Figure 1(a) shows XRD patterns of some of the Y2Mo3O12:xSm3+ powders with concentrations of 0.300, 0.100, 0.010, and 0.000 mol, from top to bottom, and Fig. 1(b) shows the patterns of [Y2Mo3O12], [Sm2MoO5], and [Sm2Mo3O12] phases from JCPDS Nos. 28-1451, 31-1214, and 25-0748, respectively. For lower Sm3+ concentrations, the observed XRD patterns follow the reference [Y2Mo3O12] phase clearly. As we increase the amount of doped Sm3+ ions, however, the XRD patterns of x ≥ 0.100 are distinctively different from those of x ≤ 0.100. These observations suggest that a phase transition occurs around x = 0.100. For higher Sm3+ concentrations, new phases samarium molybdates [Sm2MoO5] and [Sm2Mo3O12] are observed along with [Y2Mo3O12] phase. The decrease of the peak intensities of [Y2Mo3O12] phase with an increase of Sm3+ amount means that the crystal structure depends on the doped ion concentrations.
In order to show that these phases are related to the hygroscopicity, we take TGA data. Figure 1(c) shows the weight variations of the samples according to the temperature. A gradual decrease of weight corresponding to the release of adsorbed water and/or crystal water with increasing temperature is observed, which means the reduction of hygroscopicity. In other words, the newly obtained phases look responsible for the reduction of hygroscopicity of the synthesized powders.
For the morphology of the microstructure of the Y2Mo3O12:xSm3+ powders, SEM images are recorded and represented in Fig. 2 with concentrations of (a) 0.000, (b) 0.005, (c) 0.050, and (d) 0.150 mol, respectively. All images show aggregated grains, and powders have various shapes with a few tens of mm sizes regardless of the doped amount. As mentioned in Experimental Section, we have applied a solid-state reaction method for the sample preparation procedures. Thus, we have powders of aggregated particles caused by the high-temperature sintering process. All the powder samples are found to form similar shapes via SEM images. The morphology of the obtained Y2Mo3O12:xSm3+ powders do not show any significant phase variations related to the hygroscopicity with the Sm3+ concentration.
Nano-SIMS is utilized in elemental analysis, with high spatial resolution, by tracing a small amount of elements and isotopes. With high lateral resolution and sensitivity, the possibility of imaging components based on the elemental and isotopic compositions of atomic and molecular ion fragments can be offered by using a nano-SIMS. The obtained information through nano-SIMS complements other analytical results obtained by using other types of SIMS experiments [36]. Also we can compare elemental distributions clearly at a fixed position of the sample because nano-SIMS can record five atomic mass images simultaneously with a high-resolution of 50 nm. Thus, we have applied this equipment to studying phosphors and reported the element compositional analysis results in phosphor material [37]. In this study, we apply nano-SIMS to our samples rather than energy-dispersive spectrometer (EDS).
We take the elemental acquisition images of oxygen, yttrium, molybdenum, and samarium in the Y2Mo3O12:xSm3+ samples and represent them in Fig. 3(a) ~ (d), respectively. Sm3+ concentrations are 0.000, 0.005, 0.050, and 0.200 mol, respectively, from top to bottom, in each element. These images are taken at high vacuum with ion beam etching the surface of the samples and the scale bar shown in each image has different scales depending on the element. We observe that elements are distributed throughout the samples, although the obtained images provide only qualitative information. Sm3+ is observed even in a lower amount-doped sample. At first, when we prepare the samples, we think that Sm3+ ions are substituted into Y3+ sites through the sintering procedure as in Ref. [18] and in Y2Mo4O15 case [24-28] because the ionic radii of these are quite close: 104 pm for Y3+ and 109.8 pm for Sm3+ [38]. However, samarium ions are observed at the molybdenum positions rather than yttrium positions, at higher Sm3+ concentrations. As mentioned in XRD pattern results, we observe that there are newly formed samarium molybdates such as [Sm2MoO5] and [Sm2Mo3O12] in higher concentration powders. Thus, the images obtained by using a nano-SIMS strongly support the discussion of new phases in XRD patterns for the higher Sm3+ concentration samples.
For the structural aspect, we take the Raman and the FT-IR spectra of the samples and represent them in Fig. 4. Figure 4(a) shows the Raman spectra of the Y2Mo3O12:xSm3+ samples with Sm3+ concentrations 0.300, 0.150, 0.050, and 0.005 mol, respectively, from top to bottom. All the Raman modes are observed nearly at the same positions regardless of the doped amount. The spectral shapes are distinctively different for x ≥ 0.100 and x ≤ 0.100 similar to the XRD patterns, which confirms the phase variations with the Sm3+ ion concentrations clearly. The modes below 150 cm-1 are identified as translational and librational vibrations of the polyhedra YO6 and MoO4 [9,10]. The modes in the region 300 ~ 400 cm-1 are bending vibrations of the MoO4 tetrahedra [10,11]. The modes around 476 cm-1 are not identified. The ones in 700 ~ 900 cm-1 and 900 ~ 1,000 cm-1 are asymmetric stretching and internal stretching vibrations of the MoO4 tetrahedra, respectively [10,11]. The modes around 48 and 97 cm-1 become stronger gradually with the doped Sm3+ amount, which indicates that translational and librational motions related to doped-REs make a significant contribution to the Y2Mo3O12 structure. The strong modes around 1100 cm-1 which appear in samples with higher concentrations are emission bands due to the transition 4G5/2 → 6H5/2, and this will be discussed later in the emission spectra part.
Figure 4(b) represents the FT-IR spectra of the Y2Mo3O12:xSm3+ with Sm3+ concentrations 0.300, 0.100, 0.010, and 0.000 mol, respectively, from top to bottom. The effects caused by atmospheric moisture and carbon dioxide, which are assumed to have been absorbed from air during sample preparation procedures, are removed from each measurement through background subtraction. Clearly we observe the phase variations according to the doped Sm3+ amount as in Raman spectra. We represent the regions 3,800 ~ 3000 cm-1 and 1,800 ~ 400 cm-1 because we are interested in the hygroscopicity of Y2Mo3O12 and the vibrations of polyhedra MoO4 and YO4. The broad bands in the range 3,700 ~ 3,100 cm-1 are the O-H stretching vibrations depending on the environment. The sharp bands in the range 1,750 ~ 1,530 cm-1 are H2O bending/O-H bond [39]. Weak intensity bands around 1,460 cm-1 and 1,400 cm-1 are C-OH and O-H bond, respectively. We use only oxides in the synthesis procedure, thus the C-OH is originated from the atmospheric carbon dioxide. The intensities of O-H bonds are comparable to usual materials. Thus, the atmospheric moistures do not affect the prepared samples a lot regardless of the hygroscopicity. The observed peak positions below 1,000 cm-1 region remain almost the same with the doped Sm3+ amount as in Raman case. The broad band around 1,000 ~ 700 cm-1 correspond to the vibrations of polyhedra MoO4 and YO6 [9-11]. The bands at about 867 cm-1 and 936 cm-1 (represented by blue vertical dotted lines) correspond to symmetric and asymmetric stretching vibrations of the MoO4 tetrahedra, respectively [11,18]. These show clear peak position shifts and their relative intensities, respectively, with the doped amount, which means that the influences of atmospheric moisture on the stretching vibrations of the MoO4 tetrahedra become weaker [11]. As mentioned earlier, we can think that these shifts are related to the phase variations of the materials with the Sm3+ amounts. The small peaks around 600 ~ 400 cm-1 are correspond to the vibrations of polyhedra MoO4 and YO6 [2].
From the Raman and the FT-IR spectra shown in Fig. 4, we observe the structural variations as in XRD patterns with the doped Sm3+ concentrations. At room temperature Y2Mo3O12 forms in orthorhombic structure with a space group Pba2 and undergoes a phase transition, ferroelectric-ferroelastic to paraelectric-paraelastic, at higher temperature [40]. Although we have synthesized the Y2Mo3O12 samples at higher temperature, we have taken the data at room temperature. Thus, we need more study about Y2Mo3O12 samples whether the reduction of hygroscopicity is directly related to the new phases appeared at higher concentrations.
To see the chemical states of the elements in the prepared powders, we take XPS spectra. Figure 5 represents the wide scan XPS spectra of the Y2Mo3O12:xSm3+ samples with concentrations of 0.300, 0.150, 0.050, and 0.000 mol, respectively, from top to bottom. In XPS spectra, the position on the kinetic energy scale means the difference between the photon excitation energy, and the spectrometer work function corresponds to a binding energy of 0 eV with reference to the Fermi level [41]. All the peak positions are calibrated with respect to carbon 1s at 284.6 eV. Peaks of elements such as Y, Mo, O, and Sm including KLL (oxygen Auger peak) are clearly identified [41]. Other small peaks such as Y 4p/O 2s around 23~24 eV, Mo 4p around 36 eV, Mo 4s around 63 eV, and Mo 3s around 506 eV are not designated in the figure for clarity, even though they are identified. The starting oxides are burned in the atmosphere during sintering process, which means that some of the oxygen atoms from the starting materials are used during the sintering process inside of the furnace. Also, we take XPS spectra in high vacuum without etching the samples. Hence, the observed carbon peaks are from atmospheric carbon dioxide attached to the sample powders during the pre-treatment procedure for the measurements.
To see the effect of Sm3+ introducing, we take the narrow scan XPS spectra of the Y2Mo3O12:xSm3+ samples and represent them in Fig. 6. Figure 6(a) ~ (d) shows the spectra of Y 3d orbitals, Mo 3d orbitals, O 1s orbitals, and Sm 3d orbitals, respectively. The dopant concentrations are the same as those in Figure 5 from top to bottom, except for the samarium shown in (d). After introducing Sm3+, the spin-orbit splitting of Y 3d and Mo 3d are 2.0 eV and 3.2 eV, respectively, and these remain unchanged. However, the binding energies of Y 3d5/2 and Mo 3d5/2 decrease from 158.0 eV to 157.8 eV and from 232.6 eV to 232.3 eV, respectively. The differences are 0.2 eV and 0.3 eV, respectively. For oxygen, the binding energy shift is 0.2 eV after the Sm3+ introducing and the orbital feature looks symmetric. That is, binding energy shifts are quite small even though Sm3+ ions are introduced. The differences are the same regardless of the concentration variations. Compared with Ce3+-doped Y2Mo3O12 in Ref. [18], where Ce3+ ions are substituted into the Y3+ sites, we have new samarium molybdate phases [Sm2MoO5] and ]Sm2Mo3O12] at higher Sm3+ concentration samples. These newly-formed materials seem to cause the reduction of the hygroscopicity in the host material, Y2Mo3O12. Also, Sm3+ ions are clearly observed in the samples as shown in Fig. 6(d) even though the doped amounts are very small. The photoelectron lines of the observed elements Y, Mo, O, and Sm in Figs. 5 and 6 are schematically summarized in Fig. 6(e). The values of the line positions reflect the energies of the elemental peaks observed.
For optical properties, we observe quite similar features in the UV-visible absorption spectra and represent absorption spectra of the Y2Mo3O12:xSm3+ samples in Figure 7 with the dopant concentrations 0.200, 0.150, and 0.100 mol, from top to bottom. A strong absorption band in the NUV region, corresponding to the 6H5/2 → 4F7/2 transition from doped Sm3+, is observed around 405 nm. Other identified absorption bands are associated with electronic transitions such as 6H5/2 → (6P,4P)5/2 around 420 nm, 6H5/2 → 4G9/2 around 439 nm, 6H5/2 → 4I15/2 around 452 nm, 6H5/2 → 4I13/2 around 464 nm, 6H5/2 → 4I11/2 around 482 nm, and 6H5/2 → 4I9/2 around 491 nm [32]. From these bands, we observe that the luminescent features are based on the direct transitions from the ground state 6H5/2 to the excited states of Sm3+ [32].
Figure 8(a) represents the emission spectra taken by excitation of 6H5/2 → 4F7/2 transition at 405 nm in the Y2Mo3O12:xSm3+ samples. We show the samples of Sm3+ concentrations 0.100, 0.050, and 0.010 mol for clarity. The identified emission bands are 4G5/2 → 6H5/2 around 566 nm, 4G5/2 → 6H7/2 around 603 nm, and 4G5/2 → 6H9/2 around 649 nm, showing the spectral features of the intra 4f-4f transitions of Sm3+. Each band consists of several weak intensity peaks due to the splitting of the 4G5/2 state arising from the Stark effect [42], which makes the spectral shapes slightly complex. As usual, the main emission band of Sm3+ ion is red-orange-emitting part of spectra and corresponds to 4G5/2 → 6H7/2 transition around 603 nm.
We are interested in the actual intensities of the strongest emission band, 6H5/2 → 4F7/2, of the Y2Mo3O12:xSm3+ samples. Thus, we plot the variation of actual intensity with the doped amount in the inset of Fig. 8(a). Since we have prepared all samples simultaneously under the same conditions (sintering procedure and preparing pellets, etc) except for the dopant concentrations, only the Sm3+ concentration can affect the emission characteristics. As shown in the inset, the emission intensity increases upto around 0.100 mol and then decreases, i.e. the concentration quenching is active when the Sm3+ concentration becomes higher than 0.150 mol in the luminescence feature. And the quenching of the emission intensity is originated from a decrease in the number of optically active Sm3+ ions in the samples, which means that dopant pairs and/or clusters are formed at higher concentrations of the dopant.
The observed energy levels and visible emission levels shown in Figs. 7 and 8(a) are schematically drawn in Fig. 8(b) by following the calculated energy levels [43]. From this figure, we can find that Sm3+ ions are excited from the ground state 6H5/2 to the excited state 4F7/2 with an excitation of 405 nm, then the ions at 4F7/2 level relax to the 4G5/2 level through the non-radiative relaxation process. Thus, the emission bands around 566 nm, 603 nm, and 649 nm arising from the transitions 4G5/2 to 6H5/2, 6H7/2, and 6H9/2, respectively, are observed.
The intensity ratio between electric and magnetic dipole transition gives a measure for the symmetry of local environment of the trivalent 4f ions [37]. Magnetic dipole transitions obey the selection rule of DJ = 0 and +1, where J is the angular momentum, and electric dipole transitions obey the selection rule of DJ ≤ 6 unless J or J’ = 0 when DJ = 2,3,6. Thus, the transition 4G5/2 → 6H5/2 is a magnetic dipole transition because this satisfies the selection rule of DJ = +1, while the transition 4G5/2 → 6H9/2 is an electric dipole transition. The transition 4G5/2 → 6H7/2 is a partly magnetic and partly forced electric dipole transition [41]. The transition 4G5/2 → 6H11/2 is not observed in this study because of its forbidden feature DJ = 3.
Consider the ratio between the intensities of the magnetic dipole transition 4G5/2 → 6H7/2 to the electric dipole transition 4G5/2 → 6H9/2, which is a measure of the local symmetry of the coordination polyhedron of the Sm3+ ions. Hence, the larger value of this ratio means more distortion from the inversion symmetry [37]. The obtained values are almost constant between 1.33 and 1.48, which means that the Sm3+ ions are embedded in samarium molybdates [Sm2MoO5] and [Sm2Mo3O12] structures rather than in the distorted cation environment. In our previous result for CaTiO3:Sm3+ phosphors [37], the calculated values are from 1.40 to 1.53, which implies that the environment of Sm3+ ions are substituted into the host material and remain in symmetric state. Also the observed spectral widths remain similar values varying from 8.55 ~ 9.23 nm except the lower concentration samples. Therefore there is no significant structural environment change around Sm3+ ions with increase of the concentration, except the formation of new phases.
As mentioned in Introductory section, Y2Mo3O12 is known as a NTE material having disadvantage such as hygroscopicity. From the previous discussion, we have observed a phase transition for the samples of the concentrations x ≥ 0.100 and a concentration quenching for the samples of the higher concentrations. We can think that around 0.100 mol of Sm3+-doped Y2Mo3O12 is a hygroscopicity-reduced sample with highly luminescent feature. There are many studies in reducing the hygroscopicity by using the ion-substitution methods and the coating methods. As shown in this study, RE3+-doped Y2Mo3O12 can be a candidate material for reduced hygroscopicity as well as photonic application. We need more studies for the newly formed samarium molybdates whether these can also be host materials for photonic applications.