Multicolor emission in lithium-aluminum-zinc phosphate glasses activated with Dy³⁺, Eu³⁺ and Dy³⁺/Eu³⁺

Abstract

A spectroscopic analysis of Dy³⁺, Eu³⁺ and Dy³⁺/Eu³⁺ doped lithium-aluminum-zinc phosphate glasses is carried out based on absorption and photoluminescence spectra and decay time measurements. According to the CIE1931 chromaticity coordinates and correlated color temperature (CCT), neutral white and reddish-orange global emissions were observed in the Dy³⁺ and Eu³⁺ singly-doped glasses, respectively, after excitations of Dy³⁺ at 348 nm and Eu³⁺ at 393 nm. A high red color purity of 97.2% is achieved in the Eu³⁺ singly-doped glass excited at 393 nm. Upon Dy³⁺ excitation at 348 nm, the Dy³⁺/Eu³⁺ doped glass showed warm white overall emission, with CCT value of 3629 K. Upon Dy³⁺ and Eu³⁺ co-excitations at 362, 381 and 387 nm, the Dy³⁺/Eu³⁺ doped glass showed reddish-orange overall emissions, with CCT values in the 1620–2385 K range, depending on the excitation wavelength. In the Dy³⁺/Eu³⁺ doped glass excited at 348 nm, the warm white emission was achieved by a non-radiative energy transfer from Dy³⁺ to Eu³⁺ with an efficiency and probability of 0.08 and 89.51 s⁻¹, respectively. The dominant mechanism could be through an electric quadrupole–quadrupole interaction, as it is suggested from the Inokuti-Hirayama model. A back non-radiative energy transfer from Eu³⁺ to Dy³⁺ is also observed, which could also be mediated by an electric quadrupole–quadrupole interaction. The Eu³⁺ to Dy³⁺ energy transfer efficiency and probability being of 0.08 and 30.00 s⁻¹, respectively.

1 Introduction

Due to its ⁴F_9/2 → ⁶H_13/2 (~480 nm), ⁴F_9/2 → ⁶H_15/2 (~573 nm), ⁶H_11/2 → ⁶H_13/2 (~1.34 μm) and ⁶H_13/2 → ⁶H_15/2 (~3.02 μm) emission bands, the incorporation of Dy³⁺ into a great variety of hosts has received considerable attention for W-LED, solid state laser and optical amplifier applications, among others [1–3]. By adjusting the relative blue (⁴F_9/2 → ⁶H_13/2) and yellow (⁴F_9/2 → ⁶H_15/2) emission intensities, it is possible to generate white light emission in Dy³⁺ singly-activated phosphors upon near-ultraviolet (NUV) excitation [4, 5], making Dy³⁺ a very attractive ion for W-LED applications. However, the negligible red emission of Dy³⁺ singly-doped phosphors represents a drawback to modulate and produce warm white light. In order to overcome this deficiency, Eu³⁺ ions are introduced to contribute with a red component, generating by this way tunable white emission [4–6]. Such emission can be achieved by an efficient Dy³⁺ → Eu³⁺ energy transfer process upon NUV excitation, which matches well with the emission for instance of InGaN-based LED. Among the hosts tested for the incorporation of Dy³⁺, Eu³⁺ and Dy³⁺/Eu³⁺, phosphate glasses possess unique characteristics, such as low melting temperature, high transparency in the NUV–Vis range, high thermal expansion, high luminous ion concentration, large emission and absorption cross section, among others [7–9]. However, their poor chemical durability is an inconvenience for many applications [8, 10]. As a solution, it has been reported that ZnO improves the chemical stability because Zn²⁺ is able to form ZnO₄ tetrahedron and/or P–O–Zn bridges through phosphate chain linkages in the glassy structure [9]. Furthermore, the combination of lithium and aluminum in phosphate glasses offers high transparency and thermal stability. All these characteristics make lithium–aluminum–zinc phosphate glass very attractive for potential W-LED applications [9, 11]. Thus, in this work a spectroscopy study based on absorption, excitation and emission spectra, and decay time profiles of lithium–aluminum zinc phosphate glasses activated with Eu³⁺, Dy³⁺ and Dy³⁺/Eu³⁺ is presented.

2 Experimental details

The molar compositions of the lithium–aluminum–zinc phosphate glasses under investigation were 5.0 Li₂O–5.0 Al₂O₃–39.5 ZnO–50.0 P₂O₅–0.5 Dy₂O₃, 5.0 Li₂O–5.0 Al₂O₃–39.0 ZnO–50.0 P₂O₅–1.0 Eu₂O₃, 5.0 Li₂O–5.0 Al₂O₃–38.5 ZnO–50.0 P₂O₅–0.5 Dy₂O₃–1.0 Eu₂O₃, which will be referred hereafter as LAZD, LAZE and LAZDE, respectively. The glasses were synthetized by mixing stoichiometric amounts of reagent grade NH₄H₂PO₄, ZnO, Li₂CO₃, Al₂O₃, Dy₂O₃ and Eu₂O₃ in a sintered alumina crucible and melting the mixture for 6 h at 1250 °C. The melts then were quenched onto a copper plate. The glasses were annealed at 350 °C for 16 h to obtain thermal and mechanical stability. Photoluminescence spectra were recorded by a Horiba Jobin–Yvon Fluorolog 3–22 spectrofluorometer operating with a 450 W ozone-free Xe lamp in the steady mode or with a pulsed Xe lamp for decay time profile measurements. The decay times were recorded in the phosphorescence mode using a delay time of 0.01 ms after the excitation pulse (3 μs half-width) and a 5 ms sample window. The absorption spectra were recorded by a Carry 5000 spectrometer. All measurements were carried out at room temperature.

3 Results and discussion

3.1 LAZD glass phosphor

Figure 1a shows the characteristic absorption spectrum of the LAZD glass on the 300–770 nm range. It consists of several bands centered at about 324, 348, 364, 386, 425, 452, 474 and 750 nm, corresponding to Dy³⁺ ion ⁴K_15/2, (⁴M_15/2,⁶P_7/2), ⁴I_11/2, (⁴K_17/2,⁴M_19/2,21/2,⁴I_13/2,⁴F_7/2), ⁴G_11/2, ⁴I_15/2, ⁴F_9/2 and ⁶F_1/2 + ⁶F_3/2, respectively. The absorption spectrum on the 800–1800 nm range displayed in Fig. 1b, shows bands at 886, 1094, 1266 and 1677 nm, associated with the Dy³⁺ ions ⁶H_5/2 + ⁶F_7/2, ⁶H_7/2 + ⁶F_9/2, ⁶H_9/2 + ⁶F_11/2 and ⁶H_11/2 transitions from the ⁶H_15/2 ground state, respectively. The absorption edge located below 340 nm reveals a high host transparency, which fulfills the requirements for optical applications in the visible range.

Fig. 1

Absorption on the a 300–770 and b 800–1600 nm ranges and c emission spectra of the LAZD glass

The emission spectrum, exciting into the ⁶H_15/2 → (⁴M_15/2,⁶P_7/2) absorption transition at 348 nm (Fig. 1c), displays the typical Dy³⁺ emissions centered at 454, 483, 572 and 661 nm, associated with the ⁴I_15/2 → ⁶H_15/2, ⁴F_9/2 → ⁶H_15/2, ⁶H_13/2 and ⁶H_11/2 Dy³⁺ ion transitions, respectively. These emissions are originated by an initial population from the (⁴M_15/2,⁶P_7/2) state, which non-radiatively relaxes to the ⁴F_9/2 emitting level, giving rise to the blue, yellow and red Dy³⁺ emissions. The feeble ⁴I_15/2 → ⁶H_15/2 emission at 454 nm is related with a non-radiative relaxation from the ⁴I_15/2 state to the ⁴F_9/2 one, as consequence of the small energy gap (~1000 cm⁻¹) between the ⁴I_15/2 and ⁴F_9/2 levels (see Fig. 2). On the other hand, the intensity ratio of the ⁴F_9/2 → ⁶H_13/2 (yellow) and ⁴F_9/2 → ⁶H_15/2 (blue) transitions dominates the emission tonality, and it is associated with the predominant symmetry around Dy³⁺ ions into the host. It is well known that emissions with ΔJ = 0, ±1, ±3 correspond to magnetic dipole (MD) transitions, and they are insensitive to the host environment, whereas emissions, with ΔJ = ± 2, ±4, ±6, are associated with hypersensitive transitions to the host environment. So that, when Dy³⁺ is mainly located into a non-inversion symmetry site, the intensity of the ⁴F_9/2 → ⁶H_13/2 emission is significantly higher than that of the ⁴F_9/2 → ⁶H_15/2 emission. Thus, the (⁴F_9/2 → ⁶H_13/2)/(⁴F_9/2 → ⁶H_15/2) emission intensity ratio can be used to evaluate the predominant symmetry around Dy³⁺ into the lithium–aluminum zinc phosphate glass host. In this case, the (⁴F_9/2 → ⁶H_13/2)/(⁴F_9/2 → ⁶H_15/2) emission intensity ratio resulted to be 1.57, suggesting that Dy³⁺ ions are mostly distributed into non-inversion symmetry sites. Thereby, neutral white emission is obtained from the LAZD glass according to its x = 0.367 and y = 0.412 CIE1931 chromaticity coordinates and correlated color temperature (CCT) of 4547 K, as calculated by the following Eq. [12]:

Fig. 2

Dy³⁺ and Eu³⁺ energy level diagram illustrating the possible energy transfer pathways among the Dy³⁺ and Eu³⁺ ions

$$CCT=449{n^3}+3525{n^2}+6823.8n+5520.33,$$

(1)

where n = (x − 0.3320)/(0.1858 − y), and x and y are the chromaticity coordinates.

3.2 LAZE glass phosphor

The absorption spectrum of the LAZE glass (Fig. 3a) shows eight bands located at 318, 362, 375, 382, 393, 415, 465 and 533 nm, corresponding to transitions of the ⁵H₆, ⁵D₄, ⁵G_2,3,4,5, ⁵L₈, ⁵L₇, ⁵D₃, ⁵D₂ and ⁵D₁ Eu³⁺ excited states, respectively. The emission spectrum, upon 393 nm excitation into the ⁵L₇ predominant absorption, exhibits the Eu³⁺ emissions centered at 577, 591, 611, 651 and 699 nm, associated with the ⁵D₀ → ⁷F₀, ⁵D₀ → ⁷F₁, ⁵D₀ → ⁷F₂, ⁵D₀ → ⁷F₃ and ⁵D₀ → ⁷F₄ electronic transitions, respectively (Fig. 3b). Such emissions are achieved after non-radiative relaxations from the ⁵L₇ state to the ⁵D₀ one, through the (⁵D₃,⁵L₆), ⁵D₂ and ⁵D₁ intermediate levels (see Fig. 2). Similarly to the dysprosium emissions, the europium ⁵D₀ → ⁷F₂/⁵D₀ → ⁷F₁ emission intensity ratio is useful to test the dominant symmetry around Eu³⁺ [4]. If Eu³⁺ ions occupy non-inversion symmetry sites, the intensity of the ⁵D₀ → ⁷F₂ emission should be higher than that of the ⁵D₀ → ⁷F₁ emission. In this case, the ⁵D₀ → ⁷F₂/⁵D₀ → ⁷F₁ emission intensity ratio is around 2.56, suggesting that like Dy³⁺ ions, Eu³⁺ ions mainly occupy sites without inversion symmetry. The overall emission of the LAZE glass upon 393 nm excitation was also characterized by the CIE1931 chromaticity coordinates and correlated color temperature (CCT), resulting in values of x = 0.638 and y = 0.354, and 2070 K, respectively. Such color coordinates are within the reddish-orange region, with a high red color purity of 97.2%. The color purity was estimated from the the weighted average of the sample emission colour coordinates (x _s, y _s) relative to the CIE1931 Standard Source C illuminant coordinates (x _i, y _i) and the dominant wavelength coordinates (x _d, y _d) relative to the (x _i, y _i) coordinates, through the following relation [13]:

Fig. 3

a Absorption and b emission spectra of the LAZE glass

$${\text{Color}}\;{\text{purity}}=\frac{{\sqrt {{{({x_s} - {x_i})}^2}+{{({y_s} - {y_i})}^2}} }}{{{{({x_d} - {x_i})}^2}+{{({y_d} - {y_i})}^2}}} \times 100\%$$

(2)

where the coordinates (x _d, y _d) are those of the monochromatic wavelength having the same color as the light source. The dominant wavelength is determined by drawing a straight line from the (x _i, y _i) coordinates through the (x _s, y _s) coordinates until the line intersects to the outer locus of points along the spectral edge of the CIE1931 chromaticity diagram.

3.3 LAZDE glass phosphor

In order to study the effect of Dy³⁺ co-doping on the Eu³⁺ excitation, Fig. 4a displays the excitation spectra of the LAZDE and LAZE glasses monitoring the europium ⁵D₀ → ⁷F₂ emission at 611 nm, wherein Dy³⁺ does not emit (see Fig. 1c). Besides the Eu³⁺ characteristic excitation bands, the excitation spectrum of the LAZDE glass shows two additional bands centered at 348 and 452 nm, related with the dysprosium ⁶H_15/2 → (⁴M_15/2,⁶P_7/2) and ⁴I_15/2 excitations, respectively, suggesting that Dy³⁺ sensitizes Eu³⁺. This process leads to Eu³⁺ emission through the Dy³⁺ excitation. The europium excitation band intensity in the LAZDE glass is lower than in the LAZE one, which could be evidencing Eu³⁺ → Dy³⁺ energy transfer, as discussed below.

Fig. 4

Excitation spectra of the glasses a LAZE and LAZDE and b LAZD and LAZDE monitoring the Eu³⁺ and Dy³⁺ emissions at 611 and 480 nm, respectively

Figure 4b depicts the excitation spectra of the LAZDE and LAZD glasses recording the dysprosium ⁴F_9/2 → ⁶H_15/2 emission at 480 nm, wherein Eu³⁺ is not able to emit. The excitation spectrum of the LAZDE glass displays, in addition to the Dy³⁺ transitions, a shoulder mounted on the dysprosium ⁶H_15/2 → (⁴K_17/2,⁴M_19/2,21/2,⁴I_13/2,⁴F_7/2) excitation band, which matches with the europium ⁷F₀ → ⁵L₇ excitation, pointing out a feasible Eu³⁺ → Dy³⁺ energy transfer process. Also, it can be noticed that the dysprosium excitation band intensity in the co-doped glass (LAZDE) is lower than in the singly doped one (LAZD). This fact might be associated with Dy³⁺ → Eu³⁺ energy transfer according to the Dy³⁺ excitation bands observed by monitoring the isolated Eu³⁺ emission.

In order to clarify the mechanism involved in the Dy³⁺ → Eu³⁺ energy process, Fig. 5a shows the decay profiles of the isolated dysprosium ⁴F_9/2 → ⁴H_15/2 emission at 480 nm under excitation at 348 nm of the LAZD and LAZDE glasses. The emission of the Dy³⁺ singly doped glass (LAZD) follows a non-exponential decay, attributed to cross relaxation processes among Dy³⁺ ions [14]. In presence of Eu³⁺, the decay remains non-exponential and becomes shorter, suggesting a non-radiative Dy³⁺ → Eu³⁺ energy transfer process. Because both decays are not exponential the lifetime was taken as the average lifetime [15], resulting to be 0.92 and 0.85 ms for the LAZD and LAZDE glasses, respectively. So that, the Dy³⁺ → Eu³⁺ energy transfer process is accomplished with an efficiency and probability of 0.08 and 89.51 s⁻¹, respectively [16]. According to the Dy³⁺ and Eu³⁺ energy levels and overlap region of the Dy³⁺ emission and Eu³⁺ absorption portrayed in Fig. 5b, this process could be achieved through the following channels illustrated in Fig. 2 [15, 17]:

Fig. 5

Decay time profiles of a Dy³⁺ emission at 480 nm for the LAZD and LAZDE glasses. Solid lines represent the best fitting by using Eq. (3), and b overlap region of the Dy³⁺ emission and Eu³⁺ absorption

ET1: ⁴F_9/2(Dy³⁺) + ⁷F_0,1(Eu³⁺) → ⁶H_15/2(Dy³⁺) + ⁵D₂(Eu³⁺) and.

ET2: ⁴F_9/2(Dy³⁺) + ⁷F_0,1(Eu³⁺) → ⁶H_13/2(Dy³⁺) + ⁵D₀(Eu³⁺).

The dominant electrostatic mechanism involved in the Dy³⁺ → Eu³⁺ energy transfer process is inferred by fitting the emission decay at 480 nm of the Dy³⁺ and Eu³⁺ co-doped glass with the Inokuti-Hirayama model given as follows [18]:

$$I\left( t \right)={I_o}\exp \left( { - \frac{t}{{{\tau _0}}} - {\gamma _s}{t^{3/S}}} \right)$$

(3)

where I ₀ is the intensity at t = 0, $\tau _{0}$is the donor (Dy³⁺) lifetime in absence of acceptors (Eu³⁺), ${\gamma }_{\text{s}}$ is a measure of the direct energy transfer and S is the multipolar interaction parameter. The latter one can take values of 6, 8 and 10 for electric dipole–dipole (d–d), dipole–quadrupole (d–q) and quadrupole–quadrupole (q–q) interactions, respectively. The best fitting was achieved for S = 10, with ${\gamma }_{10}=$ 5.24 s^−3/10 (see Fig. 5a). Therefore, it is inferred that an electric quadrupole–quadrupole interaction might dominate the Dy³⁺ → Eu³⁺ energy transfer process, as expected from the inter-4f transitions of the interacting ions.

The R _c critical energy transfer distance and $\gamma _{{\text{S}}}$parameter are related as follows:

$$\gamma _{S} = \frac{{4\pi }}{3}\Gamma \left( {1 - \frac{3}{S}} \right)\rho _{a} \left( {\frac{{R_{c} }}{{\tau _{0}^{{{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 S}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{$S$}}}} }}} \right)^{3}$$

(4)

In this equation, ${\rho }_{\text{a}}$is the acceptor (in this case Eu³⁺) concentration (8.9  × 10¹⁹ ions cm⁻³), $\tau _{0}$ is the donor average lifetime in absence of acceptors and $\Gamma (1-\frac{3}{S})$ is the Euler´s gamma function, which takes values of 1.77, 1.43 and 1.30 for S = 6, 8 and 10, respectively. Based on this relationship, the R _c critical distance for an electric quadrupole–quadrupole interaction resulted to be 11 Å. The Dy³⁺–Eu³⁺ interaction distance for a random ion distribution [19], considering ρ _Dy + ρ _Eu ≈ 1.3 × 10²⁰ ions cm⁻³, is around 24 Å. This distance is larger than the critical distance, suggesting that the Dy³⁺ → Eu³⁺ energy transfer predominantly arises from Dy–Eu clusters instead of from the randomly distributed ions.

With regard to the Eu³⁺ → Dy³⁺ energy transfer, Fig. 6a depicts the isolated Eu³⁺ emission decay profiles at 611 nm, under excitation at 393 nm, for the LAZE and LAZDE glasses. Unlike to the Dy³⁺ emission decay, in the singly Eu³⁺ doped glass (LAZE) the Eu³⁺ emission decay exhibits an almost exponential evolution. In the co-doped glass (LAZDE) the Eu³⁺ emission decay displays a non-exponential behavior at times shorter than 1 ms and it becomes faster, which reveals Eu³⁺ → Dy³⁺ non-radiative energy transfer. The average lifetimes of the Eu³⁺ emission resulted to be 2.82 and 2.60 ms for the LAZE and LAZDE glasses, respectively. These values suggest that the Eu³⁺ → Dy³⁺ energy transfer process is achieved with efficiency and probability of 0.08 and 30.00 s⁻¹, respectively [16]. Based on the Dy³⁺ and Eu³⁺ energy level diagram and overlap region of the Eu³⁺ emission and Dy³⁺ absorption showed in Fig. 6b, such process might arise through the following pathways (see Fig. 2) [15, 17]:

Fig. 6

a Eu³⁺ emission at 611 nm for the LAZE and LAZDE glasses. Solid lines represent the best fitting by using Eq. (3), and b overlap region of the Eu³⁺ emission and Dy³⁺ absorption

ET3: ⁵D₃(Eu³⁺) + ⁶H_15/2(Dy³⁺) → ⁷F₂(Eu³⁺) + ⁴G_11/2(Dy³⁺),

ET4: ⁵D₃(Eu³⁺) + ⁶H_15/2(Dy³⁺) → ⁷F₃(Eu³⁺) + ⁴I_15/2(Dy³⁺) and

ET5: ⁵D₂(Eu³⁺) + ⁶H_15/2(Dy³⁺) → ⁷F₁(Eu³⁺) + ⁴F_9/2(Dy³⁺).

To infer the dominant electrostatic mechanism involved in the Eu³⁺ → Dy³⁺ energy transfer, the Eu³⁺ emission decay at 611 nm was fitted by Eq. (3). The best fitting was attained for S = 10 with ${\gamma }_{10}=$ 2.12 s^−3/10 (Fig. 6a), suggesting that the energy transfer from Eu³⁺ to Dy³⁺ might be mediated by a quadrupole–quadrupole electric interaction, in a similar way as the Dy³⁺ → Eu³⁺ energy transfer. In this case, the R _c distance determined by Eq. (4), for a Dy³⁺ acceptor concentration of 4.4⨉10¹⁹ ions cm⁻³, is 12 Å, which is smaller than the Eu–Dy interaction distance assuming a random ion distribution (24 Å). Therefore, the Eu³⁺ → Dy³⁺ energy transfer might take place among Eu–Dy clusters.

3.4 CIE1931 chromaticity coordinates and correlated color temperature for different excitation wavelengths

By comparing the Dy³⁺ and Eu³⁺ relative excitation intensities illustrated in Fig. 4a, b, it can be observed that excitation wavelengths at 348, 362, 381 and 387 could be useful to modulate the relative intensity of the Dy³⁺ and Eu³⁺ emissions in order to produce different light tonalities. Such excitations are within emissions of commercial AlGaN (348 nm), GaN (360 nm) and InGaN (380–405 nm) LEDs [20–22]. Figures 7 and 8 show respectively, the feature emissions and CIE1931 color chromatic coordinates of the LAZDE glass, upon the excitations above described. The emissions of the LAZD and LAZE singly doped glasses were included for comparison as well. Upon Dy³⁺ excitation [⁶H_15/2 → (⁴M_15/2,⁶P_7/2)] at 348 nm (Fig. 7a), the LAZDE glass exhibits, besides the Dy³⁺emissions a band located at 611 nm, which is due to the europium ⁵D₀ → ⁷F₂ transition. In the LAZDE glass pumped at 348 nm, wherein Eu³⁺ is hardly excited (Figs. 4a, 7a), the Eu³⁺ (⁵D₀ → ⁷F₂) red emission is achieved at expensed of Dy³⁺ → Eu³⁺ energy transfer. This fact allows to obtain warm white light emission with x = 0.406 and y = 0.409 CIE1931 chromaticity coordinates and CCT value of 3629 K, evidencing that the Eu³⁺ addition shifts the neutral white emission of the LAZD glass to warm one (Fig. 8). Under dysprosium ⁶H_15/2 → ⁴I_11/2 and europium ⁷F₀ → ⁵D₄ co-excitations at 362 nm, the dominant Eu³⁺ emission in the LAZDE glass (Fig. 7b) produces reddish-orange tonality with x = 0.513 and y = 0.377 CIE1931 chromaticity coordinates, CCT value of 1839 K and red color purity of 71.7% (Fig. 8). Upon dysprosium ⁶H_15/2 → ⁶P_3/2,5/2 and europium ⁷F₀ → ⁵L₈ co-excitations at 381 nm (Fig. 7c), the overall emission of the LAZDE glass is within the reddish-orange region with x = 0.566 and y = 0.368, CCT value of 1620 K and red color purity of 82.8% (Fig. 8). Upon dysprosium ⁶H_15/2 → (⁴K_17/2,⁴M_19/2,21/2,⁴I_13/2,⁴F_7/2) and europium ⁷F₀ → ⁵L₈ excitations at 387 nm (Fig. 7d), the overall emission of the LAZDE glass results in reddish-orange light with red color purity of 63.7% (Fig. 8), according to x = 0.471 and y = 0.392 CIE1931 coordinates and CCT value of 2385 K. In all cases, it can be observed that the emission intensity of the LAZDE glass is lower than that of the Dy³⁺ and Eu³⁺ singly doped glasses, which is associated with energy transfer between Dy³⁺ and Eu³⁺ ions. Finally, from Fig. 8 and Table 1, it can be appreciated that the overall emission of the Dy³⁺ and Eu³⁺ co-doped glass (LAZDE) can be tuned from warm white to reddish-orange light with red color purity in the 50.1–82.8% range, depending on the excitation wavelength, as consequence of the relative Dy³⁺ and Eu³⁺ emission intensities.

Fig. 7

Emission spectra of the LAZD, LAZE and LAZDE glasses upon excitations at a 348, b 362, c 381 and d 387 nm

Fig. 8

Chromaticity coordinates in CIE1931 diagram of the global emissions observed in the LAZD, LAZE and LAZDE glasses upon 348, 362, 381 and 387 nm excitations

Table 1 CIE1931 color chromaticity coordinates, correlated color temperature (CCT), dominant wavelength (λ _d), red color purity and emission tonality of the glasses LAZD excited at 348 nm, LAZE excited at 393 nm, and LAZDE excited at 348, 362, 381 and 387 nm

4 Conclusions

A spectroscopic analysis was carried out in Dy³⁺, Eu³⁺ and Dy³⁺/Eu³⁺ doped lithium-aluminum-zinc phosphate glasses from absorption, excitation and emission spectra, and decay time measurements. It was observed that the Dy³⁺ and Eu³⁺ doped glasses exhibits the absorption edge below than 350 nm, which fulfils the requirements for W-LEDs applications. Reddish-orange emission tonality was obtained in the Eu³⁺ singly doped glass exciting at 393 nm, whereas neutral white emission was generated in the Dy³⁺ singly doped glass upon 348 nm excitation. The Dy³⁺ and Eu³⁺ co-doping allowed shifting the emission tonality from neutral white to warm white, induced by the Eu³⁺:⁵D₀ → ⁷F₂ red emission contribution, upon excitation of 348 nm. Under excitations at 362, 381 and 387 nm, the emission tonality is located in the reddish-orange region due to the Eu³⁺ dominant red emission. Dy³⁺ → Eu³⁺ non-radiative energy transfer, dominated by an electric quadrupole–quadrupole interaction, allows to obtain warm white emission in the Dy³⁺/Eu³⁺ co-doped glass. Such process was achieved with efficiency and probability values of 0.08 and 89.51 s⁻¹, respectively. The decay time shortening of the Eu³⁺ emission in the Dy³⁺/Eu³⁺ co-doped glass revealed that Eu³⁺ → Dy³⁺ non-radiative energy transfer, mediated by an electric quadrupole–quadrupole interaction, occurs with an efficiency and probability of 0.08 and 30.00 s⁻¹, respectively.