Effect of Preparation Method on the Optical Properties of Novel Luminescent Glass-Crystalline Composites

Abstract

Phosphor-in-glass (PiG) composites are promising materials for applications in various fields of material engineering. There are competing methods of preparation of PiGs which result in materials with different structural and performance characteristics. The glass-crystal composites comprising tellurite-zinc-sodium glass (TZN) and perovskite LaAlO₃ doped with Eu³⁺ (LAO:Eu) are prepared using three distinct methods: remelt, direct-doping and co-sintering, in order to evaluate the impact of the preparation method on the structural, optical and luminescence properties of the novel phosphor-in-glass (PiG) composites. The composites prepared by the remelt and direct-doping method suffer from the decomposition of LAO:Eu and Eu³⁺ ion diffusion into the glass matrix. The highest rate of preservation and luminescence intensity of LAO:Eu is achieved in the composites prepared by the co-sintering method. Unfortunately, the loss of transparency is substantial. This article demonstrates the challenges and tradeoffs that are yet to be resolved in preparation of PiG composites. The preservation of the crystalline phase leads to the lower transparency of the final material.

1. Introduction

Phosphor-in-glass (PiG) is a type of composite material, where the luminescent crystalline powders/particles are bound by amorphous inorganic materials. Unlike phosphor-in-silicone (PiS) commonly used in WLEDs [1], luminescent composites consisting of crystalline materials in glass are both heat and moisture resistant, and in addition, the freedom to choose glasses allows for the adjustment of the refractive index, eliminating scattering and, as a result, obtaining transparent materials.

The phosphor-in-glass composites have been studied in recent years as promising materials for applications in photovoltaics [2,3], high-energy lighting [4] and LCD displays [5]. The motivation behind the development of PiG is usually focused on the high luminescence efficiency, whether it is in the visible range, as in LuAG:Ce/YAG:Ce in silicate glass [6,7], or near-infrared, as in CsPbBr₃ Quantum Dots embedded in Nd³⁺-activated tellurite glass [8]. To this day, the synthesis of PiG is performed with one of three methods: the remelt method [9], the co-sintering method [10] and the direct-doping method [11,12]. In the remelt method, glass is obtained by conventional melting and rapid cooling, then ground, mixed with crystalline powders, and melted again. In the co-sintering method, the mixed glass and crystalline powders are pressed into pellet and sintered in elevated temperature. In the direct-doping method, the polycrystalline powders are added to the molten glass right before the initial quenching. All of these methods have their individual parameters aiming to optimize the material’s homogeneity, and prevent crystalline dissolution and diffusion of the dopant form the crystalline to amorphous phase.

The goal of (PiG) composites science is to maximize the benefits of the multi-substance interplay and minimize the disadvantages arising from the non-homogeneity of the material. This fundamental problem is very clearly seen in the design of novel composite luminescent materials: the benefits of the efficient luminescence from the crystalline particles in the luminescent composite are met with the loss of transparency and light extraction. The latter stems predominantly from the mismatch of the refractive indices of the crystalline phosphor and the organic/amorphous matrix. This mismatch causes both the exciting and emitted light to bend at the interface of the constituent phases, which leads to random reflections and, as a result, scattering and loss of transparency [13].

By using certain glass types combined with suitable crystalline phosphor, it is possible to achieve matching refractive indices of both phases. The usability of refractive index matching has been demonstrated by Xia et al. [10] for YAG:Ce³⁺ phosphor-based PiG composite, where the glass chosen for a matrix was a bismuth borate glass with the refractive index 1.92. This value was still not ideally matched to 1.84 of the YAG, yet exhibited superior optical qualities compared to silica glass with embedded YAG:Ce³⁺. A similar attempt for Y₂O₃:Pr³⁺ embedded in germanate and silicate glass fibers was documented in [14]. The most precise attempt on refractive index matching in a PiG composite was reported for SrAl₂O₄:Eu,Dy embedded in silica glass. Yamashita et al. reported the enhanced transparency of the obtained composite, when the refractive index of the glass matrix was matched with the refractive index of the phosphor [15].

The optical and spectroscopic qualities of luminescent ions in the tellurite glasses depend on their composition, structure and the involved effects related to ion–host interactions. The tellurite glasses are considered favorable and advantageous optical materials which are characterized by a high transmittance range, quite low phonon energy, high refractive index, beneficial thermal properties and chemical durability [16,17,18,19]. The tellurite glass exhibits a refractive index of 1.97 (at 600 nm) [20]. The radiative transition effectiveness of luminescent admixtures can be enhanced in tellurite glasses as a result of the high value of the refractive index [17]. Among others, it has been recently reported that the thermal treatment of the TeO₂-ZnO-Bi₂O₃-Yb₂O₃ system leads to the surface crystals precipitation substantially increasing RE emission features [18]. Furthermore, the excellent thermal stability of Eu/Tb luminescence in co-doped tellurite glass was confirmed by V. Sangwan et al. [19]. Tellurite glass was used as a host for YAG:Ce phosphor in a recent study by Sun et al. [21] on tellurite-based PiG.

Research on nanocrystalline LaAlO₃ doped with Eu³⁺ ions (LAO:Eu) began in the 1970s and so far, there have been nearly 50 scientific publications on them. At first, Blasse G. et al. [22]. described three types of non-radiative transitions that occur during the non-radiative decay of LAO:Eu. Then, in 2006 [23] a paper appeared on research on the luminescent properties of LaAlO₃ perovskite NCs doped with Eu³⁺ ions for applications in optoelectronic systems. Most of the papers, however, concern the use of LAO:Eu as red phosphors for LED application [24,25,26] and horticultural lighting [27]. In the context of the present research, LaAlO₃ NCs were also studied as a modifier in lead-tellurite glass, where they have been found to modify the glass structure by, e.g., converting TeO₄ → TeO₃ and producing non-bridging oxygen [28]. LaAlO₃ exhibits a refractive index of 2.09 at 790 nm [29]. The pairing of tellurite glass with LAO:Eu crystalline powder is meant to study the bilateral interaction of a chemically durable and optically advantageous glass host with a phosphor that is known for its strong and stable luminescence, doped with Eu³⁺, which serves as an optical probe. Compared to other Eu³⁺-doped materials (

Table 1. Comparison of charge transfer band (CTB) energies and the electric dipole to magnetic dipole transition (ED/MD) ratio for common Eu³⁺-doped crystalline powder materials.

Compound	CTB (eV)	ED/MD Ratio	Source
La2O3	4.38	8.5	[30]
Al2O3	4.23	3.4–3.6	[31]
Y3Al5O12 (YAG)	6.64	0.7	[32,33]
LaAlO3 (LAO)	3.94	1.4	[22], this work

), LAO:Eu exhibits a low electric dipole to magnetic dipole transition ratio, which allows for its utilization as an optical probe, as well as relatively low charge transfer band (CTB) energy, which allows for excitation by 266 nm laser diode.

In this article we report the results of TZN–LAO:Eu composite preparation. This particular pairing was based on several premises, mainly the following: the low melting temperature of TZN, vast research and known luminescence properties of Eu-doped LAO, and similar refractive indices of TZN and LAO, making it a useful example for further refractive index engineering studies aimed at matching these values to obtain a transparent composite material. The composites were prepared using three methods: the remelt, direct doping and co-sintering method. The obtained composites were subjected to structural, optical and luminescence studies to investigate the impact of the crystalline phase on the properties of the glass and vice versa.

2. Materials and Methods

Glass with the composition of 70 TeO₂–20 ZnO–10 Na₂O (in mol%) was prepared using the standard melt-quenching method. The 6 g batch was prepared using TeO₂ (Alfa-Aesar, Ward Hill, MA, USA, 99.99%), ZnO (Sigma-Aldrich, Burlington, MA, USA, 99.99%), and Na₂CO₃ (Sigma-Aldrich, 99.5%) mixed using a mortar to obtain a homogenous batch. Then the batch underwent a melting process at 850 °C for 20 min preceded by 30 min of decarbonization at 650 °C. The glass melt is quenched at room temperature on a brass plate and is then annealed at 270 °C for 6 h to release residual stresses.

Polycrystalline powder La_1−xEu_xAlO₃ (x = 1 mol%) with a single perovskite structure was obtained by a conventional solid-state method. Metal oxides La₂O₃ (Alfa Aesar 99.99%), Al₂O₃ (Alfa Aesar 99.98%), and Eu₂O₃ (Alfa Aesar 99.99%) were used as raw materials. The mixture of reactants was calcined at 800 °C for 12 h and then at 1500 °C for 5 h. After each heating, the powder was ground in a mortar to increase the homogeneity of the material.

The composite materials were prepared using three different methods labeled as the remelt method (RM), direct-doping (DD) and co-sintering method (CS). In the remelt method the TZN was crushed into powder and mixed with 5 wt.% of LAO:Eu, which is a typical amount of phosphor in a PiG material [34,35]. The mixture was subjected to the same melting treatment as the as-prepared TZN, with the exception of the decarbonization step, which was unnecessary and thus omitted. In the direct-doping method, the fresh batch of the TZN was prepared and melted according to the aforementioned procedure. A total of 5 wt.% of LAO:Eu was added to the glass melt right before quenching. The quenched material was treated the same way as the as-prepared TZN. In the co-sintering method, the starting mixture was analogous to the remelt method, but instead of the heat treatment at a high temperature, the mixture was pressed into a Φ100 × 1 mm pellet and heat treated at 330 °C for 1 h. The composite samples were then fragmented as follows: parts of them were polished and the remaining material was ground into powder.

X-ray diffraction (XRD) patterns were recorded with an X’Pert PRO PANalytical (Almelo, The Netherlands) X-ray diffractometer, working in the reflection geometry, using Cu Kα radiation (λ = 1.5418 Å). The results were fitted in Match software version 3.11.5.203 and juxtaposed with fitting ICDD patterns. A scanning electron microscope (FEI NOVA NanoSEM 230, Brno, Czech Republic), equipped with an EDAX Genesis XM4 detector (Mahwah, NJ, USA) was used to characterize the morphology and chemical composition of the samples. The SEM images, as well as the EDS spectra and maps, were recorded with an accelerating voltage of 30 kV and ×700 magnification. The line elemental analysis was performed using the Genesis Maps-Linescan tool. The transmittance spectra were recorded using the Agilent Varian Cary 5E (Santa Clara, CA, USA) UV–vis-NIR spectrophotometer. The thicknesses of the TZN, RM, DD and CS samples were 2.7, 2.65, 2.5 and 1.1 mm, respectively, and the results were corrected to exhibit the value for 1 mm thick samples. Emission spectra were measured with the Hamamatsu (Hamamatsu, Japan) photonic multichannel analyzer PMA-12 equipped with a BT-CCD linear image sensor. The excitation sources were a 266 nm 6 mW laser diode (Opto Engine LLC, Midvale, UT, USA) and a Ti-sapphire Laser (JV “LOTIS TII”, Minsk, Belarus) pumped with a Nd:YAG pulse laser (JV “LOTIS TII”, Minsk, Belarus).

3. Results

The TZN is yellow-tinted and transparent. Upon different composite preparation routes, the material becomes opaque to a different degree and additional coloration occurs (Figure 1a). The RM sample remains yellow, but the transparency is restrained. The DD and CS samples become opaque and exhibit additional coloration of black and gray, respectively. Under UV irradiation the Eu³⁺ ions exhibit luminescence. Under 360 nm excitation very weak emission is observed from the RM sample and relatively strong orange emission is observed from the DD and CS samples (Figure 1b). Under 254 nm excitation only the CS sample exhibits orange luminescence (Figure 1c).

The XRD results reveal that the prepared LAO:Eu crystals are pure phase and fit the desired pattern (Figure 1d). Upon composite preparation using the RM, they are not registered and are weakly registered in the DD sample. In both of those samples there are observed traces of Al₂O₃ and La₂O₃. The broad band results from the scattering of the X-rays in the amorphous material. In the CS sample the LAO:Eu pattern is clearly detected overlayed with the amorphous band. No additional phases are found.

The transmittance spectra (Figure 1e) indicate that the as-prepared TZN exhibit moderate transparency with T = 5–10% in the whole visible range. The RM sample is slightly less transparent, especially in the blue region, which explains more yellow coloration visible in Figure 1a. The DD sample exhibits reduced transparency by more than one order of magnitude, but the shape of the transmittance spectrum is identical to the as-prepared TZN. The transmittance of the CS sample is very poor and reduced by two orders of magnitude compared to the as-prepared TZN.

The SEM images provide additional evidence of the processes indicated by the XRD results. The elemental mapping of the constituents of the LaAlO₃ crystals reveals, that in the RM sample the aluminum and lanthanum ions are totally diffused into the glass matrix and distributed almost homogeneously (Figure 2a). A few standalone aluminum clusters can be identified as an Al₂O₃ precipitates resulting from the decomposition of the LaAlO₃, also observed as weak peaks in the XRD results (Figure 1d). The elemental mapping of Al and La for the DD sample (Figure 2b) reveals that the majority of the recorded LAO crystals held their composition in the glass matrix, but the size of the crystalline areas being around 100 μm indicates the clustering of the LAO crystals. There is aluminum and lanthanum present in the glass matrix and areas of aluminum without lanthanum present, which confirms the partial dissolution of LAO, as well as Al₂O₃ precipitation during the composite preparation process. Contrary to the other preparation methods and in line with the XRD results, the elemental mapping of the CS (Figure 2c) reveals a very small amount of Al and La in the glass matrix, indicating no dissolution of LAO crystals. All Al-rich areas match the La-rich areas, which excludes the precipitation of additional crystalline phases in the obtained composite. Full elemental mapping of all elements can be found in Supplementary Figures S1–S3.

The above results are consistent with the photographs of the UV excited samples. Under 360 nm both LAO:Eu and Eu³⁺-doped Al₂O₃ precipitates can be excited to emit orange light [36,37], thus the emission can be observed from samples DD, CS and weakly from RM (Figure 1b). Under 254 nm, only LAO:Eu emission is stimulated, which occurs only for the CS sample (Figure 1c).

The cross-section of the crystalline areas in the DD and CS (Figure 3) provides additional details about the glass-crystalline interface in the samples obtained by two methods. The composition of the glass and crystals is the same in both samples (Figure 3c,d). The difference in the glass/crystal ratio can be due to the fact that the analyzed crystal is under the surface in the DD sample and on the surface in the CS sample. Although the elemental mapping indicates different stages of LAO dissolution in the glass matrix for samples prepared by DD and CS, the cross-section analysis shows that the glass/crystal interface is similarly sharp for both methods—see around the 160–170 and 260–270 μm marks in Figure 3e—contrasting with the sharp glass/crystal boarded in the CS sample—see the 105 and 250 μm marks in Figure 3f. The interdiffusion zone for both these samples was estimated to be around 30 nm (see Figure S4). This result may suggest that the thickness of the interdiffusion zone does not depend on the method, but rather is the result of the types of crystal and glass used for the composite preparation.

The composite samples exhibit the same luminescence spectrum as the LAO:Eu powders under 266 nm excitation (into the charge transfer band of LAO:Eu). The shape of the DD and CS composites’ emission spectra is exactly the same as the LAO:Eu (Figure 4a). In the case of the RM composite, due to very poor intensity, it is difficult to conclusively confirm the similarity of the emission shape (Figure 4b). The similarity of the emission shape indicates that the LAO:Eu crystals remain intact in the DD and CS samples.

The emission spectra under 393 nm (excitation into the ⁷F₀ → ⁷L₆ transition) were measured to investigate the Eu³⁺ ions diffusion into the TZN. Contrary to emission spectra under 266 nm excitation, the spectra under 393 nm differ from each other depending on the composite preparation method. This stems from the fact that the 266 nm excitation is absorbed predominantly by the LAO:Eu crystals and their residuals in the glass matrix, while the 393 nm excitation is capable of activating both the LAO:Eu and the Eu³⁺ ions in the amorphous glass host.

As the ⁵D₀ → ⁷F₁ and ⁵D₀ → ⁷F₂ transitions of Eu³⁺ are a magnetic dipole- (MD) and electric dipole-type (ED), respectively, the relative intensity of these transitions (ED/MD) can be used as an indicator of the Eu³⁺ ions’ environment. Decomposition can be evidenced as the ED/MD ratio increasing and diverging from the value for original phosphor (1.4) becoming closer to the value for TZN:Eu glass [38].

The shapes of the emission spectra exhibit different levels of resemblance to the reference LAO:Eu spectrum with ED/MD equal to 1.4 (Figure 4c): for the CS sample the spectrum is identical—with the dominant ⁵D₀ → ⁷F₁ line and ED/MD equal 1.5, for the RM the spectrum is similar to the spectrum of Eu³⁺ in the completely amorphous environment of the TZN [38]—with the dominant ⁵D₀ → ⁷F₂ line and ED/MD equal 4.1, while the spectrum of the DD sample is a sum of both the LAO:Eu spectrum and TZN:Eu glass spectrum and the ED/MD is equal to 3.2. These observations can be explained by the different level of LAO:Eu decomposition during the preparation process: for the RM the decomposition is almost total with only a small fraction of LAO:Eu crystals remaining in the glass matrix; for DD the decomposition is partial—some of the Eu³⁺ are located in crystals and some in the glass matrix; for the CS there is no apparent decomposition of the LAO:Eu.

The comparison of the emission intensity with respect to the equivalent emission of LAO:Eu (Figure 4d) reveals that under the 266 nm excitation the CS sample exhibits the strongest emission. Overall, the emission under 266 nm excitation is very low due to the poor transmittance of the glass at this wavelength (Figure 1e). The intensity of the emission under 393 nm excitation is the highest for the RM sample and it is higher even than the LAO:Eu emission, because the Eu³⁺ ions in the amorphous glass matrix can be excited more conveniently into ⁷F₀ → ⁷L₆ transition than Eu³⁺ ions in LAO. The emission intensity is lower for the DD sample and the lowest for the CS, for which no decomposition occurred. The transmittance of the glass host also plays a role in reducing the emission intensity of composites under 393 nm excitation.

4. Conclusions

The study was prepared to assess three different methods leading to the fabrication of novel luminescent composites consisting of TZN and LAO:Eu polycrystalline powders. The methods are evaluated on the merit of crystalline particles decomposition during the preparation process, as well as the optical and luminescence properties of the resulting composites. In terms of the invasiveness of the method towards the LAO:Eu structural integrity, the co-sintering (CS) method proves to be the most optimal, the direct-doping (DD) method is less beneficial and the remelt method leads to the complete decomposition of LAO:Eu. The best transparency is displayed by the sample prepared by the remelt method, but out of the methods that ensure the integrity of LAO:Eu, the direct-doping method provides the lowest transparency loss. The emission spectrum shape of the LAO:Eu polycrystals is preserved the most in the sample prepared by the co-sintering method. When excited directly into LAO:Eu’s charge transfer band, the integral emission intensity of this sample is only 10% of that from an equivalent amount of LAO:Eu powder. Moreover, this sample exhibits the lowest emission intensity, when it is excited into the f-f absorption band of Eu³⁺ (at 393 nm), due to the lowest rate of particle decomposition and diffusion of the Eu³⁺ ions into the glass matrix and the poorest transmittance. The co-sintering preparation of TZN + LAO needs to be further optimized for transparency in the visible spectrum to be applicable to producing better quality phosphor-in-glass composites. The optimization should focus on the LAO:Eu particle concentration, as well as the sintering conditions, such as temperature and duration.