Unraveling the site-specific energy transfer driven tunable emission characteristics of Eu³⁺ & Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ phosphors

Abstract

In this study we have explored Ca₁₀(PO₄)₆F₂ as host to develop a variety of phosphor materials with tunable emission and lifetime characteristics based on Eu³⁺ and Tb³⁺ as co-dopant ions and the energy transfer process involved with them. The energy transfer from the excited state of Tb³⁺ ion to the ⁵D₀ state of Eu³⁺ makes it possible to tune the colour characteristics from yellow to orange to red. Further, such energy transfer process is highly dependent on the concentration of Eu³⁺ and Tb³⁺ ions and their site-selective distribution among the two different Ca-sites (CaO₉ and CaO₆F) available. We have carried out DFT based theoretical calculation for both Eu³⁺ and Tb³⁺ ions in order to understand their distribution. It was observed that in cases of co-doped sample, Tb³⁺ ions prefer to occupy the Ca2 site in the CaO₆F network while Eu³⁺ ions prefer Ca1 site in the CaO₉ network. This distribution has significant impact on the lifetime values and the energy transfer process as observed in the experimental photoluminescence lifetime values. We have observed that for the 1^st series of compounds, wherein the concentration Tb³⁺ ions are fixed, the energy transfer from Tb³⁺ ion at Ca2 site to Eu³⁺ ion at Ca1 site is dominating (Tb³⁺@Ca2 → Eu³⁺@Ca1). However, for the 2^nd series of compounds, wherein the concentration Eu³⁺ ions are fixed, the energy transfer process was found to occur from the excited Tb³⁺ ion at Ca1 site to Eu³⁺ ions at both Ca1 and Ca2 (Tb³⁺@Ca1 → Eu³⁺@Ca1 and Tb³⁺@Ca1 → Eu³⁺@Ca2). This is the first reports of its kind on site-specific energy transfer driven colour tunable emission characteristics in Eu³⁺ and Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ phosphor and it will pave the way for the future development of effective colour tunable phosphor materials based on a single host and same co-dopant ions.

---

1 Introduction

White light-emitting diodes (WLEDs), a new generation solid-state lighting system, are dominating today's lighting industry. WLEDs are environmentally benign in nature, long-lasting and offer significant energy savings compared to the earlier fluorescent or incandescent lamps, which are now being replaced by WLEDs worldwide.^1–3 Most of the current commercial WLEDs are phosphor-converted (pc-WLEDs) using a combination of a blue LED chip and the yellow phosphor (Y₃Al₅O₁₂:Ce³⁺). However, the lack of a red light emitting component makes the color-rendering index (CRI) lowers than 80.⁴ An alternative approach is to use an ultraviolet (UV) or near-UV LED chip combined with blue, green, and red phosphors, which may generate warm white light with high CRI value after a proper adjustment of the ratio of tricolour phosphors. In such cases it is desirable that all the three phosphors have similar thermal and chemical stability and a degradation of colour of any of the three phosphors will change the ratio of three colours and result in different light instead of white light. The crystal lattice of such host matrix should not degrade and there should not be any chemical reaction on the surface due to exposure to outer atmosphere. One of the best ways to overcome such problem is to develop blue, red and green phosphors based on the same host matrix and using similar activators (such as rare earth ion), which have high thermal and chemical stability. Therefore, development of colour tunable phosphor materials based on same host and dopant ion's composition would be a promising area of research work in the field of light emitting materials.

Apatite based host matrices with general formula of M₁₀[TO₄]₆Z₂, where M is the cationic site with +1, +2 or +3 charges (e.g. K⁺, Na⁺, Ca²⁺, Sr²⁺, Ba²⁺, Pb²⁺, Mn²⁺, La³⁺, Y³⁺, Ce³⁺ etc.), and [TO₄] represents anion group (e.g. SiO₄⁴⁻, PO₄³⁻, GeO₄⁴⁻, MnO₄⁴⁻, VO₄³⁻, AsO₄³⁻, SO₄⁴⁻ etc.), and Z is anion with −1 or −2 charges (e.g. O²⁻, OH⁻ and halogen ions), possess high chemical and thermal stability and considered as an efficient host to develop a variety of phosphor materials with high luminescence efficiency.^5–7 Due to their high thermal, chemical and radiation stability such matrices are even potential candidate as host for immobilization of hazardous and highly radioactive nuclear waste.^8,9 Among them the fluorapatite based compounds with formula M₁₀(PO₄)₆F₂, (where M = Ca, Sr, Ba etc.) is advantageous owing to their abundance and high strength and thermal stability. As shown in Fig. 1, there are two different types of cationic sites in the crystal structure of Ca₁₀(PO₄)₆F₂, namely the nine-fold coordinated Ca1 site in the [CaO₉] polyhedron with C₃ point symmetry and seven-fold coordinated Ca2 site in the [CaO₆F] polyhedron with C_S point symmetry. Existence of these two different cationic sites have resulted in different emission characteristics of the same rare earth ion,¹⁰ which indicates that for the development of colour tunable phosphor materials apatite based hosts are a favoured choice, wherein by changing the distribution of the rare earth ions one can generate a variety of colour.

Fig. 1 1 × 1 × 2 supercell for Ca₅(PO₄)₃F crystal structure.

Eu³⁺ and Tb³⁺ ions are the two widely used rare earth activators whose combined emissions covered a wide wavelength range in the blue-green-red region in the visible spectra. The ⁵D₀ → 7Fj (j = 1, 2, 3 & 4) transitions of Eu³⁺ ions results in emissions in the orange and red region i.e. 580–700 nm.¹¹ On the other hand Tb³⁺ ion has emission peaks around 540 nm and hence serves as a green center.¹² The blue emission of Tb³⁺ ion in the 370–450 nm range is generally quenched through the cross-relaxation between neighbouring Tb³⁺ ions.¹³ However, owing to the efficient energy transfer from Tb³⁺ to Eu³⁺, co-doping Eu³⁺ and Tb³⁺ ions is found to be a successful strategy to develop a variety of phosphor materials with not only tunable colour characteristics in the blue-green-red region but also with tunable lifetimes values just by varying their respective concentration while keeping the host lattice intact.^14–17

In present case we have chosen Ca₁₀(PO₄)₆F₂ as host matrix to explore such tunable color characteristics. The two different Ca-sites (CaO₉ and CaO₆F) available for the dopant ions (Eu³⁺ and Tb³⁺ ion) makes it more flexible to tune the colour characteristics from yellow to orange to red. It has also been found that the F atom in the Ca-2 site made a significant difference in the lifetime and luminescence efficiency since F atom has less quenching effect compared to O atom.^8,9 Further, an energy transfer dynamics from Tb³⁺ ion to Eu³⁺ is associated with these compounds, which makes the tunable emission characteristics more exciting. Till date, no such study on site-specific energy transfer driven colour tunable emission characteristics in Ca₁₀(PO₄)₆F₂ is reported. Further, we have carried out density functional theory (DFT) based calculations in order to check the most favourable distribution of Eu³⁺ and Tb³⁺ ions among the two lattice sites available in Ca₁₀(PO₄)₆F₂ after considering a variety of possible substitutions. We have also considered the impact on any lattice ion vacancy on such distribution. We believe a comprehensive combined theoretical and experimental study on site-specific tunable colour characteristics and the associated lifetime values will pave the way for the future development of effective colour tunable phosphor materials based on a single host and same dopant ions.

2 Experimental

2.1. Synthesis of Eu³⁺ and Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ compounds

Precursor materials. Calcium carbonate (CaCO₃), europium oxide (Eu₂O₃), terbium oxide (Tb₂O₃), ammonium dihydrogen phosphate (NH₄H₂PO₄) and ammonium bi-fluoride (NH₄HF₂) were purchased from Sigma Aldrich and have been used for the compound preparation.

Synthesis procedure. Various Eu³⁺ and Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ compounds were synthesized by solid-state reaction method using appropriate amount of CaCO₃, Eu₂O₃, Tb₂O₃, NH₄H₂PO₄ and NH₄HF₂ following similar procedure as reported earlier.⁸ These compounds were taken in appropriate amount and mixed in an agate mortar and the mixture was pelletized into a 10 mm diameter pellet. The pellets of various Eu³⁺ and Tb³⁺ co-doped compounds were then heated under high pure argon gas atmosphere in a stepwise manner at 500 K for 4 hours, 800 K for 4 hours and 1200 K for 14 hours without any interruption. After the final heat treatment, the solid product was collected and grind well and characterized by powder X-ray diffraction and FTIR spectroscopy. Various Tb³⁺ and Eu³⁺ doped samples such as 0.05 mol% Eu³⁺ and 0.5 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.05–Tb_0.5:CPF), 0.1 mol% Eu³⁺ and 0.5 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.1–Tb_0.5:CPF), 0.3 mol% Eu³⁺ and 0.5 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.3–Tb_0.5:CPF), 0.5 mol% Eu³⁺ and 0.5 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.5–Tb_0.5:CPF), 0.5 mol% Eu³⁺ and 0.05 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.5–Tb_0.05:CPF), 0.5 mol% Eu³⁺ and 0.1 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.5–Tb_0.1:CPF), 0.5 mol% Eu³⁺ and 0.3 mol% Tb³⁺ doped Ca₁₀(PO₄)₆F₂ (hereinafter Eu_0.5–Tb_0.3:CPF), were prepared using the solid state reaction method.

2.2. Instrumentation

The details of the instrumentation have been provided in ESI.†

2.3. Computational details

All the spin polarized density functional theory (DFT) calculations have been carried out employing the projector augmented wave (PAW) based electronic structure code Vienna ab initio simulation (VASP).^18,19 The set of valence states chosen in the present study for the pseudo potentials are Ca (3s²3p⁶4s²), Eu (6s²5p⁶5d¹), Tb (6s²5p⁶5d¹), P (3s²3p²), F (2s²2p⁵) and O (2s²2p⁴). Perdew–Burke–Ernzerhof (PBE) functional within generalized gradient approximation (GGA) has been considered during the geometry optimization technique.^20,21 Full relaxation of both cell parameter and ionic positions for all the model structures has been considered during geometry optimization process. The energy cut off for the expansion of electronic wave function has been kept fixed at 500 eV with self-consistent energy convergence of 10⁻⁶ eV throughout the calculations. Electronic structure calculations have been carried out using the optimized geometries. Monkhorst and Pack scheme has been adopted for generating Γ-centered k-point mesh during Brillouin zone sampling.²² To overcome the limitations of conventional density functional theory we have used hybrid functional as prescribed by Heyd, Scuseria, and Ernzerhof (HSE06) during electronic structure calculations,²³ which has been shown to reproduce the experimental band gap of different semiconductor materials.²⁴ According to the HSE06 functional the exchange-correlation functional is defined by short-ranged (SR) and long-ranged (LR) parts of exchange. Short-ranged part is described by both Hartree–Fock (HF) exact exchange functional (E^SR_X) and PBE exchange functional (E^PBE,SR_X), while the long-ranged part is completely treated by PBE exchange functional (E^PBE,LR_X) as

E^HSE_XC = aE^SR_X(μ) + (1 − a)E^PBE,SR_X(μ) + E^PBE,LR_X(μ) + E^PBE_C (1)

where, ‘μ’ indicates screening parameter, defines the distance of the short range interactions and ‘a’ is the HF mixing coefficient. The value of screening parameter and quantity of HF mixing are 0.2 Å⁻¹ and 25%, respectively.

3 Results and discussion

3.1. Phase purity and elemental analysis: X-ray diffraction (XRD) and X-ray fluorescence (XRF) study

The compounds obtained by solid-state reaction were characterized by X-ray diffraction technique. The obtained XRD of various Eu³⁺ and Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ compounds are shown in Fig. 2. All the compounds showed the characteristics XRD pattern of pure fluorapatite phase with ICDD file no. 01-070-8135. This has also confirmed that Eu³⁺ and Tb³⁺ ions were successfully substituted into lattice sites of calcium fluorapatite. The XRF spectra was recorded for Eu_0.1Tb_0.5:CPF and it is provided in the ESI,† which showed the peaks of Eu, Tb, Ca and P atoms at their respective energy position.

Fig. 2 XRD pattern of (a) Eu_0.3–Tb_0.5:CPF, (b) Eu_0.5–Tb_0.1:CPF, (c) Eu_0.5–Tb_0.3:CPF (d) Eu_0.5–Tb_0.05:CPF, (e) Eu_0.1–Tb_0.5:CPF, (f) Eu_0.5–Tb_0.5:CPF.

3.2. Fourier-transform infrared (FTIR) spectroscopy study

IR spectrum of various compounds are shown in Fig. 3, which confirmed the existence of characteristics stretching and bending frequencies of phosphate groups of apatite. The transmission deep exist around 1075.81 cm⁻¹, 1024.17 cm⁻¹, 948 cm⁻¹, 589.5 cm⁻¹ and 563 cm⁻¹ are due to phosphate stretching vibration modes as reported earlier.⁸

Fig. 3 FTIR spectra of (A) Eu_0.5–Tb_0.5:CPF, (B) Eu_0.5–Tb_0.3:CPF, (C) Eu_0.3–Tb_0.5:CPF.

3.3. Geometry and electronic structure

At room temperature, Ca₁₀(PO₄)₆F₂ shows a hexagonal crystal structure (Fig. 1) with space group symmetry P6₃/m (space group no. 176). Structural framework of Ca₁₀(PO₄)₆F₂ contains two different Ca lattice sites and three different O lattice sites whereas one type of lattice site for both P and F. The unit cell of Ca₅(PO₄)₃F contains two formula units (i.e., 10 Ca, 6 P, 24 O, and 2 F). In can be noted that, the first type of Ca site (Ca1) is coordinated by nine O atoms, while second type of Ca site (Ca2) is coordinated by six O atoms and one F atoms.²⁵ The calculated lattice parameter (a = b = 9.495 Å, and c = 6.918 Å) are found to be very close to the experimentally reported values (a = b = 9.3783 Å, and c = 6.8888 Å).^26–29 This justifies the choice of the computational methodology. To investigate the electronic structure of Ca₅(PO₄)₃F, we have analysed the density of states (DOS) and partial density of states (PDOS) (Fig. 4), which shows that the valence band maximum (VBM) is mainly composed of O 2p and F 2p states with small contribution of P (3p, 3s) and Ca (3p, 4s) states. On the other hand, the conduction band minimum (CBM) is dominated by P (3p, 3s) and F 2p states, with small contributions of Ca 4s state. The calculated band gap of Ca₁₀(PO₄)₆F₂ within HSE06 hybrid functional is found to be 7.26 eV, which is very close to the experimentally reported value of 7.21 eV for Ca₅(PO₄)₃F.³⁰

Fig. 4 Density of states of Ca₁₀(PO₄)₆F₂. The vertical dashed line indicates the Fermi level.

Codoping of Eu³⁺ and Tb³⁺ in Ca₁₀(PO₄)₆F₂. To model (Eu, Tb)-codoped system, we have considered 1 × 1 × 2 supercell for Ca₁₀(PO₄)₆F₂ which consist of total 84 atoms (20 Ca, 12 P, 48 O, and 4 F), and replaced two Ca by Eu and Tb. In this case, we have considered total six configurations (Str. 1–6 as shown in Fig. 5) by varying the lattice sites of dopant elements. For example, both Eu³⁺ and Tb³⁺ occupy lattice sites of Ca1 type in case of Str. 1 (Eu, and Tb at 1^st neighboring lattice sites of each other) and Str. 2 (Eu, and Tb at 2^nd neighboring lattice sites of each other). We have also calculated defect formation energies for various such structures of (Eu, Tb)-codoped Ca₁₀(PO₄)₆F₂ using the following relationship.^31,32

ΔH_f = E_doped − E_perfect + q∑n_xμ_x (2)

where, E_doped and E_perfect represent the energy of the codoped and pristine Ca₁₀(PO₄)₆F₂, calculated with same supercell size, μ_x stands for the chemical potential of the element X and n_x is the number of elements added (q = −1) or replaced (q = +1) to form the codoped system.

Fig. 5 Structures of for (Eu³⁺, Tb³⁺) co-doped Ca₁₀(PO₄)₆F₂ with different configurations. The numbers in Str. 6 denote positions of Ca vacancies.

The formation energies of the various structures are given in Table 1. Lower the formation energy, higher the thermodynamic stability of the system. It should be noted that Str. 1 is energetically less stable by 0.24 eV in comparison to the Str. 2. On the other hand, both Eu, and Tb occupy lattice sites of Ca2 type in case of Str. 3 (Eu, and Tb at 1^st neighboring lattice sites of each other) and Str. 4 (Eu, and Tb at 2^nd neighboring lattice sites of each other). It is interesting to note that Str. 3, and Str. 4 are energetically more stable by 0.40 eV with respect to that of Str. 1. However, both types of configurations in this case are energetically comparable to each other. Finally we have considered co-doping of Eu, and Tb involving both Ca1 and Ca2 types lattice sites, and generated two different structures viz. Str. 5 (Eu at Ca2 type site, while Tb at Ca1 type site) and Str. 6 (Eu at Ca1 type site, while Tb at Ca2 type site). Interestingly, Str. 5 and Str. 6 are energetically lower than the other four structures, and Str. 6 is found to be energetically most stable among all the possible structures considered in the present study. The calculated formation energy therefore follows the order, Str. 6 < Str. 5 < Str. 4 < Str. 3 < Str. 2 < Str. 1 while the stability of the structures follow the reverse order. This indicates that the Eu and Tb prefer different types of Ca lattice sites to form co-doped system. Since, both Eu and Tb normally exist in +3 states; there may be formation of Ca-vacancy in the co-doped system due to formation of charge compensated system. Therefore, we have also investigated the influence of Ca-vacancy on the electronic structure and properties of the codoped system. For this purpose, we have considered Str. 6, which is the lowest energy structure, and removed one Ca atom. In this case, we have considered three different structures by varying the position of Ca lattice site (indicated by Ca1, Ca2, Ca3 in Str. 6, Fig. 3). As for example, Str. 7 can be generated vacancy by removing Ca from the first neighbouring lattice of Eu (Ca1 in Str. 6, Fig. 3), Str. 8 from the second neighbouring lattice of both Eu and Tb (Ca3 in Str. 6, Fig. 3), and Str. 9 from the first neighbouring lattice of both Tb (Ca2 in Str. 6, Fig. 3). Interestingly, formation energy for Str. 7 has been found to be the lowest among the three (Table 1), indicating higher formation probability.

Table 1 Calculated formation energies for the (Eu, Tb)-codoped Ca₁₀(PO₄)₆F₂ with different model structures and Ca vacancy defects

Systems	Structures	Formation energy (eV)
(Eu, Tb)-codoped Ca10(PO4)6F2	Str. 1	3.77
	Str. 2	3.53
	Str. 3	3.16
	Str. 4	3.15
	Str. 5	1.19
	Str. 6	1.10
(Eu, Tb)-codoped Ca10(PO4)6F2 with VCa	Str. 7	1.64
	Str. 8	2.64
	Str. 9	1.27

---

Electronic structure of (Eu³⁺, Tb³⁺) codoped Ca₁₀(PO₄)₆F₂ and (Eu, Tb)-codoped Ca₁₀(PO₄)₆F₂ in presence of Ca-vacancy. In this study, the electronic structure of Eu³⁺ & Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ system has been investigated by analysing the DOS and PDOS (Fig. 6a). We have considered lowest energy structure (Str. 6) for the electronic structure calculations. It has been observed that the doping of Eu³⁺ and Tb³⁺ at different type of Ca lattice sites results into introduction of discrete occupied impurity states at energy level 0.61 eV, 1.29 eV and 2.47 eV above the VBM. Analysis of PDOS indicates that the occupied impurity state is mainly composed by O (p) state with minor contribution of P (s, p), Eu (p, d) and Tb (p, d) states. Analysis of energy level indicates that the VBM and CBM levels are lowered by 0.03 eV and 0.14 eV, respectively compared to that of perfect Ca₁₀(PO₄)₆F₂. The Fermi energy is increased by 3.37 eV with respect to that Ca₁₀(PO₄)₆F₂. This is because of presence of excess electron in the (Eu, Tb)-codoped Ca₁₀(PO₄)₆F₂.

Fig. 6 Density of states of (a) Eu, Tb-codoped Ca₁₀(PO₄)₆F₂ (b) Eu, Tb-codoped Ca₁₀(PO₄)₆F₂ with Ca vacancy. The vertical dashed line indicates the Fermi level.

Interestingly, electronic structure for Eu³⁺ & Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ in presence of Ca-vacancy shows that the impurity states, as observed for (Eu, Tb)-codoped case, are completely disappear (Fig. 6b). This is due to the charge compensation system in presence of Ca-vacancy. Analysis of PDOS indicates that the VBM is mainly contributed by O (2p) states with minor contribution of F (2p) states, while the CBM is contributed by O (2s, 2p), P (3s, 3p), Eu (5d) and Tb (5d). Analysis of energy level indicates that the VBM is increased by 0.13 eV and CBM is decreased by 0.24 eV compared to that of perfect Ca₁₀(PO₄)₆F₂. The Fermi level is shifted towards the VBM by 2.28 eV with respect to that codoped system without Ca-vacancy. This is due to compensation of excess electron in presence of Ca-vacancy in the Eu³⁺ & Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂.

3.4. Photoluminescence study

Fig. 7 represents the photoluminescence excitation spectra of Eu_0.05–Tb_0.5:CFP at 618 nm and 547 nm emission wavelengths. Two excitation bands peaking at 230 nm and 250 nm were observed for both the emission wavelengths, although their intensities are different. These transitions must be originate from the charge transfer transitions from 2p orbital of O²⁻ to the vacant 5d orbital of Tb³⁺ and Eu³⁺ respectively. Since the 547 nm emission wavelength is due to transition of ⁵D₄ → ⁷F_F of the Tb³⁺ ion,³³ the highly intense excitation peak at 230 nm can be attributed to the charge transfer transition from 2p orbital of O²⁻ to the vacant 5d orbital of Tb³⁺. On the other hand at 618 nm emission wavelength, the peak intensity at 250 nm were found to be increased although 230 nm being the most intense one. The 250 nm excitation peak therefore can be attributed to the charge transfer transition from 2p orbital of O²⁻ to the vacant 5d orbital of Eu³⁺.³⁴ The fact that even in case of 618 nm emission wavelength, the excitation peak at 230 nm (which is due to Tb³⁺ ion) is present in the PLE spectrum is suggesting that there might a possibility of energy transfer from Tb³⁺ ion to Eu³⁺ ion.

Fig. 7 Excitation spectra of Eu_0.05–Tb_0.5:CFP at 618 nm and 547 nm emission wavelengths.

Fig. 8 represents the emission spectra of different Eu³⁺ and Tb³⁺ co-doped CFP compounds at 230 nm excitation wavelengths. The emission spectra consist of various emission peaks due to Eu³⁺ and Tb³⁺ ions. The emission peaks around 590, 618, and 704 nm are due to the ⁵D₀ → ⁷F_j (j = 1, 2, 3 and 4) transitions of Eu³⁺ ion³⁵ while the emission bands at 487, 547, 586 and 621 nm are due toTb³⁺ ion's ⁵D₄ → ⁷J_j (j = 6, 5, 4, 3) transitions.³³ As it is known that for Eu³⁺ ion, the emission at 590 nm (⁵D₀–⁷F₁ transition) is a magnetic dipole (MD) transition while the red color emission around 618 nm (⁵D₀–⁷F₂ transition) and at 704 nm (⁵D₀–⁷F₄ transition) are electric dipole (ED) transitions. The ED transition is hypersensitive to the local structure of Eu³⁺ ion and at an asymmetric environment, this transition is allowed one. The fact that ED transition's intensity is higher than the MD transition signifies that Eu³⁺ ion exists in an asymmetric environment.³⁴ For Tb³⁺ ion, the green emission at 547 nm i.e. ⁵D₄–⁷F₅ is magnetic dipole (MD) transitions and not sensitive to local structure or crystal field strength. However, the blue emission i.e. ⁵D₄–⁷F₆ transition around 487 nm is an electric dipole (ED) transitions, which is sensitive to the local environment although it is not hypersensitive.³³ We have calculated the respective color coordinates of various mol% of Eu³⁺ and Tb³⁺ co-doped CFP compounds at 230 nm excitation wavelengths as shown in Fig. 9. It can be seen that upon changing the concentration of any of the two dopant ion and keeping other dopant ion's concentration fixed, there is a change in color coordinated from yellow to orange to red. For the low dopant level of Eu³⁺ ion in Eu_0.05–Tb_0.5:CFP, the CIE color co-ordinates are placed in the yellow region while with increasing the concentration of Eu³⁺ ions, viz. in Eu_0.1–Tb_0.5:CFP and Eu_0.3–Tb_0.5:CFP, the color coordinates were shifted to orange and red regions. Similarly, by keeping the concentration of Eu³⁺ ion's fixed while varying the concentration of Tb³⁺ ions, the CIE color coordinates were found to be changed. It can be seen that with changing the concentration of Tb³⁺ ions viz. in Eu_0.5–Tb_0.1:CFP, Eu_0.5–Tb_0.2:CFP and Eu_0.5–Tb_0.3:CFP, the color can be tuned from orange to red. Therefore, a wide range of colored phosphors can be developed just by changing the relative concentration of Eu³⁺ and Tb³⁺ dopant ions.

Fig. 8 Emission spectra of various mol% of Eu³⁺ and Tb³⁺ co-doped CFP compounds at 230 nm excitation wavelengths.

Fig. 9 CIE colour coordinates of various mol% of Eu³⁺ and Tb³⁺ co-doped CFP compounds at 230 nm excitation wavelengths.

Fig. 10a represents the emission spectra of different Eu³⁺ and Tb³⁺ co-doped CFP compounds at 250 nm excitation wavelengths, wherein we can see that the emission spectra for all the compounds are only consist of characteristics peaks of Eu³⁺ ions and that the peaks due to Tb³⁺ ions are very weak. This confirms that there is no charge transfer process involved at 250 nm excitation and that the 250 nm excitation peak is purely due to charge transfer transition from 2p orbital of O²⁻ to the vacant 5d orbital of Eu³⁺ ion.

Fig. 10 Emission spectra of various mol% of Eu³⁺ and Tb³⁺ co-doped CFP compounds at 250 nm excitation wavelengths and their respective CIE colour coordinates.

It is worth mentioning here that these color tunable characteristics have been observed for 230 nm excitation wavelength only and for 250 nm excitation energy there is no such color tunable characteristics and all the compounds are red phosphor in nature as suggested by their calculated color coordinates represented in Fig. 10b. Therefore, the energy transfer dynamics from Tb³⁺ ion to Eu³⁺ ion at 230 nm excitation is solely responsible for such tunable color characteristics.^14–17 As shown in Fig. 11, upon 230 nm excitation (which is due to CT transition from O²⁻ to the vacant 5d orbital of Tb³⁺ ion) many Tb³⁺ ions are excited to higher energy CT state from which it come back to the lower energy level ⁵D₃ and then to lower lying ⁵D₄ meta-stable state through non-radiative pathway. From ⁵D₄ level it is relaxed to lower energy levels through ⁵D₄–⁷F_5,4,3 transitions. Few other Tb³⁺ ions at ⁵D₃ or ⁵D₄ levels transfer their energy to the ⁵D₂ level of Eu³⁺ ion which is known as the energy transfer. Now through a non-radiative decay from ⁵D₂ excited state the population of ⁵D₀ meta-stable state increases which results in emission through ⁵D₀–⁷F_J (J = 0–6) transitions.

Fig. 11 Schematic energy level diagram of Tb³⁺ and Eu³⁺ and the energy transfer dynamics.

The fact that with increasing the Tb³⁺ concentration there is an enhancement of Eu³⁺ ion's intensity while keeping it concentration unchanged as shown in Fig. 8, clearly indicates that there is a charge transfer process involved from Tb³⁺ ions to Eu³⁺ ion. Generally, the energy transfer from sensitizer (Tb³⁺) to activator (Eu³⁺) may takes place through either (i) the exchange interaction or (ii) electric multipolar transition. When the critical distance between the sensitizer and activator is shorter than 5 Å, then the energy transfer process use to take place via exchange interaction and when it is greater than 5 Å, the energy transfer from Tb³⁺ to Eu³⁺ ions happen in the way of electric multipolar interaction. The critical distance (R_c) between Tb³⁺ and Eu³⁺ can be calculated using eqn (3)^14,36

where V is the volume of the unit cell, X_c is the critical dopant concentration, and Z is the number of host cations in the unit cell. For the Ca₁₀(PO₄)₆F₂, X_c = 0.10, V = 534.35 ų and Z = 2, which gives a R_c value of 10.84 Å, which is larger than 5 Å. Therefore, energy transfer between Tb³⁺ and Eu³⁺ ions takes place due to electric multipolar interactions instead of exchange interaction.

Now let us see how the lifetime of the respective excited states of the Tb³⁺ ions to Eu³⁺ ions behave while changing their concentration and the excitation energy. This will also provide information about the energy transfer mechanism.

3.5. Photoluminescence lifetime study

PL lifetime measurements can provide detail information regarding the energy transfer dynamics in addition to the lattice site occupancy of the dopant ions.^14–17 Since, there are two lattice sites (Ca1 and Ca2 sites) available for the dopant ions and as we have observed from DFT based calculation that the distribution of Eu³⁺ and Tb³⁺ among these two lattice sites are different, it will be interesting to see which sites are involved in the energy-transfer dynamics? The photoluminescence lifetime measurement has been carried out for all the compounds at 230 nm excitation wavelength (which is attributed to Tb³⁺ ions) and with emission wavelength 547 nm (⁵D₄–⁷F₅ transition of Tb³⁺ ion) and 618 nm (⁵D₀–⁷F₂ transition of Eu³⁺ ions) as shown in Fig. 12. We have observed that the decay profiles (both at 547 nm and 618 nm emission wavelength) for all the compounds followed a bi-exponential eqn (4).where I(t) is intensity, τ₁ and τ₂ are emission decay times, and A₁ and A₂ are their relative weightage.

Fig. 12 Photoluminescence decay profile at λ_ex = 230 nm and λ_em = 547 nm & 620 nm for different Eu³⁺ and Tb³⁺ doped CFP compounds. (a and b) Eu_0.5–Tb_0.01:CFP; (c and d) Eu_0.5–Tb_0.1:CFP; (e and f) Eu_0.3–Tb_0.5:CFP (few spectra of other compounds are provided in the ESI†).

We have also calculated the percentage of a specific life-time using the formula in eqn (5)

The lifetime values are included in Table 2 for all the compounds at different emission wavelength. As mentioned before that due to the availability of two different Ca-sites (CaO₉ and CaO₆F); both Eu³⁺ and Tb³⁺ ions have a tendency to be distributed among the two lattice sites. Earlier, we have observed that the lifetime value for rare earth ion when placed at the Ca2 site (CaO₆F) which has an F-atom linkage, is more than when it is placed at Ca1 site (CaO₉). This is attributed to the F-atom, which has less quenching effect on the excited state and makes the lifetime value of the rare earth ions higher.^8,9 Therefore, the long lived component observed at 547 nm emission wavelength is due to Tb³⁺ ions present at Ca2 site while that at 618 nm emission is due to Eu³⁺ ions present at Ca2 site. On the other hand, the short-lived components for these two emission wavelength are due to Tb³⁺ and Eu³⁺ ions present at the Ca1 site, which has no F-atom linkage.

Table 2 PL lifetime values of Eu³⁺ and Tb³⁺ doped CFP compounds

Compound	PL lifetime components of Tb^3+ (λex = 230 nm and λem = 547 nm)		PL lifetime components of Eu^3+ (λex = 230 nm and λem = 618 nm)
	Short-lived (τ1) component in μs	Long-lived (τ2) component in μs	Short-lived (τ1) component in μs	Long-lived (τ2) component in μs
Eu0.5–Tb0.01:CFP	2191.98 (90%)	4054.92 (10%)	2168.83 (59%)	2992.06 (41%)
Eu0.5–Tb0.1:CFP	1867.70 (70%)	3601.64 (30%)	2423.43 (89%)	3673.61 (11%)
Eu0.5–Tb0.3:CFP	1966.44 (86%)	3379.12 (14%)	2439.29 (93%)	3739.46 (7%)
Eu0.1–Tb0.5:CFP	1384.55 (35%)	2754.84 (65%)	2098.41 (67%)	3333.81 (33%)
Eu0.3–Tb0.5:CFP	952.79 (28%)	2203.12 (72%)	1477.1 (45%)	2527.96 (55%)
Eu0.5–Tb0.5:CFP	651.77 (24%)	2216.65 (76%)	1934.22 (45%)	2807.46 (55%)
Na^+2Eu^3+2:Ca6(PO4)6F2 (ref. 8)	—	—	1870 (23%)	2420 (77%)

---

However, we have observed from our DFT calculation earlier that the stability of the Eu³⁺ and Tb³⁺ co-doped structure is more when Eu³⁺ ion is distributed at Ca1 site and Tb³⁺ ion at Ca2 site. The preference of distribution of Eu³⁺ ion is also reflected in the higher percentage of the short-lived component in many compounds viz. Eu_0.5–Tb_0.01:CFP, Eu_0.5–Tb_0.1:CFP, Eu_0.5–Tb_0.3:CFP, Eu_0.1–Tb_0.5:CFP. However, for Tb³⁺ ions such observation (higher percentage of the long-lived component) is not observed since the lifetime value of Tb³⁺ ions also depends on the energy transfer dynamics from Tb³⁺ ion to Eu³⁺ ion, which will decrease the percentage of contribution. The fact that there is an energy transfer process involved in this class of compounds can be explained based on decrease in lifetime value upon varying the concentration of Tb³⁺ ion or Eu³⁺ ion. Due to the energy transfer from the excited state of Tb³⁺ ion to the Eu³⁺ ion, the lifetime value of Tb³⁺ ion will be decreased, as observed earlier for various class of compounds.^14–17 The lifetime values of Eu³⁺ ion obtained for the Eu & Tb co-doped compounds in this work are found to be slight higher than that for pure Eu³⁺ doped CFP compounds reported earlier.⁸ This might be due to the energy transfer process involved in the present compounds. The fact that we have not seen a monotonic decrease or increase for few samples may be explained based on non-radiative pathways associated with various defect centers or vacancies, which also plays an important role in determining the lifetime of the excited state. These vacancies are generated due to the charge imbalance when Tb³⁺ or Eu³⁺ ions are substituted at a place of Ca²⁺ ion. The additional positive charge will lead to creation of negatively charged vacancies, which will have impact on the nearby Tb³⁺/Eu³⁺ ion's local structure. Further, depending on the distance from the Tb³⁺/Eu³⁺ ion, the impact will be different. We believe such vacancies are the main reason to not have any systematic decrease in the lifetime values for few samples.

Now let us consider two cases; (1) for the series of compounds where the concentration Eu³⁺ ion is fixed but that of Tb³⁺ ions is varied viz. for Eu_0.5–Tb_0.01:CFP, Eu_0.5–Tb_0.1:CFP and Eu_0.5–Tb_0.3:CFP compounds, and (2) for the series of compounds where the concentration Tb³⁺ ion is fixed but that of Eu³⁺ ions is varied viz. Eu_0.1–Tb_0.5:CFP, Eu_0.3–Tb_0.5:CFP and Eu_0.5–Tb_0.3:CFP. For the 1st series of compounds wherein the concentration of Tb³⁺ ion is varied, it is observed that the percentage of the long-lived Tb³⁺ component is small than its short-lived component. This is due to fact that the long-lived Tb³⁺ component at the Ca2 site mostly transfers its energy to nearby Eu³⁺ ions. Hence in this series, majority of the Tb³⁺ ions at Ca2 site are preferably taking part in the energy transfer dynamics and results in small contribution towards its radiative lifetime value and the radiation transition. Now for the second series of compound wherein the concentration of Tb³⁺ ion is fixed but that of Eu³⁺ ion is varied, the opposite thing has happened i.e. the percentage of the short-lived Tb³⁺ component is less than the long-lived component. Hence, we believe that for this series of compounds, majority of the Tb³⁺ ions at Ca1 site are preferably taking part in the energy-transfer process to Eu³⁺ ion. Now if we consider the lifetime values of Eu³⁺ ions, then for the 1^st series compound the percentage of contribution of the short-lived Eu³⁺ component at Ca1 site is higher than that at Ca2 site. Further, with increasing the concentration of Tb³⁺ ion, the percentage of contribution of this short-lived component is found to increase. We have already concluded from DFT calculation that for Tb³⁺ and Eu³⁺ co-doped system, Eu³⁺ ions most preferably go to the Ca1 site and results in smaller lifetime value. Since there is no energy transfer process involved from the excited state of Eu³⁺ ions to other co-dopant ion, this percentage of contribution is linked to the contribution from Eu³⁺ ions at Ca1 and Ca2 sites to the overall red light emission. Therefore, after considering all these changes in the relative contribution, we conclude that for the 1^st series of compound the energy transfer from Tb³⁺ ion at Ca2 site to Eu³⁺ ion at Ca1 site (Tb³⁺@Ca2 → Eu³⁺@Ca1) is the dominant one. Interestingly, for the 2^nd series of compounds wherein the concentration of Tb³⁺ ion is fixed, we have not observed much difference in the respective contribution of two different Eu³⁺ components and for the higher concentration of Eu³⁺ in compound like Eu_0.5–Tb_0.5:CFP, the percentage is a bit higher for the long-lived Eu³⁺ component. Therefore, we may conclude that for the 2^nd series of compounds, the energy transfer process from the excited Tb³⁺ ion at Ca1 site to Eu³⁺ ions at both Ca1 and Ca2 sites (Tb³⁺@Ca1 → Eu³⁺@Ca1 and Tb³⁺@Ca1 → Eu³⁺@Ca2) is happening with a bit more preference towards the Eu³⁺ ion at Ca2 site (Tb³⁺@Ca1 → Eu³⁺@Ca2).

4 Conclusion

Various Eu³⁺ and Tb³⁺ co-doped Ca₁₀(PO₄)₆F₂ compounds were synthesized by solid-state reaction method. Photoluminescence study showed that with varying the concentration of Tb³⁺ and Eu³⁺ ions, the colour characteristics can be tuned from yellow to orange to red. We have found that there is an energy transfer process involved from Tb³⁺ ion to Eu³⁺ , which is responsible for the tunable emission characteristics and the decay kinetics. It was observed that two different lifetime components exist for both Eu³⁺ and Tb³⁺ ions depending on their distribution among the two different Ca-sites (Ca1 in CaO₉ and Ca2 in CaO₆F network). The long-lived component for both the dopant ions are attributed to Ca2 site in CaO₆F network, wherein the F atom made the difference in lifetime of the excited state due to less quenching effect compared to O-atom. From density functional theory (DFT) based calculations we have observed that Tb³⁺ ions prefer to go to the Ca2 site while Eu³⁺ ions prefer the Ca site, when they are co-doped in Ca₁₀(PO₄)₆F₂ in a structure associated Ca vacancy. This site specific distribution was found to influence the energy transfer dynamics. It has been observed from the lifetime study that for the 1^st series of compounds, wherein the concentration Tb³⁺ ions are fixed, the energy transfer Tb³⁺@Ca2 → Eu³⁺@Ca1 energy transfer is more preferable than Tb³⁺@Ca1 → Eu³⁺@Ca2. However, for the 2^nd series of compounds, wherein the concentration Eu³⁺ ions are fixed, both the energy transfer processes Tb³⁺@Ca1 → Eu³⁺@Ca1 and Tb³⁺@Ca1 → Eu³⁺@Ca2 were involved with a bit more preference for Tb³⁺@Ca1 → Eu³⁺@Ca2.

Author contributions

Nimai Pathak: conceptualization, methodology, supervision, visualization, photoluminescence investigation, data analysis of all the experiments, writing-initial draft, reviewing and editing.

Bhagyalaxmi Chundawat: preparation of all the compound and XRD measurement.

Pratik Das: helping in preparation of compounds and FTIR measurement.

Pampa Modak: writing and helping in the theoretical study.

Brindaban Modak: VASP simulation and theoretical data analysis.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors acknowledge the contribution of Dr B. G. Vats FCD, BARC for his help in FTIR measurements and Dr K. Sanyal, FCD, BARC for the XRF measurement during the course of this project. Authors are also grateful to Dr P. K. Pujari, Group director, Radiochemistry and Isotope group for his encouragement and help during the course of this project. No external funding agency is involved with this work and the work is completely funded by BARC.

References

1. Q. Zhou, L. Dolgov, A. M. Srivastava, L. Zhou, Z. Wang, J. Shi, M. D. Dramićanin, M. G. Brik and M. Wu, J. Mater. Chem. C, 2018, 6, 2652–2671.
2. M. Xin, D. Tu, H. Zhu, W. Luo, Z. Liu, P. Huang, R. Li, Y. Cao and X. Chen, J. Mater. Chem. C, 2015, 3, 7286–7293.
3. E. F. Schubert and J. K. Kim, Science, 2005, 308, 1274–1278.
4. A. A. Setlur, W. J. Heward, Y. Gao, A. M. Srivastava, R. G. Chandran and M. V. Shankar, Chem. Mater., 2006, 18, 3314–3322.
5. N. Liu, L. Mei, L. Liao, J. Fu and D. Yang, Sci. Rep., 2019, 9, 15509.
6. G. Li, D. Geng, M. Shang, Y. Zhang, C. Peng, Z. Cheng and J. Lin, J. Phys. Chem. C, 2011, 115, 21882–21892.
7. H. Liu, L. Liao, M. S. Molokeev, Q. Guo, Y. Zhang and L. Mei, RSC Adv., 2016, 6, 24577–24583.
8. (a) P. Das, N. Pathak, P. Modak and B. Modak, J. Hazard. Mater., 2021, 411, 125025; (b) N. Pathak, P. Das, B. Chundawat, P. Modak and B. Modak, J. Hazard. Mater., 2021, 423, 126980.
9. P. Das, N. Pathak, B. Sanyal, S. Dash and R. M. Kadam, J. Alloys Compd., 2019, 810, 151906.
10. S. Unithrattil, H. J. Kim, K. H. Gil, N. H. Vu, V. H. Hoang, Y. H. Kim, P. Arunkumar and W. B. Im, Inorg. Chem., 2017, 56, 5696–5703.
11. (a) S. K. Gupta, P. S. Ghosh, N. Pathak and R. M. Kadam, RSC Adv., 2016, 6, 42923–42932; (b) M. Sahu, N. Pathak and M. K. Saxena, RSC Adv., 2021, 11, 17488–17497.
12. X. Yan, G. R. Fern, R. Withnall and J. Silver, Nanoscale, 2013, 5, 8640–8646.
13. P. Van Do, N. X. Ca, L. D. Thanh, N. Van Nghia and T. T. C. Thuy, Phys. Chem. Chem. Phys., 2020, 22, 27590–27599.
14. R. Mi, J. Chen, Y.-g. Liu, M. Fang, L. Mei, Z. Huang, B. Wang and C. Zhaob, RSC Adv., 2016, 6, 28887–28894.
15. Y. Gao, X. Sun, Z. Feng, L. Zhu, J. Zhang, W. Gao, X. Zhou, R. Cong and T. Yang, New J. Chem., 2017, 41, 2037–2045.
16. B. Zhang, H. Zou, H. Guan, Y. Dai, Y. Song, X. Zhou and Y. Sheng, CrystEngComm, 2016, 18, 7620–7628.
17. M. Yang, Y. Liang, Q. Gui, B. Zhao, D. Jin, M. Lin, L. Yan, H. You, L. Dai and Y. Liu, Sci. Rep., 2015, 5, 11844.
18. G. Kresse and D. Joubert, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 1758–1775.
19. P. E. Blöchl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 17953–17979.
20. J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson, D. J. Singh and C. Fiolhais, Phys. Rev. B: Condens. Matter Mater. Phys., 1992, 46, 6671–6687.
21. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
22. H. J. Monkhorst and J. D. Pack, Phys. Rev. B: Solid State, 1976, 13, 5188–5192.
23. J. Paier, M. Marsman, K. Hummer, G. Kresse, I. C. Gerber and J. G. Angyán, J. Chem. Phys., 2006, 124, 154709.
24. B. Modak and S. K. Ghosh, Phys. Chem. Chem. Phys., 2018, 20, 20078–20087.
25. V. Louis-Achille, L. De Windt and M. Defranceschi, Comput. Mater. Sci., 1998, 10, 346.
26. P. H. J. Mercier, Z. Dong, T. Baikie, Y. Le Page, T. J. White, P. S. Whitfield and L. D. Mitchel, Acta Crystallogr., 2007, 63, 37–48.
27. L. Calderín, M. J. Stott and A. Rubio, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 67, 134106.
28. P. Rulis, L. Z. Ouyang and W. Y. Ching, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 70, 155104.
29. C.-X. Li, Y.-H. Duan and W.-C. Hu, J. Alloys Compd., 2015, 619, 66–77.
30. Q. Zeng, H. Liang, G. Zhang, M. D. Birowosuto, Z. Tian, H. Lin, Y. Fu, P. Dorenbos and Q. Su, J. Phys.: Condens. Matter, 2006, 18, 9549–9560.
31. P. Modak and B. Modak, Phys. Chem. Chem. Phys., 2020, 22, 16244–16257.
32. B. Modak, P. Modak and S. K. Ghosh, J. Phys. Chem. C, 2017, 121, 12980–12990.
33. M. Sen, R. Shukla, N. Pathak, K. Bhattacharyya, V. Sathian, P. Chaudhury, M. S. Kulkarni and A. K. Tyagi, Mater. Adv., 2021, 2, 3405–3419.
34. (a) R. Phatak, N. Pathak, S. Muhammed, S. K. Sali and A. Das, ChemPlusChem, 2018, 83, 1144–1152; (b) R. Phatak, N. Pathak, S. Muhammed, A. Das and S. K. Sali, J. Am. Ceram. Soc., 2020, 103, 2617–2629; (c) N. Pathak, S. Mukherjee, D. Das and D. Dutta, J. Alloys Compd., 2021, 887, 161414.
35. (a) N. Pathak, S. Mukherjee, B. P. Mandal, A. K. Yadav, S. N. Jha and D. Bhattacharyya, Mater. Adv., 2020, 1, 2380–2394; (b) N. Pathak, P. S. Ghosh, S. Mukherjee and B. P. Mandal, RSC Adv., 2020, 10, 31070–31086; (c) N. Pathak, S. Mukherjee, D. Das, D. Dutta, S. Dash and R. M. Kadam, J. Mater. Chem. C, 2020, 8, 7149–7161; (d) S. Mukherjee, N. Pathak, D. Das and D. Dutta, RSC Adv., 2021, 11, 5815–5831; (e) D. Hebbar, K. S. Choudhari, N. Pathak, S. A. Shivashankar and S. D. Kulkarni, J. Alloys Compd., 2018, 768, 676–685; (f) K. Sanyal, N. Pathak, A. K. Yadav, B. Kanrar, R. M. Kadam, S. N. Jha, D. Bhattacharya and N. L. Misra, Dalton Trans., 2016, 45, 7650–7664; (g) S. K. Gupta, P. S. Ghosh, N. Pathak and R. M. Kadam, RSC Adv., 2016, 6, 42923–42932; (h) D. Hebbar, K. S. Choudhari, N. Pathak, S. A. Shivashankar and S. D. Kulkarni, Mater. Res. Bull., 2019, 119, 110544; (i) S. K. Gupta, N. Pathak and R. M. Kadam, J. Lumin., 2016, 169, 106–114; (j) S. K. Gupta, K. S. Prasad, N. Pathak and R. M. Kadam, J. Mol. Struct., 2020, 1221, 128776.
36. S. K. Gupta, N. Pathak, S. K. Thulasidas and V. Natarajan, J. Lumin., 2016, 169, 669–673.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ra04941k

---