It is advisable to consider the YSAG : Yb : Er solid solutions formed in the system of oxide compositions Y₂O₃–Er₂O₃–Yb₂O₃–Sc₂O₃–Al₂O₃ as substitutional solid solutions in the matrix of yttrium-aluminum garnet (Y₃Al₂Al₃O₁₂). Since the ionic radii of ytterbium (R (Yb_C³⁺) = 0.0985 nm) and erbium (R (Er_C³⁺) = 0.1004 nm) are close to the ionic radius of yttrium (R (Y_C³⁺) = 0.1019 nm), then they occupy dodecahedral positions. Aluminum cations are located in octahedral (Al_A³⁺) and tetrahedral (Al_D³⁺) positions. Scandium can be incorporated into dodecahedral (Sc_C³⁺) and octahedral (Sc_A³⁺) positions [25–26]. Previously, it was reported on the synthesis and study of the properties of scandium-containing garnets (YSAG : Yb : Er) compositions:

– Y_2.34Yb_0.45Er_0.09Sc_0.2Al_1.92Al₃O₁₂ (AP1),

– Y_1.66Yb_0.45Er_0.09Sc_1.0Al_1.8Al₃O₁₂ (AP2) and

– Y_0.96Yb_0.45Er_0.09Sc_1.70Al_1.8Al₃O₁₂ (AP3) [24].

In this work, the following YSAG : Yb : Er optical ceramics samples were manufactured: Y_1.86Yb_0.45Er_0.09Sc_1.0Al_1.6Al₃O₁₂ (СА20) and Y_2.26Yb_0.45Er_0.09Sc_1.2Al_1.0Al₃O₁₂ (А50), which were distinguished by increased concentrations of Sc³⁺ cations. In particular, in sample CA20, the concentration of Sc_A³⁺ = Sc_C³⁺ = 20 at.%, and in sample A50, the concentration of Sc_A³⁺ is 50 at.%, while Sc_C³⁺ content is 6.67 at.%. At the same time, in samples CA20 and A50, as well as in the samples from ref. [24], the concentrations of Yb_C³⁺ and Er_C³⁺ cations were 15 at.% and 3 at.%, respectively.

Samples of optical ceramics were fabricated through non-reactive sintering of nanocrystalline powders of yttrium-scandium-aluminum garnets doped with ytterbium and erbium cations (YSAG : Yb : Er). The precursor powders were synthesized via reverse chemical precipitation of concentrated salt solutions: AlCl₃ • 6H₂O (99.9%, Acros Organics, Belgium), ScCl₃ • 6H₂O (99.9%, Vekton, Russia), ErCl₃ • 6H₂O (99.9%, Vekton, Russia), YbCl₃ • 6H₂O (99.9%, Vekton, Russia), YCl₃ • 6H₂O (99.9%, Chemical Point, Germany). A concentrated ammonia solution (25%) with the addition of ammonium sulfate (0.45 mol./l) was used as a precipitant. The resulting precipitate was washed with ammonium sulfate solution (0.045 M) through centrifugation. The washed precipitate was dried at 60 °C for 20 h. Next, the powders were wet ground in a planetary mill for 30 min. in an aqueous medium using zirconium dioxide balls. At this stage, magnesium chloride was introduced into the precursor powders as a precursor of magnesium oxide (MgO), which was used as a sintering additive. The calculated concentration of MgO was fixed and amounted to 0.75 mol.%. Synthesis of YSAG : Yb : Er ceramic powders were carried out by calcination the precursor powder in air at a temperature of 1200 °C during 2 h.

Next, the ceramic powders were ground in an ethanol medium using zirconia balls in a Pulverisette 5 planetary mill for 20 min., followed by drying at 60 °C for 20 h.

In this way, nanocrystalline powders with crystallite size ≈ 60 ± 5 nm and specific surface area ≈ 9,2 ± 1,1 m²/g were obtained.

Small portions of powder were taken from each sample of YSAG : Yb : Er ceramic powders to obtain microcrystalline powders. These microcrystalline powders were used to check the phase composition of the studied oxide compositions. The samples were calcined in air at a temperature of 1600 °C using a high-temperature furnace NT 40/17 (Nabertherm GmbH). The duration of isothermal exposure was 2 h.

Ceramic powders were uniaxially pressed into the green bodies with a diameter of 15 mm and a thickness of 4 mm at a pressure of 50 MPa.

Sintering of ceramic samples was carried out in an “SShVE-1.2.5/25” vacuum furnace (LLC “VNIIETO”). The residual pressure in the vacuum furnace chamber did not exceed 5 · 10^-5 Pa.

The obtaining ceramic samples were subjected to double-sided grinding and mirror polishing using a Qpol-250 machine (Germany). The final thickness of all samples was 1.00 ± 0.01 mm.

The phase composition of YSAG : Yb : Er powders was determined by XRD-method. A diffractometer Empyrean (PANalytical, Netherlands) with CuK_α radiation (λ = 0.15418 nm, 2θ range 10–90°, with a step of 0.01°, scanning speed of 0.7°/min.) were used for XRD-investigations. Data evaluation and phase identification were performed in the HighScore Plus with the ICDD PDF-2 database.

The optical transmission spectra in the ranges of 100–200 nm and 1250–2500 nm were measured by using SF-56 and FSM-1211 (OKB Spektr, Russia) spectrophotometers, respectively.

Measurements aimed at determining the refractive index of YSAG : Yb : Er optical ceramic samples were carried out by using a SE-800 spectroscopic UV-VIS ellipsometer (SENTECH Instruments GmbH, Germany). The results of ellipsometry were analyzed using SpectraRay/3 software.

The luminescence spectra of the ceramic samples were measured by using a SFL-MDR spectrofluorimeter (OKB Spectr, Russia). The R928 photomultiplier tube (Hamamatsu Photonics, Japan) and G12180 semiconductor detector (Hamamatsu Photonics, Japan) were used for measuring at 300–900 nm and 900–1800 nm spectral ranges respectively.

Luminescence kinetics at wavelengths of 561 nm, 679 nm and 1029 nm was studied using SFL-MDR spectrofluorimeter (OKB Spectr, Russia). An SFL-MDR-2 spectrofluorimeter (OKB-Spectr, Russia) was used to study the luminescence kinetics in the region of 1535 nm.

Studies of the phase composition of microcrystalline powders CA20 and A50 showed that they are solid solutions with a garnet structure (Fig. 1). No impurity phases were detected. This indicates that all cations are distributed over positions of the garnet crystal lattice.

The samples differed in the values of the crystal lattice parameters (a_G) (Table 1). The highest a_G value was recorded for sample A50, whose crystal lattice parameter was 0.01400 nm greater than that of AP1 and 0.05705 nm greater than that of sample AP3. High values of a_G can be explained by the fact that the ionic radius of the Sc_A³⁺ cation (R (Sc_A³⁺) = 0.0745 nm) is greater than the ionic radius of the Al_A³⁺ cation (R (Al_A³⁺) = 0.0535 nm)) [27]. Therefore, according to Vegard’s law [28], an increase in a_G is due to a decrease in the proportion of Al_A³⁺ and an increase in the proportion of Sc_A³⁺. The comparable a_G values of samples AP1 [24] and C20 indicate mutual compensation of the effects of stretching and compression of the crystal lattice associated with the Sc_A³⁺ and Sc_C³⁺ cations, respectively.

Figure 1.

X-ray diffraction patterns for YSAG : Yb : Er: (a) CA20; (b) A50

Theoretical estimation of the crystal lattice parameters of scandium-containing garnets (a_SV) was made using the equation [25]:

a_SV = (a_S + a_V)/2,

a_S = 7.02954 + 3.31277r_C + 2.49398r_A + 3.34124r_D – 0.87758r_Cr_A – 1.38777r_Cr_D;

a_V = 10.092217 + 0.841118r_C + 0.734598r_A – 2.507813r_D + 3.133970r_Cr_D + 1.946901r_Ar_D;

r_C = [Y_C³⁺]R (Y_C³⁺) + [Yb_C³⁺]R (Yb_C³⁺) + [Er_C³⁺]R (Er_C³⁺) + [Sc_C³⁺]R (Sc_C³⁺);

r_A = [Al_A³⁺]R (Al_A³⁺) + [Sc_A³⁺]R (Sc_A³⁺);

r_D = [Al_D³⁺]R (Al_D³⁺);

[Y_C³⁺] + [Yb_C³⁺] + [Er_C³⁺] + [Sc_C³⁺] = 1;

[Al_A³⁺] + [Sc_A³⁺] = 1;

[Al_D³⁺] = 1.

where a_S is the crystal lattice parameter in Strock’s empirical formula [29]; a_V is the crystal lattice parameter in Vorobiev’s empirical formula [30]; [Y_C³⁺], [Yb_C³⁺], [Er_C³⁺], [Sc_C³⁺], [Al_A³⁺], [Sc_A³⁺], [Al_D³⁺] – concentrations of cations in the corresponding positions of the YSAG : Yb : Er crystal lattice; R (Y_C³⁺), R (Yb_C³⁺), R (Er_C³⁺), R (Sc_C³⁺), R (Al_A³⁺), R (Sc_A³⁺), R (Al_D³⁺) – Shannon ionic radii of cations [27]. Good agreement between the values of a_SV and a_G indicates that the planned and actual compositions of the garnet phases coincide. Modeling X-ray diffraction patterns using the Rietveld method allowed for an additional assessment of the concentrations of scandium cations in the dodecahedral and octahedral positions of the garnet crystal lattice. The initial coordinates of oxygen for the space group Ia3d (number 230) were taken from [31]. In the simulation, it was assumed that the 24d positions are occupied by Al_D³⁺ cations, and in positions 24c – cations Y_C³⁺, Yb_C³⁺, Er_C³⁺ and Sc_C³⁺. In position 16a, cations Al_A³⁺ and Sc_A³⁺ are present, and in positions 96 h, oxygen is located. Fig. 1 shows that the results of modeling and experimental studies of microcrystalline powders are in good agreement, which allows us to assert that the planned and actual compositions of the samples coincide.

Samples of optical ceramics were obtained by sintering compacts of nanocrystalline powders in a vacuum furnace. The optimum vacuum sintering temperatures were found to be 1850 °C and 1775 °C for A50 and CA20, respectively. The values of the optimal sintering temperatures are in good agreement with the results of previous studies of YSAG : Yb [32] and YSAG : Er [26] solid solutions with comparable concentrations of Sc³⁺ cations for A and C sites. Photographs of optical ceramic samples after polishing are shown in Fig. 2 a. The samples were characterized by high transmittance in the visible region of the spectrum (Fig. 2 b), which allows them to be classified as optically transparent.

Figure 2.

Photographs of YSAG : Yb : Er ceramic samples (a) and light transmission spectra (b)

Analysis of the transmission spectra of YSAG : Yb : Er samples showed that they are similar to the spectra of YAG : Yb : Er [12] and YSAG : Yb : Er [24], since they contain absorption bands associated with the presence of Er³⁺ and Yb³⁺ cations.

Table 2 shows the parameters of the Selmeyer model [33, 34] for the manufactured YSAG : Yb : Er samples. This model describes the refractive index dispersion (n (λ)) for the bulk layer:

$n(\lambda )={(\epsilon (\mathrm{\infty })+\frac{A{\lambda }^2}{{\lambda }^2-B^2}-C{\lambda }^2)}^{1/2}$ ,

where n (λ) is the refractive index at wavelength λ, ε(∞) is the high-frequency dielectric constant of the material, A, B and C are the parameters of the Sellmeyer model. The corresponding refractive index dispersions of the samples are presented in Fig. 3.

Figure 3.

Spectral dependences of the refractive index of samples A50 and CA20

According to the data presented in Table 2, the refractive index values (n) at a wavelength of 632.8 nm for the samples A50 (Sc_A³⁺ = 50 at.%, Sc_C³⁺ = 6.67 at.%) and СА20 (Sc_A³⁺ = Sc_C³⁺ = 20 at.%) are comparable in magnitude with the n values of the sample AP3 (Sc_A³⁺ =10 at.%, Sc_C³⁺ = 50 at.%) within the limits of error [24]. Note that for single crystals, the following n values were registered: Y₃Al₅O₁₂ n = 1.829 [35], Y₃Sc₁Al₄O₁₂ – 1.850 [36], and Y₃Sc₂Al₃O₁₂ – 1.873 [34].

It should be noted that the established "n" values in YSAG : Yb : Er samples with different concentrations of Sc_A³⁺ and Sc_C³⁺ cations are in good agreement with the results of previously conducted studies of YSAG : Cr solid solutions [37], which had comparable concentrations of scandium in the dodecahedral and octahedral positions of the crystal lattice. This circumstance allows us to conclude that an increase in the concentration of scandium at any position of the crystal lattice leads to an increase in the refractive index of a solid solution with a garnet structure.

Earlier, in [24], it was noted that the influence of scandium cations in the dodecahedral and octahedral positions of YSAG : Yb : Er on the properties of Yb³⁺ and Er³⁺ cations can be neglected at concentrations of cations Sc_C³⁺ and Sc_A³⁺ up to 4 at.%. In particular, it was shown that the absorption spectrum of Yb³⁺ cations in the YSAG : Yb : Er sample (composition AP1) and YAG : Yb [38, 39] and YAG : Yb : Er [12] ceramics is similar. As a rule, the spectrum of Yb_C³⁺ cations contains several Stark absorption bands (SB). At the same time, the bands SB(ν₀ → ν₅) and SB(ν₀ → ν₄) deserve special attention, since their absorption maxima are located in the region of laser pumping wavelengths of commercially produced InGaAs LEDs.

Figure 4 shows the absorption spectra of Yb³⁺ cations in samples CA20 and A50. For comparison, the spectrum of sample AP1 [24] is shown, for which the absorption coefficients (AC) are approximately 16 cm^-1 for the absorption bands SB(ν₀ → ν₅) and SB(ν₀ → ν₄) at wavelengths 939 nm and 968 nm, respectively.

Figure 4.

Absorption spectra of the YSAG : Yb : Er samples: (a) Yb_C³⁺ (²F_7/2 → ²F_5/2 term, 890–1065 nm), (b) Er_C³⁺ (⁴I_15/2 → ⁴I_13/2, 1430–1690 nm), (c) Er_C³⁺ (⁴I_15/2 → ⁴S_3/2, 535–565 nm), (d) Er_C³⁺ (⁴I_15/2 → ⁴F_9/2, 640–690 nm)

In sample A50, with the concentration of Sc_A³⁺ of 50 at.%, the SB(ν₀ → ν₅) band is shifted by 1.0–1.5 nm to the region of longer wavelengths, while the ratio of absorption coefficients AC(ν₀ → ν₄)/AC(ν₀ → ν₅) increased to 1.15.

Despite the fact that in sample CA20 the concentrations of Sc_A³⁺ and Sc_C³⁺ are five times higher than in sample AP1, the positions of the maxima of their SB(ν₀ → ν₄) bands and SB(ν₀ → ν₅) bands coincided. However, for sample CA20, the ratio of absorption coefficients turned out to be higher than AC(ν₀ → ν₄)/AC(ν₀ → ν₅) at approximately 1.2.

Since the Sc_C³⁺ and Sc_A³⁺ cations do not directly participate in light absorption processes, the detected changes in the spectra of the Yb_C³⁺ cations are probably associated with disordering of the crystal field caused by the partial replacement of yttrium and aluminum with scandium cations. As shown in [24], if the concentration of Sc_C³⁺ increases in the immediate environment of the Yb_C³⁺ cation, then the maximum SB(ν₀ → ν₅) shifts to the blue region of the spectrum, while the ratio AC(ν₀ → ν₄)/AC(ν₀ → ν₅) increases. In this study, it was found that if the concentration of Sc_A³⁺ increases in the immediate environment of the Yb_C³⁺ cation, then the maximum SB(ν₀ → ν₅) shifts to the red region of the spectrum. Thus, it can be argued that Yb_C³⁺ cations, in the immediate environment of which Sc_C³⁺ cations or Sc_A³⁺ cations dominate, are not equivalent.

Figure 4 b shows a section of the absorption spectrum of Er_C³⁺ cations for the ⁴I_15/2 → ⁴I_13/2 transition. Note that the ⁴I_15/2 term consists of 8 Stark energy levels, and the ⁴I_13/2 term consists of 7 levels [40]. It was found that in samples A50 and CA20, the number of distinguishable peaks is smaller compared to sample AP1. This is most clearly manifested in the wavelength range from 1440 to 1500 nm (Fig. 4 b). In our opinion, the decrease in discernible peaks clearly indicates a broadening of the absorption bands due to disordering of the crystal field. The main reason is an increase in the concentration of Sc_C³⁺ and Sc_A³⁺.

It is important to note that the Stark band has the highest absorption coefficient at a wavelength of 1532 nm. This band is associated with the transition of an electron from the lower Stark level of the ⁴I_15/2 term to the lower Stark level of the ⁴I_13/2 term. No significant changes in the position of the maximum of this band depending on the composition of the YSAG : Yb : Er samples were detected.

Relative to sample AP1, sample A50 has a group of Stark absorption bands in the long-wavelength part of the spectrum (1550–1680 nm) shifted to shorter wavelengths (inset in Fig. 4 b). In sample CA20, the shift of the bands is less pronounced. The shift of the absorption bands indicates that the Sc_C³⁺ and Sc_A³⁺ cations have an unequal effect on the properties of the Er_C³⁺ cation. If we take into account that the shift of the absorption bands in the CA20 sample relative to AP1 is minimal, we can conclude that the Sc_C³⁺ and Sc_A³⁺ cations can compensate for each other’s influence.

The broadening of the Stark absorption bands associated with Er_C³⁺ cations was also recorded in the wavelength ranges of 535–570 nm (Fig. 4 c) and 640–690 nm (Fig. 4 d). Analysis of the absorption spectra made it possible to establish that an increase in the concentration of Sc_A³⁺ leads to a blue shift of the majority of absorption bands associated with the ⁴I_15/2 → ⁴S_3/2 and ⁴I_15/2 → ⁴F_9/2 multiplets. At the same time, as shown in [24], an increase in the concentration of Sc_C³⁺ causes the opposite effect, which manifests itself in a shift of absorption bands towards longer wavelengths.

Thus, based on the analysis of the absorption spectra, we note that the energy structure of the ²F_7/2 и ²F_5/2 (Er_C³⁺) multiplets and the ⁴I_15/2, ⁴I_13/2, ⁴F_9/2 и ⁴S_13/2 (Er_C³⁺) multiplets changes nonequivalently with increasing concentration of Sc_C³⁺ or Sc_A³⁺. Consequently, the Sc_C³⁺ and Sc_A³⁺ cations have different effects on the crystal field of a solid solution with a garnet structure.

The energy structures of Yb³⁺ and Er³⁺ cations in YAG : Yb : Er solid solutions demonstrates two Stokes luminescence processes (Fig. 5) [41, 42]. The first process (H-process) describes the electron transitions, in which the Yb_C³⁺ cations absorb photons with a wavelength of ≈940 nm, and transit from ground state ²F_7/2 to the excited ²F_5/2 corresponding to Stark level ν₅. After a non-radiative transition to the Stark level ν₄, a series of transitions ν₄ → ν₃, ν₄ → ν₂ and ν₄ → ν₀ follow, accompanied by the emission of photons with wavelengths of ≈1050, ≈1030 and ≈970 nm, respectively.

Figure 5.

Scheme of the energy diagram of Yb³⁺ and Er³⁺ cations in YSAG : Yb : Er [12, 41, 42]

In the second Stokes process (D-process), the Yb_C³⁺ cations act as donors while the Er_C³⁺ cations act as acceptors. The transitions ²F_5/2 → ²F_7/2 and ⁴I_15/2 → ⁴I_11/2 have comparable values of energy, so some Yb_C³⁺ cations can transfer energy to Er_C³⁺ cations. In turn, the Er_C³⁺ (⁴I_11/2) cations can give up part of their energy and pass into the ⁴I_13/2 excited state and then, through the ⁴I_13/2 → ⁴I_15/2 transition emit photon in the wavelength range from 1450 up to 1670 nm.

According to the diagram in Fig. 5, the upconversion G-process of the Er_C³⁺ cation is described by the absorption of two photons by the excited state ⁴I_11/2, followed by its transition into the state ⁴F_7/2. In the next stage, non-radiative transitions occur from ⁴F_7/2 to ²H_11/2/⁴S_3/2 followed by the emission of photons in the wavelength range of 540–565 nm via ²H_11/2/⁴S_3/2 → ⁴I_15/2 transitions.

The upconversion R-process is realized by energy transfer through a complicated route: ²F_7/2 → ²F_5/2 (Yb) → ⁴I_11/2 (Er) → ⁴I_13/2 (Er) → ⁴F_9/2 (Er), which finalized by the luminescent transition to ⁴I_15/2 (Er) in the wavelength range of 645–685 nm. Since the introduction of the Sc_C³⁺ and Sc_A³⁺ cations affects the energy structure of the Yb_C³⁺ and Er_C³⁺ cations, then one should expect changes in the processes of conversion of radiation associated with the luminescence of erbium and ytterbium cations.

3.5.1. Stokes YSAG : Yb : Er luminescence

Figure 6 shows the Stokes luminescence spectra of Yb_C³⁺ cations as recorded upon excitation by laser radiation with a wavelength of approximately 940 nm. Stark luminescence bands associated with the H- and D-processes were detected in the wavelength ranges of 960–1080 nm and 1440–1690 nm, respectively. To compare the influence of Sc_C³⁺ and Sc_A³⁺ cations on the luminescent properties of YSAG : Yb : Er, the ratios of integral luminescence intensities (δ_H/D) for these two processes were examined:

${\delta }_{H/D}=\frac{I{s}_H}{I{s}_D}=\frac{{\int }_{\lambda 950}^{\lambda 1080}{I}_H(\lambda )\mathrm{d}\lambda }{{\int }_{\lambda 1440}^{\lambda 1685}{I}_D(\lambda )\mathrm{d}\lambda }$,

where λ₉₅₀, λ₁₀₈₀ and λ₁₄₄₀, λ₁₆₈₀ are the wavelengths of the beginning and end of the luminescence spectrum section; Is_H, Is_D are the integral luminescence intensities.

Figure 6.

Luminescence spectra of the YSAG : Yb : Er samples

Analysis of the luminescence spectra reveals that the ratio of the integrated luminescence intensities for sample A50 is approximately 0.332, which is clearly greater than for sample AP1 (δ_H/D ≈ 0.211) [24]. Taking into account the cationic compositions of these samples, it can be assumed that an increase in the concentration of Sc_A³⁺ makes the H-process of Stokes luminescence more preferable compared to the D-process. At the same time, according to the data from [24], in the case of an increase in the concentration of Sc_C³⁺, the role of the D-process increases. With a simultaneous increase in the concentration of cations Sc_C³⁺ and Sc_A³⁺, their influence is mutually compensated. Therefore, the CA20 sample has a value of δ_H/D (≈0.294) greater than that of the sample AP1 but less than sample A50.

Figure 7 a shows the luminescence spectra of the Yb_C³⁺ cation. The broadened luminescence bands SB(ν₄ → ν₂) and SB(ν₄ → ν₃) in sample CA20 distinguish it from sample AP1 (Fig. 7 a). At the same time, no significant shifts of the Stark luminescence bands were detected: maximum SB(ν₄ → ν₂) ≈ 1029 nm, maximum SB(ν₄ → ν₃) ≈ 1049 nm. In our opinion, this effect may be a consequence of disordering of the crystal field due to a simultaneous increase in the concentration of Sc_C³⁺ and Sc_A³⁺ from 4 at.% to 20 at.%.

Figure 7.

Normalized luminescence spectra of the Yb_C³⁺ (a) and Er_C³⁺ (b) cations in YSAG : Yb : Er samples

When the concentration of Sc_A³⁺ increases to 50 at.% (sample A50), there is a red shift of the SB(ν₄ → ν₂) by approximately 0.5 nm and SB(ν₄ → ν₃) by approximately 3–4 nm. It has been reported in [24] that an increase in the concentration of Sc_C³⁺ to 50 at.% leads to a red shift of the SB(ν₄ → ν₂) band by approximately 1 nm, and the SB(ν₄ → ν₃) shifts into the blue region by approximately 5–6 nm. These data show that Sc_C³⁺ and Sc_A³⁺ cations have an unequal effect on the energy structure of the Yb_C³⁺ cation.

Fig. 7 b shows the normalized luminescence spectra of the Er_C³⁺ cation. Analysis of the luminescence spectra of sample A50 reveals a blue shift of the bands and broadening relative to sample AP1. This effect is presumably due to the higher concentration of Sc_A³⁺ cations compared to Sc_C³⁺ cations. This assumption is confirmed by the fact that the positions of the maxima of the luminescence bands for samples CA20 and AP1 practically coincide. It is important to note that, according to [24], an increase in the concentration of Sc_C³⁺ leads to a red shift of these bands. Therefore, we can conclude that scandium cations in different positions of the crystal lattice have an unequal effect on the energy structure and luminescent properties of the Er_C³⁺ cation.

It is important to note that the luminescence kinetics for YSAG : Yb : Er samples with different concentrations of Sc_C³⁺ and Sc_A³⁺ have not been previously considered. Therefore, to better understand the effect of scandium on the properties of YSAG : Yb : Er, in addition to the CA20 and A50 samples synthesized in this work, samples AP1 and AP3 were studied. The composition of these samples is described in Table 1, and the preparation conditions and optical properties can be found in [24]. Analysis of the luminescence kinetics (Fig. 8 a) revealed that in the AP3 sample, the Yb_C³⁺ cations had the shortest effective lifetime (τ) of the excited states ²F_5/2, while in the A50 sample, it is the longest. Moreover, the τ in sample A50 was approximately 48% greater than in sample AP3, indicating a significant influence of the position of scandium cations in the garnet crystal lattice on the properties of Yb_C³⁺ cations.

Figure 8.

Luminescence decay kinetics of Yb_C³⁺ cations (a) and Er_C³⁺ cations (b)

Meanwhile, the luminescence decay kinetics of samples AP1 and CA20 practically coincided with each other, probably due to the mutual compensation of the effects of the Sc_C³⁺ and Sc_A³⁺ cations. The effects of scandium cations on the decay kinetics of Stokes luminescence of Er_C³⁺ cations were less dramatic. The differences in the "τ" values indicated that the introduction of scandium cations into the dodecahedral position leads to a slight decrease in the lifetime of the excited states ⁴I_13/2. At the same time, it was found that the lifetime of the excited state ⁴I_13/2 of the Er_C³⁺ cation in YSAG : Yb : Er (8.7–9.2 ms) is slightly higher than in YAG : Yb : Er (7.0–8.5 ms) [10, 41].

Taking into account that the lifetime of the excited state is proportional to the probability of emission of light quanta, it was assumed that the probability of transitions ²F_5/2 → ²F_7/2 decreases with an increasing concentration of cations Sc_C³⁺. In this case, the probability of non-radiative energy transfer from Yb_C³⁺ cations to Er_C³⁺ cations increases. Moreover, since the value of δ_H/D decreases, it can be argued that the transfer efficiency of energy from Yb_C³⁺ cations to Er_C³⁺ cations has also increased.

The lifetime values of the excited state ²F_5/2 of the Yb_C³⁺ cations in the AP1 sample were in good agreement with the "τ" values for YAG : Yb(15 at.%) : Er(1 at.%) solid solutions [41], which indicates the reliability of the obtained values. Thus, an increase in the concentration of Sc_A³⁺ was found to increase the probability of transitions ²F_5/2 → ²F_7/2 while decreasing the probability of energy transfer from Yb_C³⁺ cations to Er_C³⁺ cations. Since the probability of the transition ⁴I_13/2 → ⁴I_15/2 weakly depends on the concentration of scandium, the decrease in the values of δ_H/D detected in the series of samples A50 → AP1(CA20) → AP3 indicates an increase in the efficiency of energy transfer with a decrease in the concentration of Sc_A³⁺.

3.5.2. Anti-Stokes YSAG : Yb : Er luminescence

Figure 9 shows the anti-Stokes luminescence spectra recorded upon laser radiation excitation at a wavelength of 940 nm. It is evident that regardless of the YSAG : Yb : Er composition, the intensity of the luminescence bands in the green region of the spectrum is significantly lower than in the red region. The ratio of integral intensities δ_G/R was calculated using the expression:

${\delta }_{G/R}=\frac{I{s}_G}{I{s}_R}=\frac{{\int }_{\lambda 520}^{\lambda 570}{I}_G(\lambda )\mathrm{d}\lambda }{{\int }_{\lambda 640}^{\lambda 685}{I}_R(\lambda )\mathrm{d}\lambda }$,

where λ₅₂₀, λ₅₇₀ and λ₆₄₀, λ₆₈₀ is wavelengths of the beginning and end of the luminescence spectrum section; Is_G, Is_R is integral luminescence intensities

Figure 9.

Anti-Stokes luminescence spectra of the YSAG : Yb : Er samples

It was found that for sample СA20 – δ_G/R ≈ 0.093, and for A50 – δ_G/R ≈ 0,12. As shown in [24], for sample AP1– δ_G/R ≈ 0.109, and for AP3 – δ_G/R ≈ 0.088. In this case, an increase in the values of δ_G/R indicates that with increasing concentration Sc_A³⁺ the probabilities of the G-process decrease noticeably than the probabilities of the R-process.

Figure 10 depicts the luminescence decay kinetics observed at approximately 561 nm and 679 nm wavelengths. To investigate the luminescence kinetics, we conducted studies on the (⁴S_3/2 → ⁴I_15/2) and (⁴F_9/2 → ⁴I_15/2) bands, utilizing light excitation at approximately 357 nm and 478 nm wavelengths, respectively.

Figure 10.

Luminescence decay kinetics in the green (a) and red (b) spectral range

Analysis of the obtained dependencies indicates that the lifetime of the excited state ⁴S_3/2 of the Er_C³⁺ cation is about 12 ± 4 μs, which is in good agreement with the data for YAG : Er [44]. For the term ⁴F_9/2 τ ≈ 20 ± 5 µs. As can be seen in Fig. 10, within the error, the lifetimes of the excited states of the ⁴S_3/2 and ⁴F_9/2 terms of the Er_C³⁺ cations are practically independent of the concentration of the Sc_C³⁺ and Sc_A³⁺ cations. This circumstance indicates that the excited states of Er_C³⁺ cations are less sensitive to the presence of scandium cations than Yb_C³⁺ cations. Thus, analysis of luminescence in the visible region of the spectrum confirmed the assumption that scandium cations in dodecahedral and octahedral positions have non-equivalent effects on the properties of rare earth cations. Consequently, by changing both the total concentration of scandium in the crystal lattice and the Sc_C³⁺/Sc_A³⁺ ratio, it is possible to purposefully change the properties of YSAG : Yb : Er solid solutions.