Break the Interacting Bridge between Eu3+ Ions in the 3D Network Structure of CdMoO4: Eu3+ Bright Red Emission Phosphor

Eu3+ doped CdMoO4 super red emission phosphors with charge compensation were prepared by the traditional high temperature solid-state reaction method in air atmosphere. The interrelationships between photoluminescence properties and crystalline environments were investigated in detail. The 3D network structure which composed by CdO8 and MoO4 polyhedra can collect and efficiently transmit energy to Eu3+ luminescent centers. The relative distance between Eu3+ ions decreased and energy interaction increased sharply with the appearance of interstitial occupation of O2− ions (O″i\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${O^{\prime\prime} }_{i}$$\end{document}). Therefore, fluorescence quenching occurs at the low concentration of Eu3+ ions in the 3D network structure. Fortunately, the charge compensator will reduce the concentration of O″i\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${O^{\prime\prime} }_{i}$$\end{document} which can break the energetic interaction between Eu3+ ions. The mechanism of different charge compensators has been studied in detail. The strong excitation band situated at ultraviolet and near-ultraviolet region makes it a potential red phosphor candidate for n-UV based LED.

Scientific REPORTS | (2018) 8:5936 | DOI: 10.1038/s41598-018-24374-3 reflectance from BaSO 4 as a reference. It is shown in the UV-vis diffuse reflectance spectra in Fig. 3 that these samples have strong absorption band in UV and near UV light region. For CdMoO 4 host, the correlation between the optical band gap E gap and absorption coefficient of semiconductor oxides can be determined by the following equation 23,24 : (2) 2 where v is the photon energy, ∞ R is the reflectivity of the sample, h is the Plank constant, and n is determined by the transition type (n = 1/2, 2, 3/2 or 3 for allowed direct, allowed indirect, forbidden direct and forbidden indirect electronic transitions, respectively).
According to the calculation results of electronic band structure in Fig. 4, the semi-conducting character of CdMoO 4 has been confirmed to be allowed indirect band gap material, therefore, n is equal to 2.  , the absorption boundary is 3.316 eV. For the samples with Eu 3+ doped, the weak absorption in the visible light region is consistent with the white color of powders for the naked eye. And compared with the commercial red emitting phosphors, CdMoO 4 :Eu 3+ phosphors can effectively avoid the disadvantages such as reabsorption in the visible light region.
In order to figure out the origin of the broad band absorption in the UV-vis diffuse reflection spectra of Fig. 3, the calculation of the electronic structure for CdMoO 4 host was carried out using CASTEP code by density functional theory. The Brillouin zone path for the band structure calculation was shown in Fig. 4(d) and Table 1. The calculated electronic band structure and electron orbitals are given in Fig. 4(a). The lowest energy (2.361 eV) of conduction band (CB) located at Z site, and the valence band (VB) maximum (0.00 eV) is at G site of Brillouin zone, indicating that CdMoO 4 is an indirect band gap compound and the energy band gap is about 2.361 eV. It is well known that the DFT calculations with CASTEP program generally underestimate the band-gap energies of the semiconductor materials due to the limited dimension of the atomic cluster. Therefore, a scissors operator of 0.955 eV was introduced to widen the gap to consistent with the measured optical band gap value (3.316 eV) of CdMoO 4 host.
After first-principle calculation, the electron orbitals were shown in Fig. 4b and c. It is well known that the d orbital splits into d xy , d xz , d yz , d z 2 and − d x y 2 2 in the tetrahedral crystal field due to the Jahn-Teller effect. The electron orbitals of the Mo atoms in the tetrahedron crystal field are shown in the Fig. 4(b). It can be seen that the excited electrons enter into the d z 2 and − d x y 2 2 orbitals with lowe energy. From Fig. 4(b), the states in the conduction band region are dominated by Mo 4d orbitals. Figure 4(c) shows the position of electrons in the valance band region. The shape of electron distribution is two hemispheres which belong to the typical 2p orbitals of O atoms. The diagram of electron orbitals illustrates that the top of VB is mainly composed by 2p orbital of O atom. Therefore, the absorption bands in Fig. 3 are ascribed to electron transitions from oxygen 2p orbital to an empty molybdenum 4d orbital.  Taking into account the lower display resolution of the electron orbitals diagram, theoretical calculations on the partial density of states (PDOS) for CdMoO 4 host were carried out and the results are illustrated in Fig. 5. Photoluminescence properties. A series of Cd 1−x MoO 4 :xEu 3+ samples were synthesized to optimize the doping concentration of Eu 3+ ions. Figure 6 shows the excitation spectra of Cd 1−x MoO 4 :xEu 3+ phosphors monitored at 5 D 0 − 7 F 2 (617 nm) emission of Eu 3+ ions. The broad excitation bands ranging from 200 nm to 380 nm are related to the charge transfer transition of O 2− → Mo 6+ and O 2− → Eu 3+ . According to the reported literature, the sharp peaks at 394 and 465 nm are due to 7 F 0 − 5 L 6 and 7 F 0 − 5 D 2 transitions of Eu 3+ , respectively. When the doping concentration of Eu 3+ was less than 4%, the relative intensity of excitation bands increases with the increase of Eu 3+ concentration. A more interesting thing is that the relative intensity of excitation bands, especially at about 328 nm, enhanced sharply with the increase of Eu 3+ concentration. However, when the doping concentration of Eu 3+ was beyond 4%, the intensity of excitation bands decreased with increasing Eu 3+ concentration.
From Fig. 7, PL and PLE spectra of CdMoO 4 , Cd 0.999 MoO 4 :0.001Eu 3+ and Cd 0.96 MoO 4 :0.04Eu 3+ phosphors are presented in an immediate contrast. In Fig. 7(a), pure CdMoO 4 sample exhibits a strong green emission band with a maximum at about 504 nm under 337 nm excitation. When monitoring at 504 nm, the PLE spectrum in Fig. 7(a) indicated that the as-synthesized CdMoO 4 phosphor exhibits a broad excitation band in the range of 250-380 nm. Three obvious excitation peaks for CdMoO 4 host were observed at room temperature, which can  be further deconvoluted by assuming a Gaussian type profile into the three excitation peaks at 300 nm, 328 nm and 348 nm, which are attributed to the charge transfer transition of O 2− → Mo 6+ and O 2− → Cd 2+ transitions.
As for the Eu 3+ singly doped CdMoO 4 sample, its PL and PLE spectra are presented in Fig. 7b and c. The PL spectrum under the excitation of 337 nm has a series of sharp line emissions at 594, 617,657 and 705 nm, due to the 5 D 0 − 7 F J (J = 1, 2, 3 and 4) characteristic transitions of Eu 3+ ions. The emission peak of CdMoO 4 host could not be observed owing to the energy transfer from CdO 8 and MoO 4 polyhedra to Eu 3+ ions. In the excitation spectrum of CdMoO 4 :Eu 3+ phosphors monitoring at 5 D 0 − 7 F 2 emission of Eu 3+ ions, the broad absorption band from 200 to 380 nm which was decomposed into four components by Gaussian fitting can be assigned to the combination of the charge transfer (CT) transitions. Compared with the excitation spectra of CdMoO 4 host, the new excitation band at about 246 nm was observed. From Figs 6 and 7b and c it can be seen that the excitation band at about 328 nm enhanced sharply. This is explained by the fact that the CdO 8 and MoO 4 polyhedra in three-dimensional network structure can transfer energy to Eu 3+ ions. Figure 8 illustrates the emission spectra of Cd 1−x MoO 4 :xEu 3+ (x = 0.001, 0.003, 0.006, 0.01, 0.015, 0.02, 0.03, 0.04, 0.05) phosphors excited at about 337 nm. It can be seen from the picture that the strongest emission peak is at about 617 nm which is caused by the 5 D 0 → 7 F 2 electric dipole transition of Eu 3+ ions. For the Cd 1−x MoO 4 :xEu 3+ phosphor, the emission intensities increased with the Eu 3+ concentration up to 0.04, and then decreases with increasing concentration due to the concentration quenching.
When Eu 3+ ions are doped into the CdMoO 4 host, with the increase of Eu 3+ doping concentration, the relative distance between Eu 3+ ions will be reduced gradually. Their distance (d) can be written as:   Fig. 9.
When the doping concentration of Eu 3+ is low, Eu 3+ ions are dispersed in the CdMoO 4 host. The relative distance decreases sharply with the increase of doping concentration. When the concentration of doped Eu 3+ is more than the critical concentration, it will no longer follow the Equation (4) due to the defect reaction in Equations (5 and 6). As a result, some of Eu 3+ ions will enter into the adjacent lattice sites and share a point defect ( ″ V Cd or ″ O i ) to achieve charge balance. The defect equations are as follows: Combined with the crystal structure, we believe that the position of point defects are as shown in Fig. 9. In CdMoO 4 :Eu 3+ phosphor, Eu 3+ is an isolated emission center which has been reported that the typical critical distance is about 5 Å. That is to say, the exchange interaction becomes effective when the Eu 3+ − Eu 3+ distance is shorter than 5 Å. When the concentration of Eu 3+ is low, the interaction of Eu 3+ − Eu 3+ can be almost neglected because of the long distance between Eu 3+ ions. However, the relative distance is only about 3.8 Å when two Eu 3+ ions occupy adjacent Cd sites. This indicates that the mechanism of exchange interaction plays an important role in energy transfer between the adjacent Eu 3+ ions in CdMoO 4 :Eu 3+ phosphor. And the point defect of ″ O i is also detrimental to the luminescence of Eu 3+ ions. This is the reason why concentration quenching occurred.
The charge imbalance caused by trivalent Eu 3+ ions occupying divalent Cd 2+ sites in the CdMoO 4 host affects the intensity of PL spectra. We believe that the distance between Eu 3+ ions can be regulated by charge compensation 25 . Therefore, the variation of PL intensity with different charge compensator doping concentrations is displayed in Fig. 10. As can be observed in this picture, the effect of charge compensators, Li + , Na + , K + ions and Cd 2+ vacancy, on red emission of Eu 3+ is enhanced markedly. The photoluminescence intensity of Cd 0.96−y MoO 4 :0.04Eu 3+ , yLi + , Cd 0.96−y MoO 4 :0.04Eu 3+ , yNa + and Cd 0.96−y MoO 4 :0.04Eu 3+ phosphors reached the maximum when the charge is balanced. This indicates that the charge imbalance affects the photoluminescence properties of Eu 3+ ions in CdMoO 4 host. The maximum PL intensity is reached when the concentration of K + ions in Cd 0.96−y MoO 4 :0.04Eu 3+ , yK + is 0.02. This is caused by the presence of lattice distortion that affects the photoluminescence properties. The lattice expansion corresponds to the larger ionic radius of K + (1.51 Å) than that of Cd 2+ (1.10 Å).
From above we can see that both Li + , Na + , K + and Cd 2+ vacancies can significantly enhance the fluorescence intensity of CdMoO 4 :Eu 3+ phosphors. In order to compare the photoluminescence enhancement of different charge compensators, the excitation spectra of Cd 0.96−y MoO 4 :0.04Eu 3+ , yM (M = Li + , Na + , K + ions and Cd 2+ vacancy) phosphors were shown in Fig. 11. Selected concentrations of charge compensator are the optimum doping concentration of Li + , Na + , K + ions and Cd 2+ vacancies. As can be seen from Fig. 11, the photoluminescence intensity is increased by two times when Na + ions were introduced as charge compensator. So why a small amount of charge compensator can greatly improve the fluorescence intensity?
The Mechanism of Charge Compensation. Figure 12 shows the radius and relative position of doping ions in charge compensated CdMoO 4 :Eu 3+ system. It can be seen that the ionic radius of Eu 3+ and Cd 2+ ions are approximately equal. Therefore, Eu 3+ can successfully occupy the Cd 2+ lattice. On the basis of this theoretical development, Li + , Na + , K + ions and Cd 2+ vacancies were introduced as a charge compensators because a large number of lattice defects appeared with the introduction of Eu 3+ ions.
The possible relative position of doping ions in CdMoO 4 host and photoluminescence enhancement mechanism is shown in Fig. 13. The diagram explains how the introduced Li + , Na + , K + ions and Cd 2+ vacancy break the interaction of Eu 3+ ions. The point defect reaction of the cation sublattice was written as Equations (5 and 6). In particular, a certain amount of interstitial occupation of O atoms ( ″ O i ) is introduced into the lattice due to the charge imbalance by Eu 3+ /Cd 2+ . This leads to the increase of cell parameters and the decrease of luminescence intensity of Eu 3+ ions. Due to the smaller ion radius of Li + ions, some of interstitial occupation of doping Li + ions ( . Li i ) and O 2− ions ( ″ O i ) appear in the Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Li + sample. The point defect reaction of the cation sublattice was written as Equation (7). Therefore, the interstitial occupation of Li + ions increased the cell parameters. When the concentration of Li + ions increased more than 0.04, A large amount of ( .  appear to reduce the photoluminescence intensity of Eu 3+ ions. The point defect reaction of the K + cation sublattice may be written as Equation (8). The large ionic radius of K + ions resulted in the highest lattice distortion in the sample. As can be seen in Fig. 13, the lattice shrinkage induced by the point defect of . Eu Cd and ″ V Cd were compensated by the great lattice expansion of the ′ K Cd defect. The appropriate concentration of K + ions gives the best optical performance. Because the appearance of . Eu Cd − ″ V Cd pair can stabilize the luminescent environment of the . Eu Cd − ′ K Cd pair. The point defects equation with Cd 2+ vacancy is Equations (9 and 10). Therefore, the introduction of ″ V Cd reduced the point defect of ″ O i and thus increased the photoluminescence intensity. The ionic radius of Na + ions is most approaching to Cd 2+ , and the introduction of Na + ions can produce a negative charge. The introduction of Na + ions can also reduce the structural distortion well because of the suitable ionic radius. This results in the concentration decrease of ″ O i and thus decrease the cell parameters. Therefore, as a charge compensator, Na + ions have an ideal ion radius and charge states. The point defect reaction of the Na + cation sublattice may be written as Equation (11). Eu 3+ and Na + ions will occupy the adjacent Cd 2+ ion lattice sites in order to achieve charge balance and reduce the lattice distortion. Therefore, the introduction of Na + ions can compensate the effect of the charge imbalance well, which can significantly enhance the luminescence intensity of Eu 3+ ions.    With the purpose of application for white-light LEDs with near-UV based chips as excitation sources, photoluminescence properties of Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + and commercial red emission phosphors are compared. Figure 14a and b shows the intuitive compared PL and PLE spectra of Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + and commercial Y 2 O 3 :Eu 3+ and Sr 2 Si 5 N 8 :Eu 2+ red emitting phosphors under the same measurement conditions. According to the compared spectra, the intensity of the broad excitation band at about 345 nm and sharp emission peak at about 617 nm in Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + phosphors were high enough to be comparable to commercial red-emitting phosphors. The 5 D 0 − 7 F 2 emission intensity of Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + is about 94% of Y 2 O 3 :Eu 3+ commercial phosphor. The CIE chromaticity diagram of prepared Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + and commercial red emitting phosphors was shown in the insert of Fig. 14. It can be seen that the optimal Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + approaches the National Television System Committee (NTSC) ideal red color (x = 0.67, y = 0.33) and it is much closer to the red region on the CIE chromaticity diagram than commercial Sr 2 Si 5 N 8 :Eu 2+ and Y 2 O 3 :Eu 3+ phosphors. Thus, as-prepared Cd 0.92 MoO 4 :0.04Eu 3+ , 0.04Na + phosphor is promising red-emitting phosphor for near-UV-LEDs.

Conclusion
In summary, a series of Eu 3+ and alkali metal ions co-doped CdMoO 4 phosphors were prepared by the solid-state reaction method in air atmosphere. In CdMoO 4 crystal structure, CdO 8 and MoO 4 polyhedra were connected with each other by coplanar to form a 3D network. Eu 3+ ions occupied Cd 2+ sites will be excited with the energy collected by 3D network structure. In addition, eight adjacent MoO 4 polyhedra can also effectively transfer energy to the Eu 3+ ion. However, the interaction between Eu 3+ ions is further amplified in the 3D network structure. The introduced Li + , Na + , K + ions and Cd 2+ vacancy which can avoid ″ O i will separate the couple of Eu 3+ ions and break the interaction between them. Since the suitable ionic radius and valence states, the introduction of Na + can break the energetic interaction of Eu 3+ ions through the 3D network structure. This is because the introduction of Na + ions can decrease the lattice distortion and avoid the appearance of ″ O i point defect. This mechanism provides a model how to use charge compensator to break the energetic interaction between Eu 3+ ions which expected to be used to design efficient luminescent materials.

Experiment
Materials and Synthesis. A series of Eu 3+ doped CdMoO 4 phosphors were synthesized by a traditional high temperature solid state reaction in air atmosphere. Stoichiometric amounts of CdO (99.9%), Eu 2 O 3 (99.99%) and MoO 3 (99%) were mixed homogeneously in an agate mortar. The mixtures were put into an alumina crucible and calcined in the muffle furnace at 800 °C for 3 hours in air atmosphere, and then the white powder phosphor was obtained. In some cases, appropriate amounts of Li 2 CO 3 (99.9%), Na 2 CO 3 (99.9%) and K 2 CO 3 (99.9%) were added as charge compensators. In order to reduce the amount of Cd in the sample and increase the concentration of Cd vacancy, the stoichiometry of CdO is intentionally reduced. All samples were ground into a powder in an agate mortar and pestle for further analysis.
Experimental methods. The phase purity was verified by the powder X-ray diffraction (XRD) measurement performed on a Bruker D8 Advance X-ray diffractometer (Cu Kα 1 radiation, λ = 0.15406 nm) and high-resolution X-ray diffraction was recorded over an angular (2θ) range of 5-130° with radiation at a 0.02 o (2θ)/s scanning step. Structural refinements of X-ray diffractograms were performed by the GSAS2 program. UV-Vis diffuse reflectance spectra (DRS) were collected using a V-670 (JASCO) UV-Vis spectrophotometer. The photoluminescence excitation (PLE) and emission (PL) spectra were recorded with a Hitachi F-4600 spectrophotometer equipped with a 150 W xenon lamp as an excitation source at room temperature. DFT calculations on the CdMoO 4 host was carried out by using the CASTEP code 21 . All the measurements were performed at room temperature.
Details of calculation. The CASTEP program was employed to determine the Band structure and Orbital population by density functional theory [26][27][28] . The Vanderbilt ultrasoft pseudopotential with a cutoff energy of