Effect of BaF₂ Variation on Spectroscopic Properties of Tm³⁺ Doped Gallium Tellurite Glasses for Efficient 2.0 μm Laser

Introduction

Over the past few decades, fiber lasers operating in eye-safe 2.0 μm spectral region have attracted a great deal of attention due to strong absorption band of several chemical compounds (H₂O, CO₂, N₂O, etc.) in this region (Chen et al., 2010). Therefore, there are some potential applications in eye-safe laser radar, material processing, laser surgery, remote sensing and effective pump sources as mid-infrared lasers and optical parametric oscillators (Geng et al., 2011; Geng and Jiang, 2014; Slimen et al., 2019; Wang et al., 2019). Up to now, active ions for 2.0 μm laser have been mainly focused on Tm³⁺ and Ho³⁺ ions arising from ${\text{Tm}}^{3+}{:}^3{\text{F}}_4\to {}^3{\text{H}}_6$ and ${\text{Ho}}^{3+}{:}^5{\text{I}}_7\to {}^5{\text{I}}_8$ transition. Compared with Ho³⁺, Tm³⁺ owns very strong absorption band of ${{}^3\text{H}}_6\to {{}^3\text{H}}_4$ transition and thus can be effectively pumped by commercial high-power 808 nm laser diode. Under the pump scheme, a quantum efficiency of 200% can be expected from “two-for-one” cross relaxation process ${(}^3{\text{H}}_4{+}^3{\text{H}}_6\to 2^3{\text{F}}_4$) (Richards et al., 2010). In addition, broad emission bandwidth of ${\text{Tm}}^{3+}{:}^3{\text{F}}_4{\to }^3{\text{H}}_6$ transition about 300 nm is advantageous to the generation of femtosecond pulse (Agger et al., 2004).

In pursuit of efficient 2.0 μm laser, different glass hosts have been extensively investigated and the laser operation has been demonstrated in silicate, fluoride, germanate and tellurite glasses (Richards et al., 2010; He et al., 2013; Wang et al., 2019). Among these glass hosts, tellurite glasses own a lot of advantage such as broad infrared transmission region, lower phonon energy, high rare-earth ion solubility, high refractive index (~2) and easy fabrication with low melting temperature(Richards et al., 2010). Recently, our groups have exploited several new tellurite glass systems such as TeO₂-Ga₂O₃-BaO (TGB) and TeO₂-Ga₂O₃-ZnO (TGZ) with excellent glass-forming ability, thermal stability and 2.0 μm spectroscopic properties (Li et al., 2019; Mao et al., 2020). To further improve 2.0 μm emission properties, it is very essential to reduce the hydroxyl content in glasses because OH⁻ groups are the main energy loss channels for active ions and can result in strong 2.0 μm fluorescence quenching (Terra et al., 2006). We found that the strength of interaction between Tm³⁺ and OH⁻ (12.9 × 10⁻¹⁹ cm⁴/s) was stronger than that between Er³⁺ and OH⁻ (1.9 × 10⁻¹⁹ cm⁴/s) (Yuan et al., 2014).

Herein, based on the composition of TGB glass with good thermal stability, we systematically investigate the effects of substitution of BaF₂ for BaO on physical properties and 1.8 μm emission properties. Density, refractive index, Raman spectra, absorption spectra and emission spectra were measured along with the lifetime of Tm³⁺:³F₄ energy level. Moreover, energy transfer mechanism is proposed and emission cross section and gain coefficient of ${\text{Tm}}^{3+}{:}^3{\text{F}}_4{\to }^3{\text{H}}_6$ transition in TGB glass with 9 mol.% BaF₂ are determined.

Materials and Methods

Tm³⁺ doped gallium tellurite glasses (TGB) with the molar compositions of 80TeO₂-10Ga₂O₃-(9-x)BaO-xBaF₂-1Tm₂O₃ (x = 0, 3, 6, and 9) were prepared by the conventional melt-quenching method. TeO₂, Ga₂O₃, BaO, BaF₂ and Tm₂O₃ with 99.99% purity (Aladdin) were used as raw chemicals. Appropriate amounts of these chemicals (~20 g) were well mixed and then melted in an alumina crucible with an alumina lid at ~950°C for 30 min. Afterwards, the melts were poured onto a preheated graphite mold and further annealed at 330°C for 2 h, after which they were cooled slowly inside the furnace to room temperature. The annealed samples for the optical property measurements need to be double-sided polishing into 10 × 10 × 1.5 mm³ cylinders. Densities of glasses were determined by the Archimedes' principle using the distilled water as the medium. The refractive index of all the samples was measured by the prism coupling method (Metricon Model 2010) at 633, 1,309, and 1,533 nm with an error of ±5 × 10⁻⁴. The infrared transmittance spectra were obtained using Vector 33 Fourier transform infrared (FTIR) spectrophotometer (Bruker, Switzerland). The Raman spectra were measured by Raman spectrometer (Renishawin Via, Gloucestershire, UK) and 532 nm laser as the excitation source. Optical absorption spectra measurements were performed on a Perkin-Elmer Lambda 900/UV/VIS/NIR spectrophotometer. The fluorescence spectra were recorded by a computer-controlled Triax 320 type spectrofluorimeter (Jobin-Yvon Corp.) equipped with an InAs detector upon the excitation of an 808 nm LD. After exciting the samples with an 808 nm LD, InAs detector was used to detect the lifetime of Tm³⁺:³F₄ energy level (1.8 μm) along with a digital phosphor oscilloscope (TDS3012C, Tektronix, America) and signal generator. All of the measurements were carried out at room temperature.

Results and Discussion

Table 1 presents the refractive index (n) and density (ρ) of TGB glasses with different BaF₂ contents. It is found that the refractive index and density monotonously decrease when BaF₂ content increases from 0 to 9 mol.% in step of 3 mol.%. This indicates that the addition of BaF₂ makes glass network looser (Yang et al., 2017), which is demonstrated by the Raman spectra as shown Figure 1. It is noted that three major bands appear in TGB glasses with different BaF₂ amounts. The peak A at ~466 cm⁻¹ is assigned to the symmetrical stretching or bending vibrations of Te-O-Te linkages at corner sharing sites (Murugan and Ohishi, 2004; Jose et al., 2007). The peak B at ~682 cm⁻¹ is ascribed to the anti-symmetric stretching vibrations of Te-O-Te linkages constructed by two un-equivalent Te-O bonds containing bridging oxygens (BO) in TeO₄ trigonal bipyramid and the peak C is due to the symmetrical stretching vibrations of Te-O⁻ and Te=O bonds with non-bridging oxygens (NBO) in TeO₃ trigonal pyramid and TeO₃₊₁ polyhedra (Murugan and Ohishi, 2004; Jose et al., 2007). It is worth noting that the position of peak C slightly shifts from 769 to 787 cm⁻¹ and normalized intensity of peak B declines with the increment of BaF₂ from 0 to 9 mol.%, revealing that glass network structure is broken and more non-bridging oxygens arise. Such low phonon energy of TGB glasses is able to effectively decrease non-radiative relaxation in favor of the enhancement of 2.0 μm emission intensity.

Table 1. The refractive index and density of TGB glasses with different BaF₂ contents.

Figure 1. Normalized Raman spectra of TGB glasses with different BaF₂ amounts.

Figure 2 shows the typical absorption spectra of TGB glasses in the wavelength range from 350 to 2,100 nm. The absorption spectrum consists of five absorption bands of Tm³⁺ centered at 473, 687, 794, 1,214, and 1,700 nm, corresponding to respective transitions from the ³H₆ ground state to excited states ¹G₄, ³F₂,₃, ³H₄, ³H₅, and ³F₄. Energy levels above ¹G₄ energy level are not clearly identified because of strong intrinsic bandgap absorption in the host glass. It is also found that the position and shape of five absorption peaks are almost constant with the addition of BaF₂.

Figure 2. Absorption spectra in TGB glasses with different BaF₂ contents.

When BaF₂ is added, F⁻ ions crack O-H bond in glass network and produce HF gas so that OH⁻ content is reduced. OH⁻ content is reflected by OH⁻ absorption coefficient (α_OH) (Wang et al., 2013).

where l represents the thickness of glass samples, T₀ and T are the incident and transmitted intensity, respectively. According to FTIR spectra, OH⁻ absorption coefficient of TGB glasses is determined and presented in Figure 3. There are two absorption bands centered at 3.1 and 4.4 μm, corresponding to stretching mode of free Te-OH groups and/or stretching mode of molecular water and stretching mode of strong hydrogen-bonded Te-OH groups, respectively (Wang et al., 2019). α_OH at 3.1 μm is obviously higher than the value at 4.4 μm. Moreover, α_OH monotonically decreases from 3.4 to 2.2 cm⁻¹ with increasing BaF₂ content from 0 to 9 mol.% in step of 3 mol.%, which is beneficial to improve 1.8 μm emission properties of Tm³⁺ ions.

Figure 3. FTIR spectra of TGB glasses with different proportions of BaF₂.

Figure 4 compares the fluorescence spectra and decay curves of ${\text{Tm}}^{3+}{:}^3{\text{F}}_4{\to }^3{\text{H}}_6$ transition in TGB glasses with different BaF₂ amounts pumped by 808 nm LD. From Figure 4A, it is clear that the spectra are characterized by two emission peaks located at 1,488 and 1,808 nm, corresponding to ${{}^3\text{H}}_4\to {{}^3\text{F}}_4$ and ${{}^3\text{F}}_4\to {{}^3\text{H}}_6$ transitions, respectively. Emission intensity at 1,488 nm is obviously weaker than that at 1,808 nm, which is attributed to effective cross relaxation process ${(}^3{\text{H}}_4{+}^3{\text{H}}_6\to 2^3{\text{F}}_4$). Moreover, emission intensity at 1,488 nm remains almost unchanged and that near 1.8 μm gradually increases with the increment of BaF₂ concentration. The peak value near 1.8 μm in TGB glasses with 9 mol.% BaF₂ is 1.6 times as high as that without BaF₂ because the reduction of OH⁻ content weakens the interaction between Tm³⁺ and OH⁻ and thus enhances radiative transition probability of ${{}^3\text{F}}_4\to {{}^3\text{H}}_6$ transition. Figure 4B describes fluorescence decay curves of Tm³⁺:³F₄ energy level monitored at 1,808 nm in TGB glasses with different proportions of BaF₂. It is clearly noted that the lifetime of ³F₄ energy level gradually prolongs from 337.4 to 577.8 μs when BaF₂ content increases from 0 to 9 mol.% in step of 3 mol.%. The lifetime in TGB glass with 9 mol.% BaF₂ is 1.7 times as long as the value without BaF₂. These results mean that the addition of BaF₂ can greatly improve 1.8 μm emission properties.

Figure 4. (A) Fluorescence spectra and (B) decay curves of Tm³⁺:³F₄ energy level in TGB glasses with different proportions of BaF₂ pumped by 808 nm LD.

In general, the total decay rate (W) of Tm³⁺:³F₄ energy level is defined as the reciprocal of the measured decay lifetime (τ_m) and is described by the following equations (Zhou et al., 2010).

where A_r represents the radiative decay rate, W_OH is the energy transfer rate between Tm³⁺ and OH⁻, W_mp is the multiphonon decay rate, W_ET is the energy transfer rate between Tm³⁺ ions, N_Tm is the total concentration of Tm³⁺ ions and α_OH is OH⁻ absorption coefficient. k_OH−Tm is defined as the strength of interaction between Tm³⁺ and OH⁻ and doesn't rely on the concentrations of Tm³⁺ and OH⁻. Figure 5 represents a good linear relationship between the total decay rate and α_OH. From this fit, k_OH−Tm is determined and equals to 2.82 × 10⁻¹⁸ cm⁴/s, which is larger than k_OH−Er (1.9 × 10⁻¹⁹ cm⁴/s) (Zhou et al., 2010) and lower than k_OH−Tm (7.89 × 10⁻¹⁸ cm⁴/s) in germanate glasses (Wang et al., 2014).

Figure 5. The dependence of the total decay rate on α_OH along with the red fitting curve.

Based on above-mentioned results, Figure 6 shows energy transfer mechanism. Under excitation at 808 nm LD, Tm³⁺ ions are motivated to ³H₄ state from the ³H₆ ground state. Then, a few Tm³⁺ ions return radiatively to ³F₄ state with 1,488 nm photon. However, the majority of ions relax nonradiatively to ³F₄ state via muliphonon relaxation process and efficient cross relaxation process (CR) between two adjacent Tm³⁺ ions ${(}^3{\text{H}}_4+{}^3{\text{H}}_6\to 2^3{\text{F}}_4$). Finally, Tm³⁺ ions in the excited ³F₄ state return to the ³H₆ ground state, emitting fluorescence at 1.8 μm. Significantly, the residual OH⁻ in TGB glasses can impair 1.8 μm emission via two OH⁻ ions, indicating that it is essential to decrease the hydroxyl content for improving 1.8 μm emission.

Figure 6. The energy level diagram and energy transfer mechanism of Tm³⁺ ions.

Both absorption and emission cross sections of Tm³⁺ ions are very crucial parameters to evaluate the potential of TGB glasses as 2 μm laser material. Based on the Beer-Lambert equation and Fuchtbauer-Ladenburg equation (Chen et al., 2007), absorption and emission cross sections of ${\text{Tm}}^{3+}{:}^3{\text{H}}_6\leftrightarrow {}^3{\text{F}}_4$ transition in TGB glass with 9 mol.% BaF₂ are calculated and presented in Figure 7A. The maximum absorption cross section of Tm³⁺ reaches 5.3 × 10⁻²¹ cm² at 1,706 nm, which is higher than that of silicate glass (1.5 × 10⁻²¹ cm²) (Li et al., 2012), fluorophosphate glass (3.0 × 10⁻²¹ cm²) (Li et al., 2015), tellurium germanate glass (3.2 × 10⁻²¹ cm²) (Gao et al., 2015) and germanate glass (4.1 × 10⁻²¹ cm²) (Yu et al., 2009). Moreover, corresponding maximum emission cross section is 8.8 × 10⁻²¹ cm² at 1,814 nm, which is higher than that of silicate glass (3.6 × 10⁻²¹ cm²) (Li et al., 2012), fluorophosphate glass (5.5 × 10⁻²¹ cm²) (Li et al., 2015), tellurium germanate glass (6.8 × 10⁻²¹ cm²) (Gao et al., 2015), germanate glass (5.5 × 10⁻²¹ cm²) (Yu et al., 2009) and zinc tellurite glass (7.3 × 10⁻²¹ cm²) (Yuan and Xiao, 2018). The high emission cross section of TGB glass with 9 mol.% BaF₂ is helpful to provide high laser gain.

Figure 7. (A) Absorption, emission cross section and (B) calculated gain coefficient of TGB glass with 9 mol.% BaF₂.

Once absorption and emission cross sections are determined and it is supposed that Tm³⁺ ions are only in either the ³H₆ or ³F₄ state, the gain coefficient G(λ) of Tm³⁺ near 1.8 μm can be obtained by the following equation (Zou and Toratani, 1996).

where N represents the total concentration of Tm³⁺ ions and p is the inversion factor given by the ratio between the population of lasing upper level (³F₄) and the total concentration that ranges from 0 to 1. Figure 7B shows gain coefficient spectrum of TGB glass with 9 mol.% BaF₂. It is found that the gain peak shifts to shorter wavelength with increasing p, which is a typical feature of the quasi-three-level system. Moreover, gain coefficient starts to be greater than zero in the wavelength range from 1,824 to 2,100 nm when p ≥ 0.2 and the maximum value is 3.3 cm⁻¹ at 1,814 nm, which is larger than that of silicate glass (2.57 cm⁻¹) (Tang et al., 2016), germanate glass (2.11 cm⁻¹) (Slimen et al., 2019) and tellurite glass (0.91 cm⁻¹) (Tian et al., 2019). This means that TGB glass with 9 mol.% BaF₂ is a promising host material for efficient 2.0 μm fiber laser development.

Conclusions

In summary, the effects of BaF₂ contents on density, refractive index, Raman spectra, OH⁻ absorption coefficient and 1.8 μm spectroscopic properties of Tm³⁺ doped gallium tellurite glasses are studied in detail. When BaF₂ content increases from 0 to 9 mol.% in step of 3 mol.%, refractive index and density gradually reduce. Meanwhile, α_OH monotonically decreases from 3.4 to 2.2 cm⁻¹, which makes emission peak value near 1.8 μm in TGB glass with 9 mol.% BaF₂ being 1.6 times as large as that without BaF₂ while the lifetime becomes 1.7 times as long as the value without BaF₂. The maximum emission cross section at around 1.8 μm of TGB glass with 9 mol.% BaF₂ reaches 8.8 × 10⁻²¹ cm². In addition, positive gain coefficient in the wavelength range from 1,824 to 2,100 nm is achieved when p ≥ 0.2 and the maximum gain coefficient is 3.3 cm⁻¹ at 1,814 nm. As a result, TGB glass with 9 mol.% BaF₂ appears to be a highly promising host material for efficient 2.0 μm fiber laser development.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author Contributions

JY, PX, and WW conceived the idea. JY and PX wrote the paper. TD, YY, DO, JC, and SY advised the paper. All authors contributed to the article and approved the submitted version.

Funding

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (Grant Nos. 51902053 and 61804029), Natural Science Foundation of Guangdong province (Grant Nos. 2019A1515011988 and 2018A030310353), Guangdong Basic and Applied Basic Research Foundation (Grant No. 2019A1515110002), the Foundation for Distinguished Young Talents in Higher Education of Guangdong (Grant No. 2019KQNCX172), the Project of Foshan Education Bureau (Grant No. 2019XJZZ02), Guangdong-Hong Kong-Macao Intelligent Micro-Nano Optoelectronic Technology Joint Laboratory (Grant No. 2020B1212030010), and the Open Fund of the State Key Laboratory of Luminescent Materials and Devices (South China University of Technology, Grant No. 2020-skllmd-13).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References