Eu³⁺-based luminescence ratiometric thermometry

Abstract

Recently, luminescence ratiometric thermometry has gained ever-increasing attention due to its merits of rapid response, non-invasiveness, high spatial resolution, and so forth. For research fields relying on temperature measurements, achieving a higher relative sensitivity of this measurement is still an important task. In this work, we developed a strategy for achieving a more sensitive temperature measurement, one merely depending on the photoluminescence of Eu³⁺. We showed that using the ⁵D₁–⁷F₁ transition and the hypersensitive ⁵D₀–⁷F₂ transition of Eu³⁺ can boost the relative sensitivity compared with the method relying on the ⁵D₁–⁷F₁ and ⁵D₀–⁷F₁ transitions of Eu³⁺. The difference between these two strategies was studied and was explained by the hypersensitive ⁵D₀–⁷F₂ transition more steeply decreasing than the ⁵D₀–⁷F₁ transition with a rise in temperature. Our work is expected to help researchers design sensitive optical thermometers via proper use of this hypersensitive transition.

---

1. Introduction

Temperature, one of the most significant physical and chemical parameters, plays a quite important role in several fields, spanning from biological medicine to industrial processes and other fields affecting our daily lives.¹ For instance, an increase of 3 °C in body temperature directly threatens human health. Therefore, developing novel and practical methods or devices for measuring temperature has become very important.^2–5 Although a traditional thermocouple and thermistor are widely used nowadays, they are not very good at measuring temperature precisely, quickly, and in particular, non-invasively. Optical thermometry can overcome these problems well as it features a short response time, non-contact work model, relatively high thermal and spatial resolutions, and so forth.^6–10

So far, several approaches have been developed to realize optical thermometry. Of these, luminescence ratiometric thermometry has gained ever-increasing attention.¹¹ It depends on two emission lines and is thus strongly resistant to many influencing factors, such as the loss of optics transmission.¹² When first developed, luminescence ratiometric thermometry involved thermally coupled energy levels (TCELs) of rare earth and transition metal ions.^13–18 Sui et al. prepared a Tm³⁺/Yb³⁺-co-doped LiYF₄ single crystal and found that the intensity ratio between the ³F₂/³F₃–³H₆ and ³H₄–³H₆ transitions of Tm³⁺ could be used for temperature determination.¹⁹ Rakov et al. demonstrated that the ²F_5/2 Stark sublevels of Yb³⁺ and the ⁴I_13/2 Stark sublevels of Er³⁺ were both TCELs, based on which they successfully measured the temperature.^20,21 Marciniak et al. prepared LiYbP₄O₁₂ : Er³⁺ nanocrystals and showed that the intensity ratio between the ²H_11/2–⁴I_15/2 and ⁴S_3/2–⁴I_15/2 transitions of Er³⁺ followed the Boltzmann distribution, and could be used for measuring temperature.²² Nikolić et al. demonstrated the potential application of Dy³⁺-doped GdVO₄ as a thermographic phosphor, with this potential application based on the ⁴I_15/2 and ⁴F_9/2 states being TCELs.²³ Rai et al. observed the ³P₁–³H₅ and ³P₀–³H₅ transitions in Pr³⁺-doped lithium tellurite glass and showed the luminescence intensity ratio (LIR) between them to be a promising temperature indicator.²⁴ Back et al. and Ueda et al. have shown the TCELs of Cr³⁺ also follow the Boltzmann distribution and thus can be used for ratiometric thermometry.^25–27 Recently, luminescence ratiometric thermometry was also realized using non-TCELs.^28–30 Cui et al. prepared the first ratiometric luminescent metal–organic framework thermometer on the basis of the green-light emission of Tb³⁺ and the red-light emission of Eu³⁺.³¹ Savchuk et al. proposed a new upconversion thermometry method based on emissions of light from Ho³⁺ and Tm³⁺.³² Cortelletti et al. showed that Nd³⁺–Yb³⁺-codoped SrF₂ nanoparticles with a multi-shell architecture, operating in the biological window, were qualified temperature indicators.³³ Nearly all of these reported investigations were committed to increasing the relative sensitivity of temperature measurements, which is one of the most important goals in the field of optical thermometry.

In this work, we developed a strategy for boosting the relative sensitivity of luminescence ratiometric thermometry. This strategy was based on the combination of a common transition and a hypersensitive transition of Eu³⁺. We showed that the luminescence intensity ratio between the ⁵D₁–⁷F₁ and ⁵D₀–⁷F₁ transitions and the LIR between the ⁵D₁–⁷F₁ and ⁵D₀–⁷F₂ transitions of Eu³⁺ can be used for temperature measurements as both LIRs were found to increase gradually with increasing temperature. However, the latter LIR, compared with the former, showed a greater relative sensitivity. The discrepancy of sensitivity between these two temperature detection methods was especially obvious at relatively high temperatures. The mechanism for this phenomenon was studied in detail.

2. Materials and characterization

CaO, WO₃, and Eu₂O₃ (4N, Aladdin Co., Ltd) were used as raw materials. CaWO₄ : X% Eu³⁺ (X = 1, 10 and 20) phosphors were prepared by following a high-temperature solid state method. To synthesize the phosphors, first stoichiometric amounts of CaO, WO₃ and Eu₂O₃ raw materials were fully ground together in a 50 mL agate mortar, with the aim of forming uniform mixtures. These mixtures were then pressed into several disks for the subsequent optical tests, and the disks were sintered at 1473 K for 4 h to form the final products.

Powder X-ray diffraction (PXRD) patterns of the as-prepared samples were acquired (PANalytical X'Pert powder X-ray diffractometer) at room temperature. A continuous-wave 405 nm laser diode (ITC-510, Thorlabs) and 150 W xenon lamp (LHX-150, Zolix Instruments Co., Ltd) were used as excitation sources. A spectrometer (SBP-300, Zolix Instruments Co., Ltd) was used for wavelength discrimination. A photomultiplier (PMTH-S1-CR131, Zolix Instruments Co., Ltd) was used to collect excitation and emission spectra of the samples. The temperature of the samples was changed using a home-made heating chamber with an accuracy of ±0.3 K over the experimental temperature range of 333 to 733 K.

3. Results and discussion

3.1 PXRD patterns and crystal structure

The expected structure of the as-prepared samples was confirmed using PXRD analysis (Fig. 1a). For comparison, the reference data set of JCPDS no. 41-1431 is also presented. The PXRD patterns acquired of the as-prepared samples were observed to be quite similar to the reference. Moreover, no redundant peaks were observed for the samples. These results revealed that the CaWO₄ : X% Eu³⁺ (X = 1, 10 and 20) phosphors were prepared successfully and each formed a pure tetragonal phase. The addition of Eu³⁺ did not affect the structure of CaWO₄. Eu³⁺ successfully occupied the Ca²⁺ site in the CaWO₄ host, apparently due to the similarity of the Eu³⁺ (0.095 nm) and Ca²⁺ (0.099 nm) radii. The crystal structure of the scheelite-type CaWO₄ is also given (Fig. 1b). The Eu³⁺/Ca²⁺ ions embedded in the phosphors were each octahedrally coordinated by eight O²⁻ oxygens (Fig. 1c). Moreover, the W⁶⁺ ions in the samples were each surrounded by four O²⁻ oxygens (Fig. 1d), forming a tetrahedron.

Fig. 1 (a) PXRD patterns of the as-prepared samples. (b) Crystal structure, (c) [CaO₈] cluster and (d) [WO₄] cluster of the Scheelite-type CaWO₄.

3.2 Basic spectroscopic properties

A room-temperature photoluminescence (PL) spectrum of the CaWO₄:10%Eu³⁺ phosphors excited with 405 nm-wavelength light was acquired (Fig. 2a). Between wavelengths of 500 and 750 nm, six PL bands with peaks at wavelengths of 535, 580, 593, 615, 650 and 700 nm, respectively, were observed. The 535 nm PL band arose from the ⁵D₁–⁷F₁ transition of Eu³⁺, and the 580, 593, 615, 650 and 700 nm PL bands were attributed, respectively, to the ⁵D₀–⁷F_J (J = 0, 1, 2, 3 and 4) transitions of Eu³⁺.^34–39 The 615 nm PL band originating from the hypersensitive transition was the most intense band, a result similar to that reported previously.^34–39 By monitoring this PL band, the excitation spectrum was collected (Fig. 2a). All of these excitation bands were sharp. The attributions of these bands have been described previously.^34–39 According to a proposed model, upon excitation at 405 nm, the Eu³⁺ ions at the ground state absorbed these photons and adopted each a high-energy excited state (Fig. 2b). This process was then followed by non-radiative relaxation to the lower-energy ⁵D₁ and ⁵D₀ emitting states. Finally, radiative transitions from these two excited states to the different ground states occurred.

Fig. 2 (a) Room-temperature excitation spectrum (monitored at 615 nm) and PL spectrum (excitation at 405 nm) of the CaWO₄:10%Eu³⁺ phosphors. (b) Energy level diagram of Eu³⁺ and possible excitation–emission processes.

A series of CaWO₄ : X% Eu³⁺(X = 1, 10, 20) phosphors were prepared. For all of the samples, the characteristic PL bands of Eu³⁺ were observed. The profile and position of these PL bands remained unchanged as the concentration of Eu³⁺ was changed (Fig. 3a). However, as the Eu³⁺ concentration was increased, the PL band intensity of Eu³⁺ increased sharply, due to more Eu³⁺ ions acting as luminescence centres. However, when the Eu³⁺ concentration was increased tenfold, the enhancement of the total emission intensity of the phosphors did not keep pace (Fig. 3b). The concentration quenching effect was expected to play a significant role.

Fig. 3 (a) Room-temperature PL spectra, (b) integrated emission intensities, (c) room-temperature decay curves, and (d) lifetimes of the 615 nm PL band of the phosphors.

Room-temperature decay curves of the 615 nm PL band of the prepared phosphors were collected (Fig. 3c). All of the decay curves were modelled effectively using the single exponential function

where I is the intensity of the PL decay curve, t is time and t₀ is the lifetime. By using this equation, the lifetimes of the 615 nm PL band were calculated for the prepared phosphors at room temperature (Fig. 3d). As the Eu³⁺ concentration was increased from 1 to 20%, the lifetime of the 615 nm PL band decreased gradually from 569 to 470 μs, indicating the presence of a concentration quenching effect shortening the lifetimes of the emitting states.

Room-temperature PL spectra of the as-prepared phosphors excited with light of a wavelength of 405 nm were acquired (Fig. 4a). Note that they were normalized to the PL intensity at 615 nm. The LIR between the 615 and 593 nm PL bands was obtained (inset in Fig. 4a). As has been described previously, the 615 nm PL band arose from the ⁵D₀–⁷F₂ transition, which has been confirmed to be a hypersensitive transition. Due to this unique characteristic, the 615 nm PL band was sensitive to the symmetry property of the Eu³⁺ site in the CaWO₄ host. Therefore, the LIR between the 615 and 593 nm PL bands reflected the distortion of the crystal field environment around Eu³⁺.³⁵ There was a slight fluctuation of the LIR between the 615 and 593 nm PL bands, demonstrating a lack of any effect of Eu³⁺ concentration on the symmetry properties of the Eu³⁺ site in the CaWO₄ host. The CIE chromaticity coordinates of the prepared phosphors were also calculated based on their emission spectra (Fig. 4b). The emissions of all of these phosphors were located in the red region, as the 615 nm PL band in each case was the predominant band in the emission spectrum. Moreover, as the Eu³⁺ concentration was increased, the CIE chromaticity coordinates gradually became located closer to the deep-red region. This result may be explained as follows: as the Eu³⁺ concentration was increased, the population of the ⁵D₀ state increased, at the expense of a depopulation of the upper ⁵D₁ state, due to the concentration quenching effect; and therefore, the red transition accounted for an increasing proportion of the total emission from the phosphors with increasing Eu³⁺ concentration.

Fig. 4 (a) Normalized room-temperature PL spectra of the prepared CaWO₄ : X% Eu³⁺(X = 1, 10, 20) phosphors, and the LIR between the 615 and 593 nm PL bands as a function of Eu³⁺ concentration (inset). (b) CIE chromaticity coordinates of the prepared phosphors.

3.3 Temperature sensing properties

PL spectra were acquired of the as-prepared phosphors excited at 397 nm at various temperatures between 333 and 733 K (Fig. 5a and b). The integrated emission intensities of the 535, 593 and 615 nm PL bands were also calculated (Fig. 5c, d and e). The 535 nm PL band intensity was observed to increase as the temperature was increased from 333 to 453 K, and then decrease continuously with further increasing temperature. In contrast, the 593 and 615 nm PL band intensities each decreased markedly and continuously as the temperature was increased over the whole temperature range. This set of results can be understood by considering that the 535 nm PL band arose from the ⁵D₁ state while the other two PL bands were from the ⁵D₀ state: as there was a thermal equilibrium between the ⁵D₁ and ⁵D₀ states, the upper ⁵D₁ state became populated gradually at the expense of the lower ⁵D₀ state. Note that the decreased 535 nm PL band intensity at relatively high temperatures was likely due to an effect of thermal quenching.

Fig. 5 (a and b) PL spectra, (c–e) integrated intensities of the (c) 535 nm, (d) 593 nm, and (e) 615 nm PL bands, and (f) CIE diagram of the as-prepared phosphors excited with light of a wavelength of 397 nm, each as a function of temperature from 333 to 733 K.

The CIE chromaticity coordinates of the as-prepared phosphors were also calculated as a function of temperature from 333 to 733 K (Fig. 5e). At 333 K, the CIE chromaticity coordinates were determined to be located in the red region. As the temperature was increased, the 615 nm PL band became weaker. In contrast, the 535 nm PL band accounted for an increasing proportion of the total emission for the phosphors. Therefore, the CIE chromaticity coordinates gradually moved to the red region.

The LIR between the 535 and 593 nm PL bands (LIR-12) and that between the 535 and 615 nm PL bands (LIR-13) were obtained as a function of temperature from 333 to 733 K (Fig. 6a). Both LIRs were observed to increase monotonically with increasing temperature, with this result due to the thermal coupling between the ⁵D₁ and ⁵D₀ states. That is, both LIRs can be used for indicating temperature. According to the Boltzmann distribution, the LIR between TCELs is expressed as¹²

where A is a constant, ΔE is the energy gap between TCELs, k is the Boltzmann constant, T is the absolute temperature and B is the fitting offset. Both LIR-12 and LIR-13 were fitted well by using eqn (2) (Fig. 6a). For LIR-12, the parameters A, ΔE and B were found to be 8.4, 1538 cm⁻¹ and 0.024, respectively. The fitted energy gap was slightly smaller than the experimentally determined value. This discrepancy has been reported many times, and an explanation for it is still being pursued. For LIR-13, the fitted parameters A, ΔE and B were 1.9, 1762 cm⁻¹ and 0.027, respectively. Obviously, the 593 and 615 nm PL bands showed different responses to temperature, and the intensity of the latter decreased more sharply with increasing temperature than that of the former.

Fig. 6 (a) LIR-12, LIR-13 and (b) LIR-32 as a function of temperature from 333 to 733 K.

To confirm the discrepancy between these two PL bands, the LIR between the 615 and 593 nm PL bands (LIR-32) was also calculated as a function of temperature from 333 to 733 K (Fig. 6b). LIR-32 was observed to drop sharply as the temperature was increased. At 733 K, LIR-32 was only ∼70 percent of that at 333 K. This result revealed again a steeper response of the 615 nm PL band than of the 593 nm PL band to temperature. As mentioned above, LIR-32 can reflect the crystal field environment around Eu³⁺. Therefore, the LIR-32 results, shown in Fig. 6b, suggested an increase in the symmetry of the Eu³⁺ site in the CaWO₄ host with increasing temperature.

In order to obtain the relative sensitivity S_r, the absolute sensitivity S_a was calculated first according to the equation S_r = S_a/LIR, with the absolute sensitivity S_a defined using the equation⁴⁰

By using this equation, the absolute sensitivity S_a was calculated for LIR-12 and LIR-13 as a function of temperature from 333 to 733 K (Fig. 7a). LIR-12 showed an obviously larger S_a than did LIR-13. Based on the expression S_r= S_a/LIR, the relative sensitivity S_r was also computed for LIR-12 and LIR-13 (Fig. 7b). The S_r for LIR-13 was calculated for each temperature value tested to be larger than that for LIR-12. This result may have been due to the 615 nm PL band originating from the hypersensitive transition and thus being more sensitive to the surroundings. With an increase in temperature, the symmetry of the Eu³⁺ site also increased, which resulted in a more serious decrease in the intensity of the 615 nm PL band than of the intensity of the 593 nm PL band.

Fig. 7 (a) Absolute sensitivities, (b) relative sensitivities and (c) temperature resolutions for LIR-12 and LIR-13 as a function of temperature from 333 to 733 K.

The temperature resolution ΔT, in addition to the relative sensitivity S_r, is also very important in practice. It is defined as

where ΔLIR/LIR is the relative uncertainty of the LIR that is used for indicating temperature. The terms ΔLIR-12/LIR-12 and ΔLIR-13/LIR-13 can be expressed asandwhere in the current work I₁, I₂ and I₃ were used to denote the intensities of the 535, 593 and 615 nm PL bands, respectively. Under the same experimental setup, ΔI₃/I₃ was always lower in value than was ΔI₂/I₂, due to the higher intensity of the 615 nm PL band. Inspection of the emission spectra of the samples clearly showed a much larger value for ΔI₁/I₁ than for ΔI₂/I₂ and ΔI₃/I₃. Therefore, the value of ΔLIR-13/LIR-13 was just slightly lower than that of ΔLIR-12/LIR-12. The S_r for LIR-13 was shown above to be greater than that for LIR-12. So, the strategy of using 535 and 615 nm PL bands yielded a better temperature resolution than did the strategy depending on the 535 and 593 nm PL bands (Fig. 7c). Therefore, using the 535 and 615 nm PL bands was shown to achieve not only a higher relative sensitivity but also a better temperature resolution than that achieved using the 535 and 593 nm PL bands.

4. Conclusion

In summary, Eu³⁺-embedded scheelite-type phosphors were prepared. Upon excitation at 405 nm, the characteristic ⁵D₁–⁷F₁, ⁵D₀–⁷F₁ and ⁵D₀–⁷F₂ transitions of Eu³⁺ were observed for the as-prepared samples. It was shown that both LIR-12 and LIR-13 can be used for indicating temperature. Moreover, LIR-13 showed a larger relative sensitivity than LIR-12. The discrepancy between these two temperature detection methods was disclosed, and was ascribed to a faster decay of the ⁵D₀–⁷F₂ transition than of the ⁵D₀–⁷F₁ transition with increasing temperature, further demonstrating a gradually increasing symmetry around Eu³⁺ in the CaWO₄ host with increasing temperature. Our work, by revealing the proper use the hypersensitive transition, is expected to help researchers design sensitive optical thermometry.

Conflicts of interest

The authors declare no competing financial interests.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (NSFC) (Grant No. 81571720 & 61505045).

References

1. P. R. N. Childs, J. R. Greenwood and C. A. Long, Rev. Sci. Instrum., 2000, 71, 2959–2978.
2. M. Quintanilla and L. M. Liz-Marzán, Nano Today, 2018, 19, 126–145.
3. D. Jaque and F. Vetrone, Nanoscale, 2012, 4, 4301–4326.
4. J. Rocha, C. D. S. Brites and L. D. Carlos, Chem.–Eur. J., 2016, 22, 1–15.
5. X. Wang, O. S. Wolfbeis and R. J. Meier, Chem. Soc. Rev., 2013, 42, 7834–7869.
6. F. Vetrone, R. Naccache, A. Zamarrón, A. J. de la Fuente, F. Sanz-Rodríguez, L. M. Maestro, E. M. Rodriguez, D. Jaque, J. G. Solé and J. A. Capobianco, ACS Nano, 2010, 4, 3254–3258.
7. L. Marciniak, K. Prorok and A. Bednarkiewicz, J. Mater. Chem. C, 2017, 5, 7890–7897.
8. R. Lei, D. Deng, X. Liu, F. Huang, H. Wang, S. Zhao and S. Xu, Opt. Mater. Express, 2018, 8, 3023–3035.
9. S. Zhou, C. Duan, M. Yin, X. Liu, S. Han, S. Zhang and X. Li, Opt. Express, 2018, 26, 27339–27345.
10. B. Dong, D. P. Liu, X. J. Wang, T. Yang, S. M. Miao and C. R. Li, Appl. Phys. Lett., 2007, 90, 181117.
11. S. Balabhadra, M. L. Debasu, C. D. S. Brites, L. A. O. Nunes, O. L. Malta, J. Rocha, M. Bettinellie and L. D. Carlos, Nanoscale, 2015, 7, 17261–17267.
12. S. A. Wade, S. F. Collins and G. W. Baxter, J. Appl. Phys., 2003, 94, 4743–4756.
13. M. D. Dramićanin, Methods Appl. Fluoresc., 2016, 4, 042001.
14. M. Back, J. Ueda, J. Xu, D. Murata, M. G. Brik and S. Tanabe, ACS Appl. Mater. Interfaces, 2019, 11, 38937–38945.
15. S. Liu, H. Ming, J. Cui, S. Liu, W. You, X. Ye, Y. Yang, H. Nie and R. Wang, J. Phys. Chem. C, 2018, 122, 16289–16303.
16. H. Suo, C. Guo, J. Zheng, B. Zhou, C. Ma, X. Zhao, T. Li, P. Guo and E. M. Goldys, ACS Appl. Mater. Interfaces, 2016, 8, 30312–30319.
17. B. Dong, R. N. Hua, B. S. Cao, Z. P. Li, Y. Y. He, Z. Y. Zhang and O. S. Wolfbeis, Phys. Chem. Chem. Phys., 2014, 16, 20009–20012.
18. L. Tong, X. Li, J. Zhang, S. Xu, J. Sun, H. Zheng, Y. Zhang, X. Zhang, R. Hua, H. Xia and B. Chen, Opt. Express, 2017, 25, 16047–16058.
19. G. Sui, B. Chen, J. Zhang, X. Li, S. Xu, J. Sun, Y. Zhang, L. Tong, X. Luo and H. Xia, J. Lumin., 2018, 198, 77–83.
20. N. Rakov and G. S. Maciel, Opt. Lett., 2014, 39, 3767–3769.
21. N. Rakov and G. S. Maciel, J. Lumin., 2018, 199, 293–297.
22. L. Marciniak, K. Waszniewska, A. Bednarkiewicz, D. Hreniak and W. Strek, J. Phys. Chem. C, 2016, 120, 8877–8882.
23. M. G. Nikolić, D. J. Jovanović and M. D. Dramićanin, Appl. Opt., 2013, 52, 1716–1724.
24. V. K. Rai, D. K. Rai and S. B. Rai, Sens. Actuators, A, 2006, 128, 14–17.
25. M. Back, E. Trave, J. Ueda and S. Tanabe, Chem. Mater., 2016, 28, 8347–8356.
26. M. Back, J. Ueda, M. G. Brik, T. Lesniewski, M. Grinberg and S. Tanabe, ACS Appl. Mater. Interfaces, 2018, 10, 41512–41524.
27. J. Ueda, M. Back, M. G. Brik, Y. Zhuang, M. Grinberg and S. Tanabe, Opt. Mater., 2018, 85, 510–516.
28. D. Chen, W. Xu, S. Yuan, X. Li and J. Zhong, J. Mater. Chem. C, 2017, 5, 9619–9628.
29. L. Marciniak, A. Bednarkiewicz, J. Drabik, K. Trejgis and W. Strek, Phys. Chem. Chem. Phys., 2017, 19, 7343–7351.
30. K. Trejgis and L. Marciniak, Phys. Chem. Chem. Phys., 2018, 20, 9574–9581.
31. Y. Cui, H. Xu, Y. Yue, Z. Guo, J. Yu, Z. Chen, J. Gao, Y. Yang, G. Qian and B. Chen, J. Am. Chem. Soc., 2012, 134, 3979–3982.
32. O. A. Savchuk, J. J. Carvajal, C. D. S. Brites, L. D. Carlos, M. Aguilo and F. Diaz, Nanoscale, 2018, 10, 6602–6610.
33. P. Cortelletti, A. Skripka, C. Facciotti, M. Pedroni, G. Caputo, N. Pinna, M. Quintanilla, A. Benayas, F. Vetrone and A. Speghini, Nanoscale, 2018, 10, 2568–2576.
34. D. Tu, W. Zheng, P. Huang and X. Chen, Coord. Chem. Rev., 2019, 378, 104–120.
35. Y. Su, L. Li and G. Li, Chem. Mater., 2008, 20, 6060–6067.
36. Y. Chen, S. W. Park, B. K. Moon, B. C. Choi, J. H. Jeong and C. Guo, CrystEngComm, 2013, 15, 8255–8261.
37. X.-J. Zhang, J. Mater. Sci.: Mater. Electron., 2014, 25, 5496–5500.
38. J. Han, C. McBean, L. Wang, C. Jaye, H. Liu, D. A. Fischer and S. S. Wong, J. Phys. Chem. C, 2015, 119, 3826–3842.
39. S. P. Culver and R. L. Brutchey, Dalton Trans., 2016, 45, 18069–18073.
40. R. Lei, D. Deng, X. Liu, F. Huang, H. Wang, S. Zhao and S. Xu, Opt. Mater. Express, 2018, 8, 3023–3035.

---