A. Crystal chase and crystal structure

For Sr_2−x MgSi₂O₇:xCe³⁺, Sr_2−y MgSi₂O₇:yTb³⁺, and Sr_2−x−y MgSi₂O₇:xCe³⁺,yTb³⁺ samples, we select three representative Sr_1.99MgSi₂O₇:0.01Ce³⁺, Sr_1.97MgSi₂O₇:0.03 Tb³⁺, and Sr_1.94MgSi₂O₇:0.01Ce³⁺,0.05 Tb³⁺ samples to be detected by XRD. Figure 1 shows the representative XRD patterns of Sr_1.99MgSi₂O₇:0.01Ce³⁺, Sr_1.97MgSi₂O₇:0.03 Tb³⁺, and Sr_1.94MgSi₂O₇:0.01Ce³⁺,0.05 Tb³⁺ samples calcined at 1100 °C for 2 h. It is apparent that all the diffraction peaks of samples can be matched well with the standard card JCPDS No. 75–1736 of Sr₂MgSi₂O₇. No other phase is observed, indicating that the doping ions can not cause significant change to the crystal structure. The Sr₂MgSi₂O₇ host is a tetrahedral (space group $P\bar 421m$ ). The unit cell parameters of the host are a = b = 7.9957 Å, c = 5.1521 Å, and V = 329.38 Å respectively. The crystal structure of Sr₂MgSi₂O₇ is exhibited in Fig. 2, according to the previously reported crystallographic data (ICSD 155,330). ^{Reference Kimata26} As shown in Fig. 2(a), the Si atom is surrounded by four oxygen atoms, forming a four-coordination [SiO₄] tetrahedron structure. Two [SiO₄] tetrahedrons connect together to form [Si₂O₇] by sharing a common oxygen atom. The Mg atom, at 2a site, constitutes a [MgO₄] tetrahedron with surrounding four oxygen atoms. These two kinds of tetrahedrons connect together in a staggered way to constitute the polyhedral layers, which is shown in Fig. 2(b). The polyhedral layers are composed of [SiO₄] and [MgO₄]. The advantage of the layers is that it can isolate Sr²⁺ by acting as a shield and resist the attack of oxygen even at high temperature. ^{Reference Saradhi, Boudin, Varadaraju and Raveau27} And the Sr atom constructs an eight-coordination [SrO₈] decanedron by surrounding oxygen atoms. A number of [SrO₈] decanedrons constitute [SrO₈] _n layers between the layers which are composed of [SiO₄] and [MgO₄]. Table I lists the ionic radii of Sr²⁺, Mg²⁺, Ce³⁺, and Tb³⁺. ^{Reference Shannon28} The radii of Ce³⁺ and Tb³⁺ are close to the radius of Sr²⁺. The charges of above ions are similar, two Ce³⁺ or Tb³⁺ may substitute three Sr²⁺ sites to remain the charge balance. On the basis of ionic radii and charge balance, Ce³⁺ and Tb³⁺ tend to occupy the Sr²⁺ sites.

FIG. 1.

XRD patterns of Sr_1.99MgSi₂O₇:0.01Ce³⁺, Sr_1.97MgSi₂O₇:0.03 Tb³⁺ and Sr_1.94MgSi₂O₇:0.01Ce³⁺,0.05 Tb³⁺ samples.

FIG. 2.

(a) The coordination environment of Si²⁺, Mg²⁺, Sr²⁺. (b) Crystal structure of Sr₂MgSi₂O₇ viewed along [010]. (c) Staggered arrangements of [SiO₄] and [MgO₄] units along [001].

TABLE I.

Ionic radii (Å) of different cations for the given coordination number (CN).

B. Photoluminescence properties of Ce³⁺ doped phosphors

Figure 3(a) and (b) present the PL spectra and relative PL intensity of Sr_2−x MgSi₂O₇:xCe³⁺ (x = 0.05–0.09), respectively. Figure 3(a) shows a broad violet–blue asymmetric emission band ranging from 370 to 500 nm with excitation at 348 nm. Originally, the PL intensity of Sr_2−x MgSi₂O₇:xCe³⁺ increase apparently with the augment of Ce³⁺. While the doping concentration of Ce³⁺ continuously increases, the intensity gradually descends. Higher doping concentration is not conducive to luminescent intensity. With increasing the concentration of doping ions, the lattice imperfection of as-prepared samples caused by doping ions will increase proportionally. However, a mass of defect will result in declining luminous intensity with energy transfer. As shown in Fig. 3(b), the luminescent intensity is optimized with the doping concentration of Ce³⁺ to be 0.01. When x is beyond 0.01, the spectra shows a negative correlation between the intensity and the doping concentration of Ce³⁺. The augment of Ce³⁺ makes the luminescence intensity decrease called concentration quenching. ^{Reference Pauling29} There are three possible major factors to cause concentration: (i) The solid-solubility of activator in the host. High doping concentration of rare earth ions not only substitute the host-ions or fill the interstice but also may react with the host to form new phase. New phase will cause the decrease of luminescence intensity. As shown in Fig. 1, doping ions don’t cause the host lattice change significantly. Therefore, concentration quenching is not caused by new phase. (ii) The effect of lattice symmetry on luminescent intensity. ^{Reference Tamura and Shibukawa30} Electric-multiple-transition is forbidden to high symmetry sites. Low doping content brings lattice distortion causing lower luminescent intensity, while the distortion may appear the new symmetry to cause concentration quenching with the increase of doping content. (iii) The exchange interaction between rare earth ions. With the increasing doping content of Ce³⁺, the distance between ions gradually decreases. A certain short distance will induce tunneling effect to cause concentration quenching. ^{Reference Adane, Klein, Leger, Belin and Defourneau31} The effect of electronic multiple transition and exchange interaction can be interpreted by Dexter’s energy resonance theory. ^{Reference Dexter and Schulman32} The concentration quenching is induced by the energy transfer between activators at certain ionic distance. To investigate the concentration quenching mechanisms among Ce³⁺, the Van Uitert expression is shown as follows ^{Reference Van Uitert33} :

(1) $${I \over x} = K{\left[ {1 + \beta {{\left( x \right)}^{{Q \mathord{\left/ {\vphantom {Q 3}} \right. \kern-\nulldelimiterspace} 3}}}} \right]^{ - 1}}\quad ,$$

where x is the activator ions concentration, I/x is the emission intensity (I), and K and β are constants at the same excitation condition for the same host. The values of Q are 6, 8, and 10, corresponding to the energy transfer mechanisms of electric dipole–dipole (d–d), dipole–quadrupole (d–q), or quadrupole–quadrupole (q–q) interactions, respectively. For the Sr₂MgSi₂O₇ host, the relationship of log(I/x) and log(x) is shown in Fig. 4. The slope of fitting line is −1.23. Thus, the value of Q is 3.69 which is close to 3. And the value of Q is far less than 6, testifying that the energy transfer of Ce³⁺ is not belonging to electric dipole–dipole interactions between Ce³⁺. The result indicates that the process of energy transfer among the nearest neighboring ions is the main mechanism of concentration quenching behavior among Ce³⁺ in Sr₂MgSi₂O₇ host. Local structural variation among Ce³⁺ makes the short range exchange interaction improve with the increase of Ce³⁺ content, the energy will transfer rapidly between neighboring ions to reach the quenching center. ^{Reference Zhang, Huang, Liu, Wang, Jiang, Zhang, Wang, Kong, Xi and Ji34} In addition, the emission peaks show a slight red shift with the increasing concentration of Ce³⁺, which could be attributed to the effect of the crystal field effect surrounding the Ce³⁺. In fact, the position of the emitting level of Ce³⁺ depends on two factors ^{Reference Blasse and Bril17} : (i) the position of the center of gravity of the 5d level, which will be at lower energy if the valence increases, and (ii) the width of the crystal field splitting of the 5d level. For the first factors, the higher charge of Ce³⁺ makes the reduction of the 5d level position of the center of gravity, shrinking the energy of transition between 5d and 4f, while increasing concentration of Ce³⁺ intensifies the crystal field effect. For the second factor, it is connected with the crystal field theory. According to the crystal field theory, the crystal field strength is influenced by site symmetry, valence, ligand charge, and bond length. ^{Reference Jang, Im, Lee, Jeon and Kim35} Crystal field splitting (D _q) is expressed as follows ^{Reference McClure36} :

(2) $${D_{\rm{q}}} = {1 \over 6}Z{e^2}{{{r^4}} \over {{R^5}}}\quad ,$$

where D _q is a measure of the energy level separation, Z is the charge of the anion, e is the point charge, r is the radius of the wave function of d level, and R is the bond length. When Ce³⁺ substitute the Sr²⁺ sites, the distance between Ce³⁺ and O²⁻ will be shorter because Ce³⁺ has a smaller ionic radii (1.143 Å for CN = 8). Crystal field splitting is proportional to 1/R ⁵, the shorter bond length improves the effect of crystal field splitting. Thus, the reduction of the 5d level position and the increase of the width of crystal field cause the red shift. As shown in Fig. 5, the PLE spectrum of Sr_1.99MgSi₂O₇:0.01Ce³⁺ shows a strong and broad excitation band which consists of three peaks located at 252 nm, 282 nm, 348 nm, respectively. The broad excitation band is attributed to 4f → 5d transition of Ce³⁺ and matches well with near-UV chips. In addition, the emission band of Sr_1.99MgSi₂O₇:0.01Ce³⁺ can be fitted by two Gaussian peaks which are located at 394 nm and 426 nm, respectively. It is well known that the emission spectrum of Ce³⁺ should be composed of double bands in view of the transition from 5d to ² F _7/2 and ² F _5/2 states which are separated by crystal field splitting. ^{Reference Zhang, Huang, Yang, Lu, Wang, Xu, Zhang, Wang, Kong, Xi and Ji37} The theoretical energy difference of splitting between ² F _7/2 and ² F _5/2 of Ce³⁺ is about 2000 cm⁻¹. ^{Reference Lin and Su38} And the energy difference between 394 and 426 nm is about 1907 cm⁻¹, which is close to the theoretical value. Thus, we can demonstrate that these two Gaussian peaks derive from the ground splitting of the single Ce³⁺ emission center.

FIG. 3.

(a) PL spectra of Sr_2−x MgSi₂O₇:xCe³⁺ under 348 nm excitation. (b) Dependence of emission intensity on Ce³⁺ concentration.

FIG. 4.

The relationship between the log(I/x) and log(x) of Sr_2−x MgSi₂O₇:xCe³⁺.

FIG. 5.

PLE and PL spectra of Sr_1.99MgSi₂O₇:0.01Ce³⁺.

C. The influence of sintering temperature on photoluminescence properties

We select the Sr_1.99MgSi₂O₇:0.01Ce³⁺ samples to investigate the influence of sintering temperature on photoluminescence properties for Sr₂MgSi₂O₇ phosphors. The PL intensity of the samples is superior to other samples in the doping concentration aspect. Figure 6(a) shows the PL spectra of Sr_1.99MgSi₂O₇:0.01Ce³⁺ calcined at different temperature with excitation at 348 nm. The variation trend is demonstrated clearly in Fig. 6(b). According to the PL spectra, the optimal sintering temperature is obtained which is 1100 °C. Initially, the PL intensity increases with the temperature increasing. When temperature is beyond 1100 °C, the intensity gradually decreases. The augment of temperature contributes to the improvement of crystallinity of phosphors. ^{Reference Metz39} However, higher temperature may induce extra defects of host lattice. The defects, considered as quenching centers, consume transition energy by lattice vibration, resulting in downgrading the luminescent intensity and going against the crystallinity.

FIG. 6.

(a) PL spectra of Sr_1.99MgSi₂O₇:0.01Ce³⁺ phosphors at different sintering temperature under 348 nm excitation. (b) Dependence of emission intensity on sintering temperature.

D. The influence of sintering time on photoluminescence properties

To investigate the influence of sintering time on photoluminescence properties for Sr₂MgSi₂O₇ phosphors, we select the Sr_1.99MgSi₂O₇:0.01Ce³⁺ samples calcined at 1100 °C to research and analysis. The PL intensity of the samples is superior in doping concentration and sintering temperature. Figure 7(a) displays the PL spectra of Sr_1.99MgSi₂O₇:0.01Ce³⁺ synthesized at different sintering time monitored at 348 nm. It also shows a broad emission band ranging from 370 to 500 nm with excitation at 348 nm. Initially, increasing the sintering time strengths the emission intensity. When time is beyond 2 h, the intensity gradually decreases and then becomes stable. The variation trend of emission intensity is shown in Fig. 7(b). The trend shows that the optimum condition of sintering time is 2 h for luminescent intensity. Properly extending synthesis time contributes to improve the crystallinity, ^{Reference Kumar, Khan and Chawla40} while long sintering time may lead to Ce³⁺ producing more lattice vibration or phonons to consume energy, which makes total luminous intensity descend. In addition, long sintering time may increase the defect and distortion of crystal structure. The system will absorb extra energy to counteract the effect brought by the defect and distortion, causing the reduction of the emission intensity.

FIG. 7.

(a) PL spectra of Sr_1.99MgSi₂O₇:0.01Ce³⁺ phosphors under different sintering time at 1100 °C under 348 nm excitation. (b) Dependence of emission intensity on sintering time.

E. The morphology of Ce³⁺ doped phosphors

Figure 8 shows the morphology of a series of samples with different doping content of Ce³⁺ by field emission scanning electron microscopy (FESEM) equipped with energy dispersive spectroscopy (EDS). In general, Ce³⁺ doping does not have much influence on the micromorphology of Sr₂MgSi₂O₇ phosphor. For pure Sr₂MgSi₂O₇, irregular and dispersive particles can be observed in Fig. 9(a). The size of particles is about 1 μm, exhibiting a proper particle size and promising prospect for the WLEDs application. ^{Reference Guo, Huang, You, Yang, Liu and Zheng41} With the increase of Ce³⁺, the morphologies do not change significantly. And the EDS spectrum of doped Ce³⁺ samples reveals that the powders consist of Sr, Mg, Si, O, Ce. The peak corresponding to C is obtained in the spectra because of conductive adhesive. Table II lists the content of each element. It indicates that Ce³⁺ have been successfully doped into Sr₂MgSi₂O₇ host.

FIG. 8.

The FESEM images of Sr_2−x MgSi₂O₇:xCe³⁺ phosphors with different doping concentration and EDS result of Sr_1.99MgSi₂O₇:0.01Ce³⁺ phosphors. (a) x = 0. (b) x = 0.005. (c) x = 0.01. (d) x = 0.03. (e) x = 0.05. (f) x = 0.07. (g) x = 0.09. (h) EDS result of Sr_1.99MgSi₂O₇:0.01Ce³⁺ phosphors.

FIG. 9.

(a) PL spectra of Sr_2−y MgSi₂O₇:yTb³⁺ (y = 0.01, 0.03, 0.05, 0.07, 0.09) under 243 nm excitation. (b) Dependence of emission intensity on Tb³⁺ concentration.

TABLE II.

Each element content of EDS result.

F. Photoluminescence properties of Tb³⁺ doped phosphors

Figure 9 shows the PL spectra and relative PL intensity of Sr_2−y MgSi₂O₇:yTb³⁺ respectively. Under 243 nm excitation, the PL spectra exhibit sharp and strong characteristic emission of Tb³⁺ due to the transitions of ⁵ D ₃ → ⁷ F _J in the blue region and ⁵ D ₄ → ⁷ F _J in the green region (J = 6, 5, 4, 3, 2). Originally, the PL intensity of Sr_2−y MgSi₂O₇:yCe³⁺ increase apparently with the augment of Tb³⁺. While the doping concentration of Tb³⁺ continuously increases, the intensity gradually descends. The optimal concentration of Tb³⁺ is 0.03. When doping content of Tb³⁺ exceeds 0.03, an apparent concentration quenching appears among Tb³⁺. The variation tendency is demonstrated clearly in Fig. 9(b). To investigate the concentration quenching mechanism among Tb³⁺, the quenching mechanisms can also be acquired by Van Uitert expression. ^{Reference Van Uitert33} The values of Q are 6, 8, and 10, corresponding to the energy transfer mechanisms of electric dipole–dipole (d–d), dipole–quadrupole (d–q), or quadrupole–quadrupole (q–q) interactions, respectively. For the Sr₂MgSi₂O₇ host, the relationship of log(I/x) and log(x) is shown in Fig. 10, the slope of fitting line is −1.09. Thus, the value of Q is 3.27, which is close to 3. And the value of Q is far less than 6, testifying that the energy transfer of Tb³⁺ not belongs to electric dipole–dipole interactions between Tb³⁺. The result indicates that the process of energy transfer among the nearest neighboring ions is also the main mechanism of concentration quenching behavior among Tb³⁺ in Sr₂MgSi₂O₇ host. Figure 11(a) shows the PLE and PL spectra of Sr_1.97MgSi₂O₇:0.03 Tb³⁺ samples. The excitation spectrum (λ_em = 543 nm) exhibits two broad bands ranging from 220 to 265 nm with strong peak at 243 nm and one weak peak at 275 nm which are attributed to the 4f ⁸ –4f ⁷5d ¹ transitions. The ground state of Tb³⁺ includes high-spin (HS) ⁹ D _J and the low-spin (LS) ⁷ D _J states. ^{Reference Ning, Mak and Tanner42} The two excitation peaks correspond to two kinds of f–d transitions. One is spin-allowed with high energy, and the other spin-forbidden with low energy. ^{Reference Ten Kate, Zhang, Dorenbos, Hintzen and Van Der Kolk43} Figure 11(b) demonstrates the weak and sharp peak from 360 to 400 nm which is related to the 4f–4f intra-transitions of Tb³⁺ ions. The obtained emission spectrum consists of the f–f transition lines within the Tb³⁺ 4f ⁸ electron configuration, i.e., ⁵ D ₃ → ⁷ F ₅ (378 nm), ⁵ D ₃ → ⁷ F ₄ (414 nm), ⁵ D ₃ → ⁷ F ₃ (436 nm), ⁵ D ₄ → ⁷ F ₆ (486 nm), ⁵ D ₄ → ⁷ F ₅ (543 nm), ⁵ D ₄ → ⁷ F ₄ (579 nm), and ⁵ D ₄ → ⁷ F ₃ (618 nm), respectively. ^{Reference Zhang, Huang, Jin, Sun, Ma, Zhang, Wang, Kong, Xi and Ji44} The sharp emission peaks of Tb³⁺ also verify that Tb³⁺ are not sensitive to local crystal structure due to the f–f transition. Thus, the shape of peaks can hold a characteristic emission spectrum. ^{Reference Ye, Xiao, Pan, Ma and Zhang45,Reference Zhang, Zhou, Pang, Shi and Gong46} Comparing two situations of doping, it illustrates that silicate is a kind of host which is appropriate for phosphors with the stable physical and chemical property. Meanwhile, doping two different rare earth ions achieve the thing that we can tune the emission wave length by control doping ions and their concentration to obtain the optimal effect.

FIG. 10.

The relationship between the log(I/x) and log(x) of Sr_2−y MgSi₂O₇:yTb³⁺.

FIG. 11.

(a) PLE and PL spectra of Sr_1.97MgSi₂O₇:0.03 Tb³⁺. (b) The magnified excitation spectra (λ_em = 543 nm) of Sr_1.97MgSi₂O₇:0.03 Tb³⁺ from 320 to 400 nm.

Figure 12 shows the morphologies of as-prepared Sr_2−y MgSi₂O₇:yTb³⁺ phosphors and the EDS result of Sr_1.97MgSi₂O₇:0.03 Tb³⁺ phosphors. With the increase of Tb³⁺, the morphologies do not change significantly as well as the phosphor morphologies of doping Ce³⁺. And the EDS spectrum of doped Tb³⁺ samples reveals that the powders consist of Sr, Mg, Si, O, Tb. The peak corresponding to C is also obtained in the spectra because of conductive adhesive. Table III lists the content of each element. It proves that Tb³⁺ have been successfully doped into Sr₂MgSi₂O₇ host.

FIG. 12.

The FESEM images of Sr_2−y MgSi₂O₇:yTb³⁺ phosphors with different doping concentration and EDS result of Sr_1.97MgSi₂O₇:0.03 Tb³⁺ phosphors. (a) x = 0.01 (b) x = 0.03. (c) x = 0.05. (d) x = 0.07. (e) x = 0.09. (f) EDS result of Sr_1.97MgSi₂O₇:0.03 Tb³⁺ phosphors.

TABLE III.

Each element content of EDS result.

Figure 13 shows the CIE chromaticity coordinates diagram of Sr_2−y MgSi₂O₇:yTb³⁺ (y = 0.01, 0.03, 0.05, 0.07, 0.09) phosphors. The chromaticity coordinates (X, Y) for Sr_2−y MgSi₂O₇:yTb³⁺ phosphors are shown in Table IV. They are (0.2802, 0.6790), (0.2806, 0.6761), (0.2788, 0.6819), (0.2792, 0.6827), (0.2826, 0.6765), respectively. With increasing the concentration of Tb³⁺, the CIE chromaticity coordinate remains a steady position, which indicates doping concentration have little impact on emitting wave length shifts and CIE chromaticity coordinates are relative stable with the identical doping ions.

FIG. 13.

The CIE chromaticity coordinates diagram for Sr_2−y MgSi₂O₇:yTb³⁺ (y = 0.01, 0.03, 0.05, 0.07, 0.09) phosphors.

TABLE IV.

CIE coordinates of Sr_2−y MgSi₂O₇:yTb³⁺ samples.