1. Introduction

Trivalent rare earth ions (RE³⁺) doped fluoride upconversion (UC) luminescent materials REF₃ and ALnF₄ (A = Alkali, Ln = RE), which can achieve anti-Stokes emission, have aroused wide concerns of researchers in the last decade due to its unique optical properties in the fields of optical components, volumetric displays, vivo imaging, solar cells and non-contact temperature sensor [1–10]. Compared with the traditional oxide hosts and most inorganic matrices, the fluoride-based crystals possess numerous advantages, such as low phonon energy and toxicity, high chemical and thermal stability, long luminescence lifetimes and superior luminescent quantum yields [11, 12]. In particular, NaYF₄ was once considered to be an efficient UC matrix among various materials that have been studied [13]. With the similar crystal structure to NaYF₄ and a smaller ion radius (Y³⁺ = 0.89 Å, Lu³⁺ = 0.85 Å), NaLuF₄ is expected to be more stable and efficient as an ideal UC matrix [14]. NaLuF₄ holds two crystalline structures, including the cubic (α-) and the hexagonal (β-) phases, and the former can be transformed into the latter under certain reaction conditions. Li et al. has reported that different crystal phase could be obtained by controlling the solution PH values, F^- sources and organic additives, and the growth mechanism has been studied systematically [15]. Lin’s group demonstrated a citric acid-assisted hydrothermal method to synthesize Gd³⁺ doped NaLuF₄:Yb³⁺, Er³⁺, Tm³⁺ nano- and micro-crystals with enhanced UC emission, and the phase can be seriously affected by Gd³⁺ doping, NaF content, reaction time and temperature [16]. These works provide guidance for us to manipulate the crystal morphology. UC materials with white emission show tremendous application value in white light-emitting diodes (WLEDs) filed, which is ascribed to its superiority of long lifespan, high luminescent efficiency, low energy consumption and environmental friendliness [17]. Compared with the most commonly used blue-emitting YAG:Ce³⁺-based phosphors, multiple phosphor system has a higher color rendering index (CRI) due to its green emission, but at a much lower cost [18]. Niu et al. firstly prepared well-defined NaLuF₄:Yb³⁺, Er³⁺/Tm³⁺/Ho³⁺ micro-crystalline through a molten-salt process, and white UC light was achieved via tuning the concentration of Yb³⁺ [19]. Gao et al. synthesized well-crystalline hexagonal NaYF₄:Yb³⁺, Ho³⁺, Tm³⁺ nanoparticles by sol-gel method with multicolor emission, and white light output with the CIE color coordinate (0.325, 0.320) was obtained by adjusting RE³⁺ dopant content [20]. Their efforts present us with a method of color modulation. However, the color of UC fluorescence can be notably influenced by pump power and ambient temperature. In order to achieve stable white-light output, it is necessary to overcome the drawback of light color drift when the laser power changes or environment temperature fluctuates. Hence, crystals with excellent color stability are in urgent need of introduction. By contrast, there were few reports available on color drift of NaLuF₄ matrix.

In this paper, we adopted a facile procedure for the fabrication of α- or β-NaLuF₄:Yb³⁺, Ho³⁺, Tm³⁺ microstructures, using oleic acid as the surfactant. The possible transformation mechanism for the phase and the evolution process of morphology are discussed in detail. Furthermore, the UC properties of different prepared samples under the excitation of a 980 nm diode laser (LD) were exhaustively investigated. By tuning sensitizer (Yb³⁺) concentration, the relative intensities of blue, green and red color bands were changed and the white emission was acquired. Moreover, the chromaticity coordinates of the crystal with white-light output maintained in the white region of CIE-1931 software when the pump power and sample temperature were adjusted in a broad range, which indicated excellent stability of color for NaLuF₄-based samples. This unique property makes β-NaLuF₄:Yb³⁺, Ho³⁺, Tm³⁺ microstructure to be a potential candidate for 3D display and stable WLEDs light source.

2. Experimental

All the chemicals in the experiment were used as received without further purification. Lu(NO₃)₃, Ho(NO₃)₃, Tm(NO₃)₃ and Yb(NO₃)₃ stock solutions were prepared by dissolving the corresponding RE(NO₃)₃·6H₂O (RE = Lu, Ho, Tm, Yb) (99.99%) in deionized water. The remaining drugs and reagents used were all of analytical purity, such as NaOH, NH₄F, anhydrous ethanol and oleic acid.

2.1 Synthetic procedures

A series of NaLuF₄:Yb³⁺, Ho³⁺/Tm³⁺ samples with various dopants were prepared via an OA-assisted hydrothermal method. In a typical process, NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺, for example, 10 mL anhydrous ethanol was added into 2 mL aqueous solution containing 600 mg NaOH. Next, 15 mL OA was poured slowly into the solution. Then, 0.745 mmol Lu(NO₃)₃ (0.5M, 1.49 mL), 0.005 mmol Ho(NO₃)₃ (0.01M, 0.5mL), 0.05 mmol Tm(NO₃)₃ (0.01M, 5mL) and 0.2 mmol Yb(NO₃)₃ (0.1M, 2mL) were added into the mixture under magnetic stirring and stirred for 30 min. 9 mmol NH₄F (2M, 4.5mL) was added into the above system to form precursor solution. The precursor solution was stirred vigorously for another 30 min and then transferred to an autoclave kept at 180 °C for 12 h. The product was collected by centrifugation and washed with anhydrous ethanol and deionized water for several times before dried at 80 °C for 4 h. The dosage of different chemicals and reaction time could be adjusted according to the subjects we studied.

2.2 Characterization

The phases of prepared powder samples were identified by powder X-ray diffraction meter, model Empyrean (Netherlands). The micrographs of samples were measured on a scanning electron microscope (SEM, SU-70). UC luminescence spectra were recorded on HORIBA Jobin Yvon iHR550 Spectrometers.

2.3 Spectrum-test system

The optical path of spectral measurement system is shown in Fig. 1. Different from conventional yellow phosphors emitted by blue LED, the sample fixed on the heating stage is excited by 980 nm LD, duo to without response to 405 nm LD. The thermocouple thermometer is close to the laser spot but not directly exposed to light, so that not only the temperature derived from both the heating stage and laser beam can be accurately measured, but also the laser beam on sample is not blocked by the thermocouple probe. When measuring the pump power, the heating stage is removed and the sample is replaced by an optical power probe. The power density is easily adjusted through the moveable lens L1.

 

Fig. 1 The optical path of spectral measurement system.

Download Full Size | PDF

3. Results and discussion

3.1 Crystalline structure and morphology

In this work, the impacts of NH₄F content and reaction time on the crystal structure of NaLuF₄ were investigated, and the possible evolution of morphology was proposed.

3.1.1 Effect of NH₄F

The phase and morphology were strongly influenced by the NH₄F content. Figure 2 illustrates the typical XRD patterns of RE³⁺ (RE³⁺ = Yb³⁺, Ho³⁺) codoped NaLuF₄ samples that obtained at 180 °C for 8 h with different amount of reactant NH₄F as well as the standard data of pure cubic and pure hexagonal phases marked with vertical bars. It can be clearly seen that the phase of prepared samples undergoes a conversion from cubic to hexagonal. For sample achieved with 4 mmol NH₄F, only cubic phase can be recognized. When NH₄F is 6 mmol, the cubic phase dominates the crystalline phase but the extremely faint hexagonal phase peaks around 30.2°, 31.1°, 47.0° and 54.1° also can be distinguished. When the consumption of NH₄F is 9 mmol, the diffraction peaks of hexagonal phase are dominantly formed and the weak peaks of cubic phase just can be observed at a diffraction angle of 28.6°. When the dosage of NH₄F is increased further to 12 mmol, the peaks of α-NaLuF₄ disappear and only peaks of β-NaLuF₄ are demonstrated in XRD patterns. Thus, the pure hexagonal phase is obtained by adjusting the amount of NH₄F. The peak at 38.84° for NaF can be removed from samples by washing and it has no effect on luminescence performance.

 

Fig. 2 XRD patterns of samples prepared with different amounts of NH₄F at 180 °C for 8 h (* stands for the diffraction peaks of NaF).

Download Full Size | PDF

Correspondingly, the morphology of NaLuF₄ changed with the amount of NH₄F as shown in Fig. 3. When the NH₄F content is increased from 4 to 6 mmol, the morphologies of the samples displayed in Figs. 3(A) and 3(B) alter from irregular nanoparticles to nanospheres, and the average size of the nanospheres is slightly larger than that of the former. As the NH₄F content continuously increases, the microrods form and dominate the morphology gradually. When the amount of NH₄F is 9 mmol, the related morphology consists of a great number of microrods and nanospheres and while the dosage reaches up to 12 mmol, the nanospheres disappear gradually and may lead to the pure microrods, as shown in Figs. 3(C) and 3(D). On the basis of above analysis, it can be seen that the increase of NH₄F content facilitates the transformation of the phase from cubic to hexagonal and the morphology from nanoparticles to microrods.

 

Fig. 3 SEM images of the samples prepared with different amounts of NH₄F: (A) 4 mmol, (B) 6 mmol, (C) 9 mmol, (D) 12 mmol at 180 °C for 8 h.

Download Full Size | PDF

3.1.2 Effect of the reaction time

Similar to NH₄F, the longer reaction time also favors the prefabrication of hexagonal β-NaLuF₄. The XRD patterns and homologous SEM images are depicted in Fig. 4 and Fig. 5, respectively. As shown in Fig. 4, the sample synthesized with the reaction time of 4 h comprises both α and β phases, while the intensity of diffraction peaks for β-NaLuF₄ is obviously stronger than that of α-NaLuF₄. When prolonging the reaction time to 8 h, the intensities of diffraction peaks for cubic phases rapidly reduce while the phase of β-NaLuF₄ becomes dominant. The weak cubic phase diffraction peaks can be observed around the angle of 28.4° and 47.1°. Note that we scale each XRD pattern to different sizes to make the patterns as clear as possible, causing different absolute intensities for the samples 9 mmol in Fig. 2 and 8 h in Fig. 3 prepared at identical condition. We can find that the ratios of peaks intensities at 17° to 30° for the two patterns are all about 2:3, indicating high reproducibility for our samples. When the reaction time is increased to 12 or 20 h, pure hexagonal β-NaLuF₄ can be successfully obtained. The stray peaks at 38.84° belong to NaF, which can be removed by washing repeatedly. Thus, samples prepared at 180 °C for 12 or 20 h can be treated as pure β-NaLuF₄.

 

Fig. 4 XRD patterns of samples prepared with 9 mmol NH₄F at 180 °C for different reaction time (* stands for the diffraction peaks of NaF).

Download Full Size | PDF

 

Fig. 5 SEM images of the samples prepared with 9 mmol NH₄F at 180 °C for different reaction time: (A) 4 h, (B) 8 h, (C) 12 h, (D) 20 h.

Download Full Size | PDF

The evolution of morphology coincides with the XRD patterns. When the reaction time is 4 or 8 h, microrods attached with a mass of nanoparticles can be obtained and nanoparticles become less as the reaction time increases. If the reaction time reaches 12 or 20 h, only pure microrods can be observed. Therefore, prolonging the reaction time contributes to the synthesis of pure β-NaLuF₄ microrods.

3.1.3 Evolution mechanism of the structure

In the above study, conclusion that increasing the dosage of NH₄F and the reaction time promotes the evolution of both phase and morphology has been drawn. In terms of its essence, the change of the phase brings about the conversion of morphology to a great degree. It’s generally acknowledged that the α-NaLuF₄ is a metastable phase at higher temperature and usually formed as an intermediate product in the synthesis process of thermodynamically stable β-NaLuF₄. In experiment, certain reaction conditions are required to break the energy obstacle between the cubic and hexagonal phases when synthesizing β-NaLuF₄. For NH₄F, serving as reactant and mineralizer in the synthesis procedure, the facilitation effect in transformation process can be summed up in two aspects. On the one hand, excess F^- can reduce the chemical potential and its crystallization temperature, owing to massive F^- coated on and interacted with the metal surface, which lowers the energy obstacle exists in the cubic-to-hexagonal evolution process and make it easier to form stable hexagonal phase [21, 22]. On the other hand, the NH₄⁺ cation in the solution can be selectively absorbed on the different lattice planes. Through the efficient interactions between NH₄⁺ cations and F^- anions on the surface, the growth rate varies in different direction, causing the alteration of morphology [23]. Therefore, when NH₄F is increased from 4 to 6 mmol, due to the isotropic structure of α-NaLuF₄, the microstructure converts from nanoparticles to nanospheres for the sake of decreasing the surface energy. When NH₄F reaches 9 or 12 mmol, the energy obstacle is low enough to be broken and the self-assembly process in which α-NaLuF₄ nanoparticles form β-NaLuF₄ microrods can be realized. Besides, attached by NH₄⁺, the highly anisotropic β-NaLuF₄ samples grow into microrods. Thus, the more NH₄F added into the initial solution, the more drastically phase transforms. Analogously, prolonging reaction time also benefits the destruction of the energy obstacle standing in the transformation procedure and induces the generation of pure β-NaLuF₄ [24].

Since a higher efficiency of hexagonal phase to cubic, the samples studied in the following investigations are all pure β-NaLuF₄. Considering the above experimental results, energy conservation as well as the possibility of unexpected impurities generate with superfluous NH₄F and breakage of microrods when reaction time is too long, β-NaLuF₄ studied next are prepared with 12 mmol NH₄F at 180 °C for 20 h.

3.2 UC luminescence properties

Our study on the stability of white-light emissions for NaLuF₄-based white phosphor commences with the investigation of UC emission in RE³⁺ (RE³⁺ = Yb³⁺, Ho³⁺/Tm³⁺) binary and ternary doping system. Figure 6 displays the typical UC properties of Yb³⁺-Ho³⁺ or Yb³⁺-Tm³⁺ codoped β-NaLuF₄ samples under the excitation of 980 nm LD with an average power density of 35 mW/mm². Figure 6(A), the spectra of β-NaLuF₄:x%Yb³⁺, 1%Ho³⁺ (x = 5, 10, 15, 20), exhibits an intense green emission band centered at 541 nm and two weak red emission bands centered at 646 nm and 751 nm that magnified 5 times in Fig. 6(A). These peaks correspond to ⁵F₄/⁵S₂→⁵I₈, ⁵F₅→⁵I₈ and ⁵F₄/⁵S₂→⁵I₇ transition of Ho³⁺ and the positions do not move with the Yb³⁺ amount. Fixed the concentration of Ho³⁺ at 1%, the intensities of emission bands all enhance with the increase of Yb³⁺ content from 5% to 15%. The emission intensities decrease when the Yb³⁺ concentration is up to 20%, which ascribes to the concentration quenching effect [25]. For the Yb³⁺-Ho³⁺ codoped samples, the red peaks are too weak in comparison with the green peak due to nonradiative relaxation of ⁵F₄/⁵S₂→⁵F₅ transition and low probability of ⁵F₄/⁵S₂→⁵I₇ radiation, respectively. Figure 6(B) shows the UC luminescence spectra of β-NaLuF₄:x%Yb³⁺, 0.5%Tm³⁺ (x = 5, 10, 15, 20) samples. As demonstrated, a strong blue emission band centered at 475 nm, a weak blue emission at 450 nm, and a weak red emission at 646 nm can be recognized, which originate from ¹G₄→³H₆, ¹D₂→³F₄, and ¹G₄→³F₄ transitions, respectively. The emission intensity around 475 nm increases sharply while other bands rise slowly when Yb³⁺ doping concentration changes from 5% to 20% and the quenching effect can never be observed, differing from that of Ho³⁺ doped materials.

 

Fig. 6 UC luminescence spectra of (A) β-NaLuF₄:x%Yb³⁺, 1%Ho³⁺ (x = 5, 10, 15, 20) and (B) β-NaLuF₄:x%Yb³⁺, 0.5%Tm³⁺ (x = 5, 10, 15, 20).

Download Full Size | PDF

As is known to all, in the unsaturated UC process, for a fixed pump spot the emission intensity and the pump power satisfies the following Eq. (1):

 

Fig. 7 The dependence of UC luminescence spectra on pump power of β-NaLuF₄:Yb³⁺, Ho³⁺ and β-NaLuF₄:Yb³⁺, Tm³⁺ and ions energy level diagram together with possible electronic transition mechanisms. UC spectra of (A) β-NaLuF₄:Yb³⁺, Ho³⁺, (C) β-NaLuF₄:Yb³⁺, Tm³⁺, (B, D) corresponding log-log plots of UC emission intensity versus the pump power, (E) ions energy level diagram together with possible electronic transition mechanisms.

Download Full Size | PDF

Based on the generation of different colors light emission from Yb³⁺-Ho³⁺ and Yb³⁺-Tm³⁺ codoped NaLuF₄ samples, it is possible to realize white light output in tridoped NaLuF₄:Yb³⁺, Ho³⁺, Tm³⁺ nanoparticles. Figure 8(A) depicts the room-temperature UC spectra of NaLuF₄:x%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ (x = 5, 10, 20) ternary system with variable sensitizer concentrations under an average power density of 35 mW/mm². Compared with the codoped samples, all the luminous bands of tridoped system can be easily matched with the electronic transition processes. The blue peaks at 450 nm and 475 nm are bound up with the transitions of Tm³⁺. The green emission at 541 nm and red emission at 751 nm are associated with Ho³⁺. It is worth noting that a new red peak of Tm³⁺ appears at 696 nm originated from ³F₂/³F₃→³H₆, which is ignored in Yb³⁺-Tm³⁺ system due to its very weak intensity. The ⁵F₅→⁵I₈ transition of Ho³⁺ and ¹G₄→³F₄ of Tm³⁺ all make contributions to the red band at 646 nm, owing to the distinction of spectra shapes in contrast with binary dopant particles. With the increase of Yb³⁺ concentration, the UC fluorescence intensities of two blue bands increase on account of more Yb³⁺ ions which participate in populating ¹G₄ and ³F₂/³F₃ levels via three- and two-photon processes, respectively. Different from blue emissions, the obviously enhanced red emission at 696 nm is ascribed to the efficient CR processes (¹D₂ + ³H₆→³H₄ + ³F₂ and ¹G₄ + ³F₄→³H₄ + ³F₃) between two Tm³⁺ ions under high activator concentration [31]. The intensity of green band increases before Yb³⁺ content is up to 10% due to the energy transfer from Yb³⁺, and decreases with further Yb³⁺ concentration due to the CR process between Ho³⁺ and Yb³⁺:⁵S₂,⁵F₄ (Ho³⁺) + ⁷F_7/2 (Yb³⁺)→⁵I₆ (Ho³⁺) + ⁵F_7/2 (Yb³⁺) [32]. The red light emission at 646 nm decreases when the Yb³⁺ doping concentration increases. The possible population mechanism for 646 nm of Ho³⁺ consists of at least four parts. Firstly, a nonradiative relaxation process of ⁵S₂,⁵F₄→⁵F₅ in which the transition probability increases when raising Yb³⁺ concentration can populates ⁵F₅ level and strengthens the red emission of Ho³⁺. Secondly, ⁵I₆ level is populated with more electrons with the increase of Yb³⁺ dosage, and the CR process mentioned above between Ho³⁺ and Yb³⁺ also populates ⁵I₆ energy level. After a nonradiative transition of ⁵I₆→⁵I₇, ⁵F₅ level can be filled in electrons transferred from ⁵I₇ and the red band of Ho³⁺ enhances. Thirdly, a CR process of ³H₄ + ³H₆→³F₄ + ³F₄ between Tm³⁺ ions can populate ³F₄ level and then the electrons can be excited to ³F₂,³F₃ level to generate red emission when transferring to the ground state ³H₆. Furthermore, ⁵I₇ level of Ho³⁺ may be populated by the very efficient resonance energy transfer process of ³F₄ (Tm³⁺)→⁵I₇ (Ho³⁺) and this path also makes contributions to the emission band at 646 nm [33]. However, in the case of a high concentration of Tm³⁺, Yb³⁺ available for Ho³⁺ may be rare due to the competition between Tm³⁺ and Ho³⁺. Therefore, the process of red emission at 646 nm is extremely complex. In Fig. 8(B), all the transition routes mentioned above are presented. The UC emission color of tridoped samples can be tuned via adjusting Yb³⁺ concentration and Fig. 8(C) displays the CIE chromaticity coordinate diagram with different Yb³⁺ doping concentration. Increasing the Yb³⁺ concentration, the CIE color coordinates of the prepared samples move from yellow to white. For the white-emission sample of NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺, the coordinate is (0.333, 0.330) at room temperature, a point very close to the standard white light (0.33, 0.33).

 

Fig. 8 (A) Room-temperature UC spectra, (B) energy level diagrams of Yb³⁺, Ho³⁺ and Tm³⁺ ions together with the proposed transition paths, and (C) CIE chromaticity coordinate diagram of prepared β-NaLuF₄:x%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ (x = 5, 10, 20) nanoparticles. The arrow in CIE chromaticity diagram represents the color coordinates movement trend with the increase of Yb³⁺. The red band at 751 nm was neglected in calculation, given that it is relatively weaker, too close to near infrared (NIR) region to make contributions to light color and it overlapped with NIR band at 800 nm of Tm³⁺. The arrow represents the movement trend of the color coordinates as the temperature rises.

Download Full Size | PDF

Pump power of excitation laser plays a significant role in color adjustment. Figure 9(A) exhibits several typical room-temperature UC spectra of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ sample excited by 980 nm LD with different pump power. The area of laser spot is increased to 12.6 mm² to avoid damage for triple-doped samples. As known, the responses of different color emission bands to the pump power vary in UC materials, making it possible to tune the visual color by controlling the excitation power. From Fig. 9(A), it can be seen that the intensity of blue and green emission change scarcely under higher excitation while that of red keeps the upward trend, causing a movement towards the red region for chromaticity coordinates. Figure 9(B) shows the dependence of UC luminescence intensity on power. From the log-log diagram, the slopes for different color bands are 1.73, 1.79, 1.13, 1.26 and 1.49, indicating that two-photon process associates with the white emission. All the slopes are lower than that in binary dopant β-NaLuF₄ samples, which may be ascribed to the lower energy transfer efficiency resulting from more cross-relaxation and back-energy-transfer processes at high concentration of Tm³⁺. Figure 9(C) displays the CIE chromaticity coordinates and the corresponding trend of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ under various powers. The calculated coordinates shift from yellowish to white when power is changed from 100 mW to 255 mW and then locate in the white region when power is increased from 255 mW to 1560 mW. Compared with other materials, as shown in Table 1, the tri-doped samples we prepared exhibits satisfactory color stability, lower threshold pump power density and lower optimal excitation pump power density for white-emission. These findings reveal great potential applications in the field of display and white light-emitting diodes.

 

Fig. 9 The UC properties of β-NaLuF₄:Yb³⁺, Ho³⁺, Tm³⁺ with tunable visual color output by adjusting the pump power. (A) Room-temperature UC spectra of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ sample excited by 980 nm LD with various output, (B) corresponding log-log plots of the UC emission intensity versus the pump power in triply-doped sample, (C) corresponding CIE chromaticity coordinate diagram. The arrow in CIE chromaticity diagram represents the movement trend of the color coordinates as the pump power increases.

Download Full Size | PDF

Table 1. Pump Power Density for White Emission

View Table

Ambient temperature has dramatic effects on nonradiative relaxation process and resonance energy transfer process and further influences the population of corresponding excited states, leading the variation of color. Figure 10 shows the trend that color changes with temperature of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ sample excited by 980 nm laser with the power of 0.4 W, and the corresponding power density is about 10 W/cm². From Fig. 10(A), it can be found that all the emission intensities decrease when the temperature increases from 300 K to 448 K, due to the incremental probability of nonradiative relaxation when temperature rises. Figure 10(B) presents the ratios of the intensities of different color bands to the total visible light intensity. With the temperature increasing, the ratios of blue and green intensities reduce and that of red enhances, which results in a shift of CIE chromaticity coordinates towards the red region as is shown in Fig. 10(C). The CIE coordinates move from (0.295, 0.286) to (0.366, 0.301) and always locate in the white region. This study indicates good color stability over a wide range of temperature, which can, to a certain extent, suppress the color-drift of WLEDs.

 

Fig. 10 The UC luminescence properties of β-NaLuF₄:Yb³⁺, Ho³⁺, Tm³⁺ with tunable visual color output via manipulating the ambient temperature. (A) UC spectra and (B) the ratios of different color intensity and (C) corresponding CIE chromaticity coordinate diagram of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ sample at various ambient temperature. The arrow in CIE chromaticity diagram represents the movement trend of the color coordinates as the temperature rises.

Download Full Size | PDF

4. Conclusions

In summary, we have successfully synthesized cubic and hexagonal NaLuF₄ via a simple hydrothermal method using oleic acid as surfactant. The investigation of reaction parameters indicates that the NH₄F dosage and reaction time have a significant effect upon the phase and morphology of the products. Increasing the amount of NH₄F and prolonging the reaction time both favor the cubic to hexagonal transformation of phase and the nanoparticles-nanospheres-microrods evolution of morphology. Excited by 980 nm diode laser, Yb³⁺-Ho³⁺ codoped samples emit bright green and weak red lights while Yb³⁺-Tm³⁺ codoped samples produce intense blue and faint red emissions, respectively. The responses of different color bands to the Yb³⁺ concentration and pump power are various, which makes it possible to generate tunable multicolor emissions in Yb³⁺-Ho³⁺-Tm³⁺ triply-doped β-NaLuF₄ samples. By adjusting the Yb³⁺ content of β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ ternary dopant system, the visual color turns from yellow to yellowish and eventually falls in the white region when the Yb³⁺ concentration reaches 20%. For the white-emitting β-NaLuF₄:20%Yb³⁺, 0.5%Ho³⁺, 5%Tm³⁺ sample, as the excitation power increases, the CIE coordinates also move from yellow to white and then stay white when the power exceeds 255 mW. In addition, the emission color maintains in white region when the ambient temperature changes. The results prove great capacity of color controlling and satisfactory color stability of the white light-emitting samples, leading potential applications in numerous fields, especially the stable white-light output LED and color display.