Synthesis and thermal, optical and magnetic properties of new Mn²⁺-doped and Eu³⁺-co-doped scheelites

Abstract

A series of Mn²⁺-doped and Eu³⁺-co-doped calcium molybdato-tungstates, i.e., Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x (0 < x ≤ 0.2222 when y = 0.0200 and 0 < y ≤ 0.0667 when x = 0.1667, ⌷ represents vacancy) materials were successfully synthesized via high-temperature annealing. XRD results confirmed the formation of single, tetragonal scheelite-type phases (space group I4₁/a). A change in both lattice constants (a and c), lattice parameter ratio c/a and progressive deformation of MoO₄/WO₄ tetrahedra with increasing Eu³⁺ as well as Mn²⁺ contents were observed. The melting point of doped materials is lower than the melting point of pure matrix, i.e., CaMoO₄. New materials exhibit strong absorption in the UV range. They are insulators with the optical direct band gap (E_g) higher than 3.50 eV. The E_g values nonlinearly change with increasing dopants concentrations. EPR measurements allowed to establish the nature of magnetic interactions among Mn²⁺ ions. Additionally, EPR spectra were sensitive on both parameters: Mn²⁺ and Eu³⁺ concentration.

Introduction

Metal molybdates and tungstates were proved to be excellent host lattice of luminescent materials due to their high thermal stability and chemical tolerance. These compounds show many different types of structures, i.e., scheelite (I4₁/a, No. 88), pseudo-scheelite (Pnma, No. 62), wolframite (P2/c, No. 13), zircon (I4₁/amd, No. 141) and fergusonite (C2/c, No. 15) [1,2,3,4,5,6,7]. Most divalent metal molybdates and tungstates take scheelite-type or scheelite-related structure. Scheelite-type A(Mo,W)O₄ compounds possess eight symmetry elements, a body-centered unit cell and four molecules per one unit cell [1,2,3, 6, 7]. Each A-site divalent ion is surrounded by eight oxygen anions and forms AO₈ dodecahedra [1,2,3, 6, 7]. Each Mo⁶⁺ or W⁶⁺ ion is coordinated to four oxygen ions forming (Mo,W)O₄ tetrahedron [1,2,3, 6, 7].

Materials doped with trivalent rare-earth (RE) ions have been studied for many applications, such as lasers, solid state lightings, display panels and optical amplifiers [8]. Among of trivalent RE ions, Eu³⁺ ion is characterized by the lowest excited level ⁵D₀ of 4f⁶ configuration and displayed mainly very sharp red light emission centered at 615 nm ascribed to the ⁵D₀ → ⁷F₂ transition [9]. Divalent manganese ion plays an important role in some inorganic phosphors. It can produce a broad emission due to the transition from ⁴T₁ to ⁶A₁ level. The emission color strongly depends on a crystal field strength as well as coordination number of Mn²⁺, and it varies from green (tetrahedral coordination of manganese ions) to red (octahedrally coordinated Mn²⁺) [10,11,12,13,14]. Thus, simultaneously doping the host material with Eu³⁺ and Mn²⁺ ions allows to energy transfer from Mn²⁺ to Eu³⁺ ions and improve the luminescent efficiency of doped samples [15, 16].

In our earlier papers, we have focused on three scheelite-type matrices, i.e., CdMoO₄, PbMoO₄ and PbWO₄. Our studies have shown the existence of new scheelite-type solid solutions described by the following chemical formulas: (Cd,Pb)_1−3x⌷_xRE_2xMoO₄ and (Cd,Pb)_1−3x⌷_xRE_2x(MoO₄)_1−3x(WO₄)_3x, where RE = Pr–Yb [17,18,19,20,21,22,23,24,25,26]. The substitution of divalent Cd²⁺ or Pb²⁺ ions by trivalent rare-earth ones is connected with a charge-compensating defect. Charge balance is achieved by the formation of A-site vacancies, which is denoted by us as ⌷, i.e., 3 Cd²⁺(Pb²⁺) → 2 RE³⁺ + ⌷. Presence of these vacancies is seen in the dependence of wavenumber and integral intensity of some electronic absorption, IR and Raman bands on increasing concentration of the point defects in the RE³⁺-doped materials. The presence of vacancies results also in significant deformation of the crystal lattice, particularly around RE³⁺ ions, and which results in the emission of broadbands associated with f–f transitions in RE³⁺ ions [19,20,21]. The RE³⁺-doped molybdates or molybdato-tungstates are promising laser materials, phosphors and good microwave dielectric microcrystalline ceramics for band-pass filters, resonators and antenna switches (mobile and satellite communication) [19,20,21,22].

In this work, microcrystalline samples of new Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x solid solution with different Mn²⁺ (y = 0.0066; 0.0200; 0.0333; 0.0667) as well as Eu³⁺ (x = 0.0005; 0.0025; 0.0050; 0.0098; 0.0283; 0.0455; 0.0839; 0.1430; 0.1667; 0.1970; 0.2059; 0.2222) contents were successfully obtained by a high-temperature sintering. Their structure, thermal stability and optical properties were examined. Electron paramagnetic resonance spectroscopy (EPR) was used to determine the types of magnetic interactions and the local symmetry of Mn²⁺ ions.

Experimental

Synthesis of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x

New microcrystalline Mn²⁺-doped and Eu³⁺-co-doped calcium molybdato-tungstates with the following formula of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x were obtained by the two-step synthesis, in which both the steps consist of high-temperature sintering which is very often used in a preparation of many molybdates and tungstates [27,28,29,30,31,32]. The following reactants were used in the first step of synthesis: CaCO₃, MnO, MoO₃, Eu₂O₃ and WO₃ (all raw materials of high-purity grade min. 99.95%, Alfa Aesar or Fluka, and without thermal pre-treatment). Calcium molybdate (CaMoO₄), manganese molybdate (MnMoO₄) as well as europium tungstate Eu₂(WO₄)₃ were obtained according to the procedure used by us in previous studies [26, 32]. In the next step of synthesis, we prepared 15 ternary mixtures containing MnMoO₄, Eu₂(WO₄)₃ and CaMoO₄. The concentrations of manganese molybdate and europium tungstate in each initial MnMoO₄/Eu₂(WO₄)₃/CaMoO₄ mixture are shown in Table 1. The initial mixtures, not compacted to pellets, were heated in air, in ceramic crucibles, in several 12-h annealing stages, and at temperatures in the range of 1173–1473 K. After each heating period, obtained materials were cooled down to ambient temperature, weighed, ground in an agate mortar, followed by examination for their composition by XRD method. The mass change control of doped samples showed a slight mass loss observed for each material that did not exceed the value of 0.26%. This observation suggests that a synthesis of new materials runs practically without their mass change.

Table 1 Lattice constants, experimental and calculated X-ray density, and determined optical direct band gap, E_g of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x

Methods

Powder X-ray diffraction patterns of new doped materials were collected within the 2θ range of 10°–100° and under ambient conditions by using an EMPYREAN II diffractometer (PANalytical) and CuKα1,2 radiation (the scanning step 0.013°). High Score Plus 4.0 software was used to analyze registered XRD patterns. Lattice parameters were calculated using the procedure described and used in our previous works [17,18,19,20,21,22,23,24,25,26, 32]. Quantachrome Instruments Ultrapycnometer (model Ultrapyc 1200 e) was used to measure the density of each sample. As the pycnometric gas, nitrogen (purity 99.99%) was used. DTA-TG studies of manganese molybdate and europium tungstate were carried out on a TA Instruments thermoanalyzer (model SDT 2960) at the heating rate of 10 K min⁻¹, and in the temperature range from 293 to 1473 K (the air flow 110 mL min⁻¹). Melting point of some Mn²⁺-doped and Eu³⁺-co-doped ceramic materials was determined by using a pyrometric method. Selected samples, previously compressed into pellets, were heated in a resistance furnace, and their image was simultaneously recorded by a camera. Melting point of a sample was determined at the time when its image was not seen in a camera. IR spectra were recorded within the spectral range of 1500–200 cm⁻¹ on a Specord M-80 spectrometer (Carl Zeiss Jena) using pellets with potassium bromide. UV–Vis reflectance spectra were recorded within the range of 200–1000 nm using JASCO-V670 spectrophotometer containing an integrating sphere. EPR spectra of Mn²⁺-doped and Eu³⁺-co-doped samples were determined by an X-band Bruker ELEXSYS E 500 CW spectrometer operating at 9.5 GHz with 100 kHz magnetic field modulation. Temperature dependence of the EPR spectra of doped materials within the 78–300 K temperature range was recorded using Oxford Instruments ESP liquid nitrogen and helium cryostat.

Results and discussion

XRD analysis

The powder diffraction patterns of pure CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x ceramics with different contents of europium(III) and manganese ions, i.e., when x = 0.0005; 0.0025; 0.0050; 0.0098; 0.0283; 0.0455; 0.0839; 0.1430; 0.1667; 0.1970; 0.2059; and 0.2222 for constant Mn²⁺ content (y = 0.0200) as well as when y = 0.0066; 0.0200; 0.0333; and 0.0667 for constant Eu³⁺ concentration (x = 0.1667) are shown in Fig. 1a, b. XRD analysis shows the powder diffraction patterns of Mn²⁺-doped and Eu³⁺-co-doped calcium molybdato-tungstates consisted of diffraction lines which can be attributed to scheelite-type framework. No impurity phases, i.e., initial reactants, oxides (Eu₂O₃, MnO, CaO, MoO₃ and WO₃), and other europium tungstates or molybdates were observed with increasing Eu³⁺ concentration only up to x = 0.2222 (40.00 mol% of Eu₂(WO₄)₃ in initial MnMoO₄/Eu₂(WO₄)₃/CaMoO₄ mixtures when y = constant). No additional phases we have also observed on the powder diffraction patterns of doped materials when Mn²⁺ content was only up to y = 0.0667 (10.00 mol% of MnMoO₄ when x = 0.1667). The observed peaks attributed to scheelite-type structure shifted toward lower 2θ angle with increasing Eu³⁺ concentration (Fig. 1a) or toward higher 2θ angle with increasing Mn²⁺ content (Fig. 1b). All observed diffraction lines were successfully indexed to the pure tetragonal scheelite-type structure with space group I4₁/a (No. 88, CaMoO₄—JCPDs No. 04-013-6763). This fact confirmed the formation of new solid solution of MnMoO₄ and Eu₂(WO₄)₃ in CaMoO₄ matrix. The diffraction patterns of samples comprising initially above 40.00 mol% Eu₂(WO₄)₃ in MnMoO₄/Eu₂(WO₄)₃/CaMoO₄ mixtures (not presented in the paper) revealed simultaneously peaks which can be ascribed to the saturated solid solution, i.e., Ca_0.3134Mn_0.0200⌷_0.2222Eu_0.4444(MoO₄)_0.3334(WO₄)_0.6666 (x = 0.2222) as well as the diffraction lines characteristic of Eu₂(WO₄)₃. It means that the solubility limit of europium tungstate in structure of CaMoO₄ when the concentration of MnMoO₄ equals 3.00 mol% is not higher than 40.00 mol%. On the other hand, the diffraction patterns of samples with an initial content of manganese molybdate above 10.00 mol% showed diffraction lines attributed to Ca_0.4332Mn_0.0667⌷_0.1667Eu_0.3334(MoO₄)_0.4999(WO₄)_0.5001 (y = 0.0667) as well as the diffraction lines characteristic of MnMoO₄. It means that the solubility limit of manganese molybdate in structure of CaMoO₄ when the initial concentration of Eu₂(WO₄)₃ equals 25.00 mol% is not higher than 10.00 mol%.

Fig. 1

XRD patterns of CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x ceramics in the 2θ range of 10°–60°. Powder XRD patterns of 0 < x ≤ 0.2222 and y = 0.0200 (a) and powder XRD patterns of 0 < y ≤ 0.0667 and x = 0.1667 (b)

The lattice constants calculated on the basic of XRD data for CaMoO₄ as well as for Mn²⁺-doped and Eu³⁺-co-doped ceramics are presented in Table 1. Both unit cell parameters (a and c) of doped materials gradually increased with increasing Eu³⁺ concentration when Mn²⁺ content in samples was constant (y = 0.0200). This is not a typical situation because bigger Ca²⁺ ions (112 pm for CN = 8) in the matrix were substituted by much smaller Eu³⁺ ones (106.6 pm for CN = 8) [33]. We have already observed this phenomenon in other RE³⁺-doped and vacancied materials, i.e., Cd_1–3x⌷_xDy_2xMoO₄ [22]. The lattice parameters of all doped samples decreased gradually with increasing Mn²⁺ concentration for the materials with constant Eu³⁺ content (Table 1). It is a consequence of the substitution of big Ca²⁺ ions by much smaller Mn²⁺ ones (96 pm for CN = 8). In all scheelite-type molybdates and tungstates, Mo⁶⁺ and W⁶⁺ ions are tetrahedrally coordinated by oxygen ones and their ionic radii are 41 and 42 pm, respectively [33]. Thus, the substitution of Mo⁶⁺ by W⁶⁺ observed in Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x materials did not cause any visible changes in both lattice constants. Only a parameter as well as the volume of tetragonal cell (V) calculated for doped ceramics (when Eu³⁺ concentration was increasing and Mn²⁺ content was constant) obeys the Vegard’s law, i.e., they are nearly linear functions of x parameter (Fig. 2). The c parameter changes clearly nonlinearly with increasing Eu³⁺ content in the samples (Fig. 3, Table 1). In contrast, the lattice parameters (both a and c) as well as the volume of cell calculated for doped samples (when Mn²⁺ content was increasing and Eu³⁺ concentration was constant) obey the Vegard’s law when 0 < y ≤ 0.0667 (Figs. 4, 5). We also calculated the lattice parameter ratio c/a for two series of Mn²⁺-doped and Eu³⁺-co-doped ceramics (Figs. 3, 5). This parameter clearly nonlinearly changes with increasing x (Fig. 3) as well as practically linearly decreases with increasing y (Fig. 5). It means that in whole homogeneity range of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x solid solution we have observed a deformation of tetragonal scheelite-type cell of each doped sample in comparison with the tetragonal cell of pure CaMoO₄. The experimental density values determined for each sample under study are given in Table 1. The density of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x samples almost linearly increases with increasing Eu³⁺ as well as Mn²⁺ content (Table 1).

Fig. 2

Linear dependences of a parameter and volume V versus x parameter of 0 < x ≤ 0.2222 and y = 0.0200

Fig. 3

Dependences of c parameter and c · a⁻¹ versus x parameter of 0 < x ≤ 0.2222 and y = 0.0200

Fig. 4

Linear dependences of both a and c lattice constants versus x parameter of 0 < y ≤ 0.0667 and x = 0.1667

Fig. 5

Linear dependences of volume V and c · a⁻¹ versus x parameter of 0 < y ≤ 0.0667 and x = 0.1667

Thermal stability of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x materials

Appropriate thermal studies of new Mn²⁺-doped and Eu³⁺-co-doped microceramics have been preceded by DTA experiments for some initial reactants, i.e., MnMoO₄ and Eu₂(WO₄)₃, due to the fact that information on their thermal properties is not completely, contrary or unknown. Figure 6a shows DTA curve of MnMoO₄ recorded during controlled heating up to 1473 K. The single and very sharp endothermic effect with its onset at 1412 K is due to congruent melting of monoclinic manganese molybdate (a = 4.818(1) Å, b = 5.759(1) Å, c = 4.965(1) Å, β = 90.82(1)°, Z = 2, space group P2/c [34]). Only one endothermic peak is observed on DTA curve recorded during controlled heating of Eu₂(WO₄)₃ (Fig. 6b). This effect with its onset at 1426 K is connected with congruent melting of monoclinic europium tungstate (a = 7.676(3) Å, b = 11.463(3) Å, c = 11.396(5) Å, β = 109.63(4)°, Z = 4, space group C2/c [35]). In contrary to isostructural Gd₂(WO₄)₃, europium tungstate does not show polymorphism [35]. Crystallization process of Eu₂(WO₄)₃ from a melt observed during controlled cooling (exothermic effect) is quite significantly shifted toward lower temperatures, and it starts at 1408 K. Pure matrix of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x solid solution, i.e., CaMoO₄ melts congruently at 1753 K and this temperature has already been determined in our previous studies by using a pyrometric method [32]. Melting point of some Mn²⁺-doped and Eu³⁺-co-doped materials has also been determined by using the same method. These values are lower than the melting point of CaMoO₄, and they are as follows: 1653 K (x = 0.0455, y = 0.0200), 1578 K (x = 0.1430, y = 0.0200), 1528 K (x = 0.1970, y = 0.0200), 1503 K (x = 0.2222, y = 0.0200) and 1518 K (x = 0.1667, y = 0.0667). Figure 7 shows powder diffraction patterns of some initial Mn²⁺- and Eu³⁺-doped materials as well as resulting melts obtained after their pyrometric studies. According to XRD analysis, the patterns of each melt consisted of only diffraction lines which can be attributed to a scheelite-type framework. However, the observed peaks shifted toward lower or higher 2Θ angle in comparison with diffraction lines recorded on the powder XRD patterns of initial Mn²⁺- and Eu³⁺-doped materials.

Fig. 6

DTA curves of MnMoO₄ (a) and Eu₂(WO₄)₃ (b)

Fig. 7

XRD patterns of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x ceramics in the 2θ range of 10°–60° (x = 0.0455; x = 0.1430; x = 0.2059; x = 0.2222 and y = 0.0200 and x = 0.1667 and y = 0.0667) and their molten forms

IR spectra

Figures 8 and 9 show the IR spectra of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x microceramics with different contents of Eu³⁺ as well as Mn²⁺ ions, respectively. Molybdates and tungstates with scheelite-type structure are characterized by molecules containing XO₄ (X = Mo or W) tetrahedra showing strong covalent X–O bonds. For scheelite-type materials, the frequencies active in IR are observed within the spectral region of 900–700 cm⁻¹ (ν₁(A₁)—symmetric—as well as ν₃(F₂)—asymmetric stretching modes) and 450–250 cm⁻¹ (ν₂(E₁)—symmetric—and ν₄(F₂)—asymmetric bending modes) [36,37,38,39,40,41]. Figures 8 and 9 show IR spectra of pure CaMoO₄ and Mn²⁺-doped and Eu³⁺-co-doped materials. Strong and narrow absorption bands observed in the IR spectra of CaMoO₄ at 868 and 810 cm⁻¹ can be assigned to the stretching modes of Mo–O bonds in MoO₄ tetrahedra [36,37,38,39,40,41]. The bands with their maximum at 427 and 324 cm⁻¹ can be related to the binding modes of Mo–O bonds in MoO₄ tetrahedra [36,37,38,39,40,41]. The experimental results for calcium molybdate confirm a presence of only regular molybdate tetrahedra occupying sites of a cubic point symmetry T_d characteristic for scheelite-type structure [36,37,38,39,40,41]. The IR spectra of doped ceramics show differences in an intensity and a width of observed absorption bands compared to those ones registered for pure matrix. They are wide and exhibit very low intensity. Additionally, with increasing content of Eu³⁺ in doped materials, the absorption bands moved toward higher wave numbers (Fig. 8). This phenomenon is clearly visible for bands due to bending vibrations of Mo–O bonds in MoO₄ tetrahedra (Figs. 8, 9). It suggests the presence of tungstate tetrahedra in a crystal lattice of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1–3x(WO₄)_3x. For the samples when x ≥ 0.1430 we can observe additional absorption bands within the region of 940–915 cm⁻¹ (Figs. 8, 9). This fact suggests the presence of other types of MoO₄ and WO₄ tetrahedra, i.e., distorted molybdate and tungstate anions. Deformation of the shape of both anions, i.e., changes in X–O and O–X–O bond lengths and/or angles in XO₄ tetrahedra, compared to the shape of regular ones is a result of a presence of vacancies near corners of MoO₄ or WO₄ polyhedra. This deformation of regular tetrahedral shape of molybdate anions had already been observed in vacancied scheelite-type cadmium molybdates doped with RE ions, i.e., Cd_0.25⌷_0.25RE_0.50MoO₄, where RE = Pr, Nd, Sm–Dy as well as for Cd_1−3x⌷_xGd_2xMoO₄ [42, 43].

Fig. 8

IR spectra of CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x (x = 0.0005; x = 0.0025; x = 0.0098; x = 0.0455; x = 0.1430; x = 0.1970; x = 0.2222 and y = 0.0200)

Fig. 9

IR spectra of CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x of 0 < y ≤ 0.0667 and x = 0.1667

Optical properties of Mn²⁺-doped and Eu³⁺-co-doped ceramics

Optical energy band gap (E_g) of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x ceramics was determined by the method first used by Kubelka and Munk [44] and next applied by many authors [22,23,24,25, 32, 45,46,47,48,49]. This method is based on a transformation of diffuse reflectance spectra to estimate band gap energy values. In a case of parabolic band gap structure, an optical energy band gap as well as an absorption coefficient of material (α) can be calculated using the equation proposed by Tauc et al. [50, 51]:

$$ ahn = \, A(hn - E_{\text{g}} )^{{\text{n}}} , $$

(1)

where h is the Planck constant, ν is the light frequency, A is the proportionally coefficient characteristic for material under study, and n is the constant determined by an electronic transition type, and it can reach the following values: 1/2, 3/2, 2 and 3 for allowed directly, forbidden directly, allowed indirectly and forbidden indirectly transition, respectively [50, 51]. As it is known from literature information, all alkaline earth metal molybdates and tungstates with the tetragonal scheelite-type structure exhibit an optical absorption spectrum governed by direct absorption transition (n = 1/2). It means that an electron located in a maximum energy state in a valence band reaches a minimum energy state in a conduction band under the same point in the Brillouin zone [52,53,54,55,56]. Thus, the optical band gap of materials can be determined from the plots of (αhν)² as a function of photon energy (hν) by extrapolating the linear portion of the curve to intersect the hν axis at zero absorption [22,23,24,25, 32, 45,46,47,48,49, 52,53,54,55,56]. Figures 10 and 11 show UV–Vis absorption spectra of pure CaMoO₄ and Mn²⁺- and Eu³⁺-doped materials recorded within the spectral range of 200–800 nm. All samples show energy absorption ability in the UV region, and the absorption edge is near to 325 or 350 nm. The plots (αhν)² versus hν and the extrapolated direct energy band gap E_g values are shown in Figs. 12 and 13 as well as are given in Table 1, respectively. The E_g values of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x ceramics are lower than the adequate value for pure matrix, and they nonlinearly change with increasing Eu³⁺ content when Mn²⁺ concentration was constant. We have also observed the same nonlinear dependence of E_g with increasing Mn²⁺ content when Eu³⁺ concentration was constant. The lowest value of E_g (3.46 eV) was found for the sample with x = 0.1970 and y = 0.0200 (Table 1).

Fig. 10

UV–Vis absorption spectra of CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x (x = 0.0005; x = 0.0025; x = 0.0098; x = 0.0455; x = 0.1430; x = 0.1970; x = 0.2222 and y = 0.0200)

Fig. 11

UV–Vis absorption spectra of CaMoO₄ and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x of 0 < y ≤ 0.0667 and x = 0.1667

Fig. 12

Plots of (α·hν)² versus hν of CaMoO₄ (inset) and Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x (x = 0.1970; y = 0.0200) and determined band gap energy

Fig. 13

Plots of (α·hν)² versus hν of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x (y = 0.0667; x = 0.1667 (inset) and y = 0.0066; x = 0.1667) and determined band gap energy

EPR results

A group of doped materials with formula of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x with different combination of x and y parameters were studied using the EPR technique. Resonance spectra detected at temperature range of 78–300 K revealed an existence of multiline signal centered near 340 mT magnetic field (Figs. 14, 15). Observed resonance signal could be ascribed to the Mn²⁺ magnetic ions. Manganese metal is characterized by electronic spin value S = 5/2, giving usually a single asymmetric line in EPR experiment. But mentioned ion possesses nuclear spin value I = 5/2, resulting with characteristic splitting of the resonance signal to form six narrow lines.

Fig. 14

EPR spectra of Mn²⁺ ions detected at different manganese content represented by y parameter

Fig. 15

EPR spectra of Mn²⁺ ions detected at different europium content represented by x parameter

Figure 14 presents an arrangement of chosen EPR spectra detected at c.a. 80 K for samples with various manganese ion doping represented by y parameter, but with constant europium ion concentration x = 0.1667.

As could be seen in Fig. 14, for samples with lower Mn²⁺ doping the obtained EPR signal possesses six quite well-resolved lines originated from hyperfine interactions between electronic and nuclear magnetic moment of Mn²⁺ ions. But such model works properly only for low-density Mn²⁺ ions. With increasing manganese content, the arrangement of Mn²⁺ ions starts to be systematically disturbed, finally giving only single wide resonance line for samples with y parameter above 0.0200.

Figure 15 shows the evolution of EPR spectra under europium doping variation, represented by x parameter, while manganese ion concentration is maintained at constant value y = 0.0200. As could be seen, relatively low Eu³⁺ concentration gives no influence on the manganese ions. Thus, a well-resolved hyperfine structure of six narrow lines is observed. Increase in europium doping leads to situation, where resonance signal is systematically overleaped. For x value 0.2222 and higher, the sixfold arrangement is invisible and the manganese signal consists of only one wide line.

Influence of y parameters on the EPR signal seems to be very expectable. If concentration of manganese is maintained below some critical value, magnetic system of Mn²⁺ ions could be described with model of isolated magnetic ions with Zeeman interaction, disturbed only by hyperfine interaction of Mn²⁺ nuclear spin. Increasing concentration of manganese leads to perturbation of the simple magnetic model. Additional interactions, mainly exchange and dipolar nature, result in widening and overleaping the overall resonance signal.

Coexistence of Eu³⁺ ions apparently gives no influence on the magnetic system of Mn²⁺, as trivalent europium is non-magnetic. But results shown in Fig. 15 indicate progressive disturbance of manganese system by the increase in europium ion concentration. Due to comparable ionic size, doped Eu³⁺ ions are incorporated into the Mn²⁺ positions. It leads to generation of charge unequilibrium areas (vacancies) localized near such manganese position. Obviously, number of vacancies directly depends on the level of Eu³⁺ concentration, and it increases with increasing x parameter. Thus, vacancy states are denoted in the proposed formula as ⌷_x.

With higher number of vacancies, their localization near Eu³⁺ ions becomes less restrict. Due to charge compensation effect, vacancies could appear in other places, also near existed Mn²⁺ ions. Interaction between vacancies and manganese disturbs the Mn²⁺ electronic structure. Thus, with increasing x parameter, the EPR signal of Mn²⁺ ions starts to overleap, as is visible in Fig. 15.

Temperature dependence of the EPR signal intensity usually allows to describe the nature of interactions between responsible magnetic particles. All studied samples revealed decrease in the signal intensity (I_EPR) with increasing temperature according to the Curie–Weiss law:

$$ I_{\text{EPR}} = C/ \, (T - \theta ), $$

(2)

where C is the parameter connected with concentration of magnetic ions and θ is the parameter describing the sign of magnetic interactions.

According to Eq. 2, magnetic arrangement created by manganese ions revealed a strong antiferromagnetic (AFM) interaction proved by significant negative value of θ parameter for all samples. Generally, with increasing concentration of Mn²⁺ or Eu³⁺ the strength of AFM interaction increases, too. Figure 16a shows temperature dependence of the resonance signal and adequate fitting done using Eq. 2 for sample: Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x with x = 0.0283 and y = 0.0200. The sign of magnetic interactions could be also concluded from drawing of I_EPR.T product as a function of temperature (Fig. 16b). Positive loops of calculated values indicate negative value of θ parameter, i.e., domination of AFM interactions among responsible Mn²⁺ ions.

Fig. 16

a Intensity of the EPR signal as a function of temperature for sample x = 0.0283, y = 0.0200 (points) and fitting the result using Eq. 1 (line). b Dependence of I_EPRT product as a function of temperature for sample x = 0.0005 and y = 0.0200

As we mentioned earlier, magnetic properties of manganese ions could be described by Zeeman and hyperfine interactions. Enriched spin-Hamiltonian, including additional interactions with crystal field (fine interactions), should has the following form:

$$ H_{\text{s}} = \mu_{\text{B}} {\text{BgS}} + {\text{SDS}} + {\text{IAI}} + \sum {B_{\text{q}}^{k} O_{\text{q}}^{k} } , $$

(3)

where the first term represents the Zeeman interaction, the second zero-field splitting, the third hyperfine interaction and the last one higher-order fine interactions represented by Stevens operators.

In the case of axial or rhombic symmetry near magnetic particle, the zero-field splitting term could be expressed by standardized form using scalar D value and E parameter represented rhombic distortion near responsible magnetic ions. Thus, useful form of spin-Hamiltonian describing Mn²⁺ ions has the form:

$$ H_{\text{s}} = \mu_{\text{B}} {\text{BgS}} + D\left[ {S_{\text{z}}^{2} - \frac{1}{3}S(S + 1)} \right] + E(S_{\text{x}}^{2} - S_{\text{y}}^{2} ) + {\text{AIA}} + \sum {B_{\text{q}}^{{\text{k}}} O_{\text{q}}^{{\text{k}}} } . $$

(4)

Equation 4 is employed to simulate the EPR spectra in order to establish: g, D, E, A and B ^k_q values. Simulation was done using EPR–NMR program [57]. As we already mentioned, the proposed model seems to be proper only for lower x and y values. Figure 17 presents the best comparison between experimental and simulated spectra using Eq. 4 for sample x = 0.0098, y = 0.0200. As could be seen, adjustment is quite good. Spin-Hamiltonian values calculated for this case are given in Table 2. Other B ^k_q parameters are negligible, E = 0. The proposed model is not perfect, but values g_i and A_i are typical for Mn²⁺ ions. From the other hand, zero-field splitting parameter with D_x = D_y indicates an axial symmetry near manganese ions crystallographic positions.

Fig. 17

EPR spectrum of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x with x = 0.0098, y = 0.0200 (black line) and spectrum simulated using Eq. 3 (red line)

Table 2 Spin-Hamiltonian values calculated by EPR-NMR program for model represented by Eq. 3

Conclusions

In summary, samples of new Mn²⁺-doped and Eu³⁺-co-doped calcium molybdato-tungstates with the formula of Ca_1−3x−yMn_y⌷_xEu_2x(MoO₄)_1−3x(WO₄)_3x, where ⌷ represents vacancy with different Mn²⁺ (y = 0.0066; 0.0200; 0.0333; 0.0667) as well as Eu³⁺ (x = 0.0005; 0.0025; 0.0050; 0.0098; 0.0283; 0.0455; 0.0839; 0.1430; 0.1667; 0.1970; 0.2059; 0.2222) contents, were successfully obtained by a high-temperature sintering. The XRD patterns showed that doped materials have tetragonal, body-centered scheelite-type structure with space group I4₁/a. The melting point of each Mn²⁺-doped and Eu³⁺-co-doped material is lower than the melting of pure matrix (CaMoO₄, 1753 K), and it strongly depends on Mn²⁺ as well as Eu³⁺ contents in scheelite-type framework. IR results confirmed the presence of MoO₄ and WO₄ tetrahedra, the deformation of their shape in comparison with the tetrahedral shape of regular MoO₄ and WO₄ polyhedra as well as very weak covalent nature of Mo–O and W–O bonds in doped materials. The Tauc plots were used to extrapolate a direct band gap of new ceramics. They are insulators with the optical direct band gap higher than 3.50 eV. For all doped ceramics, E_g values are lower than the direct band gap of pure matrix and they change nonlinearly with increasing Mn²⁺ and Eu³⁺ contents. EPR spectra confirmed the existence of magnetic species in the form of Mn²⁺ ions. Manganese ions behave as isolated magnetic centers, if concentration of Mn²⁺ is low. These ions are characterized by close to axial local symmetry, with significant AFM interactions among each other. Increasing manganese ions concentration leads to significant changes in a simple magnetic arrangement, where above y = 0.0200 a complex magnetic system of manganese is created. From the other hand, magnetic arrangement should be also changed by increasing concentration of europium ions. In this case, Eu³⁺ ions have influence on magnetic system by indirect interactions with Mn²⁺ ions. Model with isolated manganese species is changed at europium concentration above x = 0.2222.