New amorphous nanophosphors obtained by evaporation of silicates and germanates REE

X-ray phase analysis and microscopy

The solid solutions Sr₂Gd₈(SiO₄)₆O₂:Eu, Cа₂Y(Gd)₈(SiO₄)₆O₂:Eu and Ca₂La₈(GeO₄)₆O₂:Eu belong to the oxyapatite structure type. According to [13], [16] in crystals of oxyapatite (sp. gr. P6₃/m, z=1) in the designation of the Wyckoff the site-symmetry of atoms Y(Gd, La)1 – 4f, Sr(Ca) – 4f, Y(Gd, La)2 – 6h, Si(Ge) – 6h, O – 6h, O – 6h, O – 12i, O – 2a. Europium, replacing atoms of Y, Gd or La, can also take positions 4f and 6h. The structure contains isolated Si(Ge)O₄ tetrahedra. Oxygen O(4) in the 2a position does not enter into the composition of the tetrahedra and forms chains parallel to the hexagonal crystal axis [9].

Figure 1a shows a characteristic diffraction pattern of the synthesized solid solutions Sr₂Gd_6.53Eu_1.47Si₆O₂₆ (I), Cа₂Y_6.57Eu_1.43Si₆O₂₆ (II), Cа₂Gd_6.55Eu_1.45Si₆O₂₆ (III) and Ca₂La_6.4Eu_1.6Ge₆O₂₆ (IV). This composition was chosen because in the series of phosphors the solid solutions at the concentration of europium 0.2 mol has the maximum integral luminescence intensity of Eu³⁺ ions. At the concentration of europium more than 0.2 mol, concentration quenching of Eu³⁺ luminescence is observed. Processing by the Rietveld method yields the following data: single phase, sp. gr. P6₃/m. The characteristics of the obtained samples are given in the Table 1. The NP produced from this solid solution is X-ray amorphous (the characteristic X-ray diffraction pattern is shown in Fig. 1b).

Fig. 1:

X-ray diffraction patterns of Sr₂Gd_6.53Eu_1.47Si₆O26, Cа₂Y_6.57Eu_1.43Si₆O₂₆, Cа₂Gd_6.55Eu_1.45Si₆O₂₆, Ca₂La_6.4Eu_1.6Ge₆O₂₆ (a) and amorphous NP (b).

Electron microscopy of the NP based on the sample (I) (Fig. 2a) reveals that the powder nanoparticles are strongly agglomerated and have an irregular shape tending to become quasispherical and are connected with each other by bridges of arbitrary length and shape (Fig. 2b). The electron diffraction data (Fig. 3a) and TEM HR pictures (Fig. 3b) show that the NP consists of amorphous particles only. The particle size in the NP generally does not exceed 20 nm. The BET value of the NP surface is S_sp=191.61 m²/g. Consequently, the average NP particle size is 5 nm.

Fig. 2:

A typical agglomerate of quasispherical NP particles (a) and TEM picture of bridges between quasispherical NP particles (b) based on Sr₂Gd_6.53Eu_1.47Si₆O₂₆.

Fig. 3:

Electron diffraction pattern of amorphous NP (a) and TEM HR-picture of amorphous NP based on Sr₂Gd_6.53Eu_1.47Si₆O₂₆.

Electron microscopy of the NP (Fig. 4а) produced on the basis of Ca₂Y_6.4Eu_1.6Si₆O_26–δ, where δ is oxygen nonstoichiometry, revealed that the nanoparticles are rather strongly agglomerated and have an irregular shape tending to form quasi-spherical particles and are connected with each other by nips of arbitrary length and shape. Oxygen nonstoichiometry appears due to the formation of Eu²⁺ ions (shown below). The data of electron diffraction analysis (the electron diffraction patterns were taken from six different regions of the sample, see the inset in Fig. 4a) and the TEM HR pictures (Fig. 4b) with different magnification confirm that the NP consists exclusively of amorphous particles. The BET value of the nanoparticle surface is S_sp=232.25 m²/g; as a consequence, the average particle size is ~5.7 nm.

Fig. 4:

The TEM pictures of the NP based on Cа₂Y_6.57Eu_1.43Si₆O₂₆: an agglomerate of nanoparticles (inset: the electron diffraction pattern of the amorphous region of the sample) (a); quasi-spherical amorphous nanoparticles (b).

In addition to that, electron microscopy of the NP obtained on the basis of Ca₂Gd_6.4Eu_1.6Si₆O_26–δ (Fig. 5a) showed that the nanoparticles are also agglomerated. The agglomerates in the NP several hundred nanometers in size generally have a dendrite-like shape and consist of irregularly-shaped nanoparticles prone to form quasi-spherical particles. The electron diffraction data and the TEM HR pictures demonstrated that the NP consists of amorphous particles only (Fig. 5b). The BET value of the nanoparticle surface is S_sp=172.63 m²/g; as a result of it, the average particle size is 5.6 nm.

Fig. 5:

The TEM pictures of the NP based on Cа₂Gd_6.55Eu_1.45Si₆O₂₆: a typical dendrite-like agglomerate of quasi-spherical nanoparticles (inset: the electron diffraction pattern of the amorphous region of the sample) (a); an image of a nanoparticle aggregate region (b).

Just as in the previous three cases of microscopy, the sample (IV) reveals that the powder nanoparticles are strongly agglomerated and amorphous (Fig. 6). The BET value of the nanoparticle surface is S_sp=105.5 m²/g. Consequently, the average particle size is 9.9 nm.

Fig. 6:

The TEM pictures of the NP based on Ca₂La_6.4Eu_1.6Ge₆O₂₆: agglomerate of quasi-spherical nanoparticles (a); electron diffraction pattern of amorphous NP (b).

Thermal analysis

In order to characterize the obtained nanoparticles and to determine their thermal stability and the beginning of crystallization, we carried out DSC-TG analysis. The sample heating rate is 10°C/min. In the temperature interval 50–500°C the batch mass decreases by ~13% owing to removal of water from NP and CO₂ release. Weak signals from water and carbon dioxide are found on the mass spectra in the temperature intervals 50–400°C (H₂O) and 230–500°C (CO₂). Adsorption of water and CO₂ on the nanoparticle surface can occur both directly during NP deposition in the evaporation chamber and after removal of the samples from the chamber since the highly-developed NP surface captures water and CO₂ vapors from ambient air. Crystallization of amorphous nanoparticles begins at a relatively low temperature and has probably two stages. At the first stage, at temperatures from 200 to 400°C, separate nanoparticles crystallize. Low thermal stability of the NP at this stage is likely to be due to the dimensional factor, since the size of the nanoparticles is ~5 nm. At the second stage (~800–970°C) agglomerated nanoparticles crystallize [13].

For the sample based on NP (II) in the temperature range 50–1000°С, the mass of the sample decreased by about 11.8% owing to removal of adsorbed crystal water and unknown carbon compounds from the NP. The signals from water and CO₂ are observed on the mass spectra in the temperature intervals 50–330°С and 50–1000°С, respectively Crystallization of amorphous nanoparticles begins at a low temperature as indicated by a strong exothermic peak in the temperature interval from 330 to 1000°С. The low thermal stability of the NP is due to the effect of the nanoparticle size factor. СО₂ gas emission is completed simultaneously with termination of amorphous nanoparticle crystallization at about 1000°С and is not observed further. During heating of the NP up to 1400°С, a small increase in the sample’s mass is observed (by 1.86%), which is accompanied by three exothermic transformations of unknown nature associated with oxidation of nanoparticles.

For the sample based on NP (III) in the temperature interval from 50 to 1400°С, an approximately 12.3% mass loss is observed, which is due to removal of adsorbed and crystal water. The signals from water are detected on the mass spectrum in the temperature range ~50−330°С. The crystallization of amorphous nanoparticles begins at ~1020°С, as indicated by a strong extended exothermic peak on the DSC heating. The large length of exothermic points to a slow kinetics of amorphous NP transformation into crystalline powder (probably, the smallest amorphous nanoparticles begin to crystallize, as evidenced by the inflection of the DSC baseline curve at about 1020°С). As the temperature increases up to 1080°С, active crystallization of amorphous nanoparticles begins, which is reflected in the strengthening of heat generation terminating at about 1350°С. Two insignificant thermal peaks observed on the DSC heating curve at ~860 and 940°С can be associated either with sublimation of different organic compounds of unknown nature (except СO₂) from the nanoparticle surface or with a complex stepwise crystallization of amorphous nanoparticles having smaller dimensions.

For the sample based on NP (IV) in the temperature interval from 50 to 600°С, an approximately 7.58% mass loss is observed, which is due to removal of adsorbed and crystal water. In the temperature interval 600−760°С, an approximately 0.5% mass loss is observed. As the temperature increases up to 810°С crystallization of amorphous nanoparticles begins. This is indicated by a small exothermic peak on DSC heating curve. As the temperature increases up to 870°С, active crystallization of amorphous nanoparticles begins, which is reflected in the strengthening of heat generation terminating at about 1370°С. As the NP is further heated up to 1400°С, a dip of the DSC heating curve is observed suggesting the termination of amorphous NP crystallization.

Thus, according to the results of thermal analysis, individual amorphous nanoparticles obtained on the basis of samples I–IV are resistant at temperatures below the following numbers for each sample: I-200, II-330, III-1020, IV-810°C.

Raman and IR spectroscopy

Raman spectra of sample I and amorphous NP based sample I are shown in Fig. 7a [13]. The frequencies ν₁, ν₂, ν₃ and ν₄ of the internal vibrations of isolated SiO₄ tetrahedra are designated for the polycrystals. The spectrum of amorphous NP exhibits the formation of a structure, which is slightly different from that of the initial polycrystalline sample. The location of lines in the NP spectrum shows that it contains mainly isolated SiO₄ groups. At the same time the appearance of frequencies in the region 600–700 cm⁻¹ points to the formation of an inconsiderable concentration of polymerized silicon-oxygen fragments in the sample [17]. The frequency at 755 cm⁻¹ relating probably to vibrations of bridge bonds Si–O–Si is an indicator of the polymerization processes [17], [18]. In the process of evaporation of sample I and subsequent NP deposition on a glass substrate, the restructuring of the anionic motif of the silicate is likely to take place and display silicon-oxygen complexes with an insignificant concentration of SiO₄ tetrahedra. This phenomenon is analogous to the polymerization effect in rapid-quenched silicate melts of crystals with isolated complexes SiO₄. Oxygen ions that do not enter into the composition of the tetrahedra take part in this process. This is due to the presence of TR ions that can form partially covalent bonds with oxygen, for example, in Ca₂Gd₈(SiO₄)₆O₂ [18]. The external vibration frequencies of SiO₄ shift towards the low-energy region (125–313 cm⁻¹). The shift of the line at 313 cm⁻¹ (E_2g symmetry) is presumably due to the decrease in the force coefficient of the E_2g component of the libration vibration of the SiO₄ complex [18]. Besides, redistribution of intensities of external vibration lines is observed. The internal vibration frequencies also shift towards the low-energy region of frequency ν₄, and the intensity of the lines at ν₁ and ν₃ decreases considerably. These effects in the spectrum can be due to unification of a large number of oxygen vacancies in nanosample that emerge during evaporation of the bulk sample. The unification of vacancies breaks the translational symmetry. Such symmetry breakdown can cause the reduction of the intensity of the line at ν₁ [19], [20].

Fig. 7:

Raman spectra of bulk and NP: Sr₂Gd_6.53Eu_1.47Si₆O₂₆ (a), Cа₂Y_6.57Eu_1.43Si₆O₂₆ (b), Cа₂Gd_6.55Eu_1.45Si₆O₂₆ (c).

Raman spectra of polycrystalline samples II and III and NP on their basis are depicted in (Fig. 7b and c). The spectrum of the amorphous NP produced by evaporation of II (Fig. 7b) shows the formation of a structure differing slightly from the structure of the initial polycrystalline sample. The position of the lines in the spectrum of the NP suggests that it contains mainly the isolated SiO₄ groups. The appearance of frequencies in the region of 600–700 cm⁻¹ bears witness to the formation of an insignificant concentration of polymerized silicon-oxygen fragments in the sample [13], [17]. The emergence of a new frequency at 716 cm⁻¹, which is likely to refer to vibrations of the Si–O–Si bridge bonds, is indicative of the polymerization processes [18]. Oxygen O(4) in the 2a position that does not enter into the compositions of SiO₄ tetrahedra probably participates in the polymerization process [13].

The frequency 138 cm⁻¹ is shifted into the high-frequency region and the frequencies of external vibrations of isolated SiO₄ complexes disappear (187–250 cm⁻¹ region). The shift of the line at 337 cm⁻¹ into the high-frequency region is presumably due to the increase in the force coefficient value of the E_2g component of the SiO₄ complex libration vibration. The frequencies of internal vibrations ν₂, ν₁ and ν₃ are also shifted into the high-frequency region. The frequency ν₄, by contrast, is displaced into the low-frequency region. A relative decrease in the intensity of lines at ν₁ and ν₂ is observed. These effects in the spectrum can be induced by amalgamation of a large number of oxygen vacancies in the nanosample, which appear during evaporation of the sample and its further solidification. The amalgamation of vacancies leads to translational symmetry violation. Concurrently, such symmetry violation can cause the reduction of the intensity of the line at ν₁ [19], [20].

The spectrum of amorphous NP produced by evaporation of III (Fig. 7c) has a structureless shape with no characteristic scattering lines. This can be also a result of translational symmetry violation in the NP and, consequently, complete lifting of the prohibition imposed by the quasi-momentum conservation law. As a result, phonons with any wave vectors are allowed [21].

The Raman spectra of Ca₂La_6.4Eu_1.6Ge₆O₂₆ and NP based thereon are not informative enough. Therefore the IR spectra of these samples have been analyzed (Fig. 8). Figure 8 displays the vibration frequencies ν₂, ν₃, ν₄ of isolated tetrahedron GeO₄ [22], [23]. The absorption bands in the NP spectra are broadened which is indicative of disordering of GeO₄ groups. The spectrum of the amorphous NP points to the formation of a structure, which is somewhat different from that of the initial polycrystalline sample. The position of bands in the NP spectrum suggests that it contains mainly the isolated GeO₄ groups. A band is formed at 392 cm⁻¹, which corresponds probably to the frequency ν₄. The internal vibration frequency ν₃ (789 cm⁻¹) of the GeO₄ groups is considerably shifted into the high-frequency region. A new band at 557 cm⁻¹ appears. This band corresponds to the Ge−O−Ge bridge vibrations [24] and is indicative of partial polarization of the GeO₄ groups. Oxygen ions that do not enter into the composition of the tetrahedra (oxygen O(4) in the 2a position) are involved in this process. During evaporation of sample IV and subsequent NP deposition on a glass substrate, restructuring of the germanate anionic motif is likely to take place, and germanium-oxygen complexes with condensation of GeO₄ tetrahedra probably appear.

Fig. 8:

IR spectra of sample Ca₂La_6.4Eu_1.6(GeO₄)₆O₂ and NP at its base.

Optical spectroscopy chosen silicates

The Eu³⁺ photoluminescence spectra of the samples in the bulk phosphors and in nanophosphors (sample I) are displayed in Fig. 9a and b. Two lines for the transition ⁵D₀→⁷F₀ for bulk and nanophosphor point to the formation of two types of optical centers formed by the Eu³⁺ ions occupying the 4f and 6h positions in the silicate structure (similarly to Sr₂Y₈(SiO₄)₆O₂:Eu phosphors [25]).

Fig. 9:

PL spectra of samples bulk I (а), NP (b) and PLE spectra of bulk I (c, d), NP (e, g) (λ_em=614 nm).

Consider the effect of nanoparticle disordering on red photoluminescence purity of the samples. For this purpose, we determine the ratio of intensities of red “R” (⁵D₀→⁷F₂ transition) and orange “О” (⁵D₀→⁷F₁ transition) luminescence for bulk and nanophosphor. It is known that the electric dipole transition ⁵D₀→⁷F₂ is supersensitive to site symmetry of Еu³⁺ ion in optical center. According to the selection rules, the ⁵D₀ → ⁷F₂ transition is allowed if Eu³⁺ ion in the optical center occupies a position that does not coincide with the inversion center. The magnetic dipole transition ⁵D₀→⁷F₁ is also allowed. The R/O ratio is equal to 5.8 (bulk) and 6.1 (NP). Therefore, the intensity of red luminescence in the nanophosphor increases. As noted in work [26], this phenomenon is due to nanoparticle disordering providing relatively lower symmetry of electrostatic field around Еu³⁺ ions than in bulk phosphor and consequently a higher probability of the ⁵D₀→⁷F₂ transition. This will increase the intensity of red luminescence.

The essential difference in the bulk and nanophosphors is increased integral luminescence intensity for the NP. The intensity of electric dipolar transitions ⁵D₀→⁷F₂ for the NP increases 10 times. The spectral lines for some agglomerated nanoparticles are broadened, which points to a disordered position of the optical centers in the NP.

Figure 9c–g shows the excitation spectra of optical luminescence centers in Sr₂Gd_6.52Eu_1.47Si₆O_26–δ polycrystals (c) and NP (e) in the interval λ_ex=250–360 nm (λ_em=614 nm). The excitation bands are due to the transition of Eu³⁺ to the charge-transfer state (Eu³⁺→O²⁻). Each band is a superposition of spectra of two optical centers. The maximum values of the excitation wave lengths are different for the bulk powder and NP. The maxima are located, respectively at λ_max≈269 and 263 nm.

This difference in the wave length maxima of the charge transfer bands can be explained by different stability of the nearest oxygen ions O²⁻ surrounding Eu³⁺ ion. The influence of ligands on the charge transfer band was described in work [27]. The oxygen ions coordinating Eu³⁺ in the optical centers are stabilized predominately by the Si⁴⁺ ions. Apparently, the О²⁻ ions in nanophosphor are stabilized by surrounding positive ions to a somewhat greater degree than in bulk phosphor and therefore more energy is required to remove electron from these ions [27]. In this connection it can be expected that the charge transfer band in nanophosphor will be shifted to the short-wave spectral region as compared with the analogous band in bulk phosphor. Besides, optical absorption (charge transfer transition) includes the oxygen p states. The p states are very sensitive to europium ion environment variation [28]. Consequently, this may suggest a slight variation of this environment in nanophosphor.

Figure 9c and e contain excitation bands corresponding to ⁸S→⁵D and ⁸S→⁶I transitions of Gd³⁺ ion. Therefore, a resonant transfer of excitation energy from the Gd³⁺ ions to the Eu³⁺ ions takes place in bulk and nanophosphors. An analogous transfer of excitation energy was observed in work [2].

Figure 9d and g also demonstrates the luminescence excitation spectra of the optical centers formed by the Eu³⁺ ions in the wave length interval λ_ex=350–550 nm for the polycrystals (Fig. 9d) and NP (Fig. 9g). The spectra were recorded for λ_em=614 nm. The excitation spectra of the bulk powder exhibit absorption lines of Eu³⁺ ions corresponding to the ⁵D_J and ⁵L_J transitions. Assignment of the transitions was carried out using the book [29]. The excitation spectra of the NP at λ_em=614 nm exhibit Eu³⁺ ion absorption lines corresponding to the ⁵L₇ and ⁵D₂ transitions. From comparison of Fig. 9d and g, it is seen that redistribution of intensities between sublevels of ⁷F₀→⁵D₂ transition takes place for bulk and nanophosphor. According to the selection rules, ⁷F₀→⁵D₂ is a supersensitive transition [30]. Therefore the redistribution of intensities is indicative of certain distortion of the local environment of Eu³⁺ ions in nanophosphor in comparison with bulk phosphor.

Consider the variation in the width of the forbidden band E_g of the samples during transition from bulk powder to NP. Figure 10a presents the absorption spectra for Sr₂Gd_6.52Eu_1.47Si₆O_26–δ and NP. The absorption edges are located, respectively at 354 and 304 nm. Consequently, in the transition from bulk to NP the value of E_g increases from 3.50 to 4.08 eV. The absorption edge of phosphors is due to the absorption in the charge transfer band (the 4f⁷2p⁻¹ state) that is analogous to Sr₂Y_6.4Eu_1.6(SiO)₆O₂ phosphor [25]. The charge transfer level in Eu³⁺ forms the upper edge of the band. In bulk phosphors, the charge transfer level has a somewhat smaller energy than in nanophosphors. Therefore the fundamental absorption edge in nanophosphors is higher than in bulk phosphors. Consequently, along with different stability of the nearest oxygen ions O²⁻ surrounding Eu³⁺ ion in bulk and nanophosphor, we have to do here with small variation of this environment in nanophosphor. The environment of europium ions can be changed owing to insignificant concentration of SiO₄ tetrahedra.

Fig. 10:

Spectra of absorption of samples: I (a), II (b), III (c), IV (d).

The band with a maximum at ~252 nm for bulk phosphor can be attributed to the ⁸S→⁵D transition of Gd³⁺ ion [2].

What is the reason of the integral luminescence intensity increase for amorphous nanophosphor in comparison with polycrystalline silicate? This effect is considered in the model of optical centers in the configuration coordinates in [13].

Consider the variation in the forbidden band E_g width of the samples when going from bulk powder to NP. Figure 10b depicts the absorption spectra of bulk and NP samples. For sample of II, the absorption edge is at ~347.5 nm, while for the NP produced by evaporation of this silicate – at 294.5 nm (Fig. 10b). Consequently, in the transition from bulk to NP, the value of E_g increases from 3.567 to 4.209 eV. Figure 10c demonstrates the absorption spectra of the samples of III and NP based on this phosphor. For bulk phosphor, the absorption edge is at 365 nm, which corresponds to E_g=3.396 eV. The band with a maximum at 256 nm can be attributed to the ⁸S→⁵D transition of Gd³⁺ ion [2]. In the region of absorption boundary there is an extended exponential tail, the so-called Uhrbach tail, typical of amorphous media [31]. The Uhrbach tail allows one only to estimate the absorption boundary value. The absorption edge is at 367 nm (3.34 eV).

From the excitation spectra of Eu³⁺ ions it is seen (Fig. 11b) that the excitation band maximum in bulk phosphors II and III is equal to 270 nm (λ_em=614 nm). For nanophosphors, the maximum is at λ_ex<260 nm (Fig. 12b). Therefore in bulk phosphors the charge transfer level has a smaller energy than in nanophosphors. The fundamental absorption edge in nanophosphors should be higher than in bulk phosphors. Consequently, along with different stability of the nearest O²⁻ ions surrounding the Eu³⁺ ion in bulk and nanophosphors, it is possible to suggest a small variation in this environment in the nanophosphor.

Fig. 11:

The PL and PLE spectra of bulk phosphors Ca₂Y_6.4Eu_1.6Si₆O_26–δ (1) and Ca₂Gd_6.4Eu_1.6Si₆O_26–δ (2): the PL spectra of samples 1 and 2 (a); the PLE spectra (λ_em=614 nm) (b) and (λ_em=443 nm) of samples 1 and 2 (c).

Fig. 12:

The PL (a) and PLE (λ_em=614 nm) (b) and (λ_em=443 nm) (c) spectra of the NP produced by evaporation of Ca₂Y_6.4Eu_1.6Si₆O_26–δ (II) and Ca₂Gd_6.4Eu_1.6Si₆O_26–δ (III). Scheme for Eu²⁺ ion in configuration coordinates of NP produced by evaporation of sample II (d)

Figure 11a exhibits the PL spectra of Ca₂Y_6.4Eu_1.6Si₆O_26–δ and Ca₂Gd_6.4Eu_1.6Si₆O_26–δ bulk phosphors. Along with the narrow luminescence lines from Eu³⁺ ions, there are broad bands (λ_max=443 nm) caused by luminescence of Eu²⁺ ions (4f⁶5d→4f⁷(⁸S_7/2) transition). The broad bands have a weak structure, which is likely to be due to the interaction of the 4f⁶ and 5d electrons [32]. Since during Y(Gd) substitution Eu³⁺ ion occupies two crystallographic positions in the structure of silicates, 4f and 6h, then Eu³⁺ forms two types of optical centers. The spectra are the total luminescence spectra of these two centers.

In our opinion, the mechanism of formation of Eu²⁺ ions can be analogous to that reported in works [13], [33]. During synthesis of Ca₂Y(Gd)_6.4Eu_1.6Si₆O_26–δ polycrystals, in the process of high-temperature annealing of the mixture of initial components,

The excitation spectra of bulk phosphors (Fig. 11b) for λ_em=614 nm contain the CTS bands and the lines induced by ⁷F₀→⁵D₄, ⁵G₂, ⁵L₇ transitions. For λ_em=443 nm, the excitation spectra (Eu²⁺) in the wavelength interval 260–305 nm are different (Fig. 11c). For the III phosphor, the absorption intensity in this wavelength interval increases, in contrast to the II phosphor. The broad bands in the interval 305– 390 nm are due to the transition from the 4f⁷(⁸S_7/2) level to the e_g level of the 5d state of Eu²⁺ ions. These broad bands for the both phosphors contain two maxima at ~351 and 369 nm, which are induced probably by splitting of the e_g level [34]. This is indicative of a rather strong interaction between the 4f⁶ electrons and the 5d electron.

Figure 12 demonstrates the PL and excitation spectra of the nanophosphors produced by evaporation of II and III samples in vacuo. The PL spectra are broad bands determined by the luminescence of Eu²⁺ ions. The luminescence of Eu³⁺ ions is quenched almost completely (Fig. 12a). For the nanophosphor obtained by evaporation of the II sample in argon atmosphere the intensity of the ⁵D₀→⁷F_{1, 2} transitions is much smaller than for the analogous sample produced by evaporation in vacuum, while the luminescence intensity of Eu²⁺ ions is greater. It can be supposed that the Eu²⁺ ions are additionally formed in the nanosamples due to radiation reduction Eu³⁺→Eu²⁺. Thus, the reduction of Eu³⁺→Eu²⁺ ions depends not only on the atmosphere, in which evaporation takes place, but also on the composition of initial bulk samples.

The luminescence intensity of Eu²⁺ ions for Ca₂Y_6.4Eu_1.6Si₆O_26–δ and for the nanophosphor produced on its basis is almost the same. We speculate that no Eu³⁺→Eu²⁺ reduction takes place in the sample. The small luminescence yield of Eu³⁺ ions can be explained with the use of the configuration diagram in the Deсkster-Click-Russel model [27].

Consider the excitation spectra of the nanophosphors. In the excitation spectra of the nanophosphors (Fig. 12b) for λ_em=614 nm, the lines determined by the ⁷F₀→⁵D₄, ⁵G₂, ⁵L₇ transitions have a much lower intensity compared with the excitation spectra of bulk phosphors (Fig. 11b). This confirms a relatively small number of Eu³⁺ ions taking part in the process of luminescence.

For the nanosample based on Ca₂Y_6.4Eu_1.6Si₆O_26–δ, the Eu²⁺ excitation spectrum (λ_em=443 nm) is almost the same as for the bulk sample, whereas for the nanosample based on Ca₂Gd_6.4Eu_1.6Si₆O_26–δ the ²e_g level becomes degenerated and additional excitation bands appear in the region of 260–320 nm (Fig. 12c). These bands are likely to correspond to the ⁸S_7/2→⁶P_J transitions of Eu²⁺ ion. Consequently, during the bulk → nano transition the ligand field around Eu²⁺ changes. This can be due (as mentioned above) to translational symmetry violation in the NP. The bonding of the 4f and 5d electrons weakens. The ²e_g level degenerates. The Eu²⁺ ion from the ⁶P_J state nonradiatively relaxes into the e_g state, from which the luminescence is generated (Fig. 12d).

Figure 12d presents a scheme for Eu²⁺ ion in configuration coordinates (r) for a nanophosphor based on sample III. The Eu²⁺ ion levels ⁸S_7/2 and ⁶P_J are shown. As a result of excitation of the ⁶P_J level, the Eu²⁺ ion exhibits a radiationless relaxation into the ²e_g state (a→a′). Then an a′→a′′→b′′ transition and radiation with λ^max=450 nm take place. At sufficient temperature, a radiationless transition a′→b′→с into the ⁸S_7/2 level is also possible.

Optical spectroscopy of selected germanates

Consider the variation in the forbidden band E_g width of samples IV during the transition from bulk powder to NP. Figure 10d demonstrates the absorption spectra of the bulk samples Ca₂La_6.4Eu_1.6Ge₆O₂₆ and NP based thereon. The absorption spectrum of the bulk sample has two bands with maxima at ~246 and 300 nm. These bands are due to a transition into the charge-transfer state for two types of optical centers formed by Eu³⁺ ions in the 4f and 6h positions. For Ca₂La_6.4Eu_1.6Ge₆O₂₆, the absorption band edge is located at ~350 nm, which corresponds to E_g=3.5 eV. In the region of the NP absorption boundary there is an extended exponential tail (Uhrbach tail) typical of amorphous media [31]. The Uhrbach tail does not allow the absorption boundary value to be estimated.

The luminescence and excitation spectra of the bulk sample Ca₂La_6.4Eu_1.6Ge₆O_26–δ are given in Fig. 13a and b. The spectral characteristics of this phosphor have been described in detail in work [14]. The Eu³⁺ ions form two types of optical centers at λ_max=577.4 and 578 nm. For the Ca₂La_6.4Eu_1.6Ge₆O_26–δ based nanosample, the luminescence spectrum changes considerably (Fig. 13c). Luminescence appears in the interval 450–750 nm (λ_ex=394 nm), which is due to Eu²⁺ ions. The mechanism of formation of Eu²⁺ ions, in our opinion, is analogous to that for the nanophosphors based on samples II and III: radiation reduction Eu³⁺→Eu²⁺. The luminescence spectrum of Eu²⁺ ions demonstrated in Fig. 13c is decomposed into two Gauss components (not shown in Fig. 13c) with λ_max=455.4 and 582.2 nm. Consequently, the Eu²⁺ ion forms two types of optical centers, probably, in the 4f and 6h positions. Figure 13d displays the nanophosphor excitation spectrum (λ_em=443 nm). A broad band (4f⁷–4f⁶5d¹ transition of Eu²⁺ ions) and lines corresponding to f–f transitions of Eu³⁺ ions are observed in the spectrum. The shape of the spectrum is indicative of the resonant interaction between Eu²⁺ and Eu³⁺ ions.

Fig. 13:

The PL spectra of phosphors Ca₂La_6.4Eu_1.6Ge₆O_26–δ (IV) (a), NP (c) and the PLE spectra (λ_em=614 nm) of samples IV (b), NP (λ_em=443 nm) (d).

The radiation of Eu³⁺ ions is superimposed on the luminescence spectrum of Eu²⁺ ions. For the ⁵D₀–⁷F₂ transition, the most intense lines are observed at 615 and 623 nm. For the bulk sample, the lines of this transition are observed at 613 and 620 nm. It is known that the electric dipole transition ⁵D₀→⁷F₂ is supersensitive to the site-symmetry of Еu³⁺ ion in the optical center. The shift of the transition lines in the NP towards higher energies and the redistribution of their relative intensities suggest a certain distortion of the local environment of Eu³⁺ ions in the nanophosphor compared with the bulk phosphor, which is likely to be due to partial polymerization of GeO₄ tetrahedra and to the formation of Ge−O−Ge bridges. The superposition of Eu²⁺ and Eu³⁺ ion radiation generates white luminescence of the phosphor with color coordinates x=0.313, y=0.320.