Effect of Gd³⁺, La³⁺, Lu³⁺ Co-Doping on the Morphology and Luminescent Properties of NaYF₄:Sm³⁺ Phosphors

Abstract

The series of luminescent NaYF₄:Sm³⁺ nano- and microcrystalline materials co-doped by La³⁺, Gd³⁺, and Lu³⁺ ions were synthesized by hydrothermal method using rare earth chlorides as the precursors and citric acid as a stabilizing agent. The phase composition of synthesized compounds was studied by PXRD. All synthesized materials except ones with high La³⁺ content (where LaF₃ is formed) have a β-NaYF₄ crystalline phase. SEM images demonstrate that all particles have shape of hexagonal prisms. The type and content of doping REE significantly effect on the particle size. Upon 400 nm excitation, phosphors exhibit distinct emission peaks in visible part of the spectrum attributed to ⁴G_5/2→⁶H_J transitions (J = 5/2–11/2) of Sm³⁺ ion. Increasing the samarium (III) content results in concentration quenching by dipole–dipole interactions, the optimum Sm³⁺concentration is found to be of 2%. Co-doping by non-luminescent La³⁺, Gd³⁺ and Lu³⁺ ions leads to an increase in emission intensity. This effect was explained from the Sm³⁺ local symmetry point of view.

1. Introduction

Lanthanide-doped inorganic materials have been attracting much attention from scientists for several decades. These materials have promising applications in medicine and technology as materials for optical devices, sensing, tumor therapy, bioimaging, drug delivery, anti-counterfeiting, optical thermometry, etc. [1,2,3,4,5,6,7,8,9].

The optical properties of these materials depend on the particles’ size and morphology, crystal symmetry, type, and concentration of rare earth ions in the host matrix [10,11,12,13,14,15,16]. Sodium yttrium fluoride is one of the best host matrices for luminescent rare earth-doped inorganic materials because this matrix has only low-frequency vibrational modes, and therefore does not quench the luminescence. In addition, NaYF₄ possesses chemical inertness, low toxicity, and the possibility to combine magnetic, optical, and radioactive properties of lanthanide ions that opens the way to prepare new theranostic agents for non-invasive therapy [17,18,19,20,21,22]. As a co-dopant, lanthanide ions play several key roles in photoluminescent materials: they may absorb light as sensitizers or emit photons as luminescence activators as well as transfer energy from the sensitizer to activator [13,14,23,24,25]. At the same time, the addition of non-luminescent dopants (e.g., alkali, alkali earth, some p-, d- and f-metal ions) in host matrix doped with luminescent ions is known to enhance the luminescence intensity [11,26,27,28]. This effect is assumed to be caused by several factors: structural changes in the crystal lattice upon doping (e.g., formation of ionic vacancies) and modification of the crystal field surrounding Ln³⁺ activators [28,29,30]. Yet, generally, it is still early to believe that the mechanism of the co-doping effect on luminescence is fully explained because there is no model to predict the impact of any dopant ions on the optical properties of such doped materials. We presumed that this is caused by the deficiency of studies. For example, to the best of our knowledge, the non-luminescent dopants are mainly chosen from non-lanthanide elements. This approach neglects the fundamentally interesting details of the mutual effect of ions on similar electronic structures. Previously we have reported the particle size and shape dependence on the nature of the doping lanthanide (III) ions NaYF₄:Ln³⁺ series and described the correlation between the obtained nanoparticle morphologies and the type and content of doping ions [10]. We found that the average diameter of particles reaches the least value for Sm³⁺, Eu³⁺, and Gd³⁺ doped materials. We have studied NaYF₄:Eu³⁺ particles co-doped with Gd³⁺ ions [11] and revealed that Gd³⁺ doping results in particle size reduction as well as the increase in emission intensity and ⁵D₀ lifetime of europium (III). We have obtained a similar effect of simultaneous size reduction and luminescence intensity enhancement for gadolinium ion-doped materials for NaYF₄:Yb³⁺, Tm³⁺/Er³⁺ up-conversion microcrystalline materials [16]. Further investigations of up-conversion materials based on NaYF₄ doped with erbium, ytterbium and co-doped with lutetium ions showed that the addition of optical inactive Lu³⁺ results in both increasing particles size and luminescence intensity [31]. In order to find out whether the luminescence intensity enhancement is the common trend upon doping with gadolinium or other non-luminescent lanthanide ions, we intended to study samarium-containing down-conversion phosphors in the current work.

Samarium compounds are of interest in medicine and the production of functional nanoparticles. For example, the decay energy of the samarium ¹⁵³Sm nuclide allows using this isotope for cancer therapy and SPECT imaging [32,33]. Sm³⁺ ions are also known to be used as a part of optically active materials because of their orange luminescence, originating from the ⁴G_5/2 → ⁶H_J/2 (J = 5, 7, and 9) transitions [14,34,35,36,37]. Nevertheless, the works devoted to the co-dopant effect on samarium-doped compounds as a way to control the luminescence properties of these materials are limited, and this effect should be studied in detail.

In this present study, we reported the effect of rare earth doping concentration on the morphology, structure, and luminescence properties of the series of NaYF₄ compounds doped with Sm³⁺ and co-doped with non-luminescent La³⁺, Gd³⁺, and Lu³⁺ ions and proposed the theoretical explanations of such effects.

2. Materials and Methods

Anhydrous chlorides of the rare earth elements (YCl₃, SmCl₃, LaCl₃, GdCl₃, LuCl₃, 99.999%) were purchased from Chemcraft (Kaliningrad, Russia), KBr, NaOH, NH₄F, citric acid, and ethanol were purchased from Sigma-Aldrich Pty Ltd. (Darmstadt, Germany), and used without additional purification.

Microcrystalline β-NaYF₄ samples co-doped with Sm³⁺, La³⁺, Gd³⁺, and Lu³⁺ were synthesized using the hydrothermal method using citric acid as a stabilizing agent, described previously [11,16]. Rare earth chlorides taken in stoichiometric amounts (total amount of rare earth chlorides was 0.75 mmol) with 3 mmol of citric acid were dissolved in distilled water to obtain 5 mL solution in total. Then, 2.5 mL of an aqueous solution containing 9 mmol of NaOH was added to the reaction mixture. After vigorous stirring for 30 min, 8 mL of aqueous solution containing 11 mmol of NaOH and 11 mmol of NH₄F was added into the above solution. The solution was maintained after vigorous stirring for 30 min at room temperature before being transferred to a Teflon-lined autoclave with an internal volume of 20 mL and heated for 17h at the temperature of 180 °C. After that, the precipitate was separated from the reaction mixture by centrifugation, washed with ethanol and deionized water, and dried at 60 °C for 24 h. The desired microcrystalline materials were obtained in the form of white powders.

In this work, we synthesized and studied four series of luminescent powders: NaY_1-xSm_xF₄ (x = 0–0.4) and NaY_0.98−ySm_0.2Ln_yF₄ (Ln = La, Gd, Lu; y = 0–0.6). Among NaY_1−xSm_xF₄ series, materials containing 2% (x = 0.02) of Sm³⁺ demonstrated the highest luminescence intensity (discussed below in the Results and Discussion section). Therefore, to follow the effect of Ln³⁺ (Ln = La, Gd, Lu) co-doping on the luminescence properties, we kept the concentration of Sm³⁺ equal to 2% in the NaY_0.98−ySm_0.2Ln_yF₄ series. The relative content of the rare earth elements in the synthesized compounds was confirmed by energy-dispersive X-ray spectroscopy. The particles’ morphology was characterized using scanning electron microscopy (SEM) on a Zeiss Merlin electron microscope (Zeiss, Oberkochen, Germany) using an energy-dispersive X-ray spectroscopy (EDX) module (Oxford Instruments INCAx-act, Oxford, UK). powder X-ray diffraction (PXRD) measurements were performed on a D2 Phaser (Bruker, Billerica, MA, USA) X-ray diffractometer using Cu Ka radiation (λ = 1.54056 Å). To carry out quantitative photoluminescence studies, the synthesized samples (20 mg) and potassium bromide (300 mg) were pressed into pellets (diameter 13 mm). The luminescence spectra were recorded on Fluorolog-3 fluorescence spectrometer (Horiba Jobin Yvon, Kyoto, Japan). Lifetime measurements were performed using the same spectrometer using a pulsed Xe lamp (pulse duration 3 µs).

3. Results and Discussion

3.1. Crystal Structure

The powder X-ray diffraction (PXRD) patterns are shown in Figure 1a–d). Analysis of PXRD patterns demonstrates that all synthesized materials of three series (NaY_1−xSm_xF₄, NaY_0.98−xSm_0.02Gd_xF₄ and NaY_0.98−xSm_0.02Lu_xF₄) have the same crystalline phase, which corresponds to the hexagonal β-NaYF₄ (JCPDS No. 16-0334). Additional diffraction peaks corresponding to the impurities are not observed. In opposition to the abovementioned series, we have found that substitution of yttrium by the lanthanum ions in NaY_0.98−xSm_0.2La_xF₄ series results in the formation of either β-NaYF₄ or LaF₃ (JCPDS No. 32-0483) crystalline phases depending on the lanthanum content. Thus, at the lanthanum content less 20 at.% and less, only β-NaYF₄ crystalline phase is formed similarly to other series. At the lanthanum content of 40 at.%, β-NaYF₄ or LaF₃ phases coexist. At the content of lanthanum of the 60 at.%, compounds precipitate exclusively in a form of LaF₃ phase.

Unit cell parameters were refined using UnitCell software [38]. This program can retrieve unit cell parameters from diffraction data using a method of least squares from the positions of the indexed diffraction maxima of the PXRD patterns (Pawley method [39]). The uncertainties of unit cell parameters are shown in parenthesis in Tables S1–S4, Supplementary Materials. The dependence of refined unit cell volumes on the sample composition is shown in Figure 2. Unit cell volume linearly depends on dopant concentration, therefore, Vegard’s law [40] obeys the studied systems; hence, Ln³⁺ (Ln = Sm, Gd, Lu, La) ions isomorphically substitutes Y³⁺ ions in the β-NaYF₄ structure. For compounds NaY_1−xSm_xF₄, the increase in Sm³⁺ content leads to unit cell volumes increase due to a higher ionic radius of Sm³⁺ ions (1.132 Å, the coordination number is nine) than the ionic radius of Y³⁺ ions (1.075 Å) [41]. Similarly, the doping of NaY_0.98Sm_0.2F₄ by lanthanide (III) ions with higher ionic radius than Y³⁺ ions (Gd³⁺: 1.107 Å; La³⁺: 1.216 Å) results in increasing the unit cell volumes. Moreover, for the NaY_0.98−xSm_0.02La_xF₄ series, unit cell volume increases significantly faster than for the NaY_0.98−xSm_0.02Gd_xF₄ one because La³⁺ ions have a larger ionic radius than Gd³⁺. Meanwhile, the unit cell volumes for NaY_0.98−xSm_0.02Lu_xF₄ series decrease upon lutetium concentration rise, which can be similarly explained by the lower ionic radius of Lu³⁺ ions (1.032 Å) than the ionic radius of Y³⁺ ions.

3.2. Morphology

A scanning electron microscope (SEM) was used to observe the shape and size of the particles in synthesized materials. SEM images of the synthesized materials are shown in Figure 3, Figure 4, Figure 5 and Figure 6. The particles have the shape of hexagonal prisms. The particle diameter was obtained from SEM images, the particle size distribution is shown in the inserts of Figure 3, Figure 4, Figure 5 and Figure 6. The average diameter of the particle was calculated from this distribution and is given in the legends in Figure 3, Figure 4, Figure 5 and Figure 6. The particle size strongly depends on the sample composition ranging from 46 to 1916 nm. In the NaY_1−xSm_xF₄ series, the size reduction is observed upon increasing the samarium content, Figure 3 and Figure 7. Thus, the NaYF₄ particles have an average size of 682 ± 41 nm, whereas NaY_0.5Sm_0.5F₄ particles are significantly smaller, 78 ± 9 nm. In the NaY_0.98−xSm_0.02Ln_xF₄ (Ln = La, Gd, Lu) series (Figure 3b, Figure 4, Figure 5 and Figure 6), the substitution of the yttrium ion by the lanthanum and lutetium ions results in particle size increase, whereas particle size reduction is observed upon gadolinium doping, Figure 7. This observation can be explained by the mechanism of crystal growth [10]. We assume that the particle size is determined by nucleation and crystal growth rates. If the nucleation rate is larger than the crystal growth rate, small single crystals are formed. In the opposite case, when nucleation is slow, but crystal growth is fast, large single crystals are formed. The crystal growth rate is significantly affected by the Cit³⁻ and Na⁺ adsorption on (1010) and (0001) facets, respectively [17,42]: adsorption of the ions on the grain facets slows down crystal growth [42,43], therefore, higher adsorption of ions on the crystal nuclei results in lower particle size. The ionic radius decreases in the row La³⁺-Sm³⁺-Gd³⁺-Y³⁺-Lu³⁺, therefore, surface charge density increases in this order. Nucleation is faster for ions with larger ionic radius, which means that this process slows down in the row La³⁺-Sm³⁺-Gd³⁺-Y³⁺-Lu³⁺. Adsorption of Cit³⁻ and Na⁺ ions is more pronounced for the particles with higher surface charge density increasing from La³⁺ to Lu³⁺. Therefore, the observed particle size reduction upon substitution of the yttrium ion to La³⁺, Sm³⁺, and Gd³⁺ ions is dominated by the decrease in crystal growth rate due to the adsorption of Na⁺ and Cit³⁻ ions inhibiting crystal growth. We assume that from Gd to Lu, the crystal growth rate changes insignificantly because the large amount of Na⁺ and Cit³⁻ ions covers the crystal grain surface, and additional Na⁺ and Cit³⁻ adsorption is not favorable anymore. At the same time, the nucleation rate monotonically decreases from La³⁺ to Lu³⁺, which explains the particle size growth upon substitution of the yttrium by lutetium ions. We found that co-doping of the large amounts of La³⁺ ions results in the formation of the two types of hexagonal particles of significantly different sizes (Figure 4e,f). Thus, the NaY_0.58Sm_0.02La_0.4F₄ compound consists of large (1517 ± 64 nm) and small (254 ± 16 nm) particles. The average size of the NaY_0.38Sm_0.02La_0.6F₄ sample also contains two sorts of particles with an average size of 1916 ± 132 and 102 ± 9 nm. The fraction of the smaller particles significantly increases from 40 to 60 at.% La³⁺, therefore, according to PXRD data, we assume that larger particles correspond to β-NaLnF₄ and smaller particles are attributed to the LnF₃ crystalline phase.

3.3. Luminescence Properties

Excitation spectra of NaY_1−xSm_xF₄ samples monitored at the ⁵G_5/2 → ⁶H_7/2 (595 nm) transition were in the spectral range of 350–500 nm, Figure 8a. One can see that spectra consist of sharp peaks attributed to the f-f electron transitions of the Sm³⁺ ion: ⁶H_5/2 → ⁴F_9/2 (361 nm), ⁶H_5/2 → ⁴D_5/2 (373 nm), ⁶H_5/2 → ⁶P_7/2 (389 nm), ⁶H_5/2 → ⁴K_11/2 (400 nm), ⁶H_5/2 → ⁶P_5/2 + ⁴M_19/2 (415 nm), ⁶H_5/2 → ⁴G_9/2 + ⁴I_15/2 (440 nm), ⁶H_5/2 → ⁴F_5/2 + ⁴I_13/2 (462 nm) and ⁶H_5/2 → ⁴I_11/2 + ⁴M_15/2 (476 nm). The ⁶H_5/2 → ⁴K_11/2 transition centered at 400 nm is dominated in the obtained spectra. Figure 8b presents emission spectra of NaY_1−xSm_xF₄ concentration series upon 400 nm excitation into the ⁶H_5/2 → ⁴K_11/2 band. Emission spectra included lines corresponding to transitions from excited ⁴G_5/2 to lower ⁶H_J levels: ⁴G_5/2 → ⁶H_5/2 (561 nm), ⁵G_5/2 → ⁶H_7/2 (595 nm), ⁴G_5/2 → ⁶H_9/2 (641, 646 nm) and ⁴G_5/2 → ⁶H_11/2 (703 nm). The most prominent transition in the spectra was the ⁵G_5/2 → ⁶H_7/2 transition. Analysis of the emission spectra has demonstrated that the spectral shape excitation and emission spectra do not depend on the Sm³⁺ content, whereas the Sm³⁺ doping concentration significantly affected the emission intensity, Figure 8a,b. The concentration dependence of integral intensities of the ⁵G_5/2 → ⁶H_7/2 emission band is presented in Figure 8c. The emission intensity non-monotonically depends on the Sm³⁺ concentration reaching the maximum at the Sm³⁺ content of 2 at.% (x = 0.02). Such type of concentration dependence can be explained by the two competitive effects in phosphors upon Sm³⁺ concentration rise [44,45]. Thus, the rise of the number of luminescent sites results in radiative emission probability increase and, as a result, the emission intensity increase. At the same time, upon Sm³⁺ concentration rise, the distance between Sm³⁺ ions decreases resulting in the nonradiative processes probability increase, which leads to the emission quenching. If doping ions occupy a single crystallographic position in the host, the energy transfer mechanism is determined by the critical energy transfer distance (R_c). This distance can be calculated by the following formula [46]:

$${\mathrm{R}}_{\mathrm{c}}=2{\left(\frac{3\mathrm{V}}{4{\pi \chi }_{\mathrm{c}}\mathrm{N}}\right)}^{\frac{1}{3}},$$

(1)

where ${\chi }_{\mathrm{c}}$ is a critical concentration of luminescent ion (0.02), $\mathrm{V}$ is unit cell volume for NaY_0.98Sm_0.02F₄ (109.64 ų), $\mathrm{N}$—number of cation sites in crystal structure (1.5 for β-NaYF₄ [47]). Using these parameters, the critical energy transfer distance ${\mathrm{R}}_{\mathrm{c}}$ in NaY_1−xSm_xF₄ is calculated to be of 19.11 Å. According to Blasse theory [46], when ${\mathrm{R}}_{\mathrm{c}}$ > 5 Å, the main contribution to non-radiative energy transfer occurs by the multipole–multipole interactions. At high samarium (III) concentration, the probability of radiative emission is constant; therefore, the energy transfer between Sm³⁺ ions in the NaYF₄ host is dominated by the multipole–multipole interactions. For the determination of interaction type, Van Uitert [48] proposed an equation, which later was modified by Ozawa and Jaffe [49]:

$$\frac{\mathrm{I}}{\chi }=\frac{\mathrm{k}}{1+{\mathsf{\beta }\chi }^{\frac{\mathsf{\theta }}{3}}},$$

(2)

where $\mathrm{I}$ is integral intensity, $\chi$ is the concentration of the luminescent ion. Assuming that ${\mathsf{\beta }\chi }^{\frac{\mathsf{\theta }}{3}}$ ≫ 1, one can build the linearized coordinates $\lg \frac{\mathrm{I}}{\chi }-\lg \chi$ (Figure 8d). Linear fitting of dependence in these coordinates gives the value $\frac{\mathsf{\theta }}{3}$ = 2.05. It is known that dipole–dipole, dipole–quadrupole, and quadrupole–quadrupole interactions correspond to $\mathsf{\theta }$ values of 6, 8, and 10, respectively [50]. For NaY_1−xSm_xF₄, $\mathsf{\theta }$ = 6, therefore nonradiative energy transfer between samarium (III) ions in the NaYF₄ host is caused by dipole–dipole interactions.

Luminescence decay curves of NaY_1−xSm_xF₄ phosphors monitored at 595 nm (⁵G_5/2 → ⁶H_7/2 transition) upon 400 nm excitation are presented in Figure 9a. All experimental decay curves displayed non-single exponential behavior and, therefore, bi-exponential models were applied for fitting (Equation (3)). The best-fit parameters are given in Table S5 (Supplementary Materials). Bi-exponential decay of small-sized materials is usually explained by the presence of two types of luminescent ions situated in the volume and on the surface of the particles, which have different decay times [51,52]. Sm³⁺ ions situated on the surface display lower lifetimes due to a higher probability of quenching.

$$\mathrm{I}\left(\mathrm{t}\right)={\mathrm{A}}_1{\mathrm{e}}^{\frac{\mathrm{t}}{{\mathsf{\tau }}_1}}+{\mathrm{A}}_2{\mathrm{e}}^{\frac{\mathrm{t}}{{\mathsf{\tau }}_2}},$$

(3)

where A₁ and A₂ are pre-exponential constants, and τ₁ and τ₂ are fitting lifetimes.

Average luminescence lifetime (τ_av), which corresponds to the ⁵G_5/2 level lifetime, was calculated according to the following equation to simplify comparison [53,54]:

$${\mathsf{\tau }}_{\mathrm{av}}=\frac{{\mathrm{A}}_1{\mathsf{\tau }}_1^2+{\mathrm{A}}_2{\mathsf{\tau }}_2^2}{{\mathrm{A}}_1{\mathsf{\tau }}_1+{\mathrm{A}}_2{\mathsf{\tau }}_2},$$

(4)

The Sm³⁺ concentration dependence of the obtained lifetimes is shown in Figure 9b. One can see a monotonic decrease in the lifetimes from 4.3 ms to 0.4 ms along with the increase in samarium concentration. Such behavior is most likely linked to the growth of the nonradiative decay rate due to the increase in spatial energy migration followed by further quenching of impurities.

Further studies were devoted to the co-doping effect of non-luminescent Gd³⁺, Lu³⁺, and La³⁺ ions on the luminescence properties of NaYF₄: Sm³⁺ powders. As was demonstrated earlier, Sm³⁺ optimum concentration is 2%, so this samarium concentration was used for samples with Gd³⁺, Lu³⁺, and La³⁺ co-doping. Emission spectra of NaY_0.98−xSm_0.02Ln_xF₄ (Ln = Gd, Lu, La) compounds upon 400 nm excitation, Figure 10a–c. One can notice that Gd³⁺, Lu³⁺, and La³⁺ co-doping affect only the emission intensity and alternate neither the positions of the emission bands corresponding to ⁴G_5/2-⁶H_J transitions nor their relative intensities. In order to estimate this effect, the integral emission intensities corresponding to the most intense ⁵G_5/2 → ⁶H_7/2 transition of Sm³⁺ ions (595 nm) were calculated and plotted in Figure 10d–f relative to the NaY_0.98Sm_0.02F₄ sample. We found that co-doping by the abovementioned rare earth ions results in an increase in the luminescence intensities. Thus, the substitution of Y³⁺ ion by Gd³⁺ results in the most emission enchantment up to 2.4 times, Figure 10d. The maximum emissions intensities are observed for the Gd³⁺ content of 0.5 and 10 at.% corresponding to the increase in the luminescence intensity at 2.4, and 2.2 times, respectively. The co-doping of NaY_0.98Sm_0.02F₄ compound by Lu³⁺ ion results in emission enchantment up to 2.1 times, the maximum effect is observed for the lutetium content of 1 at.%, Figure 10e. The least prominent effect is observed for co-doping of NaY_0.98Sm_0.02F₄ materials by La³⁺ ion, where the emission enchantment is barely noticeable, Figure 10f. Therefore, it is difficult to mention the precise position of the La³⁺ concentration corresponding to the largest effect. To reveal the mechanism of the luminescence enhancement effect by Gd³⁺, Lu³⁺, and La³⁺ co-doping, the luminescence kinetics was studied for the samples with various concentrations of co-doping ions. Luminescence decay curves of NaY_0.98−xSm_0.02Ln_xF₄ (Ln = Gd, Lu, La) phosphors monitored at 595 nm (⁵G_5/2 → ⁶H_7/2 transition) upon 400 nm excitation are presented in Figure 11. All experimental decay curves displayed non-single exponential behavior and two exponential models were applied for fitting (Equation (3)). The best-fit parameters are given in Tables S6–S8 (Supplementary Materials). The average luminescence lifetimes, which correspond to the ⁵G_5/2 level lifetimes, were calculated using Equation (4) and given in

Table 1. Lifetimes of ⁴G_5/2 excitation state of Sm³⁺ ion in NaY_0.98-xSm_0.02Ln_xF₄ (Ln = Gd, Lu, La).

Ln3+Content, at.%	Ln3+ = Gd3+	Ln3+ = Lu3+	Ln3+ = La3+
	τav, ms	τav, ms	τav, ms
0	3.54 ± 0.05	3.54 ± 0.05	3.54 ± 0.05
0.25	3.51 ± 0.05	3.58 ± 0.05	3.46 ± 0.05
0.5	3.50 ± 0.05	3.46 ± 0.05	3.55 ± 0.05
1	3.47 ± 0.05	3.42 ± 0.05	3.46 ± 0.05
7	3.46 ± 0.05	3.46 ± 0.05	3.47 ± 0.05
60	3.58 ± 0.05	3.45 ± 0.05	3.53 ± 0.05

. We revealed that co-doping of NaY_0.98Sm_0.02F₄ by Gd³⁺, Lu³⁺, and La³⁺ does not result in a change in the ⁵G_5/2 excited state lifetime. Therefore, the substitution of yttrium ions by gadolinium, lutetium, and lanthanum ions does not change the probability of the ⁴G_5/2-⁶H_J radiative transition.

Emission enhancement resulting from Gd³⁺, Lu³⁺, and La³⁺ co-doping of NaY_0.98Sm_0.02F₄ materials, in principle, can be caused by the absorption and/or emission probability increase. However, in the second case, excited state lifetimes must change, which is not observed in our experiments. Therefore, one can conclude, that doping by Ln³⁺ ions results in changing only extinction coefficients due to the changing probability of symmetry forbidden ⁶H_5/2 → ⁴K_11/2 transition. Luminescence intensity enhancement resulted from co-doping of Eu³⁺-containing materials by non-luminescent ions such as Bi³⁺, Gd³⁺, alkali, and alkali earth metal ions was reported previously [11,26,27,55,56,57,58,59,60,61,62,63,64]. The observed effect was explained by structure distortion due to the difference between radii of substituted and doping ions resulting in the increase in the emission and absorption probabilities. In our case, Gd³⁺, Lu³⁺, and La³⁺ co-doping of NaY_0.98Sm_0.02F₄ materials at low concentrations of the dopant results in symmetry lowering of Sm³⁺ local environment that leads to an increase in the absorption probability and, obviously, extinction coefficients [57,65]. Indeed, the maximum emission effect is observed when about 1% yttrium ions are substituted with gadolinium or lutetium ions. Meanwhile, compounds containing a significant amount of gadolinium ions also demonstrate larger emission intensity than NaY_0.98Sm_0.02F₄. A similar effect was observed by Martins and co-workers [66] where co-doping of Y₂O₃: Eu³⁺ by Gd³⁺ ions resulted in an increase in the emission intensity. They explain this phenomenon of partial absorption by the host Gd₂O₃ matrix followed by the energy transfer to Gd³⁺ ion, and then from Gd³⁺ to Eu³⁺ ion. It is known, that the β-NaYF₄ host absorbed light at 200–450 nm [7]. Most probably, the addition of Gd³⁺ ion results in the more prominent absorption of β-NaYF₄: Gd³⁺ matrix at the same range of UV spectrum. We propose that 400 nm excitation of NaY_0.98−xSm_0.02Gd_xF₄ promotes β-NaYF₄: Gd³⁺ host matrix into the excited state (in parallel with ⁶H_5/2 → ⁴K_11/2 transition of Sm³⁺ ion) followed by energy transfer from the host matrix to Sm³⁺ ions, which results in increases in luminescence intensities relative to NaY_0.98Sm_0.02F₄.

4. Conclusions

In the present work, four series of NaYF₄ particles doped with Sm³⁺, Gd³⁺, Lu³⁺, and La³⁺ ions, NaY_1−xSm_xF₄ (x = 0–0.5) and NaY_0.98−ySm_0.02Ln_yF₄ (Ln = Gd, Lu, La; y = 0–0.6), were synthesized by a hydrothermal method at a temperature of 180 °C using citric acid as a stabilizing agent. Analysis of PXRD patterns demonstrated that NaY_1−xSm_xF₄ and NaY_0.98−xSm_0.02Ln_xF₄ (Ln = Lu, Gd) have similar crystal structures corresponding to the hexagonal β-NaYF₄. For the NaY_0.98−xSm_0.2La_xF₄ series, the β-NaYF₄ crystalline phase is dominated at La³⁺ content up to 20%. At higher La³⁺ concentrations, the solid solutions are formed as a LaF₃ crystalline phase. Among the β-NaYF₄ phase, unit cell volumes linearly depend on dopant concentration, which demonstrates that Sm³⁺, Gd³⁺, Lu³⁺, and La³⁺ ions isomorphically substitute Y³⁺ ions in the β-NaYF₄ structure. Sm³⁺, Gd³⁺, and La³⁺ doping results in unit cell volumes increase because the Y³⁺ ion has a smaller ionic radius (1.075 Å) than Sm³⁺ (1.132 Å) and Gd³⁺ (1.107 Å) ions. The substitution of Y³⁺ ions by smaller Lu³⁺ ions (1.032 Å) leads to unit cell volume reduction. According to SEM data, particles of all synthesized compounds have the shape of hexagonal prisms and sizes ranging from 46 to 1916 nm depending on the sample composition. In the NaY_1−xSm_xF₄ series, the substitution of Y³⁺ by Sm³⁺ ions leads to the particle size reduction from 682 nm (NaYF₄) down to 78 nm (NaY_0.5Sm_0.5F₄). Co-doping of NaY_0.98Sm_0.02F₄ by La³⁺ and Lu³⁺ ions results in particle size increases due to faster growth (for La³⁺) and slower nucleation (for Lu³⁺) [10]. In contrast to La³⁺ and Lu³⁺, co-doping of these materials by Gd³⁺ ions leads to particle size reduction because the lowest growth/nucleation rates are characteristic of Gd³⁺. All synthesized compounds demonstrate photoluminescence under 400 nm excitation (⁶H_5/2 → ⁴K_11/2 transition in Sm³⁺). Experimental Sm³⁺ optimal doping concentration in β-NaYF₄ host is 2%. Further increasing of Sm³⁺ concentration leads to strong quenching due to dipole–dipole interactions between Sm³⁺ ions. We demonstrated that co-doping by different non-luminescent Ln³⁺ ions (where Ln is not only Gd, but also La and Lu) in low dopant concentration (in the range from 0 to 10 at.%) results in increasing luminescent intensities. Co-doping of NaY_0.98Sm_0.02F₄ by Gd³⁺, Lu³⁺, and La³⁺ ions does not lead to the change in ⁴G_5/2 excited state lifetimes, therefore co-doping by non-luminescent ions leads to increase in absorption probability due to the Sm³⁺ local symmetry distortion. Therefore, we discovered that enhancement of luminescence intensity as a result of co-doping by non-luminescent Gd³⁺, Lu³⁺, and La³⁺ ions is a general phenomenon and can be applied to improve the optical properties of a wide range of inorganic REE-containing phosphors.