Rare-Earth Tantalates and Niobates Single Crystals : Promising Scintillators and Laser Materials

Rare-earth tantalates, with high density and monoclinic structure, and niobates with monoclinic structure have been paid great attention as potential optical materials. In the last decade, we focused on the crystal growth technology of rare-earth tantalates and niobates and studied their luminescence and physical properties. A series of rare-earth tantalates and niobates crystals have been grown by the Czochralski method successfully. In this work, we summarize the research results on the crystal growth, scintillation, and laser properties of them, including the absorption and emission spectra, spectral parameters, energy levels structure, and so on. Most of the tantalates and niobates exhibit excellent luminescent properties, rich physical properties, and good chemical stability, indicating that they are potential outstanding scintillators and laser materials.

At present, with an increasing demand in dense, fast, and bright scintillators for high-energy physics, medical diagnostics, and security screening devices, the scintillation characteristics of the lanthanide orthotantalates and orthoniobates REBO 4 have been widely investigated for the development of new scintillator materials [9,14,15].Among REBO 4 materials, RETaO 4 are favorable for the registration of high-energy particles due to their extremely high density ranging from 7.8 g/cm 3 for LaTaO 4 to 9.75 g/cm 3 for LuTaO 4 , which are partly higher than the density of Bi 4 Ge 3 O 12 (BGO 7.1 g/cm 3 ) and PbWO 4 (PWO 8.28 g/cm 3 ).Normally, the rare-earth tantalates are studied systematically with polycrystalline samples produced by solid-state reaction [9].Although the optical absorption and luminescence of RETaO 4 crystals are described, the crystal size is too small, only 7 mm × 7 mm × 1 mm [10].It is urgent to develop the growth technology of large RETaO 4 crystals for the device application.Recently, our group has made great progress in the Cz growth of large-size tantalite crystals and obtained big and fine GdTaO 4 single crystals, which would promote the growth and practical application of RETaO 4 [16].
In addition, REBO 4 have been considered appropriately as host matrices and have also attracted wide interests as potential rare earth-doped laser hosts [17].They belong to monoclinic system and have large crystal field energy.The active ions occupy the site with C 2 symmetry (RE sites), which is advantageous to remove the parity-forbidden rule of f-f transition.Therefore, it is beneficial for the realization of new emission and tunable wavelength.Additionally, their mechanical and thermal properties are also sufficient for laser applications.
Recently, on the one hand, our group have finished a lot of research works on scintillation characteristics of RETaO 4 , such as GdTaO 4 , Tb:GdTaO 4 , LuTaO 4 , Nd:LuTaO 4 , and so on.The luminescence results indicate that GdTaO 4 and Nd:LuTaO 4 are the promising scintillator candidates.On the other hand, the emission spectra of active ions (Yb, Nd, Er, Ho) doped GdBO 4 , YBO 4 , and mixed GdYBO 4 crystals show that they can be used as the new materials for visible, near, and middle infrared lasers.Meanwhile, near infrared lasers are realized on the Yb and Nd-doped ReBO 4 successfully.

Polycrystalline Synthesis
In this paper, LuTaO 4 and Nd-doped LuTaO 4 series samples were prepared by solid-state reaction method using Lu 2 O 3 (99.99%),Ta 2 O 5 (99.99%), and Nd 2 O 3 (99.999%)as starting materials.First, the raw materials were weighed accurately according to the appropriate stoichiometric ratio and then mixed and ground in a mortar.The mixed powders were then heated to 1500 • C at a heat rate of approximately 2.3 • C/min and calcined at 1500 • C for 24 h.Finally, the calcined samples were slowly cooled to room temperature and then carefully ground for the measurements.

Single Crystal Growth
All of the crystals, mentioned in the text, were grown by the Czochralski (Cz) method.Raw materials, with high purity (>4 N), were weighed stoichiometrically, mixed thoroughly, and then pressed into disks.The disks were put into an iridium crucible and melted in a JGD-60 furnace with an automatic diameter controlled system.Using a seed, the crystals were grown in nitrogen atmosphere with a rotation speed of 3.0-10.0rpm and pulling rate of 0.35-1 mm/h.The growth process of all crystals is similar.The as-grown crystals and some samples are shown in Figure 1.
7.8 g/cm 3 for LaTaO4 to 9.75 g/cm 3 for LuTaO4, which are partly higher than the density of Bi4Ge3O12 (BGO 7.1 g/cm 3 ) and PbWO4 (PWO 8.28 g/cm 3 ).Normally, the rare-earth tantalates are studied systematically with polycrystalline samples produced by solid-state reaction [9].Although the optical absorption and luminescence of RETaO4 crystals are described, the crystal size is too small, only 7 mm × 7 mm × 1 mm [10].It is urgent to develop the growth technology of large RETaO4 crystals for the device application.Recently, our group has made great progress in the Cz growth of large-size tantalite crystals and obtained big and fine GdTaO4 single crystals, which would promote the growth and practical application of RETaO4 [16].
In addition, REBO4 have been considered appropriately as host matrices and have also attracted wide interests as potential rare earth-doped laser hosts [17].They belong to monoclinic system and have large crystal field energy.The active ions occupy the site with C2 symmetry (RE sites), which is advantageous to remove the parity-forbidden rule of f-f transition.Therefore, it is beneficial for the realization of new emission and tunable wavelength.Additionally, their mechanical and thermal properties are also sufficient for laser applications.
Recently, on the one hand, our group have finished a lot of research works on scintillation characteristics of RETaO4, such as GdTaO4, Tb:GdTaO4, LuTaO4, Nd:LuTaO4, and so on.The luminescence results indicate that GdTaO4 and Nd:LuTaO4 are the promising scintillator candidates.On the other hand, the emission spectra of active ions (Yb, Nd, Er, Ho) doped GdBO4, YBO4, and mixed GdYBO4 crystals show that they can be used as the new materials for visible, near, and middle infrared lasers.Meanwhile, near infrared lasers are realized on the Yb and Nd-doped ReBO4 successfully.

Polycrystalline Synthesis
In this paper, LuTaO4 and Nd-doped LuTaO4 series samples were prepared by solid-state reaction method using Lu2O3 (99.99%),Ta2O5 (99.99%), and Nd2O3 (99.999%) as starting materials.First, the raw materials were weighed accurately according to the appropriate stoichiometric ratio and then mixed and ground in a mortar.The mixed powders were then heated to 1500 °C at a heat rate of approximately 2.3 °C/min and calcined at 1500 °C for 24 h.Finally, the calcined samples were slowly cooled to room temperature and then carefully ground for the measurements.

Single Crystal Growth
All of the crystals, mentioned in the text, were grown by the Czochralski (Cz) method.Raw materials, with high purity (>4 N), were weighed stoichiometrically, mixed thoroughly, and then pressed into disks.The disks were put into an iridium crucible and melted in a JGD-60 furnace with an automatic diameter controlled system.Using a seed, the crystals were grown in nitrogen atmosphere with a rotation speed of 3.0-10.0rpm and pulling rate of 0.35-1 mm/h.The growth process of all crystals is similar.The as-grown crystals and some samples are shown in Figure 1.

Crystal Structure and Chemical Etching
Generally, crystal structure can be determined by X-ray diffraction (XRD).The XRD patterns of as-grown crystals are measured with a X'pert PRO X-ray diffractometer (PANalytical, Almelo, The Netherlands) employing Cu Kα radiation (λ = 1.540598Å).The diffraction data are recorded in the 2θ range of 10°-90° with a scan step of 0.033°.The structural parameters of above crystals are obtained by fitting the XRD data with the Rietveld refinement method.The structural parameters of pure GdTaO4 or GdNbO4 are taken as the initial values, the background function, lattice parameters, atomic coordinates, and isotropic temperature factors are refined with the software GSAS.The refinement results of cell parameters are shown in Table 1.Since the ionic radius of doped ions is smaller than that of the replaced ions, there is a slight decrease in the cell volume and density.

Crystal Structure and Chemical Etching
Generally, crystal structure can be determined by X-ray diffraction (XRD).The XRD patterns of as-grown crystals are measured with a X'pert PRO X-ray diffractometer (PANalytical, Almelo, The Netherlands) employing Cu Kα radiation (λ = 1.540598Å).The diffraction data are recorded in the 2θ range of 10 • -90 • with a scan step of 0.033 • .The structural parameters of above crystals are obtained by fitting the XRD data with the Rietveld refinement method.The structural parameters of pure GdTaO 4 or GdNbO 4 are taken as the initial values, the background function, lattice parameters, atomic coordinates, and isotropic temperature factors are refined with the software GSAS.The refinement results of cell parameters are shown in Table 1.Since the ionic radius of doped ions is smaller than that of the replaced ions, there is a slight decrease in the cell volume and density.It is very meaningful to research the crystal defects in the as-grown crystal.The crystal with defects may destroy the mechanical, optical, and laser properties which will restrict the use of the crystals.Generally, the chemical etching method is an important and direct technique to investigate the defect structure of a single crystal.As we all know, acid and alkali are often used as the etchant to dissolve the dislocation sites.
The crystal of Tm,Ho:GdYTaO 4 and Nd:GdYNbO 4 are etched with KOH.Their dislocation etching pit patterns on the (100), (010), and (001) crystallographic faces are shown in Figures 2 and 3. Figure 2a,c exhibit the similar triangular prism, different from Figure 2b: rhombohedral.Similarly, the shape of the etching pits of Nd:GdYNbO 4 on (100) and (001) crystallographic faces are both present in strip shape, but not completely equivalent, as shown in Figure 3. Dislocations are typical defects introduced by the lattice distortion due to the incorporation of impurities and usually found in the molecules or atoms with the weak chemical bond.In the process of chemical etching, the weakest and unstable bonds are broken and form the specific etching pits.Therefore, dislocation etching pits have close relations with the atomic arrangement and symmetry.It is very meaningful to research the crystal defects in the as-grown crystal.The crystal with defects may destroy the mechanical, optical, and laser properties which will restrict the use of the crystals.Generally, the chemical etching method is an important and direct technique to investigate the defect structure of a single crystal.As we all know, acid and alkali are often used as the etchant to dissolve the dislocation sites.
The crystal of Tm,Ho:GdYTaO4 and Nd:GdYNbO4 are etched with KOH.Their dislocation etching pit patterns on the (100), (010), and (001) crystallographic faces are shown in Figures 2 and 3. Figure 2a,c exhibit the similar triangular prism, different from Figure 2b: rhombohedral.Similarly, the shape of the etching pits of Nd:GdYNbO4 on (100) and (001) crystallographic faces are both present in strip shape, but not completely equivalent, as shown in Figure 3. Dislocations are typical defects introduced by the lattice distortion due to the incorporation of impurities and usually found in the molecules or atoms with the weak chemical bond.In the process of chemical etching, the weakest and unstable bonds are broken and form the specific etching pits.Therefore, dislocation etching pits have close relations with the atomic arrangement and symmetry.It is very meaningful to research the crystal defects in the as-grown crystal.The crystal with defects may destroy the mechanical, optical, and laser properties which will restrict the use of the crystals.Generally, the chemical etching method is an important and direct technique to investigate the defect structure of a single crystal.As we all know, acid and alkali are often used as the etchant to dissolve the dislocation sites.
The crystal of Tm,Ho:GdYTaO4 and Nd:GdYNbO4 are etched with KOH.Their dislocation etching pit patterns on the (100), (010), and (001) crystallographic faces are shown in Figures 2 and 3. Figure 2a,c exhibit the similar triangular prism, different from Figure 2b: rhombohedral.Similarly, the shape of the etching pits of Nd:GdYNbO4 on (100) and (001) crystallographic faces are both present in strip shape, but not completely equivalent, as shown in Figure 3. Dislocations are typical defects introduced by the lattice distortion due to the incorporation of impurities and usually found in the molecules or atoms with the weak chemical bond.In the process of chemical etching, the weakest and unstable bonds are broken and form the specific etching pits.Therefore, dislocation etching pits have close relations with the atomic arrangement and symmetry.

Measurements
All the absorption spectra were recorded by a VIS-NIR-IR spectrophotometer PE lambda 950 (PerkinElmer, Waltham, MA, USA).The excitation and emission spectra were measured by a fluorescence spectrometer FLSP920 (Edinburgh, UK), with an excitation source of Xenon lamp or diode laser.An Opolette 355I (Carlsbad, CA, USA) was used to measure the fluorescence decay curves.Thermal expansion behavior of crystals along a, b, and c axes were measured in the temperature range of 300-893 K using a thermal dilatometer DIL-402C (Netzsch, Selbe, Germany) with a heating rate of 5 K/min.The samples coated with graphite were used for the measurements of specific heat and thermal diffusivity along the a, b, and c axes by a laser flash apparatus LFA457 (Netzsch, Selbe, Germany)

Judd-Ofelt Calculation
The Judd-Ofelt (J-O) theory [18,19] has been successfully applied to various systems doped with rare-earth ions.The great advantage of the J-O theory is the ability to express the probability of any transition between two f n states with only three intensity parameters (Ω 2 , Ω 4 , Ω 6 ) which can be obtained from absorption spectra.Moreover, spectral parameters, such as line strength (s), oscillator strength (f ), transition probability (A), and radiative lifetime (τ), can be calculated with absorption spectrum.The experimental oscillator strength from the initial state |SLJ to the final state |S L J was calculated as follows: where α (λ) is absorption coefficient, N is the total number of active ions per unit volume, m is the mass of the electron, ε 0 is permittivity of vacuum, c is the velocity of light in vacuum, e is the electron charge, and λ is given as Equation ( 2): However, under certain simplifying assumptions, the oscillator strength (f ) can be expressed as the sum of electric-dipole (f ed ) and magnetic-dipole (f md ) oscillator strength: Additionally, the relationships of oscillator strength and line strength are shown as follow: where J is the total angular momentum, n is the refractive index, and h is Planck constant.S ed and S md can be calculated easily with Equation ( 6): In the above formulas, SLJ| U t |S L J is the doubly reduced matrix elements of unitary tensor operator U t with t = 2, 4, 6 between the state |SLJ and |S L J , L + 2S is the magnetic dipole operator.In addition, spontaneous transition probability (A), and radiative lifetime (τ) can be calculated with line strength:

Emission Cross-Section Calculation
In this paper, all the emission cross-sections σ em are calculated with the F-L formula: where c is the speed of light, τ is the radiative lifetime of the upper energy level, n is the refractive index, and β is the branching ratio.

GdTaO 4 and Nd:GdTaO 4 Crystals
In recent years there has been renewed interest in developing new scintillator materials characterized by high light yield, fast response, and high density [20][21][22][23].Scintillators with high density and high atomic number are mostly desirable, because high stopping power can reduce the needed amount of scintillator materials and, thus, reduce the volume of the detector.GdTaO 4 is an attractive host and exhibits a high density (8.94 g/cm 3 ).Attenuation length for GdTaO 4 is calculated to be 1 cm, only second to PWO (0.89 cm) [23,24].Previously, GdTaO 4 crystal has been reported in a few papers [10,25,26], where the grown crystals are either with inclusions and twins [10] or with small size [26].
In our lab, nearly ten years have been spent on the Cz growth of GdTaO 4 single crystals.At first, an iridium wire is used to pull a crystal from the melt for obtaining the seed.With the seed, GdTaO 4 and Tb:GdTaO 4 bulk single crystals are successfully grown by the Cz method [27].However, there are cracks, inclusions, and twins within these two crystals.After a long period of technological optimization, finally, a crack-free GdTaO 4 crystal with dimensions of Φ 23 mm × 30 mm was grown successfully, which is the largest size so far [16].
The luminescence and scintillation properties of GdTaO 4 crystal have been studied in detail.Under the excitation of 273 nm, the GdTaO 4 crystal shows a strong emission band from 400 to 700 nm, with a highly asymmetrical shape [16].To analyze the asymmetrical emission band further, a series of temperature-dependent luminescent spectra of GdTaO 4 crystal are measured.All of the emission spectra consist of two bands, 2.2 eV and 2.7 eV bands from 8 K to 300 K [28].The emission intensity increases slightly as the temperature rising from 8 K to 80 K, and then decrease with increasing temperature.It is quenched rapidly above 150 K with the intensity decreasing by two orders of magnitude.The intensity variation of these two bands indicates the existence of thermal activation process.The activation energy of 2.2 eV and 2.7 eV bands is determined to be 156 meV and 175 meV, respectively.These two bands originate from different luminescent centers, which are tentatively assigned to self-trapped excitons (STE) localized at TaO 4 3− groups (2.7 eV) and to relaxed excitons related to lattice imperfections (2.2 eV) [28].The photoluminescence decay shows two components, including a fast one of 30 ns with 53% and a relatively slow one of 452 ns with 47%.The scintillation decay consists of a fast component of 72.6 ns (9.5%) and a slow component of 1236.2 ns (90.5%).Meanwhile, the scintillation efficiency of GdTaO 4 is about four times as much as PWO by integrating the area of the radioluminescence spectra.The relative light yield of GdTaO 4 is calibrated as 19 p.e./mes as that of PWO.Although, the scintillation decay of GTO is inferior to PWO, the latter presents a dominant decay of a dozen nanoseconds [24], the light yield of GdTaO 4 is higher than PWO.
In addition, the Nd:GdTaO 4 single crystal with density of 8.83 g/cm 3 has been grown by Fang Peng et al. [29].Its photoluminescence decay time of 417 nm from the 4f -4f transition of Nd 3+ is 463 ns, and scintillation decay constants consist of 46.4 ns (48%) and 1199.7 ns (52%) under the excitation of 354 nm.The scintillation decay of Nd:GdTaO 4 is much faster than that of GdTaO 4 , and the faster component percent is remarkably increased.This result may be caused by defects induced by Nd 3+ doping.Moreover, because the concentration of Nd 3+ is 0.67%, a rather low value, the light yield of Nd:GdTaO 4 can be estimated to be equal to that of GdTaO 4 .However, the faster decay of Nd:GdTaO 4 is encouraging and makes it more effective in detecting high energy rays or particles than GdTaO 4 .

LuTaO 4
Lutetium tantalate (LuTaO 4 ) is an efficient luminescent host material, especially excited by occurs by ionizing radiation [30,31].LuTaO 4 also exhibits extremely high density (9.81 g/cm 3 ), which is the highest among the present luminescent host materials.Therefore, LuTaO 4 may be an excellent heavy scintillator when it is doped with appropriate active ions, such as Nd 3+ .M'-type Lu 1−x Nd x TaO 4 (x, 0.01-0.1)polycrystalline powders were synthesized by solid reaction method [32].
The emission at 418.5 nm corresponding to the 4 D 3/2 -4 I 13/2 transition of Nd 3+ is strongest, and the fluorescence lifetime of 418.5 nm emission is measured to be approximately 263.2 ns, which is faster than that of BGO (300 ns) [32].Meanwhile, LuTaO 4 can also emit directly.Therefore, LuTaO 4 can be expected as a very promising heavy substrate and heavy scintillator, which has potential applications in nuclear medicine and high energy detection.Thus, it is significant to explore the growth of LuTaO 4 and Nd:LuTaO 4 crystals.
Liu et al. [31] have spent long time on the LuTaO 4 crystal growth.However, unfortunately, it is hard to obtain sing crystal.According to the XRD analysis, there are

Laser Materials
In the following parts, we will briefly describe the spectral and laser properties of some active ion-doped othotantalates and orthoniobates.

Er:GdTaO4 Crystal
Trivalent erbium ion (Er 3+ ) has attracted wide attention due to its rich laser emission bands, such as, green laser ( 2 H11/2, 4 S3/2)→ 4 I15/2; red laser 4 F9/2→ 4 I15/2; 1.5-1.6 μm laser 4 I13/2→ 4 I15/2; 2.6-3 μm laser ( 4 I11/2→ 4 I13/2).These lasers can be applied in many different fields, including atmospheric monitoring, eye-safe laser, and medical treatment.Previously, 30 at% and 1 at% Er:GdTaO4 crystals are grown by the Cz method [34,35].According to the absorption spectra, the transition intensity parameters Ωt (t = 2, 4, 6) are calculated by J-O theory and compared with other Er 3+ -doped crystals, as shown in Table 2. Ω2 is sensitive to the symmetry between the rare-earth ions and the ligand field.The spectroscopic quality factor Ω4/Ω6 of 1 at% Er:GdTaO4 crystal is found to be 1.42, which is used to estimate the potential of active materials for laser operation when is linked to the luminescence branching ratios.The value of 1 at% Er:GdTaO4 is comparable with those in other Er 3+ -doped systems and larger than that of 30 at% Er:GdTaO4 crystal.High Er 3+ doping concentration is proposed to overcome the self-terminating "bottleneck" effect by inducing upconversion (UC) and crossrelaxation (CR) processes [36].However, the fluorescence intensity of 2 H11/2, 4 F9/2, and 4 S3/2 states is quenched at high Er 3+ concentration.The UC process depopulates the pumping level and the upper level of 1.6 μm laser and reabsorption in this wavelength increase with the increasing Erconcentration.Emission cross-section is an important parameters for evaluating the laser property of the materials, and larger cross-sections means easier laser realization.The largest emission crosssection of 30 at% Er:GdTaO4 crystal is 0.655 × 10 −20 cm 2 at 2.631 μm, which indicates that 30 at% Er:GdTaO4 crystal can be a promising laser medium around 2.6 μm [34].The largest emission crosssection of 1 at% Er:GdTaO4 crystal is 1.022 × 10 −20 cm 2 around 1.6 μm, which make it e great potential for near infrared laser generation.Furthermore, 1 at% Er:GdTaO4 crystal also can be used as the green and red laser materials under some special environment such as at low temperature and this will be an important issue for our future research.

Laser Materials
In the following parts, we will briefly describe the spectral and laser properties of some active ion-doped othotantalates and orthoniobates.

Er:GdTaO 4 Crystal
Trivalent erbium ion (Er 3+ ) has attracted wide attention due to its rich laser emission bands, such as, green laser ( 2 H 11/2 , 4 S 3/2 )→ 4 I 15/2 ; red laser 4 F 9/2 → 4 I 15/2 ; 1.5-1.6 µm laser 4 I 13/2 → 4 I 15/2 ; 2.6-3 µm laser ( 4 I 11/2 → 4 I 13/2 ).These lasers can be applied in many different fields, including atmospheric monitoring, eye-safe laser, and medical treatment.Previously, 30 at% and 1 at% Er:GdTaO 4 crystals are grown by the Cz method [34,35].According to the absorption spectra, the transition intensity parameters Ω t (t = 2, 4, 6) are calculated by J-O theory and compared with other Er 3+ -doped crystals, as shown in Table 2. Ω 2 is sensitive to the symmetry between the rare-earth ions and the ligand field.The spectroscopic quality factor Ω 4 /Ω 6 of 1 at% Er:GdTaO 4 crystal is found to be 1.42, which is used to estimate the potential of active materials for laser operation when is linked to the luminescence branching ratios.The value of 1 at% Er:GdTaO 4 is comparable with those in other Er 3+ -doped systems and larger than that of 30 at% Er:GdTaO 4 crystal.High Er 3+ doping concentration is proposed to overcome the self-terminating "bottleneck" effect by inducing upconversion (UC) and cross-relaxation (CR) processes [36].However, the fluorescence intensity of 2 H 11/2 , 4 F 9/2 , and 4 S 3/2 states is quenched at high Er 3+ concentration.The UC process depopulates the pumping level and the upper level of 1.6 µm laser and reabsorption in this wavelength increase with the increasing Er-concentration.Emission cross-section is an important parameters for evaluating the laser property of the materials, and larger cross-sections means easier laser realization.The largest emission cross-section of 30 at% Er:GdTaO 4 crystal is 0.655 × 10 −20 cm 2 at 2.631 µm, which indicates that 30 at% Er:GdTaO 4 crystal can be a promising laser medium around 2.6 µm [34].The largest emission cross-section of 1 at% Er:GdTaO 4 crystal is 1.022 × 10 −20 cm 2 around 1.6 µm, which make it e great potential for near infrared laser generation.Furthermore, 1 at% Er:GdTaO 4 crystal also can be used as the green and red laser materials under some special environment such as at low temperature and this will be an important issue for our future research.Nd:GdTaO 4 can be used not only as a heavy scintillator, but also as a new laser material.Nd:GdTaO 4 and Nd:Gd 0.69 Y 0.3 TaO 4 (Nd:GdYTaO 4 ) (1 at%) single crystals with high optical quality are grown successfully [40,41] and their luminescence and laser properties in near infrared wavelength are studied.The absorption cross-section of the Nd:GdTaO 4 crystal at 808 nm is 5.437 × 10 −20 cm 2 , and the full width at half maximum (FWHM) of this absorption band is about 6 nm.The stimulated emission cross-section at 1066 nm is 3.9 × 10 −19 cm 2 and the measured lifetime of 4 F 3/2 level is 178.4 µs.A diode end-pumped Nd:GdTaO 4 laser at 1066 nm with the maximum output power of 2.5 W is achieved in the continuous-wave mode.The optical-to-optical conversion efficiency and slope efficiency are 34.6% and 36%, respectively.In addition, the fluorescence branching ratio of 4 F 3/2 → 4 I 9/2 transition reaches 43%, indicating that Nd:GdTaO 4 may be an efficient laser medium at 920 nm.
The maximum absorption cross-section of Nd:GdYTaO 4 at 809 nm and the stimulated emission cross-section at 1066.6 nm are 6.886 × 10 −20 cm 2 and 22 × 10 −20 cm 2 , respectively.The fluorescence lifetime is 182.4 µs.An 808 nm laser diode end-pumped continuous wave (CW) laser at 1066.5 nm is realized.The maximum output power of 2.37 W is obtained, corresponding to an optical conversion efficiency of 36.5% and a slope efficiency of 38%.Compared with the slope efficiency 36% of Nd:GdTaO 4 , Nd:GdYTaO 4 shows an enhancement of the CW laser performance.Spectroscopic properties of Nd:GdTaO 4 and Nd:GdYTaO 4 are compared with other Nd-doped laser crystals, which are listed in Table 3.By comparison, Nd:GdYTaO 4 is a better novel laser crystal with low symmetry and has great potential in low to moderate level lasers.

Tm,Ho:GdYTaO 4 and Yb,Ho:GdYTaO 4 Crystal
The new wavelengths around 2.9 µm have attracted widely interests, due to their strong absorption in water, biological tissues, and vapor [42], which can be applied in medical, biological, and remote sensing.Moreover, laser wavelengths around 2.9 µm are also suitable pump sources for infrared optical parametric oscillation (OPO) or optical parametric generation (OPG) [43].One possibility of generating 2.9 µm radiation is the 5 I 6 → 5 I 7 transition of Ho 3+ , which possesses rich energy levels [44].Usually, Tm 3+ are used to sensitize Ho 3+ solid state lasers.Tm,Ho:GdYTaO 4 and Yb,Ho:GdYTaO 4 single crystals are grown and the spectral properties were studies in detail.
Their absorption spectra along a, b, and c axes are measured and the absorption coefficients are compared in Table 4. Their respective absorption coefficients along the c axis are larger than those along the other two directions.It indicates that the c-axis samples may be more beneficial for the laser performance by improving pumping efficiency.The 2.9 µm emission spectra of Tm,Ho:GdYTaO 4 and Yb,Ho:GdYTaO 4 crystals are measured, excited by a 783 nm and 940 nm LD, respectively.Their emission cross-sections are calculated with F-L formula and compared in Figure 5.The main emission peaks of Tm,Ho:GdYTaO 4 are located at 2895, 2915 and 2932 nm.Similarly, strong peaks of Yb,Ho:GdYTaO 4 are located at 2865 and 2911 nm.The maximum emission cross-section of Tm,Ho:GdYTaO 4 at 2933 nm is 37.2 × 10 −20 cm 2 , which is larger than that of Yb,Ho:GdYTaO 4 (2911 nm, 17.6 × 10 −20 cm 2 ).It indicates that Tm,Ho:GdYTaO 4 may be easier to realize laser output than Yb,Ho:GdYTaO 4 .In addition, the emission spectrum also indicates that the energy transfer between Yb 3+ -Ho 3+ and Tm 3+ -Ho 3+ ions can be realized successfully.

Tm,Ho:GdYTaO4 and Yb,Ho:GdYTaO4 Crystal
The new wavelengths around 2.9 μm have attracted widely interests, due to their strong absorption in water, biological tissues, and vapor [42], which can be applied in medical, biological, and remote sensing.Moreover, laser wavelengths around 2.9 μm are also suitable pump sources for infrared optical parametric oscillation (OPO) or optical parametric generation (OPG) [43].One possibility of generating 2.9 μm radiation is the 5 I6→ 5 I7 transition of Ho 3+ , which possesses rich energy levels [44].Usually, Tm 3+ are used to sensitize Ho 3+ solid state lasers.Tm,Ho:GdYTaO4 and Yb,Ho:GdYTaO4 single crystals are grown and the spectral properties were studies in detail.
Their absorption spectra along a, b, and c axes are measured and the absorption coefficients are compared in Table 4. Their respective absorption coefficients along the c axis are larger than those along the other two directions.It indicates that the c-axis samples may be more beneficial for the laser performance by improving pumping efficiency.The 2.9 μm emission spectra of Tm,Ho:GdYTaO4 and Yb,Ho:GdYTaO4 crystals are measured, excited by a 783 nm and 940 nm LD, respectively.Their emission cross-sections are calculated with F-L formula and compared in Figure 5.The main emission peaks of Tm,Ho:GdYTaO4 are located at 2895, 2915 and 2932 nm.Similarly, strong peaks of Yb,Ho:GdYTaO4 are located at 2865 and 2911 nm.The maximum emission cross-section of Tm,Ho:GdYTaO4 at 2933 nm is 37.2 × 10 −20 cm 2 , which is larger than that of Yb,Ho:GdYTaO4 (2911 nm, 17.6 × 10 −20 cm 2 ).It indicates that Tm,Ho:GdYTaO4 may be easier to realize laser output than Yb,Ho:GdYTaO4.In addition, the emission spectrum also indicates that the energy transfer between Yb 3+ -Ho 3+ and Tm 3+ -Ho 3+ ions can be realized successfully.5. Compared with other hosts, the Tm,Ho:GdYTaO4 and Yb,Ho:GdYTaO4 exhibit shorter lifetime of 5 I7 and longer lifetime of 5 I6, which are in favor of the population inversion and laser output.Moreover, the laser threshold of Yb,Ho:GdYTaO4 may be lower than that of Tm,Ho:GdYTaO4.The fluorescence decay times are obtained.Their lifetimes of upper level and low level are shown in Table 5.Compared with other hosts, the Tm,Ho:GdYTaO 4 and Yb,Ho:GdYTaO 4 exhibit shorter lifetime of 5 I 7 and longer lifetime of 5 I 6 , which are in favor of the population inversion and laser output.Moreover, the laser threshold of Yb,Ho:GdYTaO 4 may be lower than that of Tm,Ho:GdYTaO 4 .The intensive absorption broadband of 5 at% Yb:GdNbO 4 and 5 at% Yb:YNbO 4 between 900-1000 nm correspond to the typical transitions from 2 F 7/2 to the sublevels of 2 F 5/2 of Yb 3+ .There are three obvious absorption peaks of 5 at% Yb: GdNbO 4 , located at 936, 955, and 975 nm, respectively.However, the absorption peaks of 5 at% Yb:YNbO 4 are 933, 955, and 974 nm.Their different locations probably are due to the small difference for the radius of Gd 3+ (0.938 Å) and Y 3+ (0.9 Å).Therefore, the environment of the crystal field around Yb 3+ is a little different.The absorption cross-sections of 5 at% Yb:GdNbO 4 were calculated to 0.77 × 10 −20 cm 2 , 0.85 × 10 −20 cm 2 , 0.64 × 10 −20 cm 2 , respectively.The absorption cross-section of 5 at% Yb:YNbO 4 were calculated to be 0.73 × 10 −20 cm 2 , 1.85 × 10 −20 cm 2 , 0.86 × 10 −20 cm 2 , respectively.
Their refractive indices n were fitted with the following Sellmeier equation by the least square method: where T is crystal transmission.And the fitted results of 5 at% Yb:GdNbO The emission cross-sections of Yb:GdNbO 4 and Yb:YNbO 4 crystal are calculated, and compared with other hosts, as shown in Table 6.The emission cross-sections of Yb:YNbO 4 are larger that of Yb:GdNbO 4 , and comparable to Yb:YAG.Therefore, it will be a promising near-infrared laser material.Table 6.Absorption and emission cross-sections for Yb:YNbO 4 , Yb:YNbO 4 , and some other Yb 3+ -doped crystals [48,49] at room temperature.(σ abs : absorption cross-section; σ em : emission cross-section).

Crystal
σ abs /10 In addition, the preliminary laser experiment of Yb:GdNbO 4 is achieved.The pump source is an InGaAs LD with a maximum output powder of 25 W at around 976 nm in continuous mode.With a 3.54% transmission output coupler, the maximum laser power of 270 mW corresponding to the threshold of 6 W is obtained.The slope efficiency is 7.5%.High threshold pump power and low efficiency are mainly due to the crystal reabsorption, poor quality of the as-grown crystal, and crude plane-plane cavity.Good quality crystal and advanced cavity will improve the laser efficiency.With the absorption spectra, the J-O intensity parameters Ω 2,4,6 , fluorescence branching ratios β(J,J ), radiative times τ rad of Nd:GdNbO 4 (Nd:GNO), Nd:YNbO 4 (Nd:YNO), Nd:GdYNbO 4 (Nd:GYNO) and Nd:GdLaNbO 4 (Nd:GLNO) crystals are calculated by J-O theory [50][51][52][53].The comparisons of β(J,J ) are listed in Table 7.As we can see, Nd:GLNO crystal has a relatively large value of β 11/2 than other Nd-doped niobate crystals, which indicates that this crystal may be more easily to generate lasers at around 1.06 µm.The comparisons of Ω 2,4,6 are shown in Figure 6.
The Ω 2 value of Nd:GLNO is higher than that of Nd:GNO crystal, point to the presence in Nd:GLNO crystal of Nd optical centers with lower environment symmetries determined by the doping of La 3+ ions in GNO host, which induce a larger disorder around Nd 3+ ions.Moreover, Nd:YNO crystal possesses the largest value of Ω 2 .The reason is that although these four crystals all belong to the monoclinic system, the angle of β in YNO is the largest, which indicates a lowest macroscopic symmetry in the Nd:YNO host.However, despite the Nd 3+ ions possess a lowest environment symmetries in YNO host, which can supply a relatively strong crystal field for Nd 3+ ions, the growth of Nd:YNO crystal is much harder than that of Nd:GNO crystal.

Table 7.
Comparisons of τ rad and β(J,J ) between Nd:GNO, Nd:YNO, Nd:GYNO, and Nd:GLNO crystals.In addition, the preliminary laser experiment of Yb:GdNbO4 is achieved.The pump source is an InGaAs LD with a maximum output powder of 25 W at around 976 nm in continuous mode.With a 3.54% transmission output coupler, the maximum laser power of 270 mW corresponding to the threshold of 6 W is obtained.The slope efficiency is 7.5%.High threshold pump power and low efficiency are mainly due to the crystal reabsorption, poor quality of the as-grown crystal, and crude plane-plane cavity.Good quality crystal and advanced cavity will improve the laser efficiency.

Crystals
4.6.Nd:GdNbO4, Nd:YNbO4, Nd:GdYNbO4, and Nd:GdLaNbO4 Crystals With the absorption spectra, the J-O intensity parameters Ω2,4,6, fluorescence branching ratios β(J,J′), radiative times τrad of Nd:GdNbO4 (Nd:GNO), Nd:YNbO4 (Nd:YNO), Nd:GdYNbO4 (Nd:GYNO) and Nd:GdLaNbO4 (Nd:GLNO) crystals are calculated by J-O theory [50][51][52][53].The comparisons of β(J,J′) are listed in Table 7.As we can see, Nd:GLNO crystal has a relatively large value of β11/2 than other Nd-doped niobate crystals, which indicates that this crystal may be more easily to generate lasers at around 1.06 μm.The comparisons of Ω2,4,6 are shown in Figure 6.The Ω2 value of Nd:GLNO is higher than that of Nd:GNO crystal, point to the presence in Nd:GLNO crystal of Nd optical centers with lower environment symmetries determined by the doping of La 3+ ions in GNO host, which induce a larger disorder around Nd 3+ ions.Moreover, Nd:YNO crystal possesses the largest value of Ω2.The reason is that although these four crystals all belong to the monoclinic system, the angle of β in YNO is the largest, which indicates a lowest macroscopic symmetry in the Nd:YNO host.However, despite the Nd 3+ ions possess a lowest environment symmetries in YNO host, which can supply a relatively strong crystal field for Nd 3+ ions, the growth of Nd:YNO crystal is much harder than that of Nd:GNO crystal.The room-temperature emission spectra of the as-grown Nd-doped niobate laser crystals at around 1.06 μm are shown in Figure 7.The strongest emission wavelength of Nd:GLNO, Nd:GNO, Nd:YNO, and Nd:GYNO are located at 1065.0 nm, 1065.7 nm, 1066 nm, and 1065.9 nm, respectively.Owing to the inhomogeneous broadening in the mixed crystal, Nd:GLNO and Nd:GYNO have a broader emission band at around 1.06 μm.In addition, stimulated emission cross-section can be estimated from the emission spectra using the F-L formula.Therefore, the stimulated emission crosssection values of Nd:GLNO Nd:GNO, Nd:YNO, and Nd:GYNO crystals at around 1.06 μm are The room-temperature emission spectra of the as-grown Nd-doped niobate laser crystals at around 1.06 µm are shown in Figure 7.The strongest emission wavelength of Nd:GLNO, Nd:GNO, Nd:YNO, and Nd:GYNO are located at 1065.0 nm, 1065.7 nm, 1066 nm, and 1065.9 nm, respectively.Owing to the inhomogeneous broadening in the mixed crystal, Nd:GLNO and Nd:GYNO have a broader emission band at around 1.06 µm.In addition, stimulated emission cross-section can be estimated from the emission spectra using the F-L formula.Therefore, the stimulated emission cross-section values of Nd:GLNO Nd:GNO, Nd:YNO, and Nd:GYNO crystals at around 1.06 µm are estimated to be 18, 18.3, 22, and 20.5 × 10 −20 cm 2 , respectively.Moreover, the fluorescence lifetimes of 4 F 3/2 → 4 I 11/2 transition are fitted to be 152, 178, 156, and 176.4 µs for Nd:YNO, Nd:GNO, Nd:GYNO, and Nd:GLNO, respectively.The small emission cross-section and long fluorescence lifetime indicates that the Nd:GLNO crystal possess good energy storage capacity, which is advantageous to its application in Q-switched laser.

Calculations of Energy Levels
Using the relativistic model of ab initio self-consistent DV-Xα method [54,55] and effective Hamiltonian model [56], the crystal-field and spin-orbit parameters of Nd 3+ in GdTaO 4 and LuTaO 4 have been calculated.The parameters of Nd 3+ in GdTaO 4 and LuTaO 4 have been shown in Table 8.The experimental energy level values of Nd 3+ are obtained from the previous spectroscopic analysis.Then, the crystal-field and spin-orbit parameters from DV-Xα method and effective Hamiltonian method and other free-ion parameters from Ref [57] were used as the initial values to fit the experimental energy levels of Nd in GdTaO 4 , and LuTaO 4 with the f -shell fitting program, which is developed by M. F. Ried with the Fortran language [58].The calculated results, experimental results and the difference between the calculated values and experimental values are shown in Table 9.From Table 8, most of the calculated levels are quite consistent with experimental energy levels.In Nd 3+ :LuTaO 4 , among the 152 experimental energy levels, there are four Stark levels (4354, 13,321, 14,767, and 17,194 cm −1 ) with poor fitting quality with deviations of −29.78, −28.10, −28.59, and 25.69 cm −1 , respectively.Additionally, in Nd 3+ :GdTaO 4 , there are only two Stark levels (12,731 and 15,601 cm −1 ) with poor fitting quality with deviations of −24.05 and −22.02 cm −1 , respectively.The deviation of calculated energy levels and experimental energy levels are less than 30 cm -1 , which indicates that the fitted energy levels results are satisfactory.
It was concluded that the J-mixing effect does not play any significant role in the low-lying energy levels from the fitting results, especially for 4 I 9/2 , 4 I 11/2 , 4 I 13/2 , 4 I 15/2 , and 4 F 3/2 multiplets.However, at the higher-energy side, the J-mixing effect is clearly appreciable because the Stark components are overlapping with adjacent multiplets.On the other hand, the M-mixing effect is very clear for all energy levels.

Crystal Field Analysis
The experimental energy levels of Nd 3+ give a basic set for a reliable energy level simulation.A model with 30 parameters including 16 free-ion and 14 crystal-field parameters is used for Nd 3+ in rare earth tantalates, with the root mean square deviation (σ) of 12.66 and 14.6 cm −1 .In the simulation process, it could be stated that the relative positions of some experimental energy levels depend on α, β, γ, and T i .Such as the energy levels of 4 S 3/2 and 4 F 7/2 , 4 G 9/2 and 2 K 13/2 , 4 G 11/2 , and 2 K 15/2 [59].
The free-ions and crystal-field parameters of Nd 3+ in GdTaO 4 , LuTaO 4 , and YAlO 3 hosts are shown in Table 10.Er:GdTaO 4 and Tm,Ho:GdYTaO 4 single crystals have good visible and mid-infrared fluorescence properties and can be the potential laser materials.Nd 3+ , Yb 3+ -doped othotantalates and orthoniobates have been realized laser output, which prove that they can be used as laser matrix hosts successfully.iii.
The ab initio self-consistent DV-Xα method has been used in its relativistic model to investigate the crystal-field and spin-orbit parameters of Nd 3+ doped in rare earth tantalates.The deviations of the calculated energy levels and experimental energy levels are less than 30 cm −1 , which indicates that the energy levels fitting results are satisfactory.Through the calculation of the crystal field parameters it is shown that the crystal field is strong and is beneficial to the widen of the ion absorption band, which indicated that rare-earth orthotantalates will be promising laser crystal materials.
Lu 3 TaO 7 , M'-LuTaO 4 , M-LuTaO 4 phases in the crystal pulled from LuTaO 4 melt, which indicates that the phase transition of the system Lu 2 O 3 -Ta 2 O 5 is different from Gd 2 O 3 -Ta 2 O 5 .For designing or improving the single-crystal growth or ceramic preparation technique of LuTaO 4 , Xing et al. [33] investigated the detailed phase relations of the Lu 2 O 3 -Ta 2 O 5 system.The compounds containing 25-60 mol% Ta 2 O 5 are prepared by solid-state reaction at sintering temperature from 1350 • C to 2058 • C. The sintered compound phases are studied by XRD in details.Cubic Lu 3 TaO 7 , M'-LuTaO 4 , M-LuTaO 4 , O-Ta 2 O 5 , and T-Ta 2 O 5 are observed.With the temperature increases, there is an irreversible phase transition from M' to M-LuTaO 4 near 1770 • C in the composition of 30-52 mol% Ta 2 O 5 , and another phase transition from T-Ta 2 O 5 to O-Ta 2 O 5 at about 1685 • C when the ratio of Ta 2 O 5 is >52 and ≤60 mol%.Finally, a phase diagram of the Lu 2 O 3 -Ta 2 O 5 system in the range 0-100 mol% Ta 2 O 5 is constructed, as shown in Figure 4.These results are helpful to explain the phase transition of Lu 2 O 3 -Ta 2 O 5 system and to design the preparation technique of LuTaO 4 single crystals or ceramic scintillators.

Figure 7 .
Figure 7. Emission spectra of Nd:GNO, Nd:YNO, Nd:GYNO, and Nd:GLNO crystals excited by 808 nm at room temperature.In addition, laser performance of four crystals are operated based on a plano-plano resonator and the laser output power curves are shown in Figure 8.The transmission of output mirror in all of the laser experiment is 5.4% at 1.06 μm.The slope efficiency for Nd:YNO along b-orientation is 24.0%.The slope efficiency for Nd:GNO along three crystallographic axes (a-, b-, and c-) are 35.3%,33.7%, and 28%, respectively.Additionally, the slope efficiency for Nd:GYNO along three crystallographic axes (a-, b-and c-) are 30.4%,29.4% and 29.8%, respectively.Lastly, the slope efficiency for Nd:GLNO along c-orientation is 34.2%.Based on the above comparison of slope efficiency, Nd:GLNO is better than others.

Figure 7 .
Figure 7. Emission spectra of Nd:GNO, Nd:YNO, Nd:GYNO, and Nd:GLNO crystals excited by 808 nm at room temperature.In addition, laser performance of four crystals are operated based on a plano-plano resonator and the laser output power curves are shown in Figure 8.The transmission of output mirror in all of the laser experiment is 5.4% at 1.06 μm.The slope efficiency for Nd:YNO along b-orientation is 24.0%.The slope efficiency for Nd:GNO along three crystallographic axes (a-, b-, and c-) are 35.3%,33.7%, and 28%, respectively.Additionally, the slope efficiency for Nd:GYNO along three crystallographic axes (a-, b-and c-) are 30.4%,29.4% and 29.8%, respectively.Lastly, the slope efficiency for Nd:GLNO along c-orientation is 34.2%.Based on the above comparison of slope efficiency, Nd:GLNO is better than others.

Table 1 .
Lattice parameters of as-grown crystals.

Table 1 .
Lattice parameters of as-grown crystals.

Table 1 .
Lattice parameters of as-grown crystals.

Table 2 .
Comparison of Ω t values with other Er-doped hosts.

Table 4 .
The absorption coefficient of a, b, and c directions at the pumping wavelength.

Table 4 .
The absorption coefficient of a, b, and c directions at the pumping wavelength.

Table 9 .
Energy levels of Nd 3+ in rare earth tantalates.

Table 10 .
Comparing the free-ions and crystal-field parameters.

GdTaO 4 Nd 3+ :LuTaO 4 Nd 3+ : YAlO 3 [60]
Rare-earth orthotantalates have high density deserve to be studied as scintillators.Large volume and high-quality single crystals are grown and their properties are researched systematically by our group.The scintillation decay time of Nd:GdTaO 4 single crystal is faster than GdTaO 4 single crystal.LuTaO 4 has the highest density 9.8 g/cm3among the present luminescent material hosts, and the fluorescence lifetime of Nd:LuTaO 4 is about 263.2 ns and expected as the most promising heavy scintillator.ii.