Controlled synthesis of the monoclinic and orthorhombic polymorphs of Sr₂SiO₄ activated with Ce³⁺ or Eu²⁺

Abstract

Phase-pure orthorhombic and monoclinic Sr₂SiO₄ and its Ce³⁺- or Eu²⁺-activated versions were synthesized using H₃BO₃ or SrCl₂ fluxes. The effects of the type of dopant and its concentration, as well as the type of flux on the crystallization were revealed and discussed. The luminescence properties of the phase-pure Ce³⁺- or Eu²⁺-activated powders were characterized both at 30 K and 300 K. The presence of two luminescent sites was proved in both types of structures and in the case of both activators, and their characteristic excitation and emission spectra were presented. Energy transfer from ions emitting at higher energies to those giving luminescence at lower energies was shown to take place in both polymorphs and in the case of each of the dopants. Eu²⁺ emissions of both sites in both structures were located fully in the visible part of the spectrum, while part of the Ce³⁺ luminescence occurred in the long-wavelength UV. Room-temperature radioluminescence spectra showed that only the site emitting at longer wavelengths, in both structures and for both dopants, is active in the scintillation process. For the first time, for both dopants, luminescence from the two sites in both polymorphs of Sr₂SrO₄ has been spectroscopically characterized at 30 K and at room temperature.

---

Introduction

Properly activated silicates have been recognized as efficient luminescent materials. They have found applications as scintillators or phosphors for lighting, recently for white LEDs.^1–26 The latest application, solid state lighting (SSL), has boosted research on Sr₂SiO₄ in recent years, as activation of this host with Eu²⁺ or Ce³⁺ appears to lead to luminescence useful in this field of technology.^{1–5,10–19,24–26}

It is already about thirty years since the first reports on the crystal structure of Sr₂SiO₄ were published by Catti and his co-workers.^27–31 The existence of two polymorphs, orthorhombic α′-Sr₂SiO₄ and monoclinic β-Sr₂SiO₄, together with some data on their phase transitions, were presented then.

The orthorhombic α′-form crystallizes in the Pmnb space group, while the monoclinic β-form crystallizes in the P12₁/n₁ space group. Most studies have proved that getting phase-pure Sr₂SiO₄, undoped or activated, is a difficult task, and till now there are no known clear rules governing the phase equilibrium in Sr₂SiO₄. For example, replacing some Sr²⁺ ions with either larger Ba²⁺ or smaller Ca²⁺ seemed to stabilize the orthorhombic structure.^{1,14,23,29,30} It was suggested that the stabilization is due to changes in the lengths of the metal–oxygen bonds.^{1,5,14,16,23,29,30}

Consequently, in most of the published papers on the luminescence properties of Sr₂SiO₄-based phosphors, mixtures of both phases were obtained and investigated.^4,5,12,14,25 Nevertheless, it was reported that pure orthorhombic^{13,15–20,22,24,26} or monoclinic phases^9,11,18,21 could be synthesized when Sr₂SiO₄ was doped with Eu²⁺, Eu³⁺, Dy³⁺ or Tb³⁺ ions. Thus, it appeared that just a small amount of the impurity ions was able to stabilize one of the two structures. It was indicated that getting the orthorhombic α′-form is possible when more than 1 mol% of dopants with ionic radii similar to that of the Sr²⁺ ion are used. In fact, the problem of getting phase-pure powders was so difficult to solve that products with only a residual impurity phase were sometimes claimed to be phase-pure. A close examination of the presented XRD patterns proves, however, the presence of a small amount of the other polymorph.^{1,2,6,8,10,23,25}

The two structures, orthorhombic α′-Sr₂SiO₄ and monoclinic β-Sr₂SiO₄, present some similarities important for the spectroscopic properties of the incorporated activators. Firstly, in each of them, Sr²⁺ ions occupy two different symmetry sites. Secondly, the coordination numbers (CN) of the metal sites in both phases are the same: CN = 10 (Sr(1)) and CN = 9 (Sr(2)). In Fig. 1 we present unit cells of both structures together with the local arrangements of the two Sr sites in each of them. In Table 1 we list the Sr–O and Sr–Sr distances in the orthorhombic α′-Sr₂SiO₄ and monoclinic β-Sr₂SiO₄ which might be helpful in understanding the spectroscopic properties of these compositions when doped with Ce³⁺ or Eu²⁺ ions. We note that both the shortest and average Sr(1)–O and Sr(2)–O distances in the two structures are similar. This, combined with the identical CNs, may justify quite similar energetic positions of at least the lowest (emitting) 5d levels of Ce³⁺ or Eu²⁺ dopants occupying analogous – Sr(1) or Sr(2) – sites in the two polymorphs. However, since the 5d orbitals are the most outside located and are not screened from the influence of the surroundings, even the small changes in the distances should exert measurable effects on the energetic positions of the orbitals. In contrast, when we compare the shortest and average distances of Sr(1)–O and Sr(2)–O within the same polymorph, we see that they differ quite significantly (see Table 1). Thus we can expect that within the same structure, both in the case of Ce³⁺ and Eu²⁺, the energetic positions of the emitting states should differ quite profoundly. The Sr(1)–Sr(2) distances are short, which obviously favours energy exchange between dopants occupying different sites in either of the two hosts. We shall see that such energy transfer is indeed efficient in both hosts and for both activators.

Fig. 1 Unit cells of both (a) orthorhombic and (b) monoclinic Sr₂SiO₄ structures, together with the local arrangements of the two Sr sites in each of them.

Table 1 The Sr–O and Sr–Sr distances in the orthorhombic α′-Sr₂SiO₄ and monoclinic β-Sr₂SiO₄ ²⁷

	α′-Sr2SiO4	β-Sr2SiO4
Sr(1)–O(1^I)	2.386(7)	2.37(1)
Sr(1)–O(2)	2.771(7)	2.801(8)
Sr(1)–O(2,2^III)	2.842(6) (×2)	2.57(1)
Sr(1)–O(2,2^II)		3.11(1)
Sr(1)–O(3)	3.021(6) (×2)	2.764(8)
Sr(1)–O(4)		3.199(8)
Sr(1)–O(3,2^II)	2.846(6) (×2)	2.63(1)
Sr(1)–O(4,2^II)		2.93(1)
Sr(1)–O(3,4^V)	2.972(7) (×2)	3.35(1)
Sr(1)–O(4,4^V)		2.78(1)
Average	2.852	2.850
Sr(2)–O(1^I)	2.593(8)	2.62(1)
Sr(2)–O(2^I)	2.622(7)	2.651(8)
Sr(2)–O(2,4^I)	2.622(7)	2.624(9)
Sr(2)–O(1,2^III)	3.105(3) (×2)	3.40(1)
Sr(2)–O(1,2^II)		2.78(1)
Sr(2)–O(3,2^II)	2.609(6) (×2)	2.694(8)
Sr(2)–O(4,2^III)		2.565(8)
Sr(2)–O(3,3^VI)	2.507(6) (×2)	2.527(9)
Sr(2)–O(4,3^VIII)		2.519(9)
Average	2.698	2.709
Sr(1)–Sr(1)	3.019	4.891
	3.796	5.683
	3.951	3.866
	4.867	3.957
	5.811	5.851
	5.796
Sr(1)–Sr(2)	3.897	3.705
	3.634	3.622
	3.894	3.746
	3.757	4.083
		3.874
Sr(2)–Sr(2)	3.689	3.690
	4.799	4.558
	5.682	5.663

---

In this paper we shall show that it is possible to synthesize phase-pure Sr₂SiO₄:Ce and Sr₂SiO₄:Eu (both orthorhombic and monoclinic) phosphors using two different fluxes, SrCl₂ and H₃BO₃. It will be proved that by applying H₃BO₃ and using either of the dopants, Ce or Eu, the powders crystallize in the orthorhombic α′-phase. However, a SrCl₂ flux leads to the monoclinic β-phase in the case of Sr₂SiO₄:Ce powders or when not more than 0.1 mol% of Eu is doped. On the other hand, using 1 mol% of Eu causes crystallization in the orthorhombic α′-phase. Photo- and radioluminescence properties of the phase pure phosphors in the 30–300 K range of temperatures will also be presented.

Experimental details

Synthesis and structural characterization

All of the strontium silicate powders, undoped and Ce- or Eu-doped, were prepared by flux-aided synthesis. As the flux, either SrCl₂ or H₃BO₃ was used. The starting materials were SrCO₃ (99.994%, Alfa Aesar), SiO₂ (99%, Sigma Aldrich), CeO₂ (99.99%, Stanford Materials Corporation) or Eu₂O₃ (99.999%, Stanford Materials Corporation). 1 wt% of H₃BO₃ (99.8% Carl Roth GmbH + Co) was added to the reacting mixture, while in the case of the SrCl₂ flux (99%, Merck, hydrated salt SrCl₂·6H₂O used) it was 1, 5, 10, 15 or 50 wt%. Stoichiometric amounts of substrates were mixed with an appropriate flux, thoroughly ground in an alumina mortar and were heat-treated afterwards for 4 hours in different atmospheres (air, vacuum, active carbon with reduced access of air, or forming gas (H₂(5%)/N₂(95%) or H₂(25%)/N₂(75%))) at various temperatures between 1200–1400 °C. After reaction with SrCl₂, powders were recovered by washing out the flux with hot deionized water. In the case of Eu-doped samples obtained by heating in the presence of active carbon, additional heating at 650 °C for 0.5 h in air was undertaken to remove carbon residues. All the most important data on the synthesis parameters of specific samples are collected in Table 2, together with the information on the phase composition of the products as revealed by recorded XRD patterns.

Table 2 Processing parameters of selected samples and structural composition of the products

Processing parameters			Sr2SiO4:Ce
Temperature	Time	Atmosphere	Flux (wt%)	Dopant concentration (mol%)	Structure of the product	Composition number
1200 °C	4 h	Air	1%, 5%, 10% SrCl2	1%	Mixture of both phases	SCe1, SCe4, SCe7
			15%, 50% SrCl2	1%	Monoclinic	SCe10, SCe13
			50% SrCl2	0.1%	Monoclinic	SCe17
			1% H3BO3	0.1%, 1%, 1.5%	Mixture of both phases	SCe21, SCe24, SCe27
1300 °C	4 h	Air	1%, 5%, 10% SrCl2	1%	Mixture of both phases	SCe2, SCe5, SCe8
			15% SrCl2	1%	Monoclinic	SCe11
			50% SrCl2	0.1%	Monoclinic	SCe18
			1% H3BO3	0.1%	Mixture of both phases	SCe22
			1% H3BO3	1%, 1.5%	Orthorhombic	SCe25, SCe28
1300 °C + cooling to 850 °C, 3 °C min^−1	4 h	Air	50% SrCl2	0.1%, 1%	Monoclinic	SCe19, SCe14
1400 °C	4 h	Air	1%, 5%, 10% SrCl2	1%	Mixture of both phases	SCe3, SCe6, SCe9
			15%, 50% SrCl2	1%	Monoclinic	SCe12, SCe15
			50% SrCl2	0.1%	Monoclinic	SCe20
			1% H3BO3	0.1%	Mixture of both phases	SCe23
			1% H3BO3	1%, 1.5%	Orthorhombic	SCe26, SCe29
1400 °C + cooling to 850 °C, 3 °C min^−1	4 h	Air	15% SrCl2	1%	Monoclinic	SCe16

---

Processing parameters			Sr2SiO4:Eu
Temperature	Time	Atmosphere	Flux (wt%)	Dopant concentration (mol%)	Structure of the product	Composition number
1300 °C	4 h	Air	15% SrCl2	1%	Orthorhombic	SEu3
1400 °C	4 h	Air	15% SrCl2	1%	Orthorhombic	SEu4
			1% H3BO3	1%	Orthorhombic	SEu11
1400 °C	4 h	Active carbon	15% SrCl2	1%	Orthorhombic	SEu7
(1) 1400 °C, (2) 1400 °C	4 h	Air, H2(5%)/N2(95%), H2(25%)/N2(75%)	15% SrCl2	1%	Orthorhombic	SEu5, SEu6
(1) 1400 °C, (2) 650 °C	4 h, 0.5 h	Active carbon air	15% SrCl2	0.1%	Monoclinic	SEu1
				0.5%	Mixture of both phases	SEu2
				1%, 2%	Orthorhombic	SEu8, SEu9
			1% H3BO3	0.1%	Mixture of both phases	SEu10
				1%	Orthorhombic	SEu12

---

To monitor the completeness of the reaction and phase purity of the products, X-ray diffraction (XRD) patterns were measured with a Bruker D8 Advance diffractometer equipped with a Cu lamp. The measurements were taken in the range of 2θ = 10–90° with 2θ = 0.032° steps.

The morphology of the powders was revealed by means of scanning electron microscopy (SEM) imaging with a Hitachi S-3400N instrument, equipped with an energy-dispersive X-ray spectroscopy (EDX) EDAX analyser.

Optical spectroscopy

Photoluminescence and excitation spectra were recorded using an FLS980-sm fluorescence spectrometer from Edinburgh Instruments Ltd., using a 450 W continuous xenon arc lamp for excitation and equipped with a closed-cycle helium cryostat, ARS-4HW, from Advanced Research Systems, Inc. Both excitation and emission monochromators were of single grating Czerny–Turner type, with 300 mm focal length. A pellet of a studied powder, of 7 mm in diameter, was formed under a force of 2–3 tonnes. It was then mounted on a cold finger of the cryostat and transferred to the sample chamber of the spectrometer. A Hamamatsu R928P high-gain photomultiplier detector was used to record the luminescence light. Emission spectra were corrected for the recording system efficiency and excitation spectra were corrected for the incident light intensity. These measurements were taken in the temperature range of 30–300 K.

Room-temperature radioluminescence (RL) spectra were recorded using white X-rays from a Cu X-ray tube working under the voltage of 40 kV and the current of 10 mA. The emission photons were collected with a 74-UV lens connected to a QP600-2-SR-BX waveguide, which transferred the luminescent light to an Ocean Optics HR2000CG-UV-NIR spectrometer, controlled by the dedicated software SpectraSuit. The RL light yields of our materials were compared with the results for a commercial Lu₂SiO₅:Ce (LSO) powder kindly offered by Phosphor Technology Ltd.

Results and discussion

XRD measurements

Fig. 2 presents representative XRD patterns of Sr₂SiO₄ (Fig. 2a), Sr₂SiO₄:Ce (Fig. 2b – samples SCe15, SCe23 and SCe26) and Sr₂SiO₄:Eu (Fig. 2c – samples SEu1, SEu8 and SEu12) prepared at 1400 °C for 4 h using SrCl₂ (different amounts relative to the phosphor) or H₃BO₃ flux. A summary of observations on the phase composition of the resultant phosphors is given in Table 2. From the XRD data, the following conclusions about the crystal structure of the products could be derived:

Fig. 2 XRD patterns of Sr₂SiO₄ powders: (a) undoped, (b) Ce-doped (SCe15, SCe23 and SCe26) and (c) Eu-doped (SEu1, SEu8 and SEu12), obtained from synthesis at 1400 °C for 4 h with SrCl₂ and H₃BO₃ fluxes.

(I) Undoped Sr₂SiO₄ powders obtained from synthesis with either SrCl₂ or H₃BO₃ flux crystallize in the monoclinic system (see Fig. 2a).

(II) Sr₂SiO₄:Ce obtained using more than 10 wt% of SrCl₂ flux crystallizes in the monoclinic structure. A lower content of the flux leads to a mixture of α′- and β-phases.

(III) Sr₂SiO₄:Ce obtained using a H₃BO₃ flux tends to crystallize in the orthorhombic structure, but phase-pure powder is formed only when the Ce content is not lower than 1 mol% (with respect to Sr) and the preparation temperature is not lower than 1300 °C. Processing below 1300 °C and/or using less than 1 mol% of Ce leads to a powder composed of a mixture of both phases.

(IV) Sr₂SiO₄ doped with 0.1 mol% Eu synthesized with the SrCl₂ flux crystallizes in the monoclinic structure, while a 0.5 mol% concentration of Eu leads to a mixture of α′- and β-phases. However, higher concentrations give the orthorhombic structure.

(V) Powders containing 0.1 mol% of Eu and made with a H₃BO₃ flux contain a mixture of both phases.

(VI) Sr₂SiO₄:Eu powders crystallize in purely the orthorhombic phase when the Eu content is not lower than 1 mol%, regardless of whether a SrCl₂ or a H₃BO₃ flux is used.

Thus, using the two fluxes, Ce-activated Sr₂SiO₄ may be crystallized as a phase-pure, either orthorhombic or monoclinic, phosphor. When Eu is used as the dopant, both phases can be obtained, but crystallization depends on the Eu concentration.

Morphology

The morphology of the products was strongly dependent on the flux used in the synthesis and, to some extent, parameters of the processing, especially the cooling rate. It was practically independent of the dopant. Representative SEM images are presented in Fig. 3 and 4. In the presence of a SrCl₂ flux, Sr₂SiO₄:Ce crystallized in the monoclinic phase, as we already discussed. Its particles, as seen in Fig. 3a (sample SCe15), were generally very large, reaching about 30–40 μm. Yet a fraction of the powder consisted of much smaller grains, on the order of a few microns. The small-grains proportion was noticeably lowered if the cooling of the batch from 1400 °C to about 850 °C was not natural (faster) but controlled at the rate of 3 °C min⁻¹. This effect is presented in Fig. 3b (sample SCe16). It thus appears that the smaller particles were grown during the cooling of the product, presumably from the fraction dissolved in the SrCl₂ flux. Apparently, during faster cooling this portion crystallized giving finer particles, while the slow, controlled cooling allowed also the dissolved portion to crystallize into larger grains, at least partially. In the presence of a H₃BO₃ flux (Fig. 3c – sample SCe26), when the orthorhombic phase of Sr₂SiO₄:Ce was formed, the grains were also relatively large with sizes ranging over roughly 3–30 μm. Yet they were rather irregular in shape and mostly not monocrystalline, consisting of aggregates of smaller particles.

Fig. 3 Morphology of Sr₂SiO₄:Ce (1 mol%) prepared at (a) 1400 °C for 4 h with 50 wt% SrCl₂ (SCe15) and at (b) 1400 °C for 4 h with controlled cooling of 3 °C min⁻¹ and 50 wt% of SrCl₂ flux (SCe16) – both crystallized in the monoclinic phase, and also at (c) 1400 °C for 4 h with 1 wt% H₃BO₃ (SCe26), which crystallized in the orthorhombic phase.

Fig. 4 Morphology of Sr₂SiO₄:Eu (1 mol%) (orthorhombic phase), prepared at 1400 °C for 4 h in syntheses with (a) 15 wt% SrCl₂ (SEu8) and (b) 1 wt% H₃BO₃ (SEu12).

Sr₂SiO₄:1 mol% Eu crystallized in the orthorhombic phase irrespective of the flux used in the synthesis. Yet the morphology differed in both cases, as is presented in Fig. 4. Synthesis using the SrCl₂ flux gave slightly elongated grains reaching sizes of about 30 × 50 μm (see Fig. 4a). The H₃BO₃ flux led to powders with a much less uniform morphology, very similar to what was presented in Fig. 3c for the Ce-activated orthorhombic phase. Hence, it appears that the morphology of the Sr₂SiO₄ is defined by the flux used in the synthesis and is actually hardly affected by the type of structure in which the powder crystallizes. The monoclinic Sr₂SiO₄:Ce (Fig. 3a and b) and orthorhombic Sr₂SiO₄:Eu powders (Fig. 4a) consisted of particles quite similar in shape and size, although for the latter the small-grains fraction was practically absent even if the cooling rate was not controlled.

The average size of particles was smaller when boric acid was used as the flux. This might have resulted from the amount of this flux being very small (1 wt%) compared to SrCl₂ (15–50 wt%), although in such a small concentration H₃BO₃ could effectively mediate the chemical reaction but hardly affected the crystal growth. We observed a similar effect in making Lu₂O₃:Eu, in which the particles’ growth was strongly dependent on the amount of Li₂SO₄ flux used in the process.³²

Photoluminescence of Sr₂SiO₄:Ce

Fig. 5 presents room-temperature (RT) excitation and emission spectra of the orthorhombic (sample SCe26, Fig. 5a and b) and monoclinic (sample SCe15, Fig. 5c and d) phases of the Sr₂SiO₄:1 mol% Ce powders. In both cases, the luminescence was located in the deep violet and blue part of the spectrum. The orthorhombic α′-phase luminescence peak appeared at around 410 nm and the monoclinic β-phase emission maximum appeared around 400 nm. Clearly, at RT both the monoclinic and the orthorhombic structures produce very similar emission spectra. Thus, even for a mixture of the two phases, the PL emission would be affected only slightly in terms of its spectral distribution (colour). The excitation bands consist of numerous bands in the case of both structures indicating a low symmetry of the luminescent centre, which is consistent with the structural properties of the polymorphs as presented in Fig. 1. As discussed in the Introduction, in both structures Sr²⁺ occupies two different symmetry sites, and this should presumably also apply to the Ce³⁺ dopant. Yet the observed shifts of the excitation band positions depending on the monitored emission wavelength are only very small, and similarly insignificant are changes in the positions of the emissions excited at different wavelengths. Thus, at room temperature, the spectra did not clearly prove that Ce occupies different symmetry sites. A similar set of experiments at low temperatures was undertaken to verify this observation.

Fig. 5 Excitation and emission spectra measured at RT for Sr₂SiO₄:1 mol% Ce powders obtained in the syntheses with the following fluxes: (a) and (b) 1 wt% H₃BO₃ – orthorhombic phase (SCe26); (c) and (d) 50 wt% SrCl₂ – monoclinic phase (SCe15).

Nevertheless, at room temperature the emission bands of the orthorhombic Sr₂SiO₄:1 mol% Ce (Fig. 5a and b) appear at wavelengths longer by about 5–10 nm, and a similar red shift is seen in the position of the lowest-energy excitation bands compared to those of the monoclinic phase. Bearing in mind that the average Sr–O distances in both structures are indeed very similar (see Table 1), the very small difference in the RT luminescence band positions is quite realistic, in fact.

Low-temperature PL and PLE spectra of both phases of Sr₂SiO₄:Ce are presented in Fig. 6. Results related to the orthorhombic α′-phase are given in Fig. 6a and b (sample SCe26), and those connected with the monoclinic β-phase are presented in Fig. 6c and d (sample SCe15). Compared to the analogous RT spectra, the presence of two Ce³⁺ emitting centres is now evident in both phases, while analogous RT spectra did not give a clear answer to that question (see Fig. 5). Comparison of the RT and low-temperature results makes it apparent that Ce(Sr1) → Ce(Sr2) energy transfer occurs even at 30 K. It is thus not surprising that at RT almost all emitted radiation comes from the Ce(Sr2) centre in both phases. For both phases, in the PLE spectra of the long-wavelength emissions (Ce(Sr2)), the bands characteristic for the Ce(Sr1) centre emitting at shorter wavelengths are obviously present, confirming the Ce(Sr1) → Ce(Sr2) energy transfer. In the RT PLE spectra, this effect was more profound, as expected. This is most clearly seen at around 310 and 335 nm for the orthorhombic α′-phase (Fig. 6a) and at around 310 nm for the monoclinic β-phase (Fig. 6c). Accordingly, appropriately selected excitation wavelengths allowed the separation of the site-characteristic luminescent bands pretty fairly only at low temperature. This was more successful in the case of the monoclinic β-phase (Fig. 6d) than for the orthorhombic α′-phase (Fig. 6b), altogether, according to the data presented in Fig. 6, the presence of two Ce³⁺ emitting centres is obvious in both structures. On the other hand, since the positions of the electronic levels of the Ce(Sr1) and Ce(Sr2) centres favour energy transfer by the Förster mechanism³³ (which is proved by the excitation spectra), it is not surprising that a complete spectroscopic separation of the site-related emission and excitation bands was not achieved.

Fig. 6 Excitation and emission spectra measured at low temperature for Sr₂SiO₄:1 mol% Ce powders obtained in the syntheses with the following fluxes: (a) and (b) 1 wt% H₃BO₃ – orthorhombic phase (SCe26); (c) and (d) 50 wt% SrCl₂ – monoclinic phase (SCe15).

Photoluminescence of Sr₂SiO₄:Eu

Samples prepared in air contained only Eu³⁺ as only emission of this ion could be detected. In our case, only preparation in a reducing atmosphere led to luminescence of Eu²⁺. This is somewhat contradictory to observations made for the compositions BaMgSiO₄:Eu and Sr₄Al₁₄O₂₅:Eu, in which at least partial reduction to Eu²⁺ was observed upon preparation at 1300–1400 °C in air.^34,35 The difference may result from the different structures formed in both cases. On the other hand, in our compositions, the post-fabrication treatment at 650 °C in air did not lead to oxidation of the Eu²⁺ ions into Eu³⁺. Thus, also in our silicates the high stability of Eu²⁺ is maintained.

Excitation and emission spectra of the orthorhombic Sr₂SiO₄:1 mol% Eu are presented in Fig. 7a and b (sample SEu8), and of the monoclinic phase in Fig. 7c and d (0.1 mol% Eu, sample SEu1). Regardless of the crystallographic phase, for the Eu-activated composition, it is evident already at RT that the dopant occupies at least two distinct symmetry sites with significantly different positions of the emitting and excitation bands. Excitation at around 450 nm gave rise to a characteristic broad-band luminescence peaking at around 590 nm and 550 nm for the α′ and β-forms, respectively (see Fig. 7b and d). In the case of the orthorhombic α′-phase, excitation at around 290–330 nm produces a structured luminescent band in which the 590 nm component is accompanied by another, more energetic emission, peaking at around 480–500 nm (Fig. 7b). The latter component is less significant in intensity, most probably due to the anticipated Eu(Sr1) → Eu(Sr2) energy transfer. We observed a greater change with excitation wavelength in powders which crystallized in the monoclinic β-phase. Excitation at around 275–330 nm produces an emission band with a maximum at around 470 nm, accompanied by a component with a maximum at around 540–550 nm. Excitation at 250 nm favours luminescence peaking at around 540 nm, and this latter band is observed exclusively upon 450 nm irradiation (Fig. 7d).

Fig. 7 RT excitation and emission spectra of powders: (a) and (b) orthorhombic Sr₂SiO₄:1 mol% Eu (SEu8); (c) and (d) monoclinic Sr₂SiO₄:0.1 mol% Eu (SEu1).

Comparing the emission spectra of the 1 mol% orthorhombic (Fig. 7b) and 0.1 mol% monoclinic (Fig. 7d) samples, we see that in the former the energy more efficiently leaks to the Eu²⁺ ions emitting at longer wavelengths. This points to a more efficient Eu(Sr1) → Eu(Sr2) energy transfer in the 1 mol% orthorhombic powder. This is reasonable, as in both phases the short-wavelength emission band overlaps partially with absorption (excitation) of the long-wavelength emitting centre. Thus energy from the more energetic luminescence centre of the Eu(Sr1) may be nonradiatively drained by the Förster mechanism³² to the energetically lower levels of the Eu(Sr2). Thus, the RT experiments already prove that in both phases Eu²⁺ ions occupy the two available sites and give rise to distinct, though partially overlapping, emission bands. Below, we shall present low-temperature experiments from which the presence of Eu²⁺ ions in different sites will be even more convincing.

Low-temperature PLE and PL spectra of the Sr₂SiO₄:1 mol% Eu orthorhombic α′-phase (Fig. 8a and b – sample SEu8) and Sr₂SiO₄:0.1 mol% Eu monoclinic β-phase (Fig. 8c and d – sample SEu1) allowed a clearer exposure of the two emitting centres in each phase. The α′-phase gave two clearly separated emitting bands peaking at 470 nm, Eu(Sr1), and 550 nm, Eu(Sr2), upon 310 nm excitation. The former became relatively weaker with increasing excitation wavelength. When the excitation was performed at 400 nm or at yet lower energies, only the longer-wavelength emission band (550 nm) was present. The low-temperature PLE spectra show that the most efficient excitation of the long-wavelength emission (peaking at 550 nm) occurred at around 430 nm, while the short-wavelength luminescence (470 nm) was most effectively excited at around 330 nm. In the excitation spectrum of the 550 nm luminescence (Eu(Sr2)), the bands characteristic for the other centre (Eu(Sr1)) are clearly present, though they have low intensities. It is clear that some Eu(Sr1) → Eu(Sr2) energy transfer occurs.

Fig. 8 Low-temperature excitation and emission spectra of the following powders: (a) and (b) orthorhombic Sr₂SiO₄:1 mol% Eu (SEu8); (c) and (d) monoclinic Sr₂SiO₄:0.1 mol% Eu (SEu1).

Therefore, it is not possible to exclusively record luminescence of the high energy, Eu(Sr1), centre. A long-wavelength excitation (460 nm) produced a single band emission (Fig. 8b, red line) which was slightly red-shifted and peaked at around 600 nm. A closer inspection reveals that its shape was slightly distorted, suggesting a superposition of two components. This is surprising as the host does not offer a third site. A possibility of a small admixture of the β-phase (not recordable by XRD) cannot explain this effect, as the β-phase does not produce such a long-wavelength luminescence, as we shall see shortly. An excitation spectrum of the emission monitored at 620 nm indeed shows a noticeable long-wavelength broadening compared to the PLE of the 550 nm emission (see Fig. 8a). It is possible that certain Eu(Sr2) centres experience some structural distortion. Yet details of this effect remain unclear at present. Note that the presented spectra are normalized to the same height. In fact, the long-wavelength emission is much less intense than the two regular ones peaking at around 475 nm and 550–560 nm.

In the case of the monoclinic β-phase (Sr₂SiO₄:0.1 mol% Eu, see Fig. 8c and d), quite a similar behaviour was observed at 30 K. Two emission bands peaking at 470 nm (Eu(Sr1)) and 545 nm (Eu(Sr2)) are clearly seen. Again, only the one at longer wavelength can be generated separately. The more energetic emission is always accompanied by the less energetic one, due to the unavoidable energy transmission down to the Eu(Sr2) centre. PLE spectra of the Eu(Sr2) centre indeed contain features characteristic for the Eu(Sr1) one. Thus, the Eu(Sr1) → Eu(Sr2) energy transfer occurs also for the Sr₂SiO₄:0.1 mol% Eu β-phase, even though the Eu content is so low. The Eu(Sr2) centre may be exclusively excited in the 420–450 nm range of wavelengths (see Fig. 8c and d).

From Table 2 and the recorded crystallographic data, we know that orthorhombic Eu-activated Sr₂SiO₄ powders can be fabricated using either H₃BO₃ or SrCl₂ flux, if the dopant content is not lower than 1 mol%. It is interesting to compare the luminescence spectra of two such compositions upon excitation at 255 nm. This is presented in Fig. 9a (samples SEu8 and SEu12). It is immediately evident that under such irradiation, the luminescence of the material made using the H₃BO₃ flux contained quite a significant admixture of the Eu³⁺ emission superimposed on the Eu²⁺ band. In the case of the powder synthesized with the help of the SrCl₂ flux, the luminescence was generated by Eu²⁺ ions, exclusively. Thus, for some reason, H₃BO₃ hinders the susceptibility for chemical reduction of Eu³⁺ to the Eu²⁺ form. At this point, the reasons for that are not clear. Yet, in Fig. 9b (samples SEu8 and SEu12), which presents IR spectra of the orthorhombic versions of Sr₂SiO₄:Eu powders made with the help of both fluxes, we find information which seems to be useful in understanding this difference. Namely, in the case of the powder made using the H₃BO₃ flux, two well-resolved components are clearly seen in the 1150–1300 cm⁻¹ range of wavenumbers. They can be identified as resulting from B–O stretching vibrations.³⁶ Thus, boron enters the Sr₂SiO₄ host as an impurity. Since B³⁺ is a small ion (0.11 Å), it is expected to replace Si⁴⁺ (0.26 Å), as all other sites are occupied by greatly larger ions. If so, however, the smaller charge of B³⁺ compared to Si⁴⁺ may easily stabilize the Eu³⁺ ions (in the position of Sr²⁺)³⁷ as then the electrostatic neutrality is maintained.

Fig. 9 (a) Comparison of luminescence spectra of α′-Sr₂SiO₄:1 mol% Eu obtained in the syntheses with fluxes H₃BO₃ (sample SEu12, black line) and SrCl₂ (sample SEu8, red line) upon excitation at 255 nm; (b) infrared spectra of α′-Sr₂SiO₄:1 mol% Eu obtained in the syntheses with fluxes H₃BO₃ (sample SEu12, black line) and SrCl₂ (sample SEu8, red line).

A similar comparison for the monoclinic structure could not be done, as with the Eu activator it could not be obtained as a pure phase. Thus, we can conclude that, compared to H₃BO₃, SrCl₂ flux not only allows fabrication of powders with a better morphology, but also facilitates the reduction of Eu³⁺ to Eu²⁺. We should also note that the B–O vibrations are partially positioned at ∼1200 cm⁻¹, which indicates that part of a boron impurity exists in BO₃ groups in our composition.^38,39 This may be induced by its much smaller size than Si, as mentioned above. BO₃ groups, in contrary to BO₄, were reported to not facilitate the reduction of europium.⁴⁰ Some difficulty in efficient reduction of Eu to its 2+ state in silicates was also reported by Dobrowolska.⁴¹

Radioluminescence

Independently of its crystal structure, Sr₂SiO₄ is not a dense material (∼5 g cm⁻³), and thus it is not a very attractive host for scintillation applications. Yet we present radioluminescence (RL) spectra for both phases of Sr₂SiO₄:Ce (Fig. 10, samples SCe15 and SCe26) and Sr₂SiO₄:Eu (Fig. 11, samples SEu1 and SEu8), as they appear quite intriguing.

Fig. 10 Radioluminescence spectra of Sr₂SiO₄:1 mol% Ce obtained in the syntheses with flux 1 wt% H₃BO₃ (blue line, orthorhombic phase, sample SCe26) and 50 wt% SrCl₂ (red line, monoclinic phase, sample SCe15) at 1400 °C for 4 h in air, compared to that of commercial Lu₂SiO₅ (LSO, black line).

Fig. 11 Radioluminescence spectra of Sr₂SiO₄:x mol% Eu (x = 0.1, 1 mol%) obtained in the syntheses with 15 wt% of SrCl₂ flux at 1400 °C for 4 h in an active carbon atmosphere. Sample SEu1 (blue line) crystallizes in the monoclinic phase, while SEu8 (red line) crystallizes in the orthorhombic phase. The spectra are compared with that of commercial Lu₂SiO₅ (LSO, black line).

Striking differences are observed between the orthorhombic and monoclinic Sr₂SiO₄:Ce materials in both the spectral distribution of the RL emissions and their efficiencies. The orthorhombic Sr₂SiO₄:Ce α′-phase (Fig. 10, blue line) gives only a weak emission upon excitation with X-rays, at the level of ∼10% of that of the commercial Lu₂SiO₅:Ce (LSO). Furthermore, the RL spectrum covers a much broader range of wavelengths (∼375–650 nm) than was seen in the PL spectra (see Fig. 5b and 6b). Upon optical excitation, the spectra of this phase were much narrower, which may imply that the inefficient RL results from both Ce and a defect/defects generated upon the impact of the high-energy X-rays. In contrast, the RL of the monoclinic Sr₂SiO₄:Ce β-phase (Fig. 10, red line) gives emission whose efficiency at least meets, if not surpasses, the yield of LSO. It cannot be excluded that the already impressive RL of the monoclinic Sr₂SiO₄:Ce might be further improved if dopant content and processing parameters are further optimized. This was beyond the scope of the present research, however. Since in the photoluminescence measurements such drastic differences in intensities of the emissions from the two phases were definitely not seen (though quantitative measurements were not performed), we may conclude that in the orthorhombic α′-phase, the efficiency of the host-to-activator energy transfer is critically lower than in the case of the monoclinic β-phase.⁴²

Also the RL spectra of the Sr₂SiO₄:Eu presented in Fig. 11 surprise to some extent. In this case, Eu²⁺ gave quite significant RL from both phases. However, it appears that, in contrast with the PL spectra presented in Fig. 7b and d, in both polymorphs, only the Eu(Sr2) luminescent centre is active in the generation of light. It appears that the energy transfer from the host excited upon the impact of X-rays is preferentially transferred to the Eu²⁺ ions, giving the long-wavelength luminescent band peaking at around 580 nm in the orthorhombic phase and at around 540 nm in the monoclinic one (the 5–10 nm shift between peaks in the RL and PL spectra may result from the fact that the former spectra were not corrected for the system characteristics). The reason for such behaviour is not clear at present and its explanation would require rather extensive, specialized research. Yet a reasonable explanation/guess might be that the Eu(Sr2) ions giving the long-wavelength luminescence are more efficient hole traps compared to Eu(Sr1). Such a preferential trapping of holes by Eu(Sr2) would automatically make these ions more attractive for subsequent trapping of electrons from the conduction band, which would close the first step of harvesting the energy from the excited host by the emitting centres.^43–45 Consequently, only the Eu(Sr2) centre would be fed by the energy acquired from the incident X-ray photons and only these ions would produce luminescent photons. A strongly preferential transfer of energy from a host to a dopant located in a specific symmetry site is a known effect. For example, in Lu₂SiO₅:Ce (LSO) scintillator crystals used in positron emitting tomography (PET) cameras, also, for lower concentrations, only one Ce site gives light upon the impact of γ-rays.⁴⁶

A direct reason for that might be connected with the position of the ground states of Eu³⁺ ions occupying the two sites against the valence band, and thus the formal energy for h-trapping. From Fig. 7 and 8, we know that in both phases the long-wavelength emitting ions (Eu(Sr2)) have a lower energy of the first f–d transition. Accordingly, this can be taken as a measure of their (lower) energy for h-trapping. In other words, the holes from the valence band are energetically closer to the Eu(Sr2) ions. This may be the driving force for the preferential energy trapping in Eu(Sr2) seen in radioluminescence and described above. A similar effect was reported for storage/persistent phosphors containing two dopants acting as hole and electron traps. Depending on their relative positions against the valence (hole trap) and conduction (electron trap) bands, the energy upon releasing is emitted by only one of them.⁴⁷

Conclusion

Fabrication conditions of phase-pure monoclinic powders of Sr₂SiO₄, monoclinic and orthorhombic Sr₂SiO₄:Ce³⁺ as well as monoclinic and orthorhombic Sr₂SiO₄:Eu²⁺ were established using H₃BO₃ or SrCl₂ fluxes. It was shown that both the flux used in the fabrication process and the type and concentration of the activator affect the crystallization process, and define the type of structure formed. The single-phase powders of both phases activated with Ce³⁺ or Eu²⁺ were spectroscopically investigated at 30 K and 300 K. In both polymorphs, the dopants were shown to occupy two symmetry sites. Their emission and excitation spectra were recorded at 30 K and 300 K, and it was shown that ions emitting at shorter wavelengths transfer energy to those producing luminescence at longer wavelengths. This effect was especially significant at room temperature. For a specific activator, the emissions were located at a quite similar range of wavelengths for both polymorphs. In the case of Ce³⁺, the luminescence bands of both sites were located partially in the UV part of the spectrum, and efficient excitation was possible below ∼360 nm. Such characteristics are problematic if application of this composition in solid-state lighting is considered. Eu²⁺ produced emission bands which were located fully in the visible region. At room temperature, excitation with a blue LED (∼450 nm) gave rise to broad green luminescence from the monoclinic phase of Sr₂SiO₄:Eu and orange emission from the orthorhombic form. Efficient emission from both Eu²⁺ sites simultaneously was achieved upon excitation at around 375–400 nm for the monoclinic structure.

Acknowledgements

The work was supported by Wroclaw Research Centre EIT+ within the project “The Application of Nanotechnology in Advanced Materials” – NanoMat (POIG.01.01.02-02-002/08) co-financed by the European Regional Development Fund (Innovative Economy Operational Program 1.1.2).

References

1. X. Sun, J. Zhang, X. Zhang, Y. Luo and X. Wang, J. Rare Earths, 2008, 26(3), 421.
2. L. Pan-Li, Y. Zhi-Ping, W. Zhi-Jun and G. Qing-Lin, Chin. Phys. B, 2008, 17(3), 1135.
3. S. K. Gupta, M. Moharpatra, S. Kaity, V. Natarajan and S. V. Godbole, J. Lumin., 2012, 132, 1329.
4. Y. Qiao, X. Zhang, X. Ye, Y. Chen and H. Guo, J. Rare Earths, 2009, 27(2), 323.
5. J. H. Lee and Y. J. Kim, J. Ceram. Process. Res., 2009, 10(1), 81.
6. D. V. Sunitha, H. Nagabhushana, S. C. Sharma, B. M. Nagabhushana, B. Daruka Prasad and R. P. S. Chakradhar, Spectrochim. Acta, Part A, 2014, 127, 381–387.
7. Q. Wang, J. Qiu, Z. Song, D. Zhou and X. Xu, Chin. Phys. B, 2014, 23, 064211.
8. H. Nagabhushana, D. V. Sunitha, S. C. Sharma, B. Daruka Prasad, B. M. Nagabhushana and R. P. S. Chakradhar, J. Alloys Compd., 2014, 595, 192.
9. M. A. Tshabalala, F. B. Dejene, S. S. Pitale, H. C. Swart and O. M. Ntwaeaborwa, Phys. B, 2014, 439, 126.
10. L. C. Ju, X. Xu, L. Y. Hao, Y. Lin and M. H. Lee, J. Mater. Chem. C, 2015, 3, 1567.
11. B. Zhang, X. Yu, T. Wang, S. Cheng, J. Qiu and X. Xu, J. Am. Ceram. Soc., 2015, 98(1), 171.
12. J. Barzowska, A. Chruścińska, K. Przegiętka and K. Szczodrowski, Radiat. Phys. Chem., 2014, 104, 31.
13. N. Lakshminarasimhan and U. V. Varadaju, Mater. Res. Bull., 2008, 43, 2946.
14. Z. Song, X. Ding, S. Yang, F. Du, L. Bian, S. Duan and Q. L. Liu, J. Lumin., 2014, 152, 199.
15. X. Xu, X. Zhang, T. Wang, J. Qiu and X. Yu, Mater. Lett., 2014, 127, 40.
16. S. H. Lee, H. Y. Koo and Y. C. Kang, Ceram. Int., 2010, 36, 1233.
17. C.-H. Hsu, R. Jagannathan and C. H. Lu, Mater. Sci. Eng. B, 2010, 167, 137.
18. J. K. Han, M. E. Hannah, A. Piquette, G. A. Harita, J. B. Talbot, K. C. Mishra and J. McKittrick, J. Lumin., 2012, 132, 106.
19. D. Dutczak, A. Milbrat, A. Katelnikovas, A. Meijerink, C. Ronda and T. Jüstel, J. Lumin., 2012, 132, 2398.
20. R. Y. Yang, H. Y. Chen, S. J. Chang and Y. K. Yang, J. Lumin., 2012, 132, 780.
21. L. Zhang, Z. Lu, P. Han, J. Lu, N. Xu, L. Wang and Q. Zhang, J. Am. Ceram. Soc., 2012, 95(12), 3871.
22. L. Zhang, P. Han, K. Wang, Z. Lu, L. Wang, Y. Zhu and Q. Zhang, J. Alloys Compd., 2012, 541, 54.
23. L. C. Ju, C. Cao, Q. Q. Zhu, J. Y. Tang, L. Y. Hao and X. Xu, J. Mater. Sci.: Mater. Electron., 2013, 24, 4516.
24. H. Feng, Y. Yang, X. Zhang, Y. Xu and J. Guan, Superlattices Microstruct., 2015, 78, 150.
25. M. A. Tshabalala, H. C. Swart and O. M. Ntwaeaborwa, J. Vac. Sci. Technol., A, 2014, 32, 021401.
26. M. Pardha Saradhi, N. Lakshminarasimhan, S. Boudin, K. V. K. Gupta, U. V. Varadaraju and B. Raveau, Mater. Lett., 2014, 117, 302.
27. M. Catti, G. Gazzonni, G. Ivaldi and G. Zanini, Acta Crystallogr., Sect. B: Struct. Sci., 1983, 39, 674.
28. M. Catti and G. Gazzonni, Acta Crystallogr., Sect. B: Struct. Sci., 1983, 39, 679.
29. M. Catti, G. Gazzonni and G. Ivaldi, Acta Crystallogr., Sect. B: Struct. Sci., 1983, 39, 29.
30. M. Catti, G. Gazzonni and G. Ivaldi, Acta Crystallogr., Sect. B: Struct. Sci., 1984, 40, 537.
31. J. Liu, C. G. Duan, W. N. Mei and J. R. Hardy, J. Chem. Phys., 2002, 116(9), 3864.
32. J. Zeler, L. Jerzykiewicz and E. Zych, Materials, 2014, 7(10), 7059.
33. T. Förster, Ann. Phys., 1948, 2, 55.
34. M. Peng, Z. Pei, G. Hong and Q. Su, J. Mater. Chem., 2003, 13, 1202.
35. M. Peng, Z. Pei, G. Hong and Q. Su, Chem. Phys. Lett., 2003, 371, 1.
36. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
37. F. A. Miller and C. H. Wilkins, Anal. Chem., 1952, 24(8), 1253.
38. C. Gautam, A. K. Yadav and A. K. Singh, ISRN Ceram., 2012, 2012, 17.
39. N. J. Kreidl, Inorganic Glass-Forming Systems in Glass: Science and Technology, vol. 1 of Glass-Forming Systems, Academic Press, New York, NY, USA, 1983.
40. Z. Pei, Q. Zeng and Q. Su, J. Phys. Chem. Solids, 2000, 61, 9.
41. A. Dobrowolska and E. Zych, J. Phys. Chem. C, 2012, 116(48), 25493.
42. E. Zych, Luminescence and scintillation of inorganic phosphor materials, in Handbook of Luminescence, Display Materials, and Devices. inorganic Display Materials, ed. H. S. Nalwa and L. S. Rohwer, American Scientific Publishers, Stevenson Ranch, Calif., 2003, vol. 2, p. 251.
43. A. J. Wojtowicz, A. Lempicki, D. Wisniewski, M. Balcerzyk and C. Brecher, IEEE Trans. Nucl. Sci., 1996, 43(3), 2168.
44. E. Zych, Nanocrystalline insulating phosphors in: doped nanomaterials and nanodevices, Luminescence and applications, ed. W. Chen, American Scientific Publisher, Stevenson Ranch, 2010, vol. 1, p. 14335.
45. A. Lempicki, A. J. Wojtowicz and C. Brecher, Inorganic Scintillators in Wide-Gap Luminescent Materials: Theory and Applications, ed. S. R. Rotman, Kluwer Academic Publishers, Inc., Norwell MA, USA, 1997, p.235.
46. A. Vedda, M. Nikl, M. Fasoli, E. Mihokova, J. Pejchal, M. Dusek, G. Ren, C. R. Stanek, K. J. McClellan and D. D. Byler, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 78, 195123.
47. A. Lecointre, A. Bessiere, A. J. J. Bos, P. Dorenbos, B. Viana and S. Jacquart, J. Phys. Chem. C, 2011, 115, 4217.

---