Recent progress of effect of crystal structure on luminescence properties of Ce3+-Eu2+ Co-doped phosphors.

Currently, the mechanism of Ce3+-Eu2+ ET is frequently used to obtain color adjustable or white phosphors. Correspondingly, the ET efficiency from Ce3+ to Eu2+ becomes an important indication of the luminescent properties of phosphors. However, the ET efficiency calculated using the formula does not always match the emission spectra; the transmission efficiency of Ce3+ is high, but the emission efficiency of Eu2+ is low, depending on our investigation results. In addition to this problem, here we mainly review, on the basis of substantial examples, how to boost the actual ET efficiency of Ce3+-to-Eu2+ and thus to improve the luminescent properties of phosphors through the rational design of layered crystal structure and the way of selective occupation of activator ions. Moreover, the possible physical mechanisms are proposed. This journal is © The Royal Society of Chemistry.

Introduction

In recent years, a broad spectrum of new lighting materials has emerged, and white LED (WLED) materials have gradually replaced the traditional fluorescent lamps and incandescent lamps because of their energy-saving, environmental protection, high efficiency, long life and small size.[1] Due to these advantages, LEDs can be used in a wide range of areas, not only in solid state lighting, but also in other electronic devices such as displays. At present, researchers prefer to obtain color adjustable phosphors or white light phosphors by rational design of ET through rare earth (RE) ions co-doping such as Ce3+–Eu2+, Ce3+–Mn2+, Eu2+–Mn2+, Ce3+–Eu2+–Mn2+, Ce3+–Tb3+–Mn2+ and Ce3+–Tb3+–Eu2+ systems, and so forth.[2-14] Among these systems, both Ce3+ and Eu2+ can emit 4f–5d transitions. Because their exposed 5d orbitals are susceptible to the crystal field environment, the luminescence range of Ce3+ can vary from the violet to red, and the spectral coverage of Eu2+ can even range from the ultraviolet to near-infrared, such as that in K3LuSi2O7.[15] Therefore, rational design of Ce3+-to-Eu2+ ET has become a popular way to obtain white light phosphors.

Over the past decades, many phosphors emitting adjustable color have been prepared by doping RE ions, and we could draw a conclusion that there are two main factors affecting the luminescent properties of phosphors: the kind of activator ions and the type of hosts. Both are indispensable for phosphors to achieve good luminescent properties. First, even for the same host, different kinds of activator ions will show different phenomena.[16-19] For example, solid state synthesis and tunable luminescence of Li2SrSiO4:Eu2+/Ce3+ phosphors have been reported by Wei et al. Under excitation at 365 nm, the Ce3+ exhibits a broad blue emission ranging from 380 to 500 nm centered at 425 nm, while the Eu2+ exhibits a broad green emission in the range of 480–680 nm with center at 556 nm.[20] Synthesis, luminescence and ET of Ce3+-to-Eu2+ of SrSc2O4:Ce3+/Eu2+ phosphors have been reported by Zhao et al. Under excitation at 440 nm, SrSc2O4:Ce3+ shows a green emission peaking at ∼535 nm, and the spectral coverage ranges from 475 to 675 nm. Under excitation at 523 nm, SrSc2O4:Eu2+ shows a red emission ranging from 650 to 775 nm with center at ∼702 nm.[21] Photoluminescence, ET and tunable color of Ce3+/Tb3+/Eu2+ activated oxynitride phosphor with high brightness has been reported by Lv et al. Ca3Al8Si4O17N4:Ce3+ shows a blue emission centering at 420 nm in the range of 340–525 nm, under excitation at 330 nm. And Ca3Al8Si4O17N4:Eu2+ shows a blue emission in the range of 400–550 nm with center at 450 nm under excitation at 350 nm.[22] It can be seen that in the same host, different kinds of activator ions lead to different luminescence phenomenon.

In addition to the type of activator ions, the choice of host also has an impact on the luminescent properties of phosphors.[23-40] Eu2+ is a significant activator for luminescent materials because of its unique emission properties, so it has been researched in many hosts, as proved by the abundant examples such as K3LuSi2O7:Eu2+, NaAlSiO4:Eu2+, Ca2Al3O6F:Eu2+, K2BaCa(PO4)2:Eu2+, Sr9Mg1.5(PO4)7:Eu2+, etc.15,41–44 First of all, with the variation of hosts, the emission wavelength of Eu2+ can change from the near-ultraviolet to near-infrared. Secondly, the thermal stability of phosphor materials is different in different hosts, which can be proved by the phosphors of KBa2(PO3)5 (ref. 45) and Ca2Al3O6F (ref. 42) reported by Zhao and Xia, respectively. According to these two papers, the PL intensity of Ca2Al3O6F:Eu2+ at 150 °C is only about 39% of its initial value (25 °C), but the PL intensity of KBa2(PO3)5:Eu2+ at 150 °C retains 97% of its initial value.[42,45] In addition to the luminous color and the temperature stability of phosphors, host will also affect the quantum efficiency of phosphors. For example, in host CaAlSiN3, the quantum efficiency of Eu2+ is as high as 95%, but in host Ca7Mg(SiO4)4, the quantum efficiency of Eu2+ is only 30%.[23,29] Therefore, the effect of crystal field environment on the luminescence properties of Eu2+ is significant, including the position and half-peak width of luminescence, as well as the temperature stability and quantum efficiency of phosphors.

From the above discussion, we can conclude that host composition and crystal structure have a great effect on the luminescence properties of Eu2+-singly doped phosphors, and in fact, for Ce3+–Eu2+ co-doped phosphors, these factors also play a key role. The luminescence properties and ET mechanism of Ce3+-to-Eu2+ have been analyzed by some research groups.[46-93] For instance, the ET processes of Ce3+-to-Eu2+ in Ca2PO4Cl,[92] Ba3Ca2(PO4)3F[10] and Ba3Y(PO4)3 (ref. 93) were once studied by our group. The overlaps between the emission spectra of Ce3+ and the excitation spectra of Eu2+, the luminescence intensity and fluorescence lifetime of Ce3+ with changing the Eu2+ doping concentration were investigated in detail. After years of study in this field, we have a deeper understanding of the luminescence performance and ET mechanism of many Ce3+–Eu2+ co-doped systems, and therefore putting forward the guiding ideology of improving the application of WLED. Despite the progress, there are still some problems to be further explored. Generally, the ET efficiency of Ce3+-to-Eu2+ is calculated based on the luminescence intensity of Ce3+ together with the formula . Sometimes, this calculated result is in accord with the emission spectra of Ce3+ and Eu2+. For example, in the hosts Ba2MgSi2O7,[63] Ba3Si6O12N2 (ref. 49) and BaMg2(PO4)2,[87] their ET efficiency was calculated to be 90%, corresponding well to the emission spectra that the luminous intensity of Ce3+ decreased significantly and the luminous intensity of Eu2+ increased obviously. However, in the hosts such as LiSr4(BO3)3,[72] Ba3Y(PO4)3,[93] Na5Ca2Al(PO4)4,[56] although the ET efficiency of Ce3+-to-Eu2+ was calculated to be 80% according to the formula , it does not agree with the emission spectra of Eu2+ which did not increase correspondingly. Therefore, it is worth considering and exploring whether this phenomenon is related to the crystal structure and composition of host. If this problem can be solved, it may provide some clues and ideas when designing and selecting new Ce3+–Eu2+ and even other RE ions co-doped materials.

Subject source

So far, there have some useful reviews summarizing on the Ce3+–Eu2+ doped pc-WLED phosphors for solid-state lighting and displays such as those by Li,[94] Terraschke & Wickleder,[95] Xia,[96,97] Qiao,[98,99] Qin,[100] Wang,[101] Lin,[102] Li[103]et al. Especially, several recent reviews and articles have systematically introduced the design, the optical properties and the application of pc-WLED phosphors with the emphasis on the dependence of luminescence properties on the crystal structure and composition of host.[104-135] Different from these work, here we provide a new perspective to review the recent reports. In order to find out the reason why the ET efficiency calculated according to the formula does not match the Eu2+'s emission spectra, we have carefully investigated the crystal structure, luminescence location, quenching concentration and other related information of many hosts, as exhibited in Tables 1 and 2. As the hosts with similar crystal structure and composition have similar luminescent phenomenon, so we focus on the crystal structure of these hosts. Moreover, it was found that the actual ET efficiency of Ce3+-to-Eu2+ in the hosts with layered crystal structure is higher than those with non-layered crystal structure. Based on these results, the crystal structures of all hosts involved in this paper are then divided into two categories: layered structure and non-layered structure, in the following discussion.

Layer spacing, crystal structure, space group, and photoluminescence properties of Ce3+–Eu2+ co-doped layered structured luminescent materials

Layered structure
Compound	L (Å)	Crystal structure	Space group	λ em (Ce3+)	λ em (Eu2+)	X c (Eu2+)	R c (Å)	Mechanism of ET	Ref.
BaMg2(PO4)2	9.691	Trigonal	P1̄	412 nm	447 nm	0.05	30.7	d–d	87
Ba1.2Ca0.8SiO4	7.352	Trigonal	P3̄m2	385 nm	450 nm	0.02	17.11		80
SrSi2O2N2	7.231	Trigonal	P1̄	395 nm	530 nm	0.02	26.05	d–d	17
Ba3Si6O12N2	6.784	Trigonal	P3	393 nm	526 nm	0.01	25.11	d–d	49
Li2SrSiO4	5.522	Trigonal	P3	420 nm	570 nm	0.01	27.62		20
Ba3Si6O9N4	6.784	Trigonal	P3	420 nm	520 nm	0.005	24.86	d–d	59
Ba2MgSi2O7	5.406	Tetragonal	P4̄21m		560 nm	0.04			63
Sr2Al2SiO7	5.264	Tetragonal	P4̄21m	415 nm	510 nm	0.03	18.3	d–d	57
Ca2Mg0.25Al1.5Si1.25O7	5.05	Tetragonal	P4̄21m	409 nm	520 nm	0.025		d–d	46
Ca2Mg0.5AlSi1.5O7	5.036	Tetragonal	P4̄21m	410 nm	525 nm	0.015	26.71	d–d	4
Ca8Mg3Al2Si7O28	5.029	Tetragonal	P4̄21m		535 nm	0.01			40
BaSi2O2N2	7.592	Orthorhombic	Pbcn	390 nm	495 nm	0.05	24.87	d–d	79
Ba2SiO4	7.513	Orthorhombic	Pmcn		505 nm	0.03			39
Sr2SiO4	6.148	Orthorhombic	Pmnb	406 nm	469 nm	0.01			24
CaAlSiN3	5.636	Orthorhombic	Cmc21	580 nm	650 nm	0.008	22.75	d–d	85
Ba2ZnS3	4.21	Orthorhombic	Pnam	498 nm	660 nm	0.008	32.7	d–d	77
Ba4Si6O16	6.983	Monoclinic	P1211	442 nm	497 nm	≥0.025			66
LiBaBO3	6.416	Monoclinic	P121/c1	430 nm	530 nm	0.02	25.06		70
Ca2Si5N8	6.991	Monoclinic	C1c1	465 nm	595 nm	0.0035	22.27	d–d	50

Crystal structure, space group, and photoluminescence properties of Ce3+–Eu2+ co-doped non-layered structured luminescent materials

Non-layered structure
Compound	Crystal structure	Space group	λ em (Ce3+)	λ em (Eu2+)	X c (Eu2+)	R c (Å)	Mechanism of ET	Lattice site of Eu2+	Ref.
Ca2PO4Cl	Orthorhombic	Pbcm	370 nm	460 nm	0.07	14.7	d–d	Ca1	92
SrSc2O4	Orthorhombic	Pnam	535 nm	702 nm	0.04	14.498	d–d	Sr	21
BaSiO3	Orthorhombic	P212121	404 nm	570 nm	0.02			Ba	48
SrB2Si2O8	Orthorhombic	Pnma	373 nm	440 nm	0.01	25.82	d–d	Sr	51
Ca1.65Sr0.35SiO4	Orthorhombic	P21cn	465 nm	538 nm		17	d–d	Sr1 to Sr6	88
Ca3MgSi2O8	Monoclinic	P21/a	400 nm	470 nm	0.03	18.64	d–d	Ca	60
Ca3Si2O7	Monoclinic	P121/a1	490 nm	625 nm	0.0125	13.7	d–d	Ca1 to Ca3	75
Ca4(PO4)2O	Monoclinic	P1121	460 nm	356 nm	0.01	23.8	d–d	Ca1 to Ca8	83
Sr2LiSiO4F	Monoclinic	P21/m11	400 nm	465/520 nm	0.0075			Sr1 and Sr2	74
Ca2BO3Cl	Monoclinic	P21/c	410 nm	573 nm	0.015		d–d	Ca1 and Ca2	55
Ca3SiO4Cl2	Monoclinic	P21/c	400 nm	505 nm	0.006			Ca	67
Sr5(PO4)3Cl	Hexagonal	P63/m	370 nm	443 nm	0.1	21	d–d	Sr2	65
La5Si2BO13	Hexagonal	P63/m	380 nm	519 nm	0.05	10.66		La1 and La2	54
Ba3Ca2(PO3)3F	Hexagonal	P63/m	387 nm	508 nm	0.05	17.506	d–d	M1 and M2	10
Ca4Y6O(SiO4)6	Hexagonal	P63/m	426 nm	527 nm	0.1	18.7	d–d	—	69
Ca9Al(PO4)7	Trigonal	R3c	356 nm	445 nm	0.01	12.64	d–d	Ca	78
Sr3B2O6	Trigonal	R3̄ch	434 nm	574 nm	0.002	30.7	d–d	Sr1 and Sr2	53
Ca9Y(PO4)7	Trigonal	R3c	342 nm	493 nm	0.005	30.644	d–d	Ca1 to Ca3, Y	52
LiSr4(BO3)3	Cubic	Ia3̄d	423 nm	612 nm	0.004	29.14	d–d	Sr1 and Sr2	72
Sr3Gd(PO4)3	Cubic	I43̄d	374 nm	518 nm	0.01	32.02	d–d	Sr/Gd	62
Ba3Y(PO4)3	Cubic	I43̄d	398 nm	520 nm	0.01	19.5	d–d	Ba/La	93

As shown in Fig. 1, for the hosts with layered crystal structure, high, medium and low actual ET efficiency of Ce3+-to-Eu2+ separately make up 65%, 20% and 15% of the whole cases. In contrast, high, medium and low actual ET efficiency of Ce3+-to-Eu2+ make up 31%, 24% and 45% of all the hosts with non-layered crystal structure, respectively. It is obvious that the actual ET efficiency of Ce3+-to-Eu2+ in the hosts with layered crystal structure is significantly higher than that with non-layered crystal structure.

Fig. 1

The proportion of the host with high, medium and low ET efficiency in the hosts with layered and non-layered crystal structure.

According to previous reports, in the host with layered crystal structure, the quenching concentration is relatively large, which indicates that this structure can inhibit the quenching effect of activated ions to a certain extent. For the hosts with layered crystal structure, the cation layer is separated by other anion groups or other cationic polyhedrons, thus widening the distance between the same cationic sites. According to the diagram of concentration quenching process (Fig. 2), we found that when the doping concentration of activator ions is relatively low, the average distance of activators is relatively large. Under such conditions, it is hardly any interaction between activators, so less ions forms one luminescence center that is shown in the left part of Fig. 2. With the increase of activator ions concentration, the distance between activator ions is shortened, which thus increases the probability of ET, causing the energy loss of activator ions through ET step by step, thus multiple ions will form one equivalent luminescence center that is shown in the right part of Fig. 2.[6] Due to the particularity of crystal structure, the distance between the same cations in the hosts with layered crystal structure is far, therefore reducing the probability of ET of activator ions and make the emission efficiency of Eu2+ is higher.

Fig. 2

Diagram of concentration quenching process.

In order to find out the further explanation, the hosts with layered crystal structure listed above are investigated in detail. Owing to the special layered crystal structure, the same cations are far apart from each other, which effectively inhibits the concentration quenching effect of activator ions. So for the layered structure, the larger the spacing between layers (L) is, the more obvious the effect of inhibiting the concentration quenching of activator ions is. And then, the higher the doping concentration of activator ions is, the greater the luminous intensity is. Since the activator ion Eu2+ is sensitive to the changes of surrounding crystal field environment, the quenching concentration (QC) of activator ion is expected to change when changing the spacing between layers (L). The hosts with layered crystal structure can be divided into three categories according to crystal system: tetragonal, trigonal and orthorhombic (Fig. 3a–c). All the hosts listed in Fig. 3a are tetragonal with space group P4̄21m, those listed in Fig. 3b are trigonal with space group P3̄ or P1̄, and the hosts listed in Fig. 3c are all orthorhombic. The insets in Fig. 3a and b are the crystal structure diagrams of hosts seen from different directions. In Fig. 3c, although the hosts are all layered structures, they can be divided into the layered structures with and without connecting bonds due to different space groups. It can be seen from Fig. 3 that the QC of Eu2+ increases with the increase of spacing between layers (L) in the same crystal system. It is further proved that with the increase of distance between the cationic sites occupied by activator ions, the effect of inhibiting the concentration quenching of activator ion becomes greater. Therefore, the doping concentration of activator ions becomes higher and the luminescence property of phosphors becomes better.

Fig. 3

The relationship between the spacing between layers (L) and the QC of Eu2+ in (a) tetragonal, (b) trigonal and (c) orthorhombic host.

It should be noticed, however, that the luminescence properties of some hosts with non-layered structure are also excellent, which can be explained as follows. Some of the hosts with non-layered structure are separated by the surrounding anionic groups, while others have gaps between the same cations by means of the selective occupancy of activator ions, which is called the interstitial structure. This structure can also indirectly increase the distance between the same cationic sites, followed by the inhabitation of the quenching effect of activator ions and the increase of doping concentration of activator ions. As shown in Fig. 4, the 21 hosts listed can be divided into five categories according to their crystal systems: orthorhombic, hexagonal, cubic, monoclinic and trigonal. And the proportions of these five crystal systems are 23.81%, 19.05%, 14.29%, 28.57% and 14.29%, respectively. In Fig. 4, different colors represent different crystal systems. For the hosts within the same crystal system, the unshadowed portion represents the host with interstitial structure, while the shadowed portion represents the host whose crystal lattice is occupied by activator ions that is tightly connected (we call it the cluster structure). It can be seen from the diagram that the QC of activator ions in the interstitial structure is higher than that in the cluster structure for all crystal systems. This further demonstrates that the larger the distance between the cationic sites occupied by the activator ions is, the greater the effect of inhibiting the concentration quenching of activator ions is. And then, the greater the doping concentration of activator ions is, the better the luminescence property of phosphors is.

Fig. 4

Effect of selective occupancy of activator ions on the QC of Eu2+ in non-layered hosts.

As is well known, the emitting intensity of activator ions will be weak if the doping concentration is quite small. Therefore, inhibiting the concentration quenching of activator ions can increase the doping concentration of activator ions, thus enhancing the intensity of luminescence. Based on this conclusion, we found that the design of layered crystal structure and the selective occupation of activator ions can inhibit the concentration quenching of activator ions and improve the luminescent properties of Ce3+–Eu2+ co-doped phosphors to some extent. Next, we will take some reported hosts as an example for specific analysis.

Effects of crystal structure on luminescence properties of Ce3+–Eu2+ co-doped phosphors

Improving luminescence performance of Ce3+–Eu2+ co-doped phosphors by designing selective occupancy of activator ions

A large amount of Ce3+, Eu2+ singly-doped and Ce3+–Eu2+ co-doped phosphors have been reported. According to these reports, Ce3+ and Eu2+ are sensitive to the changes of surrounding crystal field environment. Specifically, their luminescence location, spectral range, quenching concentration and thermal stability will be different with the variations of crystal field environment. Therefore, it is a common way to adjust the luminescent properties of phosphors by changing the crystal field environment around activator ions. In particular, the method of improving the luminescent properties of phosphors by designing the occupation sites of activator ions has been intensively studied.[127-135]

Since the luminescent properties of Ce3+ or Eu2+-singly doped phosphors can be improved by the method of activator ion selective occupation, the luminescence properties of Ce3+–Eu2+ co-doped phosphors is also expected to be improved using the same way. Currently, a large number of Ce3+–Eu2+ co-activated halogenated phosphate hosts have been found. Ba3Ca2(PO4)3F[10] and Sr5(PO4)3Cl[65] are two representatives, and both of them belong to the apatite structure system. As shown in Fig. 5a and d, the crystal structure of them belongs to the hexagonal system with P63/m space group. There are two kinds of cationic sites in Ba3Ca2(PO4)3F, which are signed as M1(Ca1/Ba1) and M2(Ca2/Ba2) to facilitate the following discussion. The M1 sites are surrounded by nine oxygen anions with local symmetry C3, forming a tricapped trigonal prism. The M2 sites are seven-fold coordinate surrounded by six oxygen anions and one fluorine anion with Cs point group symmetry. While in Sr5(PO4)3Cl, the cationic sites are signed as Sr1 and Sr2. From Fig. 5a and d we can see that each Sr1/M1 sites is surrounded by anionic polyhedron and Sr2/M2, every three Sr2/M2 polyhedrons form a cluster with F− ion through the vertex angle, and each cluster is separated by Sr1/M1 polyhedron and anionic polyhedron.

Fig. 5

(a) The crystal structure of Ba3Ca2(PO4)3F. (b) The PL spectra of Ba3Ca2(PO4)3F:Eu2+ phosphors (c) the PL spectra of a series of Ba3Ca2(PO4)3F:0.1Ce3+,xEu2+ phosphors with different Eu2+ concentrations excited at 287 nm. The inset shows the emission intensity of Ce3+ and Eu2+ with the increase of Eu2+ concentration in Ba3Ca2(PO4)3F phosphors. (d) The crystal structure of Sr5(PO4)3Cl. (e) The PL spectra of Sr5(PO4)3Cl:Eu2+ phosphors. (f) The PL spectra of a series of Sr5(PO4)3Cl:0.01Ce3+,xEu2+ phosphors with different Eu2+ concentrations excited at 317 nm. The inset shows the emission intensity of Ce3+ and Eu2+ with the increase of Eu2+ concentration in Sr5(PO4)3Cl phosphors. (Reproduced with permission from ref. 10 and 65, copyright 2013, 2017, J. Mater. Chem. C, J. Alloys Compd.)

It can be seen from the structure diagram that if the doped RE ions occupy both lattice sites, the distance between activator ions is likely to be smaller, which increases the possibility of concentration quenching as discussed above. As shown in Fig. 5b and e, the emission spectrum of Ba3Ca2(PO4)3F:Eu2+ contains asymmetrically broad green-emission band from 400 to 650 nm, which is ascribed to the two different sites of Eu2+.[10] According to the previous analysis of crystal structure, this spectrum can be decomposed well into two Gaussian bands with peaks centering at 498 and 534 nm, respectively. The Eu2(M2) site will lead to a shorter wavelength band due to the presence of F− anion, so the 498 nm emission comes from the Eu2(M2) site and the 534 nm emission comes from the Eu1(M1) site.[10] In contrast, in the host Sr5(PO4)3Cl, the emission spectrum of Eu2+ shows a symmetrically narrow emission band from 400 to 500 nm, which indicates that Eu2+ is likely to enter only one of the cation sites and therefore indirectly pull the distance between the same lattices, making the same cationic lattices farther apart and reducing the probability of ET between activator ions.[65]

According to the above inference, the doping concentration of Eu2+ in Sr5(PO4)3Cl is higher than that in Ba3Ca2(PO4)3F. It can be seen from Fig. 5c and f that in Sr5(PO4)3Cl, with the increase of Eu2+ doping concentration, the luminescence intensity of Ce3+ decreases greatly, while the luminescence intensity of Eu2+ increases sharply. When the doping concentration of Eu2+ reaches x = 0.5, the emission intensity of Eu2+ still does not show a downward trend.[65] In contrast, the emission intensity of Ce3+ in Ba3Ca2(PO4)3F decreases with the increase of Eu2+ doping concentration, and the emission intensity of Eu2+ increases first and then decreases with the maximum at x = 0.05.[10] Therefore, no matter in terms of the QC of Eu2+ or the actual ET efficiency of Ce3+-to-Eu2+, Sr5(PO4)3Cl is superior to Ba3Ca2(PO4)3F. This result proves that the previous inference is correct. The distance between cation lattices does affect the quenching concentration of Eu2+ and the ET efficiency of Ce3+-to-Eu2+. When the cation lattices are far apart, the ET efficiency of Ce3+-to-Eu2+ is relatively high. It is because the probability of ET between the same activator ions is reduced and the concentration quenching effect between activator ions is restricted to some degree.

The layered structure sometimes can be formed by selective occupation of activator ions in several hosts, which thus improves the luminescent properties of phosphors to a certain extent. It will be shown using the crystal structures of Ca2PO4Cl[92] and Ca2BO3Cl[55] as given in Fig. 6a and d. The Ca2PO4Cl crystallizes in the orthorhombic system with space group of Pbcm, and the crystal structure of Ca2BO3Cl belongs to the monoclinic system with space group of P21/c. There are two kinds of cationic sites (Ca1 and Ca2 sites) in Ca2PO4Cl and Ca2BO3Cl, as shown in Fig. 6a and d. The Ca1 and Ca2 sites of Ca2PO4Cl are arranged in layers, while in Ca2BO3Cl, the Ca1 and Ca2 sites are staggered. If the doped Eu2+ enter only one of the lattice sites, the far distance between the lattice sites will lower the probability of ET between activator ions, so the concentration quenching of Eu2+ does not occur easily when the doped concentration is low.

Fig. 6

(a) The crystal structure of Ca2PO4Cl. (b) The PL spectra of Ca2PO4Cl:Eu2+ phosphors. (c) The PL spectra of a series of Ca2PO4Cl:0.03Ce3+,yEu2+ phosphors with different Eu2+ concentrations excited at 370 nm. The inset shows the ET efficiency as a function of Eu2+ concentration in Ca2PO4Cl phosphors. (d) The crystal structure of Ca2BO3Cl. (e) The PL spectra of Ca2BO3Cl:Eu2+ phosphors. (f) The PL spectra of a series of Ca2BO3Cl:0.006Ce3+,yEu2+ phosphors with different Eu2+ concentrations excited at 320 nm. (Reproduced with permission from ref. 92 and 55, copyright 2014, 2019, RSC Adv., J. Mater. Sci.: Mater. Electron.)

It can be seen from Fig. 6b that the emission spectrum of Ca2PO4Cl:Eu2+ ranges from 400 to 530 nm under excitation at 370 nm, and the spectral profile of this emission are symmetrical with a relatively narrow full width at half maximum.[92] Thus, it is very likely that Eu2+ occupies only one of the cationic lattices after entering the host, and the luminescence intensity of Eu2+ does not quench when the doping concentration is y = 0.1, as can be seen from Fig. 6c.[92] This further confirms that Ce3+–Eu2+ system has relatively high ET efficiency due to the relatively high QC of Eu2+ caused by the large distance between cationic lattices. However, the emission spectrum of Ca2BO3Cl:Eu2+ contains asymmetrically broad emission band from 450 to 700 nm, which is ascribed to the two different sites of Eu2+.[55] It can be seen from the structure diagram (Fig. 6d) that if the doped RE ions occupy both lattice sites, the distance between activator ions will be closer when the concentration of activator ions doped becomes larger. The energy will gradually dissipate due to the ET between different lattice sites and the same lattice sites, so the activator ion quenches when the doping concentration is low. In Ca2BO3Cl, the luminous intensity of Ce3+ decreased with the increase of Eu2+, and the luminous intensity of Eu2+ increased first and then decreased with its maximum at x = 0.015.[55] Therefore, no matter in terms of the QC of Eu2+ or the actual ET efficiency from Ce3+ to Eu2+, Ca2PO4Cl is superior to Ca2BO3Cl. This further shows the luminescence properties of phosphors can be improved by selective occupancy of activator ions.

Improving the luminescence properties of Ce3+–Eu2+ co-doped phosphors by designing layered crystal structure

In addition to the above mentioned way of elective occupation of activator ions, choosing the host with layered crystal structure is also a method to improve the luminescence performance of phosphors. In recent years, many Ce3+–Eu2+ co-doped hosts with layered structure have been reported, among which BaMg2(PO4)2 (ref. 87) and Ba2Mg(BO3)2 (ref. 86) are two representatives. Although the two hosts are both layered structure, the crystal structures, the composition of cationic layers, and the distances between layers are all different. As can be seen from Fig. 7c and d, the luminous intensity of Ce3+ decreased with increasing Eu2+ doping concentration, while the luminous intensity of Eu2+ increased first and then decreased. Moreover, the actual ET efficiency from Ce3+ to Eu2+ in BaMg2(PO4)2 is significantly higher than that in Ba2Mg(BO3)2.[86,87] Therefore, it is reasonable to speculate that different distances between cationic layers cause different inhibition effects on the quenching of activator ions and thus different actual ET efficiency from Ce3+ to Eu2+.

Fig. 7

(a) The crystal structure of BaMg2(PO4)2. (b) The spacing between layers (L), critical distance (Rc) and quenching concentration (Xc) of BaMg2(PO4)2 and Ba2Mg(BO3)2. (c) The crystal structure of Ba2Mg(BO3)2. (d) The photoluminescence emission (PL) spectra of a series of BaMg2(PO4)2:3%Ce3+,x%Eu2+ phosphors with different Eu2+ concentrations excited at 323 nm. The inset shows the ET efficiency as a function of Eu2+ concentration in BaMg2(PO4)2 phosphors. (e) PL spectra of a series of Ba2Mg(BO3)2:2%Ce3+,x%Eu2+,2%Na+ phosphors with different Eu2+ concentrations excited at 296 nm. (Reproduced with permission from ref. 87 and 86, copyright 2014, 2012, J. Alloys Compd., J. Rare Earths.)

In fact, in addition to the above reasons, the critical distance between different hosts also affects QC effect. As shown in Fig. 7b, the distance between the cationic layers (Ba ions and Ba ions) in BaMg2(PO4)2 is 9.691 Å, and the critical distance of ET from Ce3+ to Eu2+ system is 30.7 Å. When the doping concentration of Eu2+ is x = 0.05, the luminescence intensity of Eu2+ reaches its maximum.[87] In contrast, the distance between the cationic layers (Ba ions and Ba ions) in Ba2Mg(BO3)2 is 6.312 Å, and the critical distance between Ce3+ and Eu2+ is 27.5 Å. When the doping concentration of Eu2+ is x = 0.025, the luminescence intensity of Eu2+ reaches its maximum.[86] It is obvious that BaMg2(PO4)2 is better than Ba2Mg(BO3) in terms of the distance between cationic layers, the critical distance between Ce3+ and Eu2+ and the QC of Eu2+. In addition, the actual ET efficiency from Ce3+ to Eu2+ in BaMg2(PO4)2 is significantly higher than that in Ba2Mg(BO3)2. It further indicates that for the hosts with similar composition and crystal structure, a greater distance between layers can lower the ET probability caused by the close distance between the layers of the same activated ions, thus inhibiting the concentration quenching of activator ions and leading to higher doping concentration of activator ions. In addition, large critical distance between Ce3+ and Eu2+ ensures that the doping concentration of activator ions can be effectively increased in the range where the ET from Ce3+ to Eu2+ can happen, thus making the actual ET efficiency from Ce3+ to Eu2+ of BaMg2(PO4)2 relatively high.

Although it has been mentioned above that the actual ET efficiency from Ce3+ to Eu2+ is relatively high in the hosts with layered crystal structure, it does not mean the hosts with layered structure always ensure a high actual ET efficiency of Ce3+-to-Eu2+. For example, the crystal structures of Ca2Mg0.25Al1.5Si1.25O7 (ref. 46) and Ca2Mg0.5Al1.5Si1.5O7 (ref. 4) as shown in Fig. 8a and b are all layered structures. It can be seen from the emission spectra presented in Fig. 8c and d that the luminous intensity of Ce3+ decreased greatly with the increase of Eu2+ doping concentration, whereas the Eu2+ emission only showed a slight increase, and the total emission intensity decreased. Obviously, the ET efficiency calculated according to the formula does not match the spectra.

Fig. 8

(a) and (b) The crystal structure of Ca2Mg0.25Al1.5Si1.25O7 and Ca2Mg0.5Al1.5Si1.5O7. (c) The PL spectra of a series of Ca2Mg0.25Al1.5Si1.25O7:0.06Ce3+,yEu2+ with different Eu2+ concentrations excited at 350 nm. (d) PL spectra of a series of Ca2Mg0.5Al1.5Si1.5O7:0.06Ce3+,zEu2+ phosphors with different Eu2+ concentrations excited at 350 nm. (Reproduced with permission from ref. 46 and 4, copyright 2014, 2016, Powder Technol. Lumin.)

In order to clarify this problem, we analyzed the crystal structure and the information obtained from the reported literatures, concluding that the crystal structure of these two hosts is layered structure. Moreover, the distances between each two layers of Ca2Mg0.25Al1.5Si1.25O7 (ref. 46) and Ca2Mg0.5Al1.5Si1.5O7 (ref. 4) are separately 5.05 Å and 5.036 Å, the smallest distances of the layered structure hosts with connective bonds listed in this paper. Based on the previous discussion, this may be due to the fact that the distance between cationic layers is smaller and the distance between the same cation lattices is closer, thus increasing the probability of ET between activator ions. Therefore, the emission of Eu2+ is quenched at low doping concentration. As shown in Fig. 8c and d, the luminescence intensity of Eu2+ decreased with increasing doping concentration, which further verifies that even if the crystal structure of a host with high calculated ET efficiency from Ce3+ to Eu2+ is layered, its actual ET efficiency might be relatively low due to the small distance between layers. Therefore, not every host with layered crystal structure has good luminescence performance. When the spacing between layers is small, the distance between the same cations will be closer, thus increasing the probability of concentration quenching of activator ions. Therefore, the hosts with layered crystal structure and large spacing between layers are highly recommended when designing new phosphors in the future.

Summary and outlook

To sum up, the effect of crystal structure on the luminescence properties of phosphors is reviewed in this paper. It is found that the actual ET efficiency from Ce3+ to Eu2+ in the hosts with layered structure is higher than that with non-layered structure. It is because in these hosts with layered structure, the same cation lattice is far apart due to the special crystal structure, thus reducing the probability of ET caused by the close distance between activator ions and making the quenching concentration of activator ions relatively high. Therefore, the actual ET efficiency from Ce3+ to Eu2+ in these hosts is relatively high. Although the luminescence performance of the hosts with layered crystal structure is excellent, the ET efficiency from Ce3+ to Eu2+ in some hosts such as Ca2Mg0.25Al1.5Si1.25O7 (ref. 46) and Ca2Mg0.5Al1.5Si1.5O7 (ref. 4) is, however, not so ideal. It is because the spacing between layers is small, and the distance between the cation sites is close, which increases the probability of ET of activator ions and leads to concentration quenching when the doping concentration is small. In addition, it is also found that for some non-layered matrices, the luminescent properties can also be improved by the method of selective occupation of activator ions. By using this method, the crystal structure of hosts can be changed from cluster structure to layered or interstitial structure, which indirectly increases the distance between the same cation sites, restrains the activator ion concentration quenching, and make the emission efficiency of Eu2+ is higher, the luminous performance of phosphor material is better.

Over the past decades, great progress has been made in the field of WLED, especially the single-phase WLED excited by the near-ultraviolet radiation. Since the positions of the excited 4f65d1 state relative to the ground state are strongly influenced by host lattice, the position of emission band depends strongly on the environmental conditions of Eu2+ and Ce3+ such as coordination number, bond length, crystal field splitting, the nephelauxetic effect, the Stokes shift, the distortion of host lattice and so forth. It thus varies between the near ultraviolet and near infrared spectral range. These works have been reviewed a lot, but there is no general rule applicable to all phosphor systems at present. Exploring the structure-related luminescence mechanism is still the focus of future research.

Although a particular emphasis is placed on Ce3+–Eu2+ co-doped phosphors, we believe these theoretical approaches should be readily expandable to other lanthanide-doped systems with the characteristics of 4f–5d optical transitions, such as Ce3+–Mn2+ and Eu2+–Mn2+ co-doped systems. In addition, in some Ce3+–Eu2+ co-doped phosphor, the crystal structure of host not only affects the QC of activator ions and the ET efficiency from Ce3+ to Eu2+, but also affects the luminous color, thermal stability and quantum efficiency of phosphors. Therefore, these issues are also worthy of exploration, in order to meet different market demand for different performance LED.

It is worth noting that although the design of layered crystal structure and the selective occupation of activator ions can improve the luminescent properties of Ce3+–Eu2+ co-doped phosphors, these rules do not apply to all Ce3+–Eu2+ co-doped materials. The luminescence properties of phosphors are related not only to the crystal structure of host, but also to the overlapping degree of excitation spectrum of activator and emission spectra of sensitizer, synthesizing environment of materials, doping concentration of sensitizer and the critical distance between Ce3+ and Eu2+. In conclusion, despite the many contributions to the development of Ce3+–Eu2+ co-doped phosphors in recent years, there are still some open questions that need to make further study in the future.

Conflicts of interest

The authors declare no competing financial interest.