Effect of preparation conditions in sol-gel method on yellow phosphor with wide spectrum

Light emitting diodes (LEDs) have attracted a lot of attention in recent times as a potential candidate for new generation illumination. They are expected to replace incandescent and fluorescent lamps due to advantages such as their light weight, small size, high power efficiency, low energy consumption, long lifetime, and absence of mercury.

White LED light is achieved by mixing near-ultra-violet (NUV) or blue LED light with light from phosphors derived from their complementary colors, excited by themselves.¹ Two-band white LED light using blue LED light converted by the yellow phosphor YAG: Ce is the most popular combination to obtain white light. While it is easy to fabricate and has high efficiency,² this application suffers from low color rendering index and high correlated color temperature due to the lack of red emission. In order to overcome these problems, the three-band white LED has been developed, by applying a red phosphor material in the phosphor layer.³ While this improved the color performance, the efficiency was largely degraded due to the reabsorption between the two phosphor materials. A phosphor with a broad emission band covering yellow to red could be a solution to this problem.

Li₂SrSiO₄ is an ideal host for photoluminescence materials due to its low synthesis temperature and high thermal and chemical stability. Weak thermal quenching in luminescence intensity and tunable luminescence color by doping with different activators are its other attractive features. Many studies of Li₂SrSiO₄ have focused on the luminescent properties upon doping with Eu²⁺ and Ce³⁺. Eu²⁺-doped Li₂SrSiO₄ shows yellowish-orange emission with a maximum at ∼570 nm,^4–8 while Ce³⁺-doped Li₂SrSiO₄ shows blue emission with a maximum at 430 nm.^9–13 Li₂SrSiO₄:Eu²⁺ displays strong absorption in the range 380–460 nm and exhibits good luminescence efficacy.

While Li₂SrSiO₄ was obtained by the conventional solid-state route in most of the previous studies, the sol-gel method can provide better uniformity of ingredients and improved particle size distribution.¹⁴ However, there are very few reports of the effect of composition and synthesis conditions on the luminescence properties. Thus, the main purpose of this study was to investigate the effect of synthesis conditions on the properties of the phosphor and develop the best synthetic route for Li₂SrSiO₄:Eu²⁺.

The Li₂SrSiO₄:Eu²⁺ phosphors were synthesized by the following sol-gel method. Stoichiometric amounts of LiNO₃, Sr(NO₃)₂, and Eu₂O₃ were dissolved in deionized water; hydrolysis of tetraethyl orthosilicate (TEOS) in deionized water was the source of silicon. The hydrolysis of TEOS is expressed by the following reaction.

Further hydrolysis results in Si(OH)₄. This is followed by condensation polymerization, as shown.

TEOS alone would yield SiO₂, but the presence of LiNO₃ and Sr(NO₃)₂ resulted in the final product of the host material, Li₂SrSiO₄. In order to accelerate the slow hydrolysis reaction, an acid catalyst was added and the pH adjusted to 0.7. Due to their comparable atomic sizes, Eu can be doped instead of Sr in Li₂SrSiO₄, and the target phosphor compositions were Li₂Sr_1-xEu_xSiO₄ (x = 0.003, 0.005, 0.0075).

TEOS was first mixed with 15 mL of ethanol, dissolved in an equivalent amount of distilled water, and left at room temperature for hydrolysis. The other cation chemicals were dissolved in distilled water. As the solubility of Eu₂O₃ in water is low, a few drops of nitric acid were added. The two solutions were then mixed, the pH was adjusted to 0.7, the reaction mixture covered with foil, and stored in a drying oven. The oven temperature was set at 60 °C in order to manage the condensation rate. A transparent gel was obtained after 12 hours and dried at 80 °C in order to form a white xerogel, which was a solid form from a gel by drying with unhindered shrinkage. The xerogel was pulverized, and calcined at 800 °C for 1 h in air. As the yellow luminescence originates from Eu²⁺, reduction is required in the following step. The calcined xerogel was cooled to room temperature and reduced for 2 h (5 % H₂+95 % Ar) at 800 °C, at a heating rate of 20 °C/min. When the yellow phosphor was heated in air, the color of the material changed back to white, indicating complete oxidation.

The resultant samples were characterized by X-ray diffraction (XRD, Ultima IV MultiFlex XRD, Rigaku, Japan) using Cu Kα as the X-ray source (1.54056 Å) and a high-speed detector (D/tex Ultra). The X-ray generator was operated at 40 kV and 30 mA, and data was recorded in the 2θ range 20–70° with a 0.02° step size. Thermogravimetric analysis (TGA) of the powder precursor was carried out using a TG analyzer (TGA-50, Shimadzu, Japan) in air. Differential scanning calorimetry (DSC, DSC-6300, Hitachi, Japan) was used to analyze the phase changes and melting temperature of the resultant phosphor samples. TGA and DSC studies were carried out at a heating rate of 10 °C/min. The reflectance spectra of the phosphor samples were measured by an ultraviolet-visible (UV-vis) spectrophotometer (U-4100, Hitachi, Japan). Photoluminescence (PL) spectra were measured by a self-built integrating sphere system in which the samples were excited by a 450 nm blue InGaN LED. A field emission scanning electron microscope, FESEM (S-5200, Hitachi, Japan) was used to study the morphology of phosphor sample.

The TGA results for the Li₂SrSiO₄:Eu²⁺ phosphor synthesized under basic conditions is shown in Fig. 1. Two weight loss stages were observed: from room temperature to 570 °C, and in the range 570–680 °C. The first weight loss stage of 6.4 % was triggered by the volatilization of water and organic contaminants in the sample. The large weight loss of 46.0 % was attributed to the decomposition of the nitrate group in the polymeric linkage, since nitrogen dioxide is a red gas and easily observed. The DSC results of the Li₂SrSiO₄:Eu²⁺ (x = 0.005) phosphor and the host are shown in Fig. 2. From the DSC curve of host lattice the glass-transition, crystallization, and crystal melting temperatures are approximately 904, 907, and 916 °C, respectively. An unknown exothermic reaction is observed at 922 °C in the DSC curve of Li₂SrSiO₄: Eu²⁺ (x = 0.005). This could be due to the oxidation of europium (Eu²⁺ to Eu³⁺). The crystal melting temperature seems to increase slightly upon europium doping.

As shown in Fig. 3, all the diffraction peaks could be assigned to the trigonal Li₂SrSiO₄ (space group: P3₁21), with lattice constants a = 5.023 Å and c = 12.457 Å. In the crystal structure, tetrahedrons of SiO₄ and LiO₄ at the corners of the unit cell form vacancies where Sr/Eu is held. As a result, each Sr/Eu has eight neighboring oxygen atoms. A peak corresponding to α-$Sr2SiO4$ at 34° was also observed. This phase forms when there is an excess of SiO₂ and SrO, which suggests the volatilization of Li₂O upon calcination. Since the ionic radius of Eu (Eu²⁺ = 1.25 Å, CN = 8) is similar to that of Sr (Sr²⁺ = 1.26 Å, CN = 8), and is very different from lithium or silicon, Eu²⁺ was preferentially substituted for Sr²⁺, and formed the isostructural Li₂EuSiO₄. Samples with different Eu concentrations revealed nearly identical XRD patterns, indicating that Eu doping does not affect the crystal structure (Figs. 3(a, d, e)). The intensity of the impurity phase peak decreased at higher calcination temperatures, suggesting higher crystallinity and lesser impurity phase as shown in Figs. 3(a, b, c).The impurity diffraction peak intensity increased when the sample was calcined for longer duration, due to the volatilization of Li₂O from the lattice (Figs. 3 (a, f, g)). The condensation temperature and pH value of the precursor solution would directly affect the formation of cross linkages and the homogeneity of ions. Figs. 3(a, i, h) shows the XRD patterns of the samples prepared at different gelation temperature. While there are no obvious differences in XRD spectra, the higher gelation temperature must reduce the gelation time from sol to gel. The XRD patterns of the samples prepared at different pH values (0.7 to 1.0 and 1.5), revealed that the crystallinity and phase purity declined as the pH value increased (Figs. 3(a, j, k)). At lower pH values, the hydrolysis and polymerization reaction of the metal alkoxide is accelerated, while at higher pH value the reaction is mild, resulting in homogeneity.¹⁶ Thus, we intuitively expect that higher pH would lead to higher crystallinity, the results are contrary, and we are unable to explain it presently. However, if the reaction is too slow and requires longer gelation time, the resulting gel is found to include water and alcohol byproducts.

Eu²⁺ emissions are triggered by the 4f⁶5d¹ to 4f⁷ transition.¹⁵ The emission intensity of d-f transitions are usually much larger than that of f-f transitions, and exhibit wide band emission. Also the d-orbital in Eu²⁺ is easily affected by ligands which result in emission shifts. Thus, the emission can range from blue to red depending on the crystal fields of substituting site in host lattice. Fig. 4 is the emission spectrum of Li₂Sr_1-xEu_xSiO₄ (x = 0.003, 0.005, 0.0075) excited by 450 nm. The emission peak was observed at 573 nm indicating yellow-orange light and the full width at half maximum (FWHM) was about 86 nm. This was due to the strong crystal fields of the solitary Sr²⁺ site, which is a consequence of the asymmetric bonds and high coordination number. The Eu²⁺ concentration did not affect the peak position or FWHM, but the quantum yield did show a variation, with the highest intensity observed at x = 0.005. The higher concentration shortened the distance between two activators, increasing the possibility of energy transfer between the activators, and the possibility of irradiation relaxation. This phenomenon is also known as concentration quenching effect of phosphor materials.

Figure 5 showed the PL spectra of the phosphors synthesized at different calcination temperatures, calcination times, condensation temperatures, and pH values for gelation. It was found that a higher calcination temperature resulted in higher photoluminescence intensity (Fig. 5a). When we take into account the results of XRD as well, it is evident that higher temperatures result in higher crystallinity and efficient photoluminescence. A calcination time of 2 h yields the best photoluminescence intensity (Fig. 5b). Increasing calcination times invites Li defects which decrease efficiency (Fig. 3). Considering the Arrhenius behavior often observed in reactions, increased calcination temperature must also induce Li volatilization, and it is hence necessary to identify the optimum temperature. While both gelation temperature and pH value affect gelation time, gelation temperature has a greater influence. It is expected that high temperatures lead to inhomogeneity, and thus lower efficiency. In this experiment the pH value was fixed at 0.7, so at low gelation temperatures, it is possible that the water formed during condensation (dehydration) remained inside the 3-dimensional SiO₄ framework. On the other hand, at higher pH values, hydrolysis was so slow that water and alcohol byproducts remained in the gel.

Li₂Sr_1-xEu_xSiO₄ (x = 0.003–0.0075) phosphors were synthesized by the sol-gel process. The 4f⁶5d¹ to 4f⁷ transition of Eu²⁺ in Li₂SrSiO₄:Eu²⁺ yields a yellowish-orange luminescence in the range 500–700 nm, peaking at 573 nm. While slower reactions are better for crystallization, this reaction involved hydrolysis and condensation as well, so it was important to find a balance of all the reactions. At higher calcination temperatures, Li₂O volatilization was found to be significant. Hence, in this study, we obtained phosphors for broad yellow luminescence and the optimum synthesis conditions were identified.