Investigating the impact of molybdenum content on the room-temperature optical properties of Li₂WO₄ crystals grown by low-thermal-gradient Czochralski technique

This article is part of a special issue on current research in crystal growth and related characterization

Li₂WO₄ single crystals and Mo-doped Li₂W_1–xMo_xO₄ (x = 0.0125, 0.05) were grown by the low-thermal-gradient Czochralski technique. Optimal crystallization conditions were determined for obtaining single crystals with a length of 50 mm and a diameter of 30 mm: crystallization rate of 0.5 mm h⁻¹ and rotation rate of 5 rpm. The crystal structure was investigated using powder X-ray diffraction, which confirmed the single-phase nature and trigonal structure (space group R3) for all compositions. Differential scanning calorimetry did not reveal phase transitions in the temperature range 303–1153 K. The influence of partial substitution of W with Mo on the luminescence properties is demonstrated and a correlation between Mo concentration and emission characteristics is established.

1. Introduction

The current pace of technological development of makes obtaining new effective crystal materials one of the top scientific priorities. The problem of obtaining new crystalline compounds of high optical quality is relevant in such areas as humidity sensors, catalysis, luminescent materials and ion conductors. Among the compound classes of high interest is tungstate crystals, that have found applications as scintillating and optical materials in medicine, industrial ecology, energy conservation, and other fields due to variety of their physical–chemical properties (Kornilov et al., 2005; Chernov & Dyakov, 2019; Barinova & Kirsanova, 2008; Ryadun et al., 2025). In particular, one of the significant tasks of applied research is the improvement of scintillating detectors for elementary particle physics (Gurachevsky, 2014; Grigorieva et al., 2022; Pandey, Cho et al., 2018).

In the tungstate family, compounds with light alkaline cations are of particular interest in the field of rare events physics (Grigorieva et al., 2022; Pandey, Cho et al., 2018; Aartsen et al., 2017; Garrett & Duda, 2011; Pandey, Kim et al., 2018). The advantage of Li₂WO₄ lies in the significant difference in the mass of the cation and anion nuclei, which makes it promising for detecting weakly interacting massive particles (WIMP), which are considered one of the most likely constituents of cold dark matter, contributing about a quarter to the calculated total density of our universe (Aartsen et al., 2017). Particles of this type could have originated thermally in the early universe and are predicted by many theoretical extensions of the Standard Model (Duda & Garrett, 2011). According to the H. J. Kim et al. (Pandey, Kim et al., 2018), a scintillator based on Na₂W₂O₇ is considered promising for WIMP registration due to the significant difference in the mass of tungsten and sodium nuclei. Using a cation with an even smaller radius, such as lithium, makes it possible to compact the crystal lattice even more to significantly increase the likelihood of the scintillator interaction with WIMP.

Moreover, the Li₂WO₄ crystal is considered a promising scintillator for studying coherent elastic neutrino-nucleus scattering (CEvNS), a process predicted within the framework of the Standard Model (Lubashevskiy, 2023; Patton et al., 2012; Coloma et al., 2023; Kumpan, 2023; Bernardi, 2023; Agnes et al., 2018). This process is being studied to refine the magnetic moment of neutrinos (Patton et al., 2012; Coloma et al., 2023). The CEvNS processes on atomic nuclei are related to the uncertainty ratio. For small values of the pulse transmitted to the nucleus, the neutrino does not interact with individual nucleons, but with the entire nucleus at once (Kumpan, 2023). CEvNS was recorded experimentally for the first time in 2017 in the COHERENT international experiment (Bernardi, 2023) at the Spallation Neutron Source accelerator complex at Oak Ridge National Laboratory (USA) using 14.6 kg caesium iodide (CsI) scintillation detectors. After the first observation of the CEvNS process on the I⁻ nucleus, it became clear that further research requires more sensitive detectors in the lower-energy region, characteristic of reactor antineutrinos (Agnes et al., 2018). This field is the most difficult, but on the other hand, it is also the most interesting from a scientific and practical point of view. Since the existence of the CEvNS phenomenon has already been proved, for detailed studies with great statistical significance detectors with higher density and mass are required.

However, at the present time, the Li₂WO₄ crystal is not a widely applied scintillator as growth of large single crystals is hindered by the presence of polymorphic modifications and a phase transition from trigonal to orthorhombic structure in the 773 K region (Yamaoka et al., 1973; Pistorius, 1975). Tabero & Frackowiak (2017) reported the existence of four polymorphic modifications of Li₂WO₄: rhombo­hedral Li₂WO₄-I of phenacite structure stable at atmospheric pressure, tetragonal Li₂WO₄-II obtained at 300 MPa, ortho­rhombic Li₂WO₄-III obtained above 300 MPa and at high temperatures, and monoclinic Li₂WO₄-IV existing at temperature and pressure above those of Li₂WO₄-III.

During the growth of Li₂WO₄ from the melt by conventional high-temperature-gradient methods, upon cooling, phase transitions can lead to defects, cracks, and destruction of the grown crystals (Barinova et al., 2016). One of the ways to solve this issue is to partially substitute W with Mo. The crystal structure of α-Li₂WO₄ has the trigonal syngony of the R3 phenacite family (Tabero & Frackowiak, 2017; Barinova et al., 2016; Warr, 2021; Barinova et al., 2008). Lithium molybdate, Li₂MoO₄, isostructural to α-Li₂WO₄, has no polymorphic modifications, the only characteristic crystal lattice being phenacite-type R3 (Barinova et al., 2001). Due to this, the partial substitution of W by Mo should stabilize the trigonal structure and suppress phase transitions. Most molybdates and tungstate compounds with the same cation are isostructural and their Me_x(Mo,W)_yO_3y+x/2 mixtures form solid solutions (Khramtsova et al., 2024; Bondareva et al., 2024).

First, Li₂(Mo,W)O₄ crystal growth experiments were conducted at Mendeleev University of Chemical Technology (Moscow, Russia) (Barinova et al., 2016) with 5% and 15% Mo/W substitution. It was determined that the distribution coefficient of molybdenum in lithium tungstate lattice is above 1. The Mo/W concentration in the crystals grown by conventional Czochralski technique was higher than in the initial melt: 8% Mo/W in the crystal grown from 5% Mo/W melt, and approximately 20% in the crystal grown from 15% Mo/W melt. It should be noted that, for Mo/W concentrations above 8%, the Mo distribution within the crystal was non-uniform, varying from 12% to 35%.

Moreover, the effect of Mo on the scintillation properties of Li₂WO₄ requires thorough investigation, as even small concentrations of Mo can significantly change the luminescent characteristics (Khramtsova et al., 2024; Bondareva et al., 2024; Matskevich et al., 2021). In addition, due to their chemical similarity, residual amounts of Mo remain in W compounds.

In this work, Li₂WO₄ crystals were grown with low Mo concentrations (5 and 1.25 mol%) using the low-thermal-gradient Czochralski (LTG Cz) technique, with the aim of obtaining pure Li₂WO₄ single crystals while maintaining structural stability. The luminescent properties of the resulting crystals were then compared to assess the effect of Mo incorporation on Li₂WO₄ emission. The LTG Cz technique was designed for obtaining bulk crystals of high structural quality by the elimination of local overheating and decreasing the amount of thermoelastic stresses in the grown crystals, making it particularly suitable for the growth of compounds with phase transitions and incongruent melting type (Barabash et al., 2014; Grigorieva et al., 2025).

2. Experimental

2.1. Li₂WO₄ solid state synthesis

To ensure the accuracy of the studies, WO₃ and MoO₃ powders used as precursors were preliminarily subjected to purification according to the procedure described by Grigorieva et al. (2018). The initial charge for Li₂WO₄ crystal growth was obtained by solid-phase synthesis from Li₂CO₃ extra-pure powder (20-2 TC 6-09-4757-84) and purified WO₃ and MoO₃ by sintering at 723 K in accordance with reactions (1) and (2). Solid-phase synthesis was carried out in the same LTG Cz crystal growth set-up, in which crystal growth was subsequently carried out using platinum tooling. The completeness of the reaction was controlled by the change in mass during volatilization of CO₂.

2.2. Li₂WO₄ crystal growth

The LTG Cz technique was developed at the Institute of Inorganic Chemistry of the Siberian Branch of the Russian Academy of Sciences (Novosibirsk, Russia) for growing large oxide crystals for scintillator applications. Unlike the conventional Czochralski method with a temperature gradient of more than 100 K cm⁻¹, the LTG Cz technique achieves values of 0.5–1 K cm⁻¹, which allows the temperature range during crystallization to be controlled with high accuracy, significantly reducing the amount of thermoelastic stresses in growing crystals, thus increasing their uniformity and mechanical strength (Borovlev et al., 2001; Shlegel et al., 2017). Low-temperature gradients are most suitable for the growth of crystals that exhibit incongruent melting or polymorphism. They maintain the growing crystal within the metastable region and preserve the high-temperature modification during cooling. For example, an incongruent ZnMoO₄ single crystal for neutrinoless double beta-decay experiments was previously obtained by the LTG Cz technique (Barabash et al., 2014; Shlegel et al., 2017). Another advantage of the LTG Cz technique is a crucible with a high nozzle lid acting as a diffuse gate to suppress the decomposition and evaporation of volatile WO₃ and MoO₃ components.

During melting of the charge, the temperature in the growth furnace was elevated at a rate of 100 K h⁻¹ to 1033 K and was kept at this temperature for 5 h until the melt was completely homogenized. Before seeding, the temperature was decreased to 1006 K. The first growth experiments were carried out on a platinum seed holder, and subsequently on monocrystalline seeds cut from the obtained Li₂WO₄ crystals. The crystallization rate was 0.5–1.5 mm h⁻¹, the rotation rate was 5–10 mm h⁻¹, and the cooling rate was 60 K h⁻¹. Substitution of W for Mo by 1.25 mol% and 5 mol% did not result in a change in growth parameters and conditions, except for a decrease in the total crystallization temperature by 3 K and 5 K, respectively.

2.3. Powder X-ray diffraction analysis

Powder X-ray diffraction (XRD) patterns of Li₂WO₄ monocrystalline samples were obtained on a TD3700 (Tongda, China) diffractometer using Cu Kα radiation in Bragg–Brentano geometry. The unit-cell parameters of Li₂WO₄ samples were refined by the Le Bail method (Ait ahsaine et al., 2015; Rietveld, 1969) within the GSAS-II software (Scardi, 2020; Young, 1993). Literature data were used as starting models for the refinements with the replacement of Mo/W in the corresponding site of the crystal structure (Pistorius, 1975; Tabero & Frackowiak, 2017).

2.4. Differential scanning calorimetry

Differential scanning calorimetry (DSC) curves for Li₂WO₄ samples were measured with the scanning thermal analyzer 449 F5 Jupiter (Netzsch, Germany) in shared research facilities for multielemental and isotope study of the SB RAS under state assignment of IGM SB RAS No.122041400031-2. Li₂WO₄ powder samples (50 mg) were heated in a platinum crucible in N₂ atmosphere from room temperature to 1173 K at the rate of 20 K min⁻¹. An empty Pt crucible was used as the standard.

2.5. Optical properties

Transmission spectra were recorded using a UV-2501 PC Shimadzu spectrometer in the UV to near IR on a 3 mm-thick plate samples.

Photoluminescence (PL), PL excitation (PLE) spectra and PL decay kinetics were recorded on a Fluoro­log 3 (Horiba Jobin Yvon) spectrophotometer with a 450 W ozone-free xenon lamp as an excitation source. PL and PLE spectra were corrected for the source intensity (lamp and grating) and emission spectral response (detector and grating) using standard correction curves. The relative wavelength calibration of the monochromators was accurate within 0.05 nm. A xenon flash lamp (150 W) was used for lifetimes measurements. All measurements were made at room temperature.

Spectra for the X-ray excited luminescence (XRL) were recorded on a 1 kW X-ray set-up URS 55 (USSR) equipped with a BSV2-W tungsten tube (40 kV, 20 mA). An MDR2 monochromator and a cooled FEU83 photomultiplier were used to detect the XRL emission.

3. Results and discussion

3.1. Li₂WO₄ crystals grown by the LTG Cz technique

The first Li₂WO₄ crystal with dimensions (mm) 21.9 × 13.9 × 8.2 was obtained by spontaneous crystallization on platinum (Fig. 1). The grown crystal contains three large single crystal blocks. Despite the irregular shape, a concave shape of the crystallization front can be observed.

Figure 1 Li2WO4 crystal spontaneously grown on a platinum seed holder.

Li₂WO₄ crystal 70 mm in length and 27 mm in diameter was grown with crystallization rate 1.5 mm h⁻¹ and rotation rate 5 rpm on a seed made from a transparent part of crystal No. 1 [Fig. 2(a)]. Large transparent blocks formed in the upper part of the crystal, however, starting from 30 mm, the formation of macrostructural defects and an abrupt decrease in crystal quality were observed. The morphology of the crystal was heterogeneous: a wave-like shape was observed on the lateral surface, characteristic of the effect of over-regulation of the weight-temperature feedback. The crystal front was composed of a large number of small facets 1–2 mm² in size, which means that there are no specific directions with high reticular density where large facets could form. This indicates the absence of inclination to the layered growth mechanism and the formation of a faceted crystallization front, which coincides with the growth patterns of isostructural Li₂MoO₄ (Grigorieva et al., 2019). In this case, another approach should be used, consisting of the alteration of growth conditions for the implementation of a normal growth mechanism and the formation of a convex rounded crystallization front.

Figure 2 Li2WO4 crystals grown by the LTG Cz technique at different crystallization rates: (a) crystallization rate 1.5 mm h^−1, rotation rate 5 rpm; (b) crystallization rate 0.5 mm h^−1, rotation rate 5 rpm.

In subsequent growth processes, the crystallization rate was reduced to 0.5 mm h⁻¹. The temperature distribution over three independent heating zones was changed to give the isotherm a slightly convex shape towards the bottom of the crucible. For this purpose, the temperatures in the lower and upper zones were lowered, and the temperature in the middle zone was increased to compensate for the overall temperature change and slow down the crystallization process. Li₂WO₄ single crystal with a length of 48 mm and a diameter of 30 mm with a slightly convex front and a concave edge was obtained [Fig. 2(b)]. The crystal was transparent, visually defect-free, and the cross section was close to rounded.

The formation of a convex-concave shape of the front is possible for several reasons, one of the main ones is the release of heat during the exothermic crystallization process and the formation of a local overheated zone at the front of the growing crystal. The standard solution to this problem is to increase the rotation rate to increase the mixing of the melt. Therefore, in the next process, the rotation rate was increased from 5 rpm to 10 rpm. Li₂WO₄ crystal with length 70 mm and diameter 30 mm with a convex crystallization front was obtained [Fig. 3(a)]. However, an increase in the rotation rate led to the development of lateral pseudo-edges, and the crystal's cross-section changed from rounded to pseudo-hexagonal. On the crystallization front, there is also a manifestation of the relief of `fish scales' and screw dislocations.

Figure 3 Li2WO4 crystals grown by the LTG Cz technique at different rotation rates and varying controlled temperature gradients between three heating zones: (a) crystallization rate of 0.5 mm h^−1, rotation rate 10 rpm; (b) crystallization rate of 0.5 mm h^−1, rotation rate 5 rpm

Since an increase in the rotation rate leads to a drastic increase in the amount of structural defects, a different approach was used to control the shape of the crystallization front by further changing the isotherm. To obtain a convex front, the temperature of the lower zone was lowered by 3 K to form a slight hypothermia at the bottom of the crucible. A single crystal of Li₂WO₄ with length 50 mm and diameter 30 mm was obtained [Fig. 3(b)]. The crystal was transparent, without inclusions, with a circular cross-section and a slightly convex crystallization front. Samples from this crystal were used for further investigation of luminescent properties.

To study the effect of introducing molybdenum in the crystal lattice on the Li₂WO₄ luminescence properties, crystals of the Li₂W_1–xMo_xO₄ composition (x = 0.0125, 0.05) were grown (Fig. 4). The growth conditions corresponded to the growth conditions of pure Li₂WO₄ crystals. The grown crystal with 1.25% substitution of tungsten for molybdenum [Fig. 4(a)] is visually indistinguishable from a Li₂WO₄ crystal: transparent, visually defect-free, with a convex crystallization front.

Figure 4 Molybdenum-doped Li2W1–xMoxO4 crystals (x = 0.0125, 0.05) grown using the LTG Cz technique: (a) Li2W0.9875Mo0.0125O4 (1.25 mol% Mo); (b) Li2W0.95Mo0.05O4 (5 mol% Mo).

The obtained Li₂WO₄ single crystal with 5 mol% molybdenum content was 65 mm in length and 40 mm in diameter, transparent in the upper part, however, after the first 25 mm of the crystal, the formation of macrostructural defects, the appearance of a yellowish color, and a general decrease in crystal quality were observed [Fig. 4(b)]. Presumably, this was caused by the heterogeneity of the distribution of substitutive molybdenum in the crystal structure. The crystal had well formed lateral pseudo-edges forming a hexagonal cross-section. The crystallization front is concave with a helical dislocation outlet. A sharp increase in the cross-section of the growing crystal is presumably associated with a shift along the liquidus line during growth due to an increase in the proportion of lower-temperature molybdate.

3.2. XRD

Fig. 5(a) shows the XRD patterns of Li₂WO₄ crystal and crystals with partial substitution of [WO₄]²⁻ by [MoO₄]²⁻: 1.25 mol% and 5 mol%. All patterns show good agreement with the Li₂WO₄ reference and the absence of extraneous phases. Elementary unit-cell parameters difference is below registration accuracy and are equal to a = 14.6674 (7) Å, c = 9.7732 (8) Å. The indistinctiveness of unit-cell parameters in pure Li₂WO₄ and the ones doped with 1.25 mol% and 5 mol% Mo could be explained by the similarity of W and Mo ionic radii (rW⁶⁺ = 0.58±0.02 Å, rMo⁶⁺ = 0.56±0.02 Å (https://www.webelements.com/tungsten/atom_sizes.html, https://www.webelements.com/molybdenum/atom_sizes.html). For all refinements the reduced χ² value was approximately 1.5, indicating a satisfactory fit between observed and calculated diffraction patterns, confirming the single-phase nature of our samples and the reliability of the obtained structural parameters.

Figure 5 XRD patterns of Li2WO4 and Li2WO4 substituted with 1.25% and 5% [MoO4]^2−. (b) Li2WO4 crystal structure: green spheres – lithium, red spheres – oxygen, gray tetrahedra – [WO4]^2− anions.

The XRD data confirmed the trigonal symmetry, space group R3, of the obtained Li₂WO₄ crystals. The structure contains two inequivalent lithium positions, each surrounded by four oxygen atoms, forming [LiO₄]⁷⁻ tetrahedra. These tetrahedra connect at their corners with other [LiO₄]⁷⁻ tetrahedra and with [WO₄]²⁻ tetrahedra, forming a three-dimensional network. Li—O bond lengths vary from 1.97 Å to 2.03 Å. The [WO₄]²⁻ tetrahedra form the primary structural motif. These [WO₄]²⁻ tetrahedra are distorted: W—O bond lengths range from 1.74 to 1.79 Å (average value 1.76 Å), and O—W—O angles vary from 106° to 113°, indicating a reduction in the tetrahedron's symmetry from T_d to C_3v. As a result, the crystal field splits the degenerate electronic levels of the [WO₄]²⁻ complex, leading to a modification of energy terms and probabilities of electronic transitions. Thus, different oxygen positions can contribute differently to the formation of molecular orbitals and, consequently, influence the electronic transitions that determine the material's luminescence properties.

3.3. DSC

The DSC curve of the Li₂WO₄ crystal shows the presence of single melting and crystallization peaks, confirming the absence of phase transitions (Fig. 6). The melting temperature is 736.9 °C. On the cooling curve a single peak at 656.4 °C, outlining the boundaries of the supercooled melt existence region. It is interesting to note that while the substitution of W with Mo practically does not affect the position of the heating curve, it significantly shifts the cooling curve peak to the low-temperature region, increasing the stability range of the supercooled melt.

Figure 6 DSC curves of Li2WO4 and Li2WO4 substituted with 1.25 mol% and 5 mol% [MoO4]^2−.

3.4. Transmission spectra

The optical quality of the obtained Li₂WO₄ crystals was assessed on plated samples with thickness 3 mm by analyzing their transmission spectra in the ultraviolet and visible regions. Transmission spectra provide information about the material's band gap width, the presence of defects and impurities, and optical transitions occurring in the crystal upon interaction with electromagnetic radiation. Fig. 7 presents the transmission spectra for pure Li₂WO₄ and Li₂W_1–xMo_xO₄ crystal samples. No significant decrease in crystal transmission was observed in doped samples compared to the pure Li₂WO₄ one. Moreover, no distinctive minimum in the 390 nm region characteristic to the O vacancies defects in the crystals grown in high-temperature-gradient conditions was registered (Barinova et al., 2016).

Figure 7 UV–Vis transmission spectra showing the influence of Mo doping on Li2WO4 crystals.

It is worth noting that in the 400–700 nm region, a slight decrease in transmission is observed for the doped samples, particularly for the sample with 5 mol% Mo. This is presumably due to the non-ideal parallelism of the plate surfaces, which can lead to additional losses from reflection and light scattering, especially noticeable in the transparent spectral region where the material's intrinsic absorption is low.

3.5. Photoluminescence

Fig. 8(a) presents normalized photoluminescence excitation (PLE) spectra for Li₂W_1–xMo_xO₄ samples (x = 0, 0.0125, 0.05) in the wavelength range from 240 nm to 470 nm. The main excitation band shows a clear redshift with increasing Mo concentration: the maximum was observed around 256 nm for pure Li₂WO₄ (black curve), shifting to 270 nm for the 1.25% Mo sample (red curve) and to 275 nm for the 5% Mo sample (blue curve). This shift is consistent with a decrease in the band gap width, indicating the influence of molybdenum doping on the electronic structure. These peaks correspond to charge transfer from the oxygen ion to the tungsten ion within the [WO₄]²⁻ complex (Kröger, 1947).

Figure 8 Li2W1–xMoxO4 luminescence spectra: (a) normalized PLE spectra, λem = 500 nm; (b) normalized PL spectra, λex = 256 nm; (c) normalized luminescence spectra, λex = 375 nm.

With increasing Mo concentration, a significant change was observed in the ratio of the intensities of the peaks at 256 nm and ∼375 nm.

Fig. 8(b) presents the photoluminescence (PL) spectra of Mo-doped Li₂WO₄ crystals in the wavelength range from 350 to 700 nm under 256 nm excitation. As seen in Fig. 8(b), the luminescence spectrum of pure Li₂WO₄ (black curve) is characterized by a broad band with a maximum around 500 nm. With increasing Mo concentration to 1.25 mol% (red curve) and 5 mol% (blue curve), the overall emission maximum remains virtually unchanged. Fig. 8(c) presents the luminescence spectra under 375 nm excitation. All three samples exhibit a broad emission band with a maximum in the region 540–550 nm. Notably, the normalized spectra for pure Li₂WO₄ and samples doped with 1.25 mol% Mo and 5 mol% Mo almost completely overlap. This indicates that under 375 nm excitation, the shape of the emission band and the position of its maximum do not undergo significant changes with increasing Mo concentration, as it is attributed to intrinsic transitions within [MoO₄]²⁻ complexes.

Analysis of the luminescence decay curves showed that the process is described by a mono-exponential function (Fig. 9). The lifetimes and relative contributions of the components change with increasing Mo concentration. For pure Li₂WO₄, the lifetimes are τ = 40 µs. For the sample with 1.25% Mo, the lifetimes are τ = 90 µs. For the sample with 5% Mo, the lifetimes are τ = 122 µs. A tendency towards an increase in the lifetime is observed with increasing Mo concentration. This may indicate that Mo influences the ratio between various recombination channels, for example, by enhancing recombination through intrinsic defects and Mo-related centers.

Figure 9 Li2W1–xMoxO4 luminescence lifetimes spectra at 300 K, λex = 256 nm, λem 500 nm.

3.6. X-ray luminescence

X-ray luminescence (XRL) study showed an intense broad emission band in Li₂WO₄ crystal [Fig. 10(a), black curve] in the visible spectral region with a maximum around 420 nm. Mo doping significantly decreased the intensity of XRL: for samples with 1.25 mol% Mo (red curve) and 5 mol% Mo (blue curve), the luminescence intensity is an order of magnitude lower than that of pure Li₂WO₄. Concurrently, a visible shift of the emission maximum towards the green spectral region was observed with increasing Mo concentration, indicating a change in the nature of luminescence centers or relaxation mechanisms. The low luminescence intensity in doped crystals at room temperature is likely due to thermal or concentration quenching effects.

Figure 10 (a) XRL spectra of Li2W1–xMoxO4 at room temperature; (b) Gaussian decomposition XRL spectra of pure Li2WO4.

Fig. 10(b) presents the decomposition of the XRL spectrum of pure Li₂WO₄ crystal. Observed spectra are complex as they consist of several components. For further analysis, the spectra were decomposed into Gaussian components. The decomposition was done in the energy scale. For convenience of comparison with literature data, the components are presented in Fig. 10(b) in the wavelength scale. Experimental data (gray points) are well approximated by the sum of two Gaussian peaks (green curve). The first, more intense peak (blue curve) has a maximum 420 nm, while the second, less intense and broader peak (red curve) is located in the ∼500–520 nm region. The shape of the XRL band of pure Li₂WO₄ is close to that of the PL band, suggesting that the luminescence mechanisms under X-ray and ultraviolet excitation share a common nature and are associated with exciton transitions in [WO₄]²⁻ tetrahedral complexes. The observed shift of the XRL maximum towards the green region for doped samples, despite their low intensity, is consistent with photoluminescence data and indicates the appearance or enhanced contribution of Mo–O complexes [MoO₄]²⁻ under X-ray excitation. For a more complete understanding of XRL mechanisms in Li₂WO₄, especially in doped samples, further investigation at lower temperatures is necessary, which will likely allow for clearer observation of the emission and its components, as well as an assessment of the contribution of various centers.

4. Discussion

In this work, Li₂WO₄ and Li₂W_1–xMo_xO₄ (x = 0.0125, 0.05) single crystals were obtained by the LTG Cz technique. Investigation of the influence of growth parameters (crystallization rate, rotation rate, temperature profile) allowed to optimize the crystallization front shape and preclude defect formation. XRD analysis confirmed the formation of single-phase compounds of Li₂W_1–xMo_xO₄ and the preservation of the trigonal structure (space group R3) upon substitution of W with Mo. Refinement of unit-cell parameters revealed that the introduction of molybdenum leads to a slight decrease in the unit-cell volume, which may be related to the difference in ionic radii of W⁶⁺ and Mo⁶⁺ (manifestation of anionic isomorphism) and defect formation. DSC results showed the absence of phase transitions in the obtained crystals within the studied temperature range, indicating the stabilization of the R3 structure under controlled growth conditions. The observed shift of the crystallization peak to lower temperatures upon doping also confirms the increased Mo content in the crystal.

For studying the effect of Mo content on optical properties, the absorption and luminescence spectra of the obtained crystals were investigated. A comparison of data from absorption spectra (Fig. 7), excitation spectra [Fig. 8(a)], and PL emission spectra [Figs. 8(b) and 8(c)] allows us to conclude that molybdenum doping leads to significant changes in the electronic structure of Li₂WO₄. The decrease in the band gap width, observed in the transmission spectra (Fig. 7), is consistent with the redshift of the excitation band in the PLE spectra [Fig. 8(a)]. These changes indicate that incorporation of Mo ions into the Li₂WO₄ crystal structure creates new energy levels near the valence band edge and the bottom of the conduction band. It should be noted that non-stoichiometry and defect formation, accompanying both the formation of solid solutions and crystal growth processes, can significantly affect the optical properties of tungstates.

Additional insights are provided on Fig. 8(c), which shows the luminescence spectra obtained under 375 nm excitation. At this excitation wavelength, all samples exhibit a broad emission band with a maximum in the 540–550 nm region. Notably, the shape of the normalized spectra for all investigated Mo concentrations practically coincides, and this emission is attributed to intrinsic transitions within [MoO₄]²⁻ complexes. This indicates that under 375 nm excitation, luminescence primarily originates from Mo–O centers, and their spectral characteristics do not depend on the overall Mo concentration. However, the absolute luminescence intensities under 375 nm excitation increase with increasing Mo concentration (1 for 0 mol% Mo; 1.1 for 1.25 mol% Mo; 1.2 for 5 mol% Mo).

Thus, even in `pure' Li₂WO₄ crystals grown by the LTG Cz technique from deeply purified WO₃, it is difficult to completely exclude the presence of small amounts of molybdenum impurity. Likeliness of tungsten oxide and molybdenum oxide chemical and physical properties makes it difficult to separate one from another. Uncontrolled presence of Mo, even at low concentrations, can lead to the formation of molybdate complexes [MoO₄]²⁻, contributing to luminescence by creating additional excitation bands or altering the efficiency of energy transfer. Similar effects were observed in CdWO₄, where Mo doping led to the appearance of new luminescence bands and quenching of intrinsic luminescence, which was attributed to defect formation and changes in the electronic structure (Pinatti et al., 2019). It is important to note that this involves not merely the replacement of one type of luminescence center with another, but also a modification of energy migration and capture mechanisms (including processes related to charge transfer between tungstate and molybdate anions), which in turn affects the kinetic characteristics of luminescence.

On the other hand, the possibility of other defects not directly related to Mo should be considered. As shown in the study of Ag₂WO₄, the introduction of small amounts of alkali metals, such as Li, can compensate for defects and improve luminescence properties (Pinatti et al., 2019). This effect may be related to cationic isomorphism, where ions alter the local geometry and electronic structure of the lattice, and also compensate for the charge arising from heterovalent substitution. This indicates that cationic vacancies, anionic vacancies, and other point defects can play an important role in shaping the luminescence properties of tungstates.

It has been shown that Li₂WO₄ exhibits luminescence properties under X-ray excitation, which is a necessary condition for using the material as a scintillator. The comparison of the obtained spectral characteristics of the Li₂WO₄ samples with literature data (Cova et al., 2025; Degoda et al., 2022) allows us to draw conclusions about the effect of the growth conditions and doping technology on the XRL intensity and the content of nonradiative competitive recombination centers. For a more complete understanding of XRL mechanisms in Li₂WO₄, it is necessary to consider the possible involvement of free excitons in the energy transfer process to luminescence centers. Specifically, X-ray radiation can create free electrons and holes, which in turn form electron-hole excitons. These excitons can migrate through the crystal and transfer energy to luminescence centers associated with [WO₄]²⁻ complexes. In this case, the efficiency of such energy transfer may depend on the concentration of defects in the crystal, which can act as exciton trapping centers. However, it should be noted that at cryogenic temperatures, exciton mobility increases, and the probability of trapping by defects decreases, which allows for a significant increase in energy transfer efficiency and, consequently, an enhancement in scintillation yield.

Furthermore, it is not excluded that the XRL process involves not only exciton transitions in [WO₄]²⁻ tetrahedra but also other mechanisms related to defects and impurities in the structure. For example, X-ray radiation can ionize lithium or tungsten ions, creating Frenkel or Schottky defects. It is assumed that at low temperatures, the influence of these defects on luminescence quenching may be weakened, which would also contribute to an increase in scintillation efficiency.

5. Conclusions

A series of visually defect-free Li₂WO₄ single crystals 50 mm in length and 30 mm in diameter, both undoped and with partial substitution of [WO₄]²⁻ by [MoO₄]²⁻ (1.25 mol% and 5 mol%), were successfully grown by LTG Cz technique. Optimal growth parameters for Li₂WO₄ were determined, ensuring a normal growth mechanism and the formation of a rounded crystallization front: crystallization rate 0.5 mm h⁻¹; rotation rate 5 rpm.

Powder XRD and DSC methods confirmed the single-phase nature of the materials and the absence of phase transitions. All samples crystallized in the trigonal space group R3.

Investigation of photoluminescence spectra at room temperature showed that [MoO₄]²⁻ ions are successfully incorporated into the Li₂WO₄ crystal structure substituting [WO₄]²⁻ ions and actively participating in the formation of electronic states and the determination of the optical properties of the solid solutions. The obtained results highlight the potential for modifying the structural and optical properties of Li₂WO₄ through controlled Mo doping, opening new possibilities for the development of materials with tailored characteristics. Recorded XRL spectra, despite the relatively low efficiency at room temperature, indicate that Li₂WO₄ is of interest as a promising scintillation material for elementary particle physics, especially at cryogenic temperatures.