Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ : a novel apatite green-emitting phosphor for near-ultraviolet excited w-LEDs of Materials Chemistry

A novel apatite phosphor Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ was prepared by conventional high-temperature solid-state reaction. Phase purity was examined by XRD and XPS analysis. The crystal structure information, such as space group, cell parameters and atomic coordinates, were refined by the Rietveld method, revealing that Eu 2+ occupied the sites of Ca 2+ ions. Moreover, low-temperature experiments, including low-temperature PL spectra and low-temperature decay curve, were used to prove the existence of two luminescence centers in Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ . With the increase in doping concentration of Eu 2+ , the emission wavelength shows a red shift from 498 nm to 510 nm, which is mainly caused by the increase in crystal-field splitting by Eu 2+ . The optimized concentration of Eu 2+ was confirmed to be 0.01, the R c was calculated to be 20.09 Å and the energy transfer between Eu 2+ was demonstrated to be by exchange interaction. Moreover, good thermal stability has been proved by a temperature-dependence experiment; it shows that the phosphor can maintain 55% of emitting intensity at 150 1 C compared to that at room temperature. Finally, the Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ phosphor was fabricated with commercial red (CaAlSiN 3 :Eu 2+ ) and blue (BAM:Eu 2+ ) phosphor coating on a n-UV chip. This proves that this green phosphor has the potential to be used in a w-LED lamp.


Introduction
Nichia assembled a white light-emitting diode (w-LED) by means of a blue InGaN LED chip coated with yellow phosphor Y 3 Al 5 O 12 :Ce 3+ (YAG:Ce 3+ ) in 1996. [1][2][3] At that time the w-LED was regarded as the next generation light source due to many advantages such as significant power reduction, higher luminous efficiency, longer lifetime and environmental friendliness. 4 Accordingly, enormous research efforts have been focused on the preparation of high quality w-LEDs. 5,6 Compared with Nichia's assembly method mentioned above, which shows high correlated color temperature (CCT E 7750 K) and a low color rendering index (CRI E 75), 7 packaging multicolor phosphors on a near-ultraviolet (n-UV) chips can provide relatively high quality white light because the emission spectra can be distributed over the entire visible range (400-800 nm). [8][9][10] This means that the study of high-performance single-color phosphor is meaningful.
Researchers have paid great attention to green-emitting phosphors, for instance the cordierite structure phosphor K 0.5 Mg 1.5 Al 4 Si 5 O 13 :Eu 2+ , 11 the garnet structure phosphors Ca 2 LaZr 2 Ga 3 O 12 :Ce 3+ , 12 Ca 2 YZr 2 Al 3 O 12 :Ce 3+ , 13 and the commercial green phosphor (Ba,Sr)SiO 4 :Eu 2+ . However, there is still room for improvement in some aspects such as wider emission region, better thermal stability and higher quantum efficiency. It is well known that the apatite compound has been utilized as a highly efficient host material for phosphor due to its adjustable crystal structure and its excellent thermal and physicochemical stability. [14][15][16] Recently, our group has reported a series of apatite solid solution phosphors Ca 2+x La 8Àx (SiO 4 ) 6Àx (PO 4 ) x O 2 :Eu 2+ (x = 0, 2, 4, and 6) whose emitting color can be adjusted from green to blue by changing the ratio of Ca to La. One of these compounds, Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ , show good potential as a commercial green phosphor. Therefore, we consider it worth investigating the crystal structure and luminescence properties of Ca 6 La 4 (SiO 4 ) 2 -(PO 4 ) 4 O 2 :Eu 2+ in more detail.
As discussed above, the novel apatite phosphors Ca 6 La 4 -(SiO 4 ) 2 (PO 4 ) 4 O 2 :xEu 2+ (x = 0.05-0.1) have been synthesized by high-temperature solid state reaction for the first time. The phase purity was examined by XRD and XPS and the crystal structure information was analyzed by Rietveld refinement using XRD profiles. In addition, it is the first time that the low-temperature PL and PLE spectra and decay curves were utilized to demonstrate the inclusion of Eu 2+ in the apatite phosphor. Moreover, the luminescence properties, such as PLE and PL spectra, lifetimes and thermal stability, are discussed in detail. Finally, the novel green phosphor Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ was packaged on a u-UV LED chip (excitation wavelength 385 nm) with commercial red and blue phosphors, demonstrating the potential for it to be utilized in a w-LED.

Characterization
The X-ray diffraction (XRD) data of the series of phosphors were measured on an X-ray powder diffractometer (D/max-IIIA, Rigaku, Japan) with a stepwise scanning mode over the 2y range 101-1101 using Cu-K a radiation (1.5406 Å) with operating voltage 40 kV and current 100 mA. The XRD patterns, used in Rietveld refinement, were acquired with a step size of 0.021 and counting time of 2 s per step. X-ray photoelectron spectroscopy (XPS) measurements were collected with a Kratos Axis Ultra DLD, employing MCP stack & delay-line photoelectron detector with scanned & snapshot spectroscopy modes. The photoluminescence emission (PL) spectrum and the photoluminescence excitation (PLE) spectrum at 298 K and 8 K were measured using a Hitachi F-4600 fluorescence spectrophotometer (Japan) equipped with a 150 W Xe lamp as the excitation source. The temperature-dependent luminescence properties were measured on the same spectrophotometer, which was assembled with a computer-controlled electric furnace and a self-made heating attachment. The room-temperature luminescence decay curves were obtained from a spectrofluorometer (Horiba, Jobin Yvon TBXPS) using tunable pulsed laser irradiation (nano-LED) as the excitation. Quantum efficiency was measured by a Fluoromax-4 spectrofluorometer with an integral sphere at room temperature (Horiba, Jobin Yvon).

Photoluminescence characteristics
The PLE spectrum of Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :0.01Eu 2+ (monitored at 500 nm) and the PL spectra of Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :xEu 2+ (x = 0.005-0.1) (under 365 nm excitation) are displayed in Fig. 6a and b, respectively. Due to the 4f 7 -4f 6 5d 1 transition of Eu 2+ , the PLE spectrum depicts a broad excitation band ranging from 280 nm to 400 nm with a maximum value at 330 nm, indicating that the Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ phosphor can be suitably excited by a n-UV LED chip. 21 The PL spectrum consists of an asymmetric broad emission band peaking at 500 nm, which is ascribed to the electric dipole-allowed transition of Eu 2+ from the lowest level of the 5d excited state to the 4f ground state. 22 In general, the asymmetric emission band is due to the different coordination environments of Eu 2+ . As shown in Fig. 6c, the asymmetric emission band of Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :0.01Eu 2+ has been resolved      In theory, the width of the crystal-field splitting of the Eu 2+ 5d level is the main influence on the position of the Eu 2+ emitting energy level, 24 i.e., the lower the lowest crystal-field position, the lower the emission energy. 25 The major difference between the 4f and 6h site is that the 6h site has a free oxygen ion, which does not belong to any silicate group. Therefore, the binding strength is not saturated compared with the 4f site resulting in the average covalency of the 6h site being higher than that of the 4f site. Consequently, Eu 2+ will have lower emission bands when located at the 6h site than when at the 4f site and the higher energy emission band (492 nm) can be assigned to Eu 2+ (4f) while the lower energy emission band (539 nm) is derived from Eu 2+ (6h). 16 This inferred conclusion can be further demonstrated by an empirical relationship proposed by Van Uitert: 15 where E represents the position of the d-band edge energy for rareearth ions (cm À1 ), Q is the position energy for the lower d-band edge for the free ion (34 000 cm À1 for Eu 2+ ); V is the activator ion valence (for Eu 2+ V = 2), n represents the number of anions in the immediate shell about the Eu 2+ ion, r is the host cation radius that is replaced by the Eu 2+ ion (in Å), and E a is the electron affinity of the atoms that form anions (in eV). In general, due to the complexity of the local crystal structure, the exact energy levels of Eu 2+ at a specific site are hard to calculate precisely. 26 However, it can be deduced approximately that the value of E is proportional to the quantity of n and r. 4 For the as-prepared phosphors, r is equal because both of the substitutions occurring at cation sites are that Ca 2+ is replaced by Eu 2+ . Therefore, the band peak at 492 nm is ascribed to the 4f 6 5d 1 -4f 7 transition of Eu 2+ occupying the M(I) site with nine-coordination, whereas the longer wavelength band can be attributed to Eu 2+ occupying the M(II) site with seven-coordination.
In addition, as shown in Fig. 6b, with the increase of doping concentration of Eu 2+ , the peak emission wavelength shows an obvious red-shift from 498 to 510 nm. The red-shift phenomenon is mainly due to the variation in the crystal-field splitting of Eu 2+ . In this regard, the crystal-field splitting of Eu 2+ can be determined as obeying: 1 where Dq is a measure of the energy level separation, z represents the charge or valence of the anion, R is the distance from the central ion to its ligands, e is the charge of an electron, and r is the radius of the d wave function. For the d (Eu-O) -orbital, while z, e and r are equal, then Dq is merely a function of 1/R 5 . Along with the small Ca 2+ substituted by the larger Eu 2+ , the distance between Eu 2+ and O 2À becomes shorter and the magnitude of the crystal-field strength increases. 27 In consequence, the crystalfield splitting of the Eu 2+ ion increases, which causes a decrease in the emission energy from the 5d excited state to the 4f ground state and a resultant red-shift. 1,13,28 The concentration dependence of the PL intensity of the as-prepared samples Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :xEu 2+ (x = 0.005-0.1) are depicted in Fig. 8. With the increase of Eu 2+ content, the PL intensity shows an increase at first and it reaches the maximum value at x = 0.01. Subsequently, the intensity tends to decrease gradually due to the concentration quenching effect. Thus, the optimized concentration of Eu 2+ was confirmed to be 0.01.
In general, concentration quenching can occur in a compound by obeying two mechanisms: exchange interaction and multipolar interaction. 25,29,30 In consequence, as shown in Fig. 9, the concentration quenching originates from energy loss at a 'killer' center owing to the transfer of excitation energy among activators at a relative higher concentration. 17,31 The mechanism of energy transfer between the Eu 2+ can be expressed by the following equation: 32 where x is the concentration of activator, k and b are constants for each interaction for a given host lattice. Van Uitert has defined y = 3 for the exchange interaction and y = 6, 8, 10 correspond to dipole-dipole, dipole-quadrupole, quadrupole-quadrupole interactions, respectively. As depicted in the inset of Fig. 8, one can observe that the log(I/xEu 2+ ) showed a relatively linear relationship with log(xEu 2+ ) and that the slope was calculated to be À1.168. In consequence, the value of y was found to be approximately 3, demonstrating the energy transfer type is exchange interaction. Therefore, as proposed by Blasse, the critical energy transfer distance (R c ) can be calculated as follows: 4,13 where V is the volume of the unit cell, x c is the critical concentration of Eu 2+ ions and N is the number of total Eu 2+ sites in the unit cell. For Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :xEu 2+ , V = 552.5 Å 3 and N = 13. Thus, the R Eu-Eu distance was determined to be 25.32, 20.09, 15.95, 12.66, 11.06, 10.05, 9.33 and 8.78 Å when x = 0.005, 0.01, 0.02, 0.04, 0.06, 0.08 and 0.1, respectively. Accordingly, this indicates that with the increase in doping concentration of Eu 2+ , R Eu-Eu decreases and the crystal-field splitting of the Eu 2+ 5d bands increased, leading to a continuous increase in the red-shift with the Eu 2+ dopant concentration. 33 The critical concentration (x c ) of Eu 2+ was determined to be 0.01 mol, therefore, the R c value for energy transfer was calculated to be 20.09 Å.
Room temperature decay curves of the as-prepared phosphors, under excitation at 365 nm, were measured and are plotted in Fig. 10. Each decay curve can be well fitted with a second-order exponential function by the following equation: 4 where I presents the luminescence intensity; A 1 and A 2 are constants; t is time and t 1 and t 2 are the lifetimes for the exponential components. Furthermore, the average lifetime constant (t*) can be calculated as follows: 4 The measured lifetimes monitored at 500 nm were calculated to be 412.9, 429.1, 383.4, 336.5, 257.5, 125.3 and 54.1 ns with Eu 2+ contents = 0.005, 0.01, 0.02, 0.04, 0.06, 0.08 and 0.1, respectively. One can observe that with the increase of Eu 2+ concentration, the measured lifetimes of Eu 2+ increase to a maximum when x = 0.01 and then decrease rapidly. This phenomenon is a typical sign of energy transfer and concentration quenching. 22 The measured lifetime is also related to the total relaxation rate by: where t 0 is the radiative lifetime; A nr presents the non-radiative rate ascribed to multi-phonon relaxation and P t is the energy transfer rate between Eu 2+ ions. The distance between Eu 2+ ions decreases with the increase of Eu 2+ concentration, resulting in the increases of energy transfer rate between Eu 2+ and the probability of energy transfer to luminescent 'killer' sites. Consequently, the lifetimes show a gradually decreasing trend with the increase of Eu 2+ concentration. 34 Moreover, the two decay components (t 1 and t 2 ) were detected in both the room-temperature and lowtemperature decay curves, further demonstrating that Eu 2+ occupies two different Ca 2+ sites in Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 host. In general, temperature is the factor that has a great influence on the emitting intensity of a phosphor. 12 Therefore, thermal stability is one of the important characteristics that needs to be taken into consideration before recommending the phosphor for potential applications. 15 The PL spectra of the temperature dependence experiment are shown in Fig. 11. As we can observe from Fig. 11, the luminescence intensity shows a gradual decrease when the sample was heated and the phosphor still maintains 55% of the emitting intensity at 150 1C compared with the initial intensity   The increase of FWHM can be explained by a configuration coordinate diagram, a physical model, that electron-phonon interaction will enhanced and the population of higher vibration levels will increase when the temperature rises, 18 and then, the excited luminescent center is thermally activated and release non-radiative through the crossover between the excited states and the ground states with the help of phonons. In consequence, the luminescence is quenched due to the enhanced population density of phonons, which also causes the broadening of the PL spectra. 4,17 On the other hand, the PL spectra at different temperatures show a slight blue-shift, which is attributed to thermally-active phonon-assisted tunneling from the excited states of the lowenergy emission band to the excited states of the high-energy emission band. 11,12 As mentioned before, there are two emission centers for the Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ phosphor. The thermal back transfer from lower energy centers to higher energy centers will be promoted at higher temperature. 23,35 Consequently, the higher energy emission is strengthened and the blue-shift phenomenon is observed. In addition, the expanded host lattice can also cause a reduction in the crystalfield splitting and lead to a higher energy emission when the temperature increases. 4,36 Luminescence efficiency is an important technological parameter for the application of a phosphor. The internal quantum efficiency (QE) of Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :0.01Eu 2+ was measured and calculated to be 57.73%. It is higher than that of K 0.5 Mg 1.5 Al 4 Si 5 O 13 :Eu 2+ (48.3%) 11 and Ca 2 LaZr 2 Ga 3 O 12 :Ce 3+ (35.2%), but lower than the commercial green phosphor (Ba,Sr)SiO 4 (79%). 12 In general, the QE can be further optimized by improving the preparation conditions because the QE depends closely on crystalline defects, particle sizes and the morphology of the phosphor. 8,37 To demonstrate the potential application, the electroluminescent spectrum of w-LED lamp is displayed in Fig. 12. The self-made lamp was fabricated by coating Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ , BAM:Eu 2+ and CaAlSi-N 3 :Eu 2+ phosphors on a n-UV chip (l ex = 385 nm). The CIE color coordinates, CCT and Ra of the self-fabricated w-LED lamp were calculated to be (0.37, 0.37), 4253 K and 86, respectively. The relatively high Ra value (86) and appropriate CCT value (4253 K) demonstrate that the Ca 6 La 4 (SiO 4 ) 2 (PO 4 ) 4 O 2 :Eu 2+ can be a promising candidate for a green-emitting phosphor for the application of w-LEDs.