Optical Properties Investigation of Upconverting K₂Gd(PO₄)(WO₄):20%Yb³⁺,Tm³⁺ Phosphors

Abstract

Nowadays, scientists are interested in inorganic luminescence materials that can be excited with UV or NIR radiation and emit in the visible range. Such inorganic materials can be successfully used as luminescent or anti-counterfeiting pigments. In this work, we report the synthesis and optical properties investigation of solely Tm³⁺ doped and Yb³⁺/Tm³⁺ co-doped K₂Gd(PO₄)(WO₄) phosphors. The single-phase samples were prepared using a solid-state reaction method. The Tm³⁺ concentration was changed from 0.5% to 5%. Downshifting and upconversion emission studies were performed under 360 nm and 980 nm excitation, respectively. Yb³⁺ ions were used as sensitizers in the K₂Gd(PO₄)(WO₄) phosphors to transfer the captured energy to Tm³⁺ ions. It turned out that under UV excitation, phosphors emitted in the blue spectral area regardless of the presence or absence of Yb³⁺. However, a very strong deep-red (~800 nm) emission was observed when Yb³⁺ and Tm³⁺-containing samples were excited with a 980 nm wavelength laser. It is interesting that the highest upconversion emission in the UV/Visible range was achieved for 20% Yb³⁺, 0.5% Tm³⁺ doped sample, whereas the sample co-doped with 20% Yb³⁺, 2% Tm³⁺ showed the most intensive UC emission band in the NIR range. The materials were characterized using powder X-ray diffraction and scanning electron microscopy. Optical properties were studied using steady-state and kinetic downshifting and upconversion photoluminescence spectroscopy.

1. Introduction

The lanthanides doped inorganic upconversion (UC) materials with excellent optical properties have significant applications in wide fields, for example, temperature sensing [1], solar light conversion [2], optical sensors [3], or security applications [4], bioimaging [5], optical heating [6], optogenetics [7], nanoscopy [8], nanoscale optical writing [9], etc. Inorganic phosphors, compared with organic dyes or quantum dots, have several advantages, including a longer excited state lifetime, sharp emission bandwidths, and cheaper and more environmentally friendly synthesis [4]. UC is a process where at least two low-energy photons produce high-energy photons. Usually, the inorganic upconverting luminescent materials contain two incorporated RE³⁺ ions: one as a sensitizer, typically Yb³⁺, and another as an emitter, for example, Ho³⁺ [10,11], Tm³⁺ [12,13], etc. Yb³⁺ possesses a simple energy level structure, including one ground (²F_7/2) and one excited (²F_5/2) state level. Additionally, the excited state of Yb³⁺ is higher than the metastable energy levels of emitters in UC materials, for example, Er³⁺, Ho³⁺, or Tm³⁺. In this case, the energy of ²F_7/2 → ²F_5/2 transition is capable of exciting other rare-earth (RE³⁺) elements, and UC emitters can release emissions.

The UC process in the phosphors could be achieved through co-doped RE³⁺ ions into a suitable host lattice. The K₂Gd(PO₄)(WO₄), as a novel phosphate-tungstate compound, has been suitable for application in important optical material fields [14]. Tungstate and phosphate could be used as hosts due to their excellent structure and high thermal stability. Thus, in tungstate, the average distance between the luminescent centers is larger, which may cause a reduced concentration quenching of RE³⁺ [15], which indicates that this host lattice has a high tolerance for heavily doped RE³⁺ [16]. Thus, the RE³⁺ concentration could be an important factor for the luminescence properties, and the proper ratio between a sensitizer and an emitter should be chosen.

The purpose of achieving blue light under 980 nm excitation inspired us to synthesize materials co-doped with Yb³⁺-Tm³⁺ because this RE³⁺ pair is the best combination for blue emission because, among the rare-earth ions, Tm³⁺ is one of the most studied RE³⁺ for blue emission based upon upconversion mechanism. In this work, co-doped crystalline K₂Gd(PO₄)(WO₄):20%Yb³⁺,Tm³⁺ phosphors are synthesized by a simple solid-state reaction and analyzed for the first time to the best of our knowledge. Further study of K₂Gd(PO₄)(WO₄) samples exhibits both downshifting and upconversion photoluminescence under 360 nm and 980 nm excitation, respectively. The obtained results of the synthesized materials show great potential for NIR-excited security pigments application.

2. Materials and Methods

Two series of K₂Gd(PO₄)(WO₄) (further abbreviated as KGPW) samples were solely doped with Tm³⁺ and co-doped with 20% Yb³⁺ and Tm³⁺. The Tm³⁺ concentration was 0%, 0.5%, 1%, 2%, and 5% with respect to Gd³⁺. All samples were prepared by the solid-state reaction method. The stoichiometric amounts of reagents, namely, Gd₂O₃ (99.99% Tailorlux, Münster, Germany), K₂CO₃ (99+% Acros Organics, Geel, Belgium), NH₄H₂PO₄ (99% Reachem Slovakia, Petržalka, Slovakia), WO₃ (99+% Acros Organics), Yb₂O₃ (99.99% Alfa Aesar, Haverhill, MA, USA), and Tm₂O₃ (99.99% Alfa Aesar) were weighed, poured to an agate mortar, and moistened with some acetone. The moist mixture of the materials was homogenized until all the acetone evaporated. The dry homogenous powders were placed in the porcelain crucibles and sintered for 10 h at 873 K. To obtain single-phase compounds, the sintering procedure was repeated two more times.

The phase purity of the prepared compounds was checked using a Rigaku MiniFlexII diffractometer (Tokyo, Japan). The XRD patterns were collected in the 2θ range of 5° to 80°. IR spectra were obtained within the range of 4000 to 400 cm⁻¹ using a Bruker Alpha ATR spectrometer (Ettlingen, Germany) with a resolution of 4 cm⁻¹. The samples were also examined using the FE-SEM SU-70 microscope from Hitachi (Tokyo, Japan). Room temperature and temperature-dependent optical properties of the synthesized compounds were investigated using the FLS980 spectrometer from Edinburgh Instruments (Livingston, UK). The used spectrometer settings are given in Tables S1–S4.

The lattice parameters of the synthesized compounds were calculated from the XRD patterns using the Rietveld refinement method. FullProf Suite software (version 2 December 2022) was used for calculations. The background was set as a 24-term Chebychev-type function. A pseudo-Voigt peak shape was used for the peak profiles. The scale factor, instrument zero, unit cell parameters, preferred orientation, atomic coordinates, and the peak shape (u, v, w, γ₀, and γ₁) parameters were also refined. K₂Ho(PO₄)(WO₄) (PDF-4+ (ICDD) 04-015-9304) compound, reported by Terebilenko et al. [17], was used for Rietveld fits.

3. Results and Discussion

Overall, two different series of samples were prepared. One contained KGPW doped solely with Tm³⁺, whereas the second one contained samples co-doped with Yb³⁺ and Tm³⁺. The Tm³⁺ concentration in both series varied from 0.5% to 5%. The concentration of Yb³⁺ in the samples was 20%. Tm³⁺-containing samples could be directly excited by UV radiation (~360 nm) (³H₆ → ¹D₂ transition (blue upward arrow in Figure 1). The Yb³⁺ co-doped samples, on the other hand, could also be indirectly excited with a 980 nm wavelength laser. Here, the Yb³⁺ ions absorb the laser radiation (⁷F_7/2 → ²F_5/2 transition) and transfer the captured energy to Tm³⁺, which, after receiving several quants of energy, emits in the UV and visible (Vis) spectral areas. The schematic diagram showing the main Yb³⁺ and Tm³⁺ energy levels involved in the downshifting (DS) and upconversion (UC) processes is depicted in Figure 1. Due to the rather large amount of energy levels, Tm³⁺ can emit in a wide range, i.e., from UV to deep-red and even infrared (IR) [18,19].

The phase purity of all synthesized compounds was checked by recording their powder XRD patterns. The lattice parameters of the synthesized compounds were calculated from the XRD patterns using the Rietveld refinement method. The Rietveld refinement of undoped KGPW, KGPW:5%Tm³⁺, KGPW:20%Yb³⁺, and KGPW:20%Yb³⁺,5%Tm³⁺ specimens are given in Figure 2. The calculated unit cell parameters of undoped KGPW, KGPW:5%Tm³⁺, KGPW:20%Yb³⁺, and KGPW:20%Yb³⁺,5%Tm³⁺ specimens are summarized in Table S5. The unit cell parameters decrease with increasing Yb³⁺ and Tm³⁺ concentration since both ions are smaller than Gd³⁺. The recorded XRD patterns matched well with the reference pattern, and no additional peaks were observed; therefore, we can conclude that single-phase materials were obtained. K₂Gd(PO₄)(WO₄) is isostructural with the reported K₂Ho(PO₄)(WO₄) compound. The crystal structure of the prepared materials is orthorhombic, and the space group is Ibca (#73) [17]. The crystal structure of these compounds is built from isolated PO₄ and WO₄ tetrahedral units and K⁺ and Gd³⁺ eight-fold coordinated polyhedra. Considering the same charge and very close ionic radii, we assumed that _VIIIYb³⁺ (0.985 Å) and _VIIITm³⁺ (0.994 Å) replaced _VIIIGd³⁺ (1.053 Å) ions [20].

The SEM images of KGPW:5%Tm³⁺, KGPW:20%Yb³⁺, and KGPW:20%Yb³⁺,5%Tm³⁺ samples are shown in Figure S1. The SEM images show that samples consist of irregularly shaped and agglomerated particles. No obvious changes in particle shape and size with changing the dopant concentration were observed.

The IR spectra of KGPW and KGPW:20%Yb³⁺ are shown in Figure S2. Both spectra possess several sets of absorption lines in the range of 1100–400 cm⁻¹. The three sharp absorption lines at 650–450 cm⁻¹ are assigned to the bending vibrations of PO₄. The strong absorption band ranging from 900 to 700 cm⁻¹ is attributed to the stretching vibrations of Mo−O within MoO₄ tetrahedral units. The strong absorption band at ca. 1075 is ascribed to asymmetric vibrations of PO₄ tetrahedral units, whereas the band at ca. 935 cm⁻¹ is ascribed to symmetric ones [21].

The reflection spectra of undoped KGPW, KGPW:20%Yb³⁺, and KGPW:5%Tm³⁺ are presented in Figure 3. The reflection spectrum of KGPW:20%Yb³⁺,5%Tm³⁺ is identical to the one of KGPW:5%Tm³⁺; therefore, it was not shown. All samples possessed a white body color showing that the samples do not absorb in the VIS range. It also should be mentioned that the reflectance at longer wavelengths is almost 100% showing low defect concentration in the synthesized materials.

The reflection spectra were measured in a 250–750 nm range. The reflection spectra of Tm³⁺ doped samples contain three typical Tm³⁺ absorption lines, i.e., ³H₆ → ¹D₂ (ca. 355–370 nm), ³H₆ → ¹G₄ (ca. 460–486 nm), and ³H₆ → ³F_2,3 (ca. 655–715 nm) [22]. The broad absorption band in the UV range (around 250–300 nm) could be assigned to the charge transfer from O²⁻ to W⁶⁺ transition in the host lattice [23].

The excitation (λ_em = 450 nm) spectra of KGPW:Tm³⁺ and KGPW:20%Yb³⁺,Tm³⁺ (where the Tm³⁺ concentration is changed from 0.5% to 5%) samples were recorded from 250 to 430 nm and are shown in Figure 4a,c, respectively. The excitation spectra contain one band at 360 nm originating from the typical Tm³⁺ ground state ³H₆ absorption to the excited state ¹D₂. In both cases, the highest intensity was achieved with a 5% Tm³⁺ doped sample. Relatively lower excitation intensity in co-doped samples could be explained due to Tm³⁺ → Yb³⁺ energy transfer [24]. The same tendency was observed in emission spectra (λ_ex = 360 nm). The highest emission intensity was observed for KGPW:5%Tm³⁺ and KGPW:20%Yb³⁺,5%Tm³⁺ samples (please refer to Figure 4b,d, respectively). The insets in Figure 4b,d show the normalized integrated emission intensity of the prepared samples and reveal that the emission intensity increases with increasing Tm³⁺ concentration and reaches maximum intensity when Tm³⁺ concentration is the highest. There are few sets of emission lines in the down-conversion emission spectra: the intense blue emission at 440–463 nm corresponds to the ¹D₂ → ³F₄ transition, whereas much weaker emission lines at 463–485 nm, 650–670 nm, 740–760 nm, and 780–800 nm correspond to the ¹G₄ → ³H₆, ¹G₄ → ³F₄, ¹D₂ → ³F₃, and ³H₄ → ³H₆ transitions, respectively [25]. The most intense emission line observed at 450 nm could be explained due to the directly excited ¹D₂ energy level with 360 nm excitation. Upon 360 nm excitation, phosphors doped with Tm³⁺ emit intense blue emission through ¹D₂ → ³F₄ transition and that corresponds well with CIE 1931 chromaticity coordinates, depicted in Figure S3. The color coordinates of all the synthesized samples under 360 nm are located in the blue spectral region and are near the perimeter of the CIE 1931 color space diagram. This indicates high color purity. The precise values of color coordinates are tabulated in Table S6.

The influence of Tm³⁺ concentration on the UC emission intensity was also investigated. The UC emission spectra of KGPW:20%Yb³⁺,Tm³⁺ (where Tm³⁺ concentration is 0.5%, 2%, and 5%) samples under 980 nm laser excitation are given in Figure 5. The observed emission bands in the UV, visible, and near-IR range can be attributed to ¹D₂ → ³H₆ (at 355–365 nm), ¹D₂ → ³F₄ (at 445–458 nm), ¹G₄ → ³H₆ (458–496 nm), ¹G₄ → ³F₄ (at 625–670 nm, respectively), and ³H₄ → ³H₆ (at 755–844 nm) transition. The substantial change in UC emission spectra was observed as a function of Tm³⁺ concentration. Interestingly, the intensity of the bands in the UV and visible range do not follow the same trend as the emission band in the NIR range. The 0.5% Tm³⁺ doped sample showed the strongest emission in the UV/Visible range. On the other hand, the 2% Tm³⁺ doped sample yielded the most intensive UC emission band in the NIR range. This sample also showed the strongest overall UC emission, as shown in the inset graph of Figure 5. When Tm³⁺ concentration increases, the Yb³⁺ → Tm³⁺ energy becomes more efficient because the average distance between these ions decreases. However, a further increase in Tm³⁺ concentration leads to a decrease in ³H₄ → ³H₆ emission intensity. Typically, decreasing emission intensity with increasing Tm³⁺ concentration is attributed to the concentration quenching [26]. Furthermore, with increasing Tm³⁺ concentration the probability of energy back transfer from Tm³⁺ to Yb³⁺ increases (¹G₄ (Tm³⁺) + ²F_7/2 (Yb³⁺) → ³H₅ (Tm³⁺) + ²F_5/2 (Yb³⁺)), thus, depopulating the ¹G₄ (Tm³⁺) level and reducing blue emission [27]. Such energy back transfer also increases the population of ³F₄ (Tm³⁺) level due to non-radiative relaxation from the ³H₅ (Tm³⁺). The populated ³F₄ level can again receive the energy from Yb³⁺ and be excited to ³F₂ level (Tm³⁺) (²F_5/2 (Yb³⁺) + ³F₄ (Tm³⁺) → ²F_7/2 (Yb³⁺) + ³F₂ (Tm³⁺)). Then ³F₂ energy level can populate the ³H₄ level through non-radiative relaxation, thus, increasing the ³H₄ → ³H₆ emission in the NIR region [28].

For a further understanding of the DC process, the PL decay curves for the most intense DC transition ¹D₂ → ³F₄ (λ_ex = 360 nm, λ_em = 450 nm) as a function of Tm³⁺ concentration were recorded. The mono exponential PL decay curves for KGPW:Tm³⁺ and KGPW:20%Yb³⁺,Tm³⁺ samples were obtained, as depicted in Figure 6a,b, respectively. With increasing Tm³⁺ concentration, the PL decay curves get steeper, showing that the effective PL lifetime (τ_eff) values decrease. This, indeed, was confirmed after calculating the τ_eff values [29]:

$${\tau }_{eff}{}_{}=\frac{{\displaystyle {\int }_0^{\infty }I(t)tdt}}{{\displaystyle {\int }_0^{\infty }I(t)dt}}$$

(1)

Here, I(t) is emission intensity at a given time t. The increasing Tm³⁺ concentration leads to decreasing τ_eff values which, in turn, also indicate the decreasing internal quantum efficiency of Tm³⁺.

For a better understanding of the UC process, the PL decay curves of two main Tm³⁺ emission peaks, namely, ¹G₄ → ³H₆ (λ_ex = 980 nm, λ_em = 478 nm) (see Figure 7) and ³H₄ → ³H₆ (λ_ex = 980 nm, λ_em = 800 nm) (see Figure 8), were measured. The bi-exponential PL decay curves were observed in both cases.

With increasing Tm³⁺ concentration, the UC PL decay curves of the main emission peaks become steeper, suggesting that τ_eff values decrease. The calculated τ_eff values decreased from 201 μs to 107 μs for ¹G₄ → ³H₆ transition and from 267 μs to 101 μs for ³H₄ → ³H₆ transition when Tm³⁺ concentration was increased from 0.5% to 5%. The calculated rise time values for ¹G₄ → ³H₆ transition decreased from 73 μs to 38 μs when Tm³⁺ concentration was increased from 0.5 % to 5%. Energy transfer from Yb³⁺ to Tm³⁺ increases with increasing Tm³⁺ concentration, which is reflected by a shorter rise time in the KGPW:20% Yb³⁺ with a higher concentration of Tm³⁺. The calculated PL rise and τ_eff values are tabulated in Table S7.

To evaluate the Tm³⁺ concentration-dependent PL lifetime values of Yb³⁺, the samples were excited with a 980 nm laser, and the PL decay curves for Yb^{3+ 2}F_5/2 → ²F_7/2 transition were recorded. The Yb³⁺ emission was monitored at 1050 nm. The recorded PL decay curves as a function of Tm³⁺ concentration are depicted in Figure 9. As it was expected, the PL lifetime values of the mentioned Yb³⁺ transition drastically decreased from 1312 ± 24 μs to 416 ± 4 μs when Tm³⁺ concentration was increased from 0% to 5%. Such an expected decrease in Yb^{3+ 2}F_5/2 → ²F_7/2 transition PL lifetime with increasing Tm³⁺ concentration is caused by Yb³⁺ → Tm³⁺ energy transfer. The Yb³⁺ → Tm³⁺ energy transfer efficiency (η_tr) was determined from the Yb³⁺ PL lifetime values using this equation [30]:

$${\eta }_{tr}=\left(1-\frac{{\tau }_{Yb-Tm}}{{\tau }_{Yb}}\right)\times 100\%$$

(2)

Here, τ_Yb−Tm and τ_Yb are Yb³⁺ PL lifetime values of ²F_5/2 → ²F_7/2 transition in the presence and absence of Tm³⁺, respectively. The η_tr coherently increases from 44% to 68% when changing Tm³⁺ concentration from 0.5% to 5%, respectively. The exact PL lifetime values, together with calculated η_tr values, are given in Figure 9 insets and Table S8.

Finally, in order to represent an emission color of the synthesized material under different excitation wavelengths, the response of the standard human eye should be considered. For this reason, the color coordinates (in CIE 1931 color space) of the samples, excited with 360 and 980 nm wavelength radiation, were calculated and are given in Figure 10a,b, respectively. Additionally, the exact color coordinate values are given in Table S6. The color coordinates are invariant upon the Tm³⁺ concentrations but have a slight dependence on the excitation wavelength. The color coordinates of KGPW:20%Yb³⁺,Tm³⁺ samples excited with 980 nm wavelength laser shift upwards if compared to color coordinates obtained for 360 nm excitation. The shift is caused by the fact that the strongest emission lines at 360 nm excitation are at 450 nm (¹D₂ → ³F₄ transition), whereas for 980 nm wavelength laser excitation, the ¹G₄ → ³H₆ transition at ~475 nm is the strongest one. Color coordinates for samples excited with 360 nm radiation are closer to the perimeter of the CIE 1931 color space diagram showing higher color purity. However, color coordinates slightly shift to the center of the color space diagram if samples are excited with a 980 nm wavelength laser. The shift is caused by relatively strong emission from ¹G₄ → ³F₄ transition at ~650 nm if compared to samples excited with 360 nm (please refer to Figure 4d and Figure 5).

4. Conclusions

In summary, we have successfully synthesized single-phase K₂Gd(PO₄)(WO₄):Tm³⁺ and K₂Gd(PO₄)(WO₄):20%Yb³⁺,Tm³⁺ powders, where Tm³⁺ concentration varied from 0.5% to 5%. Under 360 nm excitation, there was no concentration quenching (at least up to 5% Tm³⁺) in solely Tm³⁺ doped samples. The highest upconversion emission in the UV/Visible range was achieved with 20%Yb³⁺,0.5%Tm³⁺ concentration. However, the sample co-doped with 20%Yb³⁺,2%Tm³⁺ shows the most intense UC emission band in the NIR range, which could be explained by a more efficient Yb³⁺ → Tm³⁺ energy transfer. Thus, the color coordinates are invariant upon the Tm³⁺ doping concentrations and take place in the blue region.