³⁺ could be adequately altered, the RGB light emissions output from a single activator (Tm³⁺) would be achievable. Furthermore, white UC light can also be obtained by controlling the intensities and ratios of the RGB light emissions in the Yb³⁺/Tm³⁺ co-doped systems.

The RGB light emissions of Tm³⁺ are usually originated from the excitation of high energy levels (¹D₂ and ¹G₄). However, the sensitization for high energy level transition is seldom reported in the literature.^16–19 Notably, the influence of the sensitizer (Mn²⁺) on the UC emission behaviors has been discussed in detail by Tian's group and Wang's group.^16,17 The results reveal that there is an efficient energy transfer between the RE³⁺ (RE = Er, Tm, Ho) and Mn²⁺. Therefore, the increase in NIR emission transition of Tm³⁺ can be attributed to the non-radiative energy transfer from the ¹D₂ and ¹G₄ levels of Tm³⁺ to the ⁴T₁ level of Mn²⁺, followed by a back-energy transfer to the ³F_2,3 level of Tm³⁺. Because the ratios of light emissions show a linear relation with the sensitized abilities,¹¹ it is highly possible that the intensities and ratios of RGB light emissions could be modulated by introducing the sensitizer (Mn²⁺) into the Yb³⁺/Tm³⁺ co-doped KMnF₃ system.

To that end, we have synthesized a series of Yb³⁺/Tm³⁺ co-doped KMnF₃ (KMnF₃:Yb, Tm) crystals utilizing the molten salt synthesis (MSS) method, followed by a detailed study on UC luminescence properties. The research results show that the reaction time is a crucial factor to affect the UC luminescence performance of KMnF₃:Yb, Tm system. It is noteworthy that a unique change has taken place in the ¹D₂ level transition of Tm³⁺ as the reaction time was extended. First, a strong green emission is observed in such materials, which is originated from the excitation of the activator (Tm³⁺). In addition, the novel transition process of Tm³⁺ has been found, which increases the low transition probability and makes it possible for obtaining all the RGB light emissions solely from the activator (Tm³⁺). Therefore, it appears that white UC light can be rationally designed by controlling the relative intensities of the RGB light emissions in the Yb³⁺/Tm³⁺ co-doped KMnF₃ material, which could be readily achieved by varying the reaction time used. To our knowledge, it is the first example of obtaining white UC light via employing a single Tm³⁺ activator in the RE³⁺-doped luminescence materials. The change in the transition probabilities provided a new route for the color-tunable UC fluorescence. White UC light was just one aspect of the application, more potential applications will be found in the future.

2. Experimental section

All the reagents were obtained from commercial sources and directly used without purification. For a typical synthesis, KCl, MnCl₂·4H₂O, NH₄HF₂, Yb(NO₃)₃·5H₂O, and Tm(NO₃)₂·6H₂O were added to an autoclave in a molar ratio of 1 : 1 : 10 : 5% : 0.5%. The autoclave was sealed and maintained at 180 °C for different reaction time. After naturally cooling down to room temperature, the resulting products were washed with deionized water and dried overnight at 60 °C.

All the samples were characterized by powder X-ray diffraction (XRD, Cu Kα radiation, Rigaku D/max2550VB, Japan) at a scanning rate of 8° min⁻¹ within the 2θ range of 10° to 80°. The morphology and particle size of the samples were determined by scanning electron microscopy (SEM, JSM-6700F, JEOL, Japan). The UC emission spectra were measured on an Edinburgh Instruments FLS920 spectrophotometer with an external 0–600 mW adjustable continuous-wave 980 nm laser as the excitation source. The particle surface was characterized by X-ray photoelectron spectroscopy (XPS, Al Kα radiation, America, ESCALAB 250). All the optical measurements were performed under ambient conditions.

3. Results and discussion

The crystalline structures of the as-synthesized products are identified by powder X-ray diffraction (XRD). Fig. 1 shows the XRD patterns of KMnF₃ prepared at 180 °C with different reaction times (30 min, 6 h, 12 h, and 24 h), as well as the standard XRD data of pure cubic KMnF₃ phase as a reference. The XRD results show that all the samples are well crystallized. All the diffraction peaks can be indexed to the cubic phase of KMnF₃ (JCPDS: 17-0116), indicating the high purity of the as-synthesized products.

Fig. 1 XRD patterns of the KMnF₃ samples prepared at 180 °C with different reaction times (A) 30 min, (B) 6 h, (C) 12 h, and (D) 24 h, and the standard data of cubic KMnF₃ (JCPDS: 17-0116).

Fig. 2 shows the scanning electron microscopy (SEM) images of the KMnF₃:Yb, Tm samples prepared at different reaction times (30 min, 6 h, 12 h, and 24 h), in which the resulting crystals are clearly cube-shaped. It is found that the size of the crystals significantly increases with the increase in reaction time. When the reaction is terminated after 30 min, the size of crystals is less than 1 μm, whereas if the reaction is allowed to proceed for 24 h, the crystal size turns out to be 5 μm.

Fig. 2 SEM micrographs of the Yb/Tm co-doped KMnF₃ samples prepared with different reaction times (A) 30 min, (B) 6 h, (C) 12 h, (D) 24 h.

The UC emission spectra of the KMnF₃:Yb, Tm crystals prepared with different reaction times (30 min, 6 h, 12 h, and 24 h) are exhibited in Fig. 3. The NIR UC emission at 800 nm corresponds to the ³H₄ → ³H₆ transition (Fig. 3A), which is due to the excitation of the ³F_2,3 level, followed by the non-radiative relaxation from the ³F_2,3 level to the ³H₄ level. Notably, two weak bands at 475 nm and 650 nm are observed corresponding to the ¹G₄ → ³H₆ and ¹G₄ → ³F₄ transition of Tm³⁺, respectively. In addition, no ¹D₂ level transition is observed, and two ¹G₄ level transitions (¹G₄ → ³H₆ and ¹G₄ → ³F₄) seem to be weak in Fig. 3A. These results could be attributed to the non-radiative energy transfer from the ¹D₂ and ¹G₄ levels of Tm³⁺ to the ⁴T₁ level of Mn²⁺, followed by a back-energy transfer to the ³F_2,3 level of Tm³⁺. When the reaction time is increased to 6 h, the band at 650 nm seems to be substantially enhanced, as illustrated in Fig. 3B. Further extending the reaction time to 12 h, two new bands at 450 nm and 550 nm clearly appeared, corresponding to the ¹D₂ → ³F₄ and ¹D₂ → ³H₅ transition of Tm³⁺, respectively (Fig. 3C). Therefore, we conclude that the ¹D₂ and ¹G₄ level transitions can be simply enhanced by extending the reaction time, as illustrated in Fig. 3A to C. In other words, the sensitization of Mn²⁺ can be substantially decreased during this process. As shown in Fig. 4, the exchange-energy transfer process from the ¹D₂ level of Tm³⁺ to the ⁴T₁ level of Mn²⁺ is more efficient than the one from the ¹G₄ level to the ⁴T₁ level. When the reaction time is increased to 24 h, the ¹D₂ level emission of Tm³⁺ at 550 nm is found to be stronger, whereas the band at 450 nm seems to be disappeared, as illustrated in Fig. 3D. Generally, when the ¹D₂ level of Tm³⁺ is excited, the most likely transition process should be the ¹D₂ → ³F₄ transition at 450 nm; other probable transitions, however, would be fairly weak or even negligible due to their low transition probabilities. Surprisingly, we observe an obvious band at 550 nm (¹D₂ → ³H₅) in the emission spectra of KMnF₃:Yb, Tm samples prepared at 24 h (Fig. 3D), which has never been reported in the literature and no reliable mechanism with it. The result suggests that an unprecedented alteration has occurred in the ¹D₂ level transition of Tm³⁺. In general, ¹D₂ → ³F₄ (450 nm, blue) transition is probability almost 220 times that of ¹D₂ → ³H₅ (550 nm, green),¹⁵ the strong green light emission from Tm³⁺ was due to the change of the ¹D₂ level transition probabilities ratio between them. The synthesis conditions of the KMnF₃ : Yb, Tm crystals in our work are try to keep the same with the traditional solid-phase synthesis except the NH₄HF₂ flux, the excess F⁻ are the most probable reason, but still need more evidence to prove. Similar results can be observed in the Yb³⁺/Tm³⁺ co-doped β-NaYF₄ system (Fig. 5), which has been prepared in the absence of Mn²⁺ by the MSS method in our research group. On the basis of the data collected from these two examples, we conclude that the abnormal change of transition probabilities could originate from the influence of the crystal structure and micrometer size, rather than the introduction of Mn²⁺. In addition, it is proved that the RGB light emissions can be simultaneously obtained by a single Tm³⁺ activator in the Yb³⁺/Tm³⁺ co-doped system, which are red (¹G₄ → ³F₄, at 650 nm), green (¹D₂ → ³H₅, at 550 nm), and blue (¹D₂ → ³F₄, at 450 nm and ¹G₄ → ³H₆, at 475 nm) UC emissions. As a matter of fact, the intensities and ratios of the RGB light emissions can be solely modulated by controlling the reaction time, affording white light emission.

Fig. 3 Room temperature UC emission spectra of the KMnF₃:Yb, Tm samples prepared with different reaction times (A) 30 min, (B) 6 h, (C) 12 h, (D) 24 h. The samples were excited with a 980 nm laser operating at 600 mW (The insets are the magnified image for the visible light area).

Fig. 4 Energy level diagrams of the Yb³⁺, Tm³⁺, and Mn²⁺, and the efficient transfer process from the ¹D₂ level of Tm³⁺.

Fig. 5 Room temperature UC emission spectra of the Yb³⁺/Tm³⁺ co-doped β-NaYF₄ sample prepared at 180 °C for 24 h. The samples were excited with a 980 nm laser operating at 600 mW.

All the UC emission spectra of the KMnF₃:Yb, Tm crystals can be readily converted to the CIE-1931 chromaticity diagrams, as plotted in Fig. 6. The corresponding chromaticity coordinates of crystals have been calculated based on the emission spectra, which are (0.4435, 0.2428), (0.4081, 0.3216), (0.3532, 0.3499) and (0.3151, 0.6018), respectively. As shown in the chromaticity diagrams, the color tone of the crystals gradually shifts from red to white, and ultimately to green with the prolongation of the reaction time. Therefore, it appears that the white light emission can be obtained only by varying the reaction time utilized in the synthetic process. Indeed, when the reaction time is prolonged from 6 h to 12 h, the chromaticity coordinates could locate in the white light range as expected. By further extending the reaction time to 24 h, the chromaticity coordinate could move into the green light area. Differing from the previously reported color-tunable methods,^20–22 the approach reported in this work can produce red, green and white fluorescent materials based on the KMnF₃:Yb, Tm system by varying the reaction time, and no changes to any other reaction conditions are needed.

Fig. 6 CIE chromaticity diagram of the KMnF₃:Yb, Tm samples prepared with different reaction times (A) 30 min, (B) 6 h, (C) 12 h, (D) 24 h.

The X-ray photoelectron spectroscopy (XPS) spectrum (Fig. 7) shows that the concentration of Mn²⁺ on the surface of the crystal is decreased with the extension of reaction time, whereas the concentration of Yb³⁺ is increased. It is noted that the energy exchange interaction between Mn²⁺ and Tm³⁺ is weakened and between Yb³⁺ and Tm³⁺ is strengthened, resulting in a gradual increase in the ¹D₂ level transitions in Tm³⁺. In addition, the Yb³⁺/Er³⁺ co-doped KMnF₃ crystals have also been synthesized at the same reaction condition by our research group, similar results are also obtained from the UC emission spectra, as shown in Fig. 8.

Fig. 7 XPS spectra for the KMnF₃:Yb, Tm samples (A) Mn²⁺ 2p; (B) Yb³⁺ 4d.

Fig. 8 Room temperature UC emission spectra of the Yb³⁺/Er³⁺ co-doped KMnF₃ samples prepared at 180 °C with different reaction times (A) 6 h, (B) 24 h. The samples were excited with a 980 nm laser operating at 600 mW.

4. Conclusions

We have successfully synthesized the Yb³⁺/Tm³⁺ co-doped KMnF₃ crystals by a MSS method. Subsequent studies indicate that the UC luminescence properties of such materials can be effectively modulated by varying the reaction time. Specifically, it appears that the ¹D₂ level transition of Tm³⁺ undergoes unprecedented changes while the reaction time is prolonged, which results in a bright green light emission from Tm³⁺ is. The RGB light emissions have been simultaneously obtained in the presence of a single activator (Tm³⁺). Moreover, it is found that the intensities and ratios of RGB light emissions can be controlled by varying the reaction time, producing white light emission in the KMnF₃:Yb, Tm system. To our knowledge, white UC light is successfully obtained for the first time in the Yb³⁺/Tm³⁺ co-doped system. Overall, this work not only provides a new route for the development of white UC light emitting materials, but also demonstrates that the colorful emissions can be readily modulated in a broad range by varying the reaction conditions employed in the synthesis. Currently, our efforts are primarily focused on the preparation of more diversified RE³⁺-doped luminescence materials, and the studies to understand the intrinsic mechanism involved in the change of transition probability are still ongoing.

Acknowledgements

This work was supported by the National Sciences Foundation of China (no. 21271082 and 21301066).

References

1. F. Auzel, Chem. Rev., 2004, 104, 139.
2. E. Downing, L. Hesselink, J. Ralston and R. Macfarlance, Science, 1996, 273, 1185.
3. E. D. Rosa, P. Salas, H. Desirenca, C. Angeles and R. A. Rodríguez, Appl. Phys. Lett., 2005, 87, 241912.
4. F. Van de Rijke, H. Zijlmans, S. Li, T. Vail, A. K. Raap, R. S. Niedbala and H. J. Tanke, Nat. Biotechnol., 2001, 19, 273.
5. M. S. Wang, S. P. Guo, Y. Li, L. Z. Cai, J. P. Zou, G. Xu, W. W. Zhou, F. K. Zheng and G. C. Guo, J. Am. Chem. Soc., 2009, 131, 13572.
6. N. Guo, H. P. You, Y. H. Song, M. Yang, K. Liu, Y. J. Huang and H. J. Zhang, J. Mater. Chem., 2010, 20, 9061.
7. K. Z. Zheng, D. S. Zhang, D. Zhao, N. Liu, F. Shi and W. P. Qin, Phys. Chem. Chem. Phys., 2010, 12, 7620.
8. T. Passuello, F. Piccinelli, M. Pedroni, M. Brttinelli, F. Mangiarini, R. Naccache, F. Vetrone, J. A. Capobianco and A. Speghini, Opt. Mater., 2011, 33, 643.
9. C. M. Zhang, P. A. Ma, C. X. Li, G. G. Li, S. S. Huang, D. M. Yang, M. M. Shang, X. J. Kang and J. Lin, J. Mater. Chem., 2011, 21, 717.
10. M. M. Shang, C. X. Li and J. Lin, Chem. Soc. Rev., 2014, 43, 1372.
11. J. Wang, R. R. Deng, M. A. MacDonald, B. Chen, J. K. Yuan, F. Wang, D. Z. Chi, T. S. A. Hor, P. Zhang, G. K. Liu, Y. Han and X. G. Liu, Nat. Mater., 2013, 13, 157.
12. E. M. Chen, G. Han, J. D. Goldberg, D. J. Gargas, A. D. O. P. J. Schuck, B. E. Cohen and D. J. Milliron, Nano Lett., 2012, 12, 383.
13. F. Wang and X. G. Liu, J. Am. Chem. Soc., 2009, 130, 5642.
14. K. Z. Zhang, D. S. Zhang, D. Zhao, N. Liu, F. Shi and W. P. Qin, Phys. Chem. Chem. Phys., 2010, 12, 7620.
15. G. Poirier, V. A. Jerez, C. B. de Araújo, Y. Messaddeq, S. J. L. Ribeiro and M. Poulain, J. Appl. Phys., 2003, 93, 1493.
16. G. Tian, Z. J. Gu, L. J. Zhou, W. Y. Yin, X. X. Liu, L. Yan, S. Jin, W. L. Ren, G. M. Xing, S. J. Li and Y. L. Zhao, Adv. Mater., 2012, 24, 1226.
17. J. Wang, F. Wang, C. Wang, Z. Liu and X. G. Liu, Angew. Chem., Int. Ed., 2011, 50, 10369.
18. N. Guo, H. P. You, Y. H. Song, M. Yang, K. Liu, Y. H. Zheng, Y. J. Huang and H. J. Zhang, J. Mater. Chem., 2010, 20, 9061.
19. J. Yang, C. M. Zhang, C. X. Li, Y. N. Yu and J. Lin, Inorg. Chem., 2008, 47, 7262.
20. J. Yang, C. M. Zhang, C. Peng, C. X. Li, L. L. Wang, R. T. Chai and J. Lin, Chem.–Eur. J., 2009, 15, 4649.
21. K. Z. Zheng, D. S. Zhang, D. Zhao, N. Liu, F. Shi and W. P. Qin, Phys. Chem. Chem. Phys., 2010, 12, 7620.
22. J. W. Wang and P. A. Tanner, J. Am. Chem. Soc., 2010, 132, 947.

---