Multicolor and Warm White Emissions with a High Color Rendering Index in a Tb³⁺/Eu³⁺-Codoped Phosphor Ceramic Plate

Abstract

A series of Tb³⁺/Eu³⁺-codoped phosphor ceramic plates with a high color rendering index (CRI) for a near-ultraviolet light emitting diode (LED) were fabricated. Color emission can be tuned from green to reddish as a function of Eu³⁺ concentration. By doping only 0.15 mol% of Eu³⁺ concentration, a comfortable warm white emission is promoted as a result of simultaneous emissions of Tb³⁺ and Eu³⁺ ions. A theoretical model is proposed to calculate the contributions of the emitted color of the donor (Tb³⁺) and acceptor (Eu³⁺) ions in terms of europium concentration. The energy transfer from Tb³⁺ to Eu³⁺ ions is corroborated by the luminescence spectra and decay time of Tb³⁺, with a maximum energy transfer efficiency of 76% for 28 mol% of Tb³⁺ and 14 mol% of Eu³⁺. Warm white LEDs were constructed using a 380 nm UV chip and showed a CRI of 82.5, which was one of highest values reported for Tb³⁺/Eu³⁺-codoped samples. Color-correlated temperature (CCT), color coordinate (CC), and luminous efficacy (LE) were utilized to know the potentials as a phosphor converter in solid-state lighting.

1. Introduction

In the last decade, phosphor-converted light emitting diodes (pcLEDs) have been introduced into the market to conquer the deficiencies of fluorescent and incandescent lamps. Current pcLEDs are basically a mixture of a phosphor converter and a silicon binder placed in the top of a blue LED chip (InGaN), and the combination of both colors yields white light. However, silicon binders that support the phosphor still present innate issues, such yellowing problems when exposed to a high density of energy caused by their low thermal conductivities [1]. In that sense, luminescence glasses, glass ceramics, phosphor in glass, and ceramic phosphor plates have emerged as good candidates to replace the silicon binder in high-power LEDs [2,3,4,5]. Among these approaches, special attention has concentrated on ceramic phosphor plates because of their superior thermal stability over different approaches for solid-state lighting [6]. It exhibits robustness, a homogenous luminescence emission, chemical stability, ageing resistance, and good thermal conductivity. The last property is key for the development of stable high-power LEDs and can be controlled through network formers, network modifiers, and network intermediators [7,8]. Even though this system shows attractive thermal and optical properties, most of the high-power systems using one single phosphor do not meet the characteristics for satisfactory color rendering index (CRI) and low color-correlated temperature (CCT) for residential lighting applications [9,10]. In that sense, a second phosphor to compensate the weak red emission of YAG:Ce³⁺ has been introduced, and the combination of both phosphors leads to warm white devices with low CCT and improved CRI [11].

Terbium (Tb³⁺) and europium (Eu³⁺) ions show prominent emission peaks in green and red regions, respectively. Such ions have been often used as phosphors in cathode ray tubes, plasma display panels, X-ray scintillation screens, and fluorescent lamps [12]. Recently, Eu³⁺-activated phosphors have been proposed as a new red emitter to compensate for the weak red emission of YAG:Ce³⁺-based pcLEDs. Results show that by choosing a proper host matrix, Eu³⁺ presents superior photometric properties compared to commercial CASN:Eu²⁺ phosphors under blue and UV excitation [13,14]. The narrow reabsorption of Eu³⁺ is the main advantage over nitride phosphors.

Tb³⁺/Eu³⁺-codoped materials such as glass, glass ceramic, and phosphor powder have been demonstrated to produce warm and cold white emission through energy transfer from Tb³⁺ to Eu³⁺ ions. Indeed, color emission and full conversion from green to red can be achieved by changing the ratio of the Tb³⁺ and Eu³⁺ concentration [15,16,17,18,19,20,21]. However, energy transfer efficiency and radiative decay probability depend strongly on the phonon energy of the host. Therefore, different promising host materials, such as oxides, oxysulfides, oxynitrides, chloride, and fluoride, have been investigated to develop efficient phosphors [22,23,24,25,26]. In that sense, special attention should be given to inorganic fluorides because of their low phonon energies that promote a decrease in the nonradiative decay probability. Presently, an NaYF₄ fluoride crystal is one of the most efficient materials to produce upconversion emission as well as good chemical stability. However, tetragonal-phase LiYF₄ crystals exceed the optical properties of NaYF₄ because they promote stronger upconversion emissions and generate UV emission lines [27]. The crystalline phases of several LiYF₄ samples have been synthetized by thermal decomposition, sol-gel, and Bridgman methods [28,29,30]. However, with these approaches it is impossible to sinter a phosphor plate. In that sense, a melt-quenching method shows greater potential than the methods mentioned above to synthetize phosphor plates with the desired shape and crystalline phase. These properties make LiYF₄ an excellent host for Tb³⁺/Eu³⁺ ions to be used in solid-state lighting applications.

In spite of several works that show that Tb³⁺/Eu³⁺-codoped materials are good candidates for solid-state lighting applications, most of the published reports only present spectroscopic and structural properties. However, luminous efficacy, lighting properties, and device prototypes based in Tb³⁺/Eu³⁺-codoped material for solid-state lighting have been seldom reported [31]. In this work, the color tuning of a Tb³⁺/Eu³⁺ ceramic phosphor plate as a function of Eu³⁺ concentration has been investigated. Color coordinate and color rendering indexes were measured to evaluate the potential of the phosphor converter in solid-state lighting.

2. Materials and Methods

Several LiY_1−x−yF₄:yTb³⁺/xEu³⁺ ceramic phosphor plates were prepared by a melt-quenching technique (where y = 28 mol% and x = 0.15, 0.3, 1.5, 8.5, and 14 mol%). In the composition, 5 mol% extra of TeO₂ was used as a sintering additive to promote the formation of a ceramic plate. The total lanthanide concentration in the host matrix and other optical properties are show in

Table 1. Decay time, energy transfer, color coordinate (CC), color-correlated temperature (CCT), color rendering index (CRI), and luminous efficacy (LE) of the ceramic phosphor plate as a function of Eu³⁺ content.

Tb3+/Eu3+ (mol%)	τTb(5D4) (ms)	ET (%)	x	y	CCT (K)	CRI	LE (lm/W)
28	4.3	0	0.369	0.602	5497	34.06	13.08
28/0.15	4.28	0.4	0.459	0.521	3658	74.81	9.22
28/0.3	4.28	0.4	0.482	0.500	3136	82.54	8.13
28/1.5	3.44	20	0.534	0.453	2225	82.59	7.09
28/8.5	1.42	66	0.614	0.381	1136	42.81	7.80
28/14	0.93	78	0.629	0.368	1294	35.54	6.04

. All reactants were 99.99% pure and used as received. The reactants were tellurium oxide (TeO₂), lithium fluoride (LiF), yttrium fluoride (YF₃), terbium fluoride (TbF₃), and europium fluoride (EuF₃). Calculated quantities of the chemicals were mixed in an agate mortar for 30 min and melted in a PDI electric furnace at 1100 °C for 1 h in alumina crucibles so that a homogeneously mixed melt was obtained. The reaction performance was 96%, indicating a low volatility of fluoride compounds. The samples were subsequently annealed at a temperature of 280 °C. The time to finish the annealing process took around 18 h. Samples were cut and then polished to 800 µm-thick slabs for different measurements. Photoluminescence characterizations were performed using a xenon lamp, monochromator 2300i from Acton Research (Trenton, NJ, USA) and R955 Hamamatsu photomultiplier tube (Hamamatsu, Japan). The decay profile (lifetime) corresponding to 545 nm and 613 nm was recorded using a pulsed UV LED (Opulent Americas, Raleigh, NC, USA) centered at 365 nm and a Teledyne LeCroy oscilloscope (HDO 5055, New York, NY, USA). Special care was taken to maintain the alignment of the set-up in order to compare the intensity of the visible signal between different characterized samples. The ceramic phosphor plate was placed on a commercial UV LED with an emission wavelength centered in 380 nm. An integrating sphere 1 m in diameter (Labsphere Co., North Sutton, NH, USA) was used to measure the CCT, CIE (Comission Internationale de l´Éclairage 1931) chromaticity diagram, color coordinates, and luminous efficacy (LE) with a bias current of 20 mA.

The crystalline structures of the samples were characterized using the X-ray diffraction (XRD) of Bruker instrument (D2 Phaser, Bruker, Billerica, MA, USA) equipment with Cu Kα radiation at 1.54184 Å. The recorded XRD diffractograms were obtained from 10 to 70° 2θ range with increments of 0.02° and a sweep time of 0.5 s. The SEM images were performed using a SEM of JEOL (JSM-7800F, Tokyo, Japan).

3. Results and Discussion

3.1. X-ray Diffraction (XRD) Characterization and Photoluminescence Properties

The composition and phase purity of Tb³⁺-doped and Tb³⁺/Eu³⁺-codoped samples with different rare earth concentrations were analyzed by XRD. As presented in Figure 1a, almost all diffraction peaks from the ceramic phosphor plate could be indexed to the standard tetragonal LiYF₄ phase (PDF#77-0816) with low impurity phases, indicating that the samples were successfully crystallized, and TeO₂ concentrations did not cause significant changes in the host structure. According to Figure 1a, it was confirmed that LiYF₄ samples possessed a tetragonal crystal structure with space group I41/a. The obtained results were comparable to those reported by Kim et. al., where the tetragonal phase appeared as the main crystalline phase when the Eu³⁺ concentration was below 40 mol% in the LiYF₄ matrix [28]. For this study, the total lanthanide concentrations in LiYF₄ samples were 28, 28.15, 28.3, 29.5, 36.5, and 42 mol%. Additional impurity peaks appeared from 28 to 29.5 mol%, which were associated with a Y₂Te₆O₁₅ (PDF#37-1393) phase, whereas from 36.5 to 42 mol% the peak centered at 2θ = 28.2° disappeared, and the purity of the tetragonal phase increased. The lanthanide concentration also modified the diffraction peaks, shifting to a high angle side, when the lanthanide increased from 28 mol% to 42 mol%. Such a fact was associated with the substitution of larger ionic radii of Tb³⁺ and Eu³⁺ by a smaller Y³⁺ ionic radius. An SEM image of 28 mol% of Tb³⁺, 28 mol% of Tb/0.15 mol% of Eu, and 14 mol% of Eu³⁺ are shown in Figure 1b. The increase in Eu³⁺ concentration did not promote a significant change on the surface of the phosphor plate.

Figure 2a shows the photoluminescence excitation (PLE) from 325 to 500 nm in Tb³⁺- and Eu³⁺-doped ceramic plates monitored at 544 and 614 nm, respectively. The spectrum shows four main excitation bands centered at 353, 373, 378, and 485 nm, which were assigned to ⁷F₆→⁵D₂, ⁷F₆→⁵L₁₀, ⁷F₆→⁵D₃, and ⁷F₆→⁵D₄ transitions of Tb³⁺ respectively [29]. The ⁷F₆→⁵D₃ transition showed two excitation peaks at 373 and 378 nm, with the 378 nm shoulder being slightly weaker. From 331 to 338 nm there was a continuous excitation, while from 388 to 475 nm no excitation peaks were observed in Tb³⁺-doped samples. Figure 2b shows the photoluminescence emission (PL) of Tb³⁺-doped and Eu³⁺-doped ceramic phosphor plates under 373 and 396 nm excitation wavelengths. The bands centered at 490, 545, 589, and 622 nm were assigned to ⁵D₄→⁷F₆, ⁵D₄→⁷F₅, ⁵D₄→⁷F₄, and ⁵D₄→⁷F₃ transitions of Tb³⁺ respectively. These visible bands were the result of the well-known down-conversion process, and their emitted colors depended on the concentration of Tb³⁺ ions. In this work, the Tb³⁺ concentration was varied systematically (not show here) from 2 to 28 mol%, and results showed that there was no quenching concentration indicium. These results were similar to those found by other research groups, where 40 mol% of dopant ions were incorporated, and the green emission (545 nm) was the dominant color for the LiYF₄ matrix. The inset in Figure 2a shows the picture of the opaque ceramic phosphor plate.

To distinguish the PLE spectra of Tb³⁺ and Eu³⁺, a 14 mol% of Eu³⁺-doped sample was synthesized. Figure 2a shows the excitation spectra of the Eu³⁺-doped ceramic phosphor plate recorded at 614 nm emission. The sample showed five dominant bands centered at 362, 380, 393, 414, and 464 nm, which were assigned to ⁷F₀→⁵D₄, ⁷F₀→⁵L₇, ⁷F₀→⁵L₆, ⁷F₀→⁵D₃, and ⁷F₀→⁵D₂ transitions of Eu³⁺ respectively. Upon excitation at 393 nm, a magenta color appeared, with a spectral range from 585 to 710 nm, which are shown in Figure 2b. The 592, 614, 653, and 701 nm bands were attributed to ⁵D₀→⁷F₁, ⁵D₀→⁷F₂, ⁵D₀→⁷F₃, and ⁵D₀→⁷F₄ transitions of Eu³⁺, respectively, where the 592 and 614 nm emission bands were the feature emissions for LiYF₄:Eu³⁺. Among these transitions, the electric dipole ⁵D₀→⁷F₂ transition was the most intense, followed by less intense magnetic dipole ⁵D₀→⁷F₁ transitions. This indicated that Eu³⁺ ions were located at noninversion symmetric sites [32]. An attractive detail of the emission spectrum was observed at the 701 nm band, which typically was 40% less intense than the 614 nm band in the LiYF₄ matrix [28]. However, in this work the 701 nm band was debilitated, as it was 78% weaker than 614 nm band. For solid-state lighting applications, the 701 nm band is a waste of energy because the eye sensitivity is zero, whereas 614 nm bands are considered the optimal red emission to obtain high luminous efficacy and high CRI in warm white LEDs [33].

Figure 3 shows the excitation spectra of the Tb³⁺/Eu³⁺-codoped ceramic phosphor plate. By adding only 0.15 mol% of Eu³⁺ to the Tb³⁺-doped sample, the intensity ratios between splitting peaks at 373 and 378 nm changed slightly, where the shoulder at 378 nm was more intense. This fact was due to the spectral overlapping of the excitation bands at 373 and 380 nm of Tb³⁺ and Eu³⁺ ions, respectively. Such bands became more pronounced when the Eu³⁺ concentration increased from 0.05 to 5 mol%. From the point of view of solid-state lighting, the red shift of the excitation bands became significant because the optical power in the commercial LED chip at 380 nm was higher than 365 nm, and there was a lower probability of dispersion at larger wavelengths. In addition, upon 378 nm excitation, both Tb³⁺ and Eu³⁺ were excited efficiently (vertical gray line), producing simultaneously green and red emission bands. As a result, a higher luminous flux was obtained from a phosphor converter device. As presented in Figure 2b, the excitation bands at 378, 393, 464, and 485 nm for Tb³⁺/Eu³⁺-codoped samples increased with the Eu³⁺ concentration; however, the intensity of the band at 393 nm depended strongly on Eu³⁺ content rather than the other bands.

Figure 4 shows the emission spectra of Tb³⁺/Eu³⁺-codoped samples as a function of Eu³⁺ concentration. Simultaneous emissions from Tb³⁺ and Eu³⁺ were observed under 378 nm, which indicated the existence of energy transfer between Tb³⁺ and Eu³⁺. Intensity of emission bands of Tb³⁺ at 490, 545, and 589 nm decreased monotonically as the Eu³⁺ concentration increased from 0.15 mol% to 14 mol%. Among these bands, the 545 nm emission was the most influenced band by Eu³⁺ content, which decreased 36% of the initial intensity with only 0.15 mol% of Eu³⁺. Concurrently, an increment of 33% was observed for the 614 nm emission band of Eu³⁺ ions. Figure 4 shows that almost all energy was transferred from Tb³⁺ to Eu³⁺ when the Eu³⁺ reached 14 mol%, where red was the main color.

3.2. Rate Equation Model and Energy Transfer

To clarify the emissions corresponding to Tb³⁺ and Eu³⁺ ions, the following simplified model was proposed. (1) First, some Tb³⁺ ions were promoted from the ground state ⁷F₆ to the excited ⁵D₃ level as a result of pumping at 373 nm. The absorption rate was denoted by A₀₂ (s⁻¹). (2) Once some Terbium ions were in the ⁵D₃ level, they relaxed nonradiatively to the ⁵D₄ level; this multiphonon relaxation was denoted by A₂₁ (s⁻¹). Subsequently, two processes could occur, phonon relaxation A₁₀ (s⁻¹) or energy transfer from Tb³⁺ to Eu³⁺ ions (W). (3) Tb³⁺ emission wavelength peaks were 490, 545, 589, and 622 nm. (4) The energy transfer promotes some Eu³⁺ ions from the ⁷F₀ ground state to ⁵D₀ level where the emission rates B₁₀ (s⁻¹) of Eu³⁺ occurred in the 592, 614, 653, and 701 nm peaks, as is shown in Figure 5.

Therefore, the following ratio equations are proposed:

$$\frac{d{N}_{Tb}^2}{dt}=A_{02}{N}_{Tb}^0-A_{21}$$

(1)

$$\frac{d{N}_{Tb}^1}{dt}=A_{21}{N}_{Tb}^2-A_{10}{N}_{Tb}^1-W{N}_{Eu}^0{N}_{Tb}^1$$

(2)

$$\frac{d{N}_{Eu}^1}{dt}=-B_{10}{N}_{Eu}^1+W{N}_{Eu}^0{N}_{Tb}^1$$

(3)

where $N_{Tb}^2$, $N_{Tb}^1$, and $N_{Tb}^0$ (ions/cm³) are the Tb³⁺ ion populations in the ⁵D₃, ⁵D₄, and ⁷F₆ energy levels, respectively. $N_{Eu}^1$ and $N_{Eu}^0$ (ions/cm³) are the Eu³⁺ ion populations in the ⁵D₀ and ¹F₀ levels, respectively. For low-excitation pumping, the ground populations are proportional to the nominal concentration, i.e., $N_{Tb}^0\approx N_{Tb}$ and $N_{Eu}^0\approx N_{Eu}$. In stationary conditions the solutions are:

$$N_{Tb}^1=\frac{A_{02}{N}_{Tb}}{\left(W{N}_{Eu}+A_{01}\right)}$$

(4)

$$N_{Eu}^1=\frac{W{A}_{02}{N}_{Tb}{N}_{Eu}}{B_{10}\left(W{N}_{Eu}+A_{01}\right)}$$

(5)

$$N_{Tb}^2=\frac{N_{Tb}{A}_{02}}{A_{21}}$$

(6)

Then, ion populations are related to the emission spectrum by:

$$N_{Tb}^1=k{{\displaystyle \int }}^{\text{}}{I}_{Tb}d\lambda$$

(7)

$$N_{Eu}^1=k{{\displaystyle \int }}^{\text{}}{I}_{Eu}d\lambda$$

(8)

where I_Tb and I_Eu are the emission spectra related to the Tb³⁺ and Eu³⁺ emission transitions, and k is a proportional constant. Thus, to obtain the I_Tb and I_Eu emission spectra, deconvolution of the spectra was performed for both ions, as is shown in Figure 6. Then, we define:

$$P_{Tb}=\frac{N_{Tb}^1}{N_{Tb}^1+N_{Eu}^1}=\frac{B_{10}}{W{N}_{Eu}+B_{10}}$$

(9)

$$P_{Eu}=\frac{B_{21}{N}_{Eu}^1}{N_{Tb}^1+N_{Eu}^1}=\frac{W{N}_{Eu}}{W{N}_{Eu}+B_{10}}$$

(10)

$$P_{Tb}=\frac{{{\displaystyle \int }}^{\text{}}{I}_{Tb}d\lambda }{{{\displaystyle \int }}^{\text{}}{I}_{Tb}d\lambda +{{\displaystyle \int }}^{\text{}}{I}_{Eu}d\lambda }$$

(11)

$$P_{Eu}=\frac{{{\displaystyle \int }}^{\text{}}{I}_{Eu}d\lambda }{{{\displaystyle \int }}^{\text{}}{I}_{Tb}d\lambda +{{\displaystyle \int }}^{\text{}}{I}_{Eu}d\lambda }$$

(12)

On the other hand, the dynamic solution for Tb³⁺ in the level ⁵D₄ is:

$$N_{Tb}^1\left(t\right)=\frac{A_{21}{A}_{02}{N}_{Tb}^0}{A_{21}-A_{10}-N_{Eu}^{0}W}\left(e^{-\left(A_{01}+W{N}_{Eu}^0\right)t}-e^{-A_{21}t}\right)$$

(13)

This equation has two terms related to the rise time and lifetime. In this way, the lifetime is expressed by:

$$\tau =\frac{1}{A_{01}+W{N}_{Eu}^0}$$

(14)

Figure 7 shows the experimental lifetime curves, and the inset graph is the experimental fitting of Equation (14), with A₀₁ = 226.8/s and W = 53.4/s mol%. In our samples, the lifetimes of the ⁵D₄ level of Tb³⁺ showed values of 4.3, 4.28, 4.28, 3.44, 1.42, and 1.01 ms when Eu³⁺ concentration increased to 0, 0.15, 0.3, 1.5, 8.5, and 14 mol%, respectively (see Table 1).

Figure 8 shows the experimental normalized emission spectra according to Equations (11) and 12. Additionally, adjustment of the experimental data was carried out using Equations (9) and (10), with B₁₀ = 131.2/s. The model simultaneously adjusted the emission spectrum and lifetime curves. Luminescent efficiency can be defined by the ratio of population loss by emission and the population gain of the level:

$$\eta =\frac{A_{10}{N}_{Tb}^1}{A_{21}{N}_{Tb}^2}$$

(15)

Substituting Equations (4) and (5) we find:

$$\eta =\frac{A_{01}}{A_{01}+W{N}_{Eu}}$$

(16)

Energy transfer efficiency can be defined by the ratio of population loss by energy transfer and the population gain of the level:

$$ET=\frac{W{N}_{Tb}^1{N}_{Eu}^0}{A_{21}{N}_{Tb}^2}=\frac{W{N}_{Eu}}{A_{01}+W{N}_{Eu}}$$

(17)

Equations (16) and (17) fulfill the following relationship:

$$ET+\eta =1$$

(18)

Then, energy transfer can be rewritten as:

$$ET=1-\frac{\tau }{{\tau }_0}$$

(19)

where $\tau ={\left(A_{01}+W{N}_{Eu}\right)}^{-1}$ is the lifetime, and ${\tau }_0={\left(A_{01}\right)}^{-1}$ is the radiative lifetime. In this work, $\tau ={\tau }_{Tb-Eu}$ and ${\tau }_0={\tau }_{Tb}$ were the fluorescence lifetimes of the ⁵D₄ level of Tb³⁺ for Tb³⁺-doped and Tb³⁺/Eu³⁺-codoped ceramic phosphor plates. The calculated ET efficiency increased from 5 to 76% when the Eu³⁺ concentration increased from 0.15 to 14 mol%. ET increased rapidly to 30% with the addition of 1.5 mol% of Eu³⁺; after this concentration, no big changes were observed, and the ET was kept almost constant at 14 mol% of Eu³⁺.

3.3. White Light Device Fabrication

Figure 9 shows multicolor light devices that were constructed using the Tb³⁺/Eu³⁺-codoped ceramic phosphor plate and the 380 nm UV LED chip.

The electroluminescence of fabricated devices as a function of Eu³⁺ concentration with a bias current of 20 mA is shown in Figure 10. Representative samples with 0.15, 0.3, and 1.5 mol% of Eu³⁺ clearly showed 380 (UV LED), 544 (Tb³⁺), 592 (Eu³⁺), and 614 nm (Eu³⁺) bands, where warm white was the feature color emissions of these devices. The maximum luminous efficacy was measured to be 13.08 lm/W for 28 mol% of Tb³⁺, whereas a decrement from 9.22 to 6.04 lm/W was observed as the Eu³⁺ concentration increased (see Table 1). The low values of luminous efficacy were associated with the opacity of the samples, the low efficiency of UV LED (InGaN), as well as the poor contribution of color from LED.

Figure 11 shows the color coordinates of the samples under study. The values were located on the edge of the chromaticity diagram, predominantly in yellow and red regions. The emitted color of the device changed from green to red through warm white by keeping the Tb³⁺ concentration and changing the Eu³⁺ content. The obtained results showed that the optical properties of LED were strongly influenced by Eu³⁺ content. Interestingly, by doping with only 0.15 mol% of Eu³⁺, it was possible to modify both CRI and CCT. CRI showed an increase from 34 to 74.81, and CCT diminished from 5497 to 3658 K; these features were very adequate for indoor lighting. The samples with 0.3 and 1.5 mol% of Eu³⁺ gave the highest CRIs of approximately 82.5 and 82.6 with CCTs of 3136 and 2225 K, respectively. Although the luminous efficacy of the devices was low, we expected to further increase such values by introducing Tb³⁺ and Eu³⁺ in an adequate matrix or by codoping with Ce³⁺ to increase the absorption strength. The obtained results showed that by choosing properly the Eu³⁺ concentration, it was possible to produce comfortable white light devices for vivid applications in daily life.

4. Conclusions

In summary, the fabrication of LiYF₄ ceramic phosphor plates doped with Tb³⁺/Eu³⁺ ions is reported. The maximum excitation peak of a single Tb³⁺ was at 373 nm; however, this changed to 378 nm when LiYF⁴:10Tb³⁺ was codoped with Eu³⁺ ions. Based on the experimental results, it was concluded that intensity ratios between emission bands could be tuned by choosing properly the ion concentrations of both Tb³⁺ and Eu³⁺ ions. By placing the LiYF₄:28Tb³⁺/yEu³⁺ (mol%) ceramic phosphor plates on the top of 380 nm LED chip, green, warm, and red color emissions were obtained. It was found that warm white was achieved by adding only 0.15 mol% of Eu³⁺ without serious detriment to the luminous efficacy. However, when the concentration of Eu³⁺ increased to 0.3 mol% in LiYF₄:28Tb³⁺/yEu³⁺ (mol%), a CRI of 82.3 and a CCT of 3136 K were measured. With an increase in Eu³⁺ concentration, the yellow and red bands were improved, but the blue and green bands were reduced. Then, it was necessary to compromise CRI, CCT, and luminous efficacy to define the ion concentration. The obtained CRI was one of the highest reported in the literature for the Tb³⁺/Eu³⁺ system, which, in combination with low CCT, made LiYF₄ ceramic phosphor plates a good candidate for solid-state lighting applications.