Energy transfer and color tuning in the Ba₉Sc₂Si₆O₂₄:Ce³⁺,Eu²⁺,Mn²⁺ phosphor

Abstract

A series of single phase Ba_9−x−ySc_2−zSi₆O₂₄:xCe,yEu,zMn phosphors was synthesized by high-temperature solid-state reactions. The luminescent properties of single-doped, co-doped and tri-doped Ce³⁺/Eu²⁺/Mn²⁺ in the host Ba₉Sc₂Si₆O₂₄ were studied on the basis of the photoluminescence excitation/emission spectra, photoluminescence decay behaviors and time resolved emission spectra. Two types of luminescent centers of Eu(1)/Eu(2,3) as well as Ce(1)/Ce(2,3) occupying different crystallography sites were confirmed, and the sensitizing effects of Ce³⁺ → Eu²⁺, Ce²⁺ → Mn²⁺ and Eu²⁺ → Mn²⁺ were clearly observed. In the tri-doped samples, Eu²⁺ competed with Mn²⁺ for the energy transferred from Ce³⁺ and also transferred its observed energy to Mn²⁺. Furthermore, the tri-doped Ba₉Sc₂Si₆O₂₄ phosphors emitting white light with various correlated color temperatures were obtained by adjusting the relative content of Ce³⁺, Eu²⁺ and Mn²⁺.

---

1. Introduction

White-light-emitting diodes (w-LEDs) have been a focus as the fourth generation light source, because they offer the advantages of compactness, energy saving, high brightness, long operational lifetime, and environmental friendliness.^1–3 Commercial w-LEDs usually consist of blue emitting LED chips (460–470 nm) and a Y₃Al₅O₁₂:Ce³⁺ (YAG:Ce³⁺) yellow phosphor. However, this type of w-LEDs has the deficiencies of a low color rendering index (CRI) and high correlated color temperature (CCT), which were induced by the scarcity of red emission component in the devices. To solve these problems, w-LEDs fabricated by ultraviolet (UV) LED chips, i.e. AlInGaN chips (300–360 nm) coated with blue/green/red (RGB) tri-emission-band phosphors have attracted increasing attention.^4,5 The emission color of these UV excited w-LEDs depends only on the phosphors, thus they have the advantages of good color stability and color reducibility.^6,7 However, it is a challenge to find appropriate phosphors that absorb UV LED emission and efficiently emit broad spectra.

Recent studies have found that rare earth ions Ce³⁺- and Eu²⁺-doped silicates can absorb UV LED emission and exhibit blue and green bands. In addition, Mn²⁺ co-doped with Ce³⁺/Eu²⁺ can create efficient red band emission with the assistance of energy transfers, Ce³⁺ → Mn²⁺ and Eu²⁺ → Mn²⁺. Herein, Ce³⁺ or Eu²⁺ that transfer their absorbed energy to other ions is called a sensitizer, whereas Mn²⁺ receiving energy is called the activator. Owing to the difference between Ce³⁺ and Eu²⁺ on their valences and electronic configurations, the energy of the d → f transition of Ce³⁺ is usually higher than that of Eu²⁺, when they are doped into the same lattice site in a specific phase.⁸ In this case, it is possible that Ce³⁺ can also transfer its absorbed energy to Eu²⁺.^9,10 As a result, energy transfer may occur in the ion pairs Ce³⁺ → Eu²⁺, Ce³⁺ → Mn²⁺, and Eu²⁺ → Mn²⁺, and tri-doping Ce³⁺, Eu²⁺ and Mn²⁺ into one host may cause multiple energy transfer processes Ce³⁺ → Eu²⁺ → Mn²⁺ and result in tri-band emissions of Ce³⁺, Eu²⁺ and Mn²⁺.¹¹

Ba₉Sc₂Si₆O₂₄, as a novel phosphor host, was first studied in 2009 by Nakano et al.¹² In our previous study, some efforts have been made to synthesize Ce³⁺ or Eu²⁺ singly-doped Ba₉Sc₂Si₆O₂₄ and Ce³⁺/Mn²⁺ or Eu²⁺/Mn²⁺ co-doped Ba₉Sc₂Si₆O₂₄ phosphors for w-LED applications.^13–16 However, to the best of our knowledge, there has been no research on Ce³⁺/Eu²⁺ co-doped or Ce³⁺/Eu²⁺/Mn²⁺ tri-doped Ba₉Sc₂Si₆O₂₄. In this study, to clearly analyze the energy transfer processes in the Ce³⁺/Eu²⁺/Mn²⁺ tri-doped phosphors, we synthesized Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ phosphors with different values of x, y and z and analyzed their photoluminescence (PL) excitation/emission spectra, PL decay behaviors and time resolved emission spectra. Herein, we revealed the three energy transfer processes, Ce³⁺ → Eu²⁺, Ce³⁺ → Mn²⁺, and Eu²⁺ → Mn²⁺, calculated their energy transfer efficiencies, and then a tri-band emission phosphor of Ba₉Sc₂Si₆O₂₄:Ce³⁺,Eu²⁺,Mn²⁺ with white emission color was obtained, which might have good performance for w-LED applications.

2. Experimental

Powder samples of Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ (including singly-doped samples: Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺, Ba_8.91Sc₂Si₆O₂₄:0.09Eu²⁺, Ba₉Sc_1.91Si₆O₂₄:0.09Mn²⁺; co-doped samples: Ba_8.82Sc₂Si₆O₂₄:0.09Ce³⁺,0.09Eu²⁺, Ba_8.91Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.36Mn²⁺, Ba_8.91Sc_1.64Si₆O₂₄:0.09Eu²⁺,0.36Mn²⁺; tri-doped samples: Ba_8.81−ySc_1.64Si₆O₂₄:0.09Ce³⁺,yEu²⁺,0.36Mn²⁺, y = 0.03–0.27) were synthesized by conventional high-temperature solid state reactions. Raw materials, BaCO₃ (analytical reagent, A.R. for short), Sc₂O₃ (99.9%), SiO₂ (A.R.), CeO₂ (99.9%), Eu₂O₃ (99.9%) and MnO₂ (A.R.), were weighed in suitable amounts and mixed thoroughly in an agate mortar. In each sample, 2 wt% H₃BO₃ (A.R.) was added as flux. The mixtures were pressed into pellets and fired at 1350 °C for 3 h in a reducing atmosphere of 5% H₂/95% N₂.

The phase purity of the samples was checked using a Rigaku Dmax 2000 X-ray diffractometer (XRD, 45 kV, 100 mA) with Cu Kα radiation (λ = 0.15418 nm). The PL excitation/emission spectra and internal quantum efficiencies (IQE) were measured using an Edinburgh Instruments FLS920 lifetime and a steady state fluorescence spectrometer equipped with a continuous Xenon lamp (450W). To measure the PL temperature dependence, a subcooled liquid Helium cooling system was employed. The color coordinates (x, y) and the correlated color temperature (CCT) of the samples were calculated using the emission spectrum data. The PL decay curves and time resolved emission spectra were also recorded on a FLS920 spectrometer with a nanosecond flash lamp as the light source for the Ce³⁺ and Eu²⁺ emissions, whereas for Mn²⁺ emission, a microsecond flash lamp was employed. The lifetimes and energy transfer efficiencies (η) were calculated using the PL decay data.

3. Results and discussion

3.1 Phase analysis

Some selected XRD patterns for the Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ series phosphors are presented in Fig. 1. Most of the patterns agree well with the standard patterns of Ba₉Sc₂Si₆O₂₄ (JCPDS 82-1118) shown in Fig. 1(f). Some patterns contain a few weak peaks from a minor phase of Ba₂SiO₄, which may not significantly interfere with the luminescence efficiencies. The XRD results indicate that doping Ce³⁺, Eu²⁺ and Mn²⁺ ions does not cause any significant change in the host structure, which clearly suggests that these doped ions as sensitizers and activators are incorporated in the crystal lattice.

Fig. 1 Selected XRD patterns in the Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ series phosphors and the pattern for JCPDS file 82-1119.

Ba₉Sc₂Si₆O₂₄ has a rhombohedral unit cell with the space group R3̄ (no. 148) and cell parameters a = 9.8716(2) Å and c = 21.9376(7) Å.¹⁰ The crystal structure of Ba₉Sc₂Si₆O₂₄ is depicted in Fig. 2(a) and (b) on different directions. In the structure, the Ba²⁺ ions have three crystallographic sites [Fig. 2(c)]: Ba(1) is 12-fold coordinated by O with an average Ba–O bond-length of 3.047 Å; Ba(2) and Ba(3) are 9-fold and 10-fold coordinated by O with the average Ba–O bond-lengths 2.899 Å and 2.903 Å, respectively. Ba(2)–O and Ba(3)–O have similar bond lengths, but they are clearly shorter than that of Ba(1)–O; thus, it is convenient to classify the Ba(2) and Ba(3) sites together as Ba(2,3) (Fig. 2c). Herein, when the rare earth ions (R), i.e., Ce³⁺ or Eu²⁺ ions, are doped into the lattice, which may occupy the Ba sites, they may be classified as two types luminescent centers R(1) and R(2,3), respectively. However, Sc³⁺ takes one crystallographic site 6-fold coordinated by O [Fig. 2(a) and (b)]. According to the similarity of the ionic radius, it is expected that Ce³⁺ (r = 1.39 Å, CN = 10) and Eu²⁺ (r = 1.49 Å, CN = 10) randomly occupy the Ba²⁺ (r = 1.66 Å, CN = 10) sites, and Mn²⁺ (r = 0.81 Å, CN = 6) prefers to occupy the Sc³⁺ (r = 0.85 Å, CN = 6) site in Ba₉Sc₂Si₆O₂₄, whereas the charge may be balanced by cations or O vacancies.¹⁷

Fig. 2 Structure of Ba₉Sc₂Si₆O₂₄ along the b axis (a) and c axis (b), and the coordination geometries of Ba(1), Ba(2) and Ba(3) sites (c).

3.2 Luminescence of Ce³⁺, Eu²⁺ or Mn²⁺ single-doped Ba₉Sc₂Si₆O₂₄

Fig. 3 shows the excitation and emission spectra of Ce³⁺ and Eu²⁺ singly doped phosphors Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺ and Ba_8.91Sc₂Si₆O₂₄:0.09Eu²⁺, respectively. The broad emission band peaking at 390 nm is attributed to the 5d → 4f transition of Ce³⁺ and that at 510 nm was due to the 5d → 4f transition of Eu²⁺. As we know, Ce³⁺ has a doublet ground level (²F_5/2 and ²F_7/2, ΔE ≈ 2000 cm⁻¹), which should have a shouldered emission band when doped in one lattice site. However, for simplification, the emission band of each luminescence center was fitted approximately using a single Gaussian band. The energy separation of the two fitted bands a (26 670 cm⁻¹) and b (23 530 cm⁻¹) is about 3140 cm⁻¹, clearly larger than 2000 cm⁻¹. Therefore, it is reasonable to assign the two bands to the different luminescent centers rather than to the doublet ground level. It is mentioned above that the bond-length of Ba(1)–O is larger than that of Ba(2,3)–O. Thus, the coordination field for the Ce(1) center is weaker than that for the Ce(2,3) center, which result in the lower edge of the excitation d band for the Ce(1) center being higher than that for the Ce(2,3) center. Therefore, the fitting bands a and b were assigned to the emissions of Ce(1) (∼375 nm) and Ce(2,3) (∼425 nm), respectively. In addition, as Eu²⁺ has only one ground level, fitting bands c and d are simply assigned to the emissions of Eu(1) (∼460 nm) and Eu(2,3) (∼515 nm). The Ba(2) and Ba(3) sites have a similar coordination number and a similar Ba–O bond-length, thus the emissions of Ce³⁺ and Eu²⁺ in these two sites cannot be well separated due to the similar wavelength. Therefore, Ce(2,3) and Eu(2,3) are used to represent the rare earth activators occupying these two sites. The suitable spectral overlap of the excitation band of Eu²⁺ to the emission band of Ce³⁺ indicates the possibility of energy transfer from Ce³⁺ to Eu²⁺. The emission of the Mn²⁺ singly-doped Ba₉Sc₂Si₆O₂₄ is very weak to be detected because of its forbidden d–d transition (⁴T₁ → ⁶A₁). According to our previous study and other recent reports,^15,18,19 some Mn²⁺ have may have emission at ∼620 nm (red) when excited in the range 350–550 nm (near-UV to green), which is overlapped with the emission bands of Ce³⁺ and Eu²⁺. Herein, the Ce³⁺/Eu²⁺/Mn²⁺ tri-doped Ba₉Sc₂Si₆O₂₄ phosphor might show three-band emissions, owing to the different luminescent centers and the energy transfer processes among them.

Fig. 3 Excitation (EX) and emission (EM) spectra of Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺ (dash lines) and Ba_8.91Sc₂Si₆O₂₄:009Eu²⁺ (solid lines). The dotted lines show the fitting bands of two activators at different sites: Ce(1) and Ce(2,3) as well as Eu(1) and Eu(2,3), respectively.

Fig. 4 shows the time resolved emission spectra of Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺ and Ba_8.91Sc₂Si₆O₂₄:0.09Eu²⁺, as well as the emission spectra (normalized) of the Ce³⁺ and Eu²⁺ doped samples at selected decay times. Fig. 4(a) and (c) show that for the Ce³⁺ doped sample, the emission intensity decreases with decay time and the spectrum profile changes. It is observed that the emission in the longer wavelength range decays slower, which means that the luminescence of the Ce(2,3) center has a longer lifetime than that of the Ce(1) center. Two Gaussian bands [peaking at 375 nm for Ce(1) and at 425 nm for Ce(2,3)] were used to fit the Ce³⁺ spectrum at 70 ns. The lifetime of the Ce(1) and Ce(2,3) centers were calculated from the decay data at 350 nm and 450 nm, respectively. At each selected wavelength, the emission was contributed mainly by one luminescent center. The decay behaviors can be fitted by a single-exponential equation:^20,21

I(t) = I₀ exp(−t/τ) (1)

where I(t) and I₀ are the luminescence intensities at t and 0 time, and τ is the lifetime. The results indicate that the lifetime of Ce(1) is 21 ns and that of Ce(2,3) is 34 ns. Similarly, the emission intensity of the Eu²⁺ doped sample also decreased and the spectrum profile changed with the decay time, as shown in Fig. 4(b) and (d). The emission in the longer wavelength decay was slower, which also indicates that the emission of the Eu(2,3) center has a longer lifetime than that of the Eu(1) center. The calculated lifetimes were ∼230 ns for Eu(1) (measured at 460 nm) and ∼571 ns for Eu(2,3) (at 550 nm). Two Gaussian bands (peaking at ∼460 nm and ∼515 nm) were used to fit the Eu²⁺ spectrum at 196 ns, which are correlated to the luminescent centers Eu(1) and Eu(2,3), respectively.

Fig. 4 Time resolved emission spectra of Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺ (a) and Ba_8.91Sc₂Si₆O₂₄:0.09Eu²⁺ (b); the emission spectra (normalized) for the Ce³⁺ (c) and Eu²⁺ doped (d) samples at selected decay time. Gaussian fittings were applied to the Ce³⁺ spectrum at 70 ns and Eu²⁺ spectrum at 196 ns (fitted curves belong to Ce(1) and Ce(2,3) as well as Eu(1) and Eu(2,3), respectively).

3.3 Luminescence of Ce³⁺/Eu²⁺, Ce³⁺/Mn²⁺ and Eu²⁺/Mn²⁺ co-doped Ba₉Sc₂Si₆O₂₄

Fig. 5 shows the excitation and emission spectra as well as the time resolved emission spectra of Ba_8.82Sc₂Si₆O₂₄:0.09Ce³⁺,0.09Eu²⁺, Ba_8.91Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.36Mn²⁺, and Ba_8.91Sc_1.64Si₆O₂₄:0.09Eu²⁺,0.36Mn²⁺, respectively. For easy comparison and analysis, all the PL emission spectra and time resolved emission spectra were excited with 330 nm UV light. In Fig. 5(a), the weak band peaking at 390 nm belongs to the emission of Ce³⁺ and the strong band peaking at 470 nm belongs to the emission of Eu²⁺ in the Ba_8.82Sc₂Si₆O₂₄:0.09Ce³⁺,0.09Eu²⁺ phosphor, which means that in the Ce³⁺/Eu²⁺ co-doped phosphor, the Ce³⁺ emission is clearly reduced. On the other hand, the emission band of Ce³⁺, as shown in Fig. 3, overlaps with the excitation band of Eu²⁺ for the Ce³⁺ and Eu²⁺ singly doped phosphors. Therefore, energy transfer from Ce³⁺ to Eu²⁺ may occur. The time resolved emission spectra in Fig. 5(b) show that the Eu²⁺ emission (460 nm) increases at first and then decreases with decreasing Ce³⁺ emission (390 nm), which also supports the energy transfer from Ce³⁺ to Eu²⁺. In the Ce³⁺ and Eu²⁺ co-doped phosphor, the Eu²⁺ emission mainly shows the character of the Eu(1) center. The excitation spectra of Mn²⁺ have similar profiles to that of Ce³⁺ and Eu²⁺ in the Ce³⁺/Mn²⁺ and Eu²⁺/Mn²⁺ co-doped phosphors shown in Fig. 5(c) and (e), indicating an energy transfer process from Ce³⁺ and Eu²⁺ to Mn²⁺. Owing to its forbidden transition (⁴T₁ → ⁶A₁), Mn²⁺ normally has a much longer lifetime than Ce³⁺ and Eu²⁺. Herein, only the decay time prolongs to 2 ms, and the Mn²⁺emission band peaking at ∼620 nm can be observed. However, at such long decay time, the emissions of Ce³⁺ and Eu²⁺ all decayed very weakly [see Fig. 5(d) and (f)], which further supports the energy transfer from Ce³⁺ and Eu²⁺ to Mn²⁺. The decay rates of Ce³⁺ and Eu²⁺ were much faster than that of Mn²⁺. In other words, the energy transfer rates from Ce³⁺ and Eu²⁺ to Mn²⁺ are much faster than the decay rate of Mn²⁺. Therefore, we could not observe the phenomena that the Mn²⁺ emissions increase with decreasing Ce³⁺ and Eu²⁺ emissions.

Fig. 5 Excitation (EX) and emission (EM) spectra as well as time resolved emission spectra of Ba_8.82Sc₂Si₆O₂₄:0.09Ce³⁺,0.09Eu²⁺ (a and b), Ba_8.91Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.36Mn²⁺ (c and d) and Ba_8.91Sc_1.64Si₆O₂₄:0.09Eu²⁺,0.36Mn²⁺ (e and f), respectively.

As discussed previously, when Ce³⁺ or Eu²⁺ is doped into the Ba₉Sc₂Si₆O₂₄ host, it occupies three types of Ba crystallographic sites randomly, and forms three luminescent centers divided into two groups, such as Ce(1) (∼375 nm) and Ce(2,3) (∼425 nm) or Eu(1) (∼460 nm) and Eu(2,3) (∼515 nm).^13,14 To calculate the energy transfer efficiency of the co-doped Ba₉Sc₂Si₆O₂₄ phosphors, the decay curve of each group of the luminescent centers for the corresponding sample is detected at its particular wavelength under 330 nm excitation (data not shown). The lifetimes of each luminescent centers were calculated using eqn (1). The energy transfer efficiency (η) for the related luminescent centers can be calculated by²²

where τ_s0 is the decay time of sensitizer in the absence of an activator, and τ_s is the decay time of sensitizer in the presence of an activator. Table 1 lists the lifetime obtained by curve fitting with eqn (1) and energy transfer efficiency calculated by eqn (2) for the related centers in the selected samples. To reduce the interference of emission band overlap, the detected wavelength for some centers were not chosen at their peak maximum. For singly doped samples, the lifetime of Ce(1) is similar to that of Ce(2,3) and much shorter than those of Eu(1) and Eu(2,3). The sensitizers with a short lifetime can easily transfer their absorbed energy to the activators with long lifetime. For the Ce³⁺/Eu²⁺ and Ce³⁺/Mn²⁺ co-doped samples, the lifetimes of Ce(1) and Ce(2,3) become shorter than the Ce³⁺ singly doped sample, which demonstrates the energy transfer processes from Ce(1) and Ce(2,3) to Eu²⁺ and Mn²⁺. In addition, due to the energy transfer from Eu²⁺ to Mn²⁺, the lifetime of Eu²⁺ become shorter in the Eu²⁺ and Mn²⁺ co-doped sample. The data of the energy transfer efficiency indicate that it is easier for Ce³⁺ to transfer energy to Eu²⁺ [Ce(1) → Eu²⁺: η = 70%] than to Mn²⁺ [Ce(1) → Mn²⁺: η = 20%]. On the other hand, the energy transfer efficiencies of Ce(1) to Eu²⁺ and Mn²⁺ are higher than those of Ce(2,3); moreover, the energy transfer efficiency of Eu(1) to Mn²⁺ is higher than that of Eu(2,3). These phenomena are due to the shorter lifetime of Ce(1)/Eu(1) than that of Ce(2,3)/Eu(2,3).²³

Table 1 Lifetime of Ce³⁺, Eu²⁺ and Mn²⁺ at different sites in Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ (monitored at different wavelength λ_M) under 330 nm excitation and calculated energy transfer efficiency

Sample composition	Ce(1)		Ce(2,3)		Eu(1)		Eu(2,3)		Mn
	λM: 350 nm		λM: 450 nm		λM: 460 nm		λM: 550 nm		λM: 620 nm
	τ (ns)	η (%)	τ (ns)	η (%)	τ (ns)	η (%)	τ (ns)	η (%)	τ (ms)
x = 0.09, y = 0.00, z = 0.00	21		34
x = 0.00, y = 0.09, z = 0.00					230		571
x = 0.09, y = 0.09, z = 0.00	7	70	24	30	221		575
x = 0.09, y = 0.00, z = 0.36	19	20	28	15					23
x = 0.00, y = 0.09, z = 0.36					212	25	561	2	23
x = 0.09, y = 0.09, z = 0.36	6		14		183		575		23

---

3.4 Luminescence of Ce³⁺/Eu²⁺/Mn²⁺ tri-doped Ba₉Sc₂Si₆O₂₄

The data in previous sections show that in the Ba₉Sc₂Si₆O₂₄ host, Ce³⁺ has violet-bluish emission and can transfer its absorbed energy to Eu²⁺ and Mn²⁺; Eu²⁺ has green emission and can transfer its energy to Mn²⁺; and when co-doped with Ce³⁺ or Eu²⁺, Mn²⁺ has red emission. It is possible that if Ce³⁺, Eu²⁺ and Mn²⁺ are triply doped into the Ba₉Sc₂Si₆O₂₄ host, a three-emission-band phosphor with a white color may be obtained. Fig. 6 represents the excitation and emission spectra of Ba_8.82Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.09Eu³⁺,0.36Mn²⁺. The emission spectrum excited by 330 nm UV radiation has three bands peaking at 390 nm, 510 nm and 620 nm corresponding to Ce³⁺, Eu²⁺ and Mn²⁺, respectively. The excitation spectra monitored at 510 nm for Eu²⁺ emission and at 620 nm for Mn²⁺ emission have a clear profile of the excitation spectra for Ce³⁺ and Eu²⁺, which indicate the energy transfer of Ce³⁺ → Eu²⁺, Ce³⁺ → Mn²⁺ and Eu²⁺ → Mn²⁺, just like the situations for the co-doped phosphors discussed in Section 3.3. The lifetimes of Ce³⁺ and Eu²⁺ in the tri-doped phosphors are clearly shorter than those in the single-doped and co-doped phosphors (except for the lifetime of the Eu(2,3) centers) represented in Table 1, which also support the energy transfer of Ce³⁺ → Eu²⁺, Ce³⁺ → Mn²⁺ and Eu²⁺ → Mn²⁺. In addition, the excitation spectra of three emission bands coincide with the emission of the AlIGaN UV chips (300 360 nm). These results show that the Ce³⁺/Eu²⁺/Mn²⁺ tri-doped Ba₉Sc₂Si₆O₂₄ phosphors have suitable features for w-LED applications assembled with UV-LEDs.

Fig. 6 Excitation (EX) and emission (EM) spectra of Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺.

Fig. 7 shows the time resolved emission spectra of Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu³⁺, 0.36Mn²⁺ and the emission spectra of Ce³⁺ (∼390 nm), Eu²⁺ (∼510 nm) and Mn²⁺ (∼620 nm) can be detected at different decay times. The pure Ce³⁺ spectrum was firstly detected at 200 ns. At 250 ns, the Ce³⁺ emission began to decrease, but the Eu²⁺ emission appeared, which supports the energy transfer of Ce³⁺ → Eu²⁺. Owing to the very long lifetime, the Mn²⁺ emission appears very late, at 2 M ns (2 ms). The time resolved emission spectra reveal the processes that Ce³⁺ transfers energy to Eu²⁺, and Eu²⁺ transfers energy to Mn²⁺.

Fig. 7 Time-resolved emission spectra of Ba_8.82Sc_1.64Si₆O₂₄:0.09Ce³⁺, 0.03Eu²⁺, 0.36Mn²⁺ excited by 330 nm UV radiation; inset represents some typical spectra with various decay times.

Although the energy transfer processes of Ce³⁺ → Eu²⁺, Ce³⁺ → Mn²⁺ and Eu²⁺ → Mn²⁺ occur in tri-doped phosphors, the relationship between the relative intensities of three bands and the doping contents is complex, because three dopants compete to absorb the excitation energy. Fig. 8 shows the emission spectra of the selected samples in the system Ba_8.91−ySc_1.64Si₆O₂₄:0.09Ce³⁺,yEu²⁺,0.36Mn²⁺ (y = 0–0.24) under the excitation of 330 nm UV radiation. Without the Eu²⁺ dopant (y = 0), the phosphor shows the Ce³⁺ band (390 nm) and Mn²⁺ band (620 nm). With the Eu²⁺ dopant (y = 0.03), the Ce³⁺ band is reduced dramatically and the Eu²⁺ band appears at 510 nm. The three band spectrum gives the phosphor a white luminescence. With increasing Eu²⁺ content, the emission intensity of Ce³⁺ decreases further, and the emission intensity of Mn²⁺ also decreases, but the emission intensity of Eu²⁺ does not change very much. Therefore, when the Eu²⁺ content increases, the luminescence color is modified to a higher correlated color temperature (CCT), which may meet the various requirements of w-LEDs. The color coordinates and CCT of some tri-doped phosphors were calculated and the results are listed in Table 2 and illustrated in Fig. 9. For comparison, the color coordinates and CCT for the single-doped phosphors Ba_8.91Sc₂Si₆O₂₄:0.09Ce³⁺ and Ba_8.91Sc₂Si₆O₂₄:0.09Eu²⁺, as well as the co-doped phosphor Ba_8.91Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.36Mn²⁺ are also represented. The internal quantum efficiency (IQE) of the phosphors Ba_8.91−ySc_1.64Si₆O₂₄:0.09Ce³⁺,yEu²⁺,0.36Mn²⁺ were measured under 330 nm excitation and are listed in Table 2. A comparison of the co-doped sample no. 3 with the singly doped sample no. 1 showed that doping Mn²⁺ could increase the IQE of the phosphor, which might be due to the flux effect of the MnO doping. By carefully analyzing their PL emission data, we found that similar phenomena also appeared in Ca₉Gd(PO₄)₇:Eu²⁺,Mn²⁺ and Ca₉La(PO₄)₇:Eu²⁺,Mn²⁺ phosphors.^24,25 For tri-doped sample no. 4, multi-step energy transfer processes reduce the IQE due to energy loss during the transfers. Because IQE is strongly related to the preparation process, we will try to improve the IQE by extending the sintering time, adding different fluxes and post-processing the phosphors.

Fig. 8 Emission spectra of the selected phosphors in the Ba_8.91−ySc_1.64Si₆O₂₄:0.09Ce³⁺,yEu²⁺,0.36Mn²⁺ system.

Table 2 Color coordinates, CCT and QE of some selected phosphors in the Ba_9−x−ySc_2−zSi₆O₂₄:xCe³⁺,yEu²⁺,zMn²⁺ system

No.	Composition	Color coordinates		CCT (K)	IQE (%)
		x	y
1	x = 0.09, y = 0.00, z = 0.00	0.180	0.078		30.3
2	x = 0.00, y = 0.09, z = 0.00	0.164	0.349		19.3
3	x = 0.09, y = 0.00, z = 0.36	0.498	0.259		56.7
4	x = 0.09, y = 0.03, z = 0.36	0.439	0.352	2466	30.4
5	x = 0.09, y = 0.12, z = 0.36	0.407	0.372	3302	21.0
6	x = 0.09, y = 0.18, z = 0.36	0.371	0.389	4315	17.4
7	x = 0.09, y = 0.21, z = 0.36	0.353	0.399	4872	16.3

---

Fig. 9 Positions of luminescence color of the phosphor samples 1–4 on the chromaticity diagram (1931); the upper right bar represents the CCT of samples 4–7. The images of related phosphor samples under the excitation of 360 nm radiation are also shown. Sample no. is listed in Table 2.

One of the key requirements for good phosphors is the satisfied PL maintenance at the operating temperature of the device. To characterize the temperature-dependence of the PL properties, the PL emission spectra of the Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺ phosphor at various temperatures in the range 11 to 400 K were measured and are shown in Fig. 10. Owing to thermal quenching, the emission intensities decreased with increasing temperature. The emission bands of Ce³⁺ and Eu²⁺ showed almost no shift, whereas the Mn²⁺ band had slight blue-shift with increasing temperature. When the temperature approached 300–400 K (27–127 °C in the operating temperature range of LED devices), the Mn²⁺ emission band shifted to 610–620 nm, which is in the good red wavelength range. In addition, this blue-shift may lead to less reduction of the PL intensity because it shifts to the more sensitive wavelength of the human's eye. The inset in Fig. 10 shows the variations of the PL intensity of each activator and the total PL intensity with temperature. Below 250 K, the temperature-dependence of the total PL intensity was controlled mainly by the Eu²⁺ behavior; whereas above this temperature, it is mainly controlled by the Mn²⁺ behavior. Compared to the room temperature (300 K) data, at 400 K, the total PL intensity of the Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺ phosphor decreased by ∼33%; this slight decay indicates that this tri-doped phosphor can be a potential phosphor for high-powder w-LED applications.

Fig. 10 Temperature-dependent emission spectra of Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺ samples, and the inset shows the integrated intensities.

To better understand the PL temperature-dependent features of the Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺ phosphor, the PL data of each activator was analyzed by the Arrhenius equation:²⁶

We mainly analyzed the data in the temperature range 200–400 K. Actually, the data were analyzed (curve fittings) using the following modified format:

where I₀ is the initial PL intensity of the phosphor at 200 K and I_T is the intensity at a given temperature T (K). ΔE is the activation energy for thermal quenching; c is pre-exponential factor, a constant only decided by the nature of the luminescent centers; and k is Boltzmann's constant (8.62 × 10⁻⁵ eV). Only the data above 200 K were analyzed and the results are represented in Fig. 11. The Eu(1) and Eu(2,3) centers have a relatively high ΔE, which indicates that Eu(1) and Eu(2,3) might have good thermal stability. Nevertheless, the c value may also affect the thermal stability of the PL intensity. Because Eu(1) and Eu(2,3) have much higher c values than the Ce(1)/Ce(2,3) centers and Mn²⁺ center, their PL intensities decrease faster than the others.

Fig. 11 Plots of ln(I₀/I_T − 1) versus 1/kT for the Ba_8.88Sc_1.64Si₆O₂₄:0.09Ce³⁺,0.03Eu²⁺,0.36Mn²⁺ samples.

4. Conclusions

We synthesized a series of Ce³⁺/Eu²⁺/Mn²⁺ single-doped, co-doped and tri-doped Ba₉Sc₂Si₆O₂₄ phosphors and examined the PL excitation/emission spectra, the decay time and the time resolved emission spectra to confirm the energy transfer among them. The Ce(1) and Ce(2,3) centers with the shortest lifetime have the highest transfer energy efficiency to Eu(1)/Eu(2,3) for the wide range of overlapping emission spectra of Ce³⁺ to the excitation spectrum of Eu²⁺. Ce³⁺ and Eu²⁺ occupying Ba(1) always have a higher transfer efficiency to Mn²⁺ than that occupying Ba(2) and Ba(3). In the tri-doped phosphors, under UV (300–350 nm) excitation, Ce³⁺, Eu²⁺, Mn²⁺ give three emission bands at 390 nm, 510 nm, and 620 nm, respectively. Herein, Eu²⁺ can receive energy from Ce³⁺ and partly transfer its absorbed energy to Mn²⁺. In tri-doped phosphors, white light with a different CCT from 4527 K to 3172 K were obtained. The results suggest that the Ce³⁺/Eu²⁺/Mn²⁺ tri-doped Ba₉Sc₂Si₆O₂₄ phosphors can be used as UV pumped white LED phosphors.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (No. 21371015).

References

1. A. Kitai, Luminescent Materials and Applications, Wiley, New York, 2008.
2. A. Mills, III–Vs Rev., 2005, 18, 32.
3. M. Y. Chen, Z. G. Xia and Q. L. Liu, J. Mater. Chem. C, 2015, 3, 4197.
4. V. Adivarahan, A. Chitnis, J. P. Zhang, M. Shatalov, J. W. Yang, G. Simin, M. Asif Khan, R. Gaska and M. S. Shur, Appl. Phys. Lett., 2001, 79, 4240.
5. W. W. Wu and Z. G. Xia, RSC Adv., 2013, 3, 6051.
6. C. K. Chang and T. M. Chen, Appl. Phys. Lett., 2007, 90, 161901.
7. Y. S. Tang, S. F. Hu, C. C. Lin, N. C. Bagkar and R. S. Liu, Appl. Phys. Lett., 2007, 90(151), 108.
8. P. Dorenbos, J. Electrochem. Soc., 2008, 155(6), 143.
9. P. Dorenbos, J. Phys.: Condens. Matter, 2003, 15, 4797.
10. N. Lakshminarasimhan and U. V. Varadaraju, J. Electrochem. Soc., 2005, 152, H152.
11. S. Ye, X. M. Wang and X. P. Jing, J. Electrochem. Soc., 2008, 155(6), 143.
12. T. Nakano, Y. Kawakami, K. Uematsu, T. Ishigaki, K. Tada and M. Sato, J. Lumin., 2009, 129, 1654.
13. L. Bian, T. L. Zhou, J. J. Yang, Z. Song and Q. L. Liu, J. Lumin., 2012, 132, 2541.
14. L. Bian, F. F. Du, S. S. Yang, Q. Y. Ren and Q. L. Liu, J. Lumin., 2013, 137, 168.
15. L. Bian, T. Wang, Z. Song, Z. H. Liu, J. X. Li and Q. L. Liu, Chin. Phys. B, 2013, 22(7), 77801.
16. L. Bian, T. Wang, S. J. Liu, S. S. Yang and Q. L. Liu, Mate. Res. Bull., 2015, 64, 279.
17. R. D. Shannon, Acta Crystallogr., 1976, 32, 751.
18. H. K. Liu, L. B. Liao and Z. G. Xia, RSC Adv., 2014, 4, 7288.
19. F. L. Yang, W. An, W. D. Zhuang, G. S. Tian and X. P. Jing, J. Lumin., 2014, 146, 269.
20. C. J. Duan, Z. J. Zhang, S. Rősler, S. Rősler, A. Delsing, J. T. Zhao and H. T. Hintzen, Chem. Mater., 2011, 23, 1851.
21. G. Blasse, Philips Res. Rep., 1969, 24, 131.
22. T. L. Barry, J. Electrochem. Soc., 1970, 117, 381.
23. J. Y. Sun, H. Y. Du and W. X. Hu, Solid State Luminescent Material, Chemical Industry Press, Beijing, 2003, 1st edn, p. 152.
24. C. H. Huang, W. R. Liu and T. M. Chen, J. Phys. Chem. C, 2010, 114, 18698.
25. C. H. Huang and T.-M. Chen, Opt. Express, 2010, 18(5), 5090.
26. S. Bhushan and M. V. Chukichev, J. Mater. Sci. Lett., 1988, 7, 319.

---