Introduction

Recently, mid-infrared (MIR) lasers operating at around 3 μm have triggered increasing interest in coherent MIR light sources. This is mainly due to the strong absorption band by water at this wavelength and ~3 μm MIR lasers cover a great number of important molecular characteristic spectral lines, which make MIR fiber lasers appropriate in the fields of medical surgery, remote sensing, environmental monitoring and so forth1,2,3,4,5. Er3+ is an ideal emission candidate for 3 μm emission due to the transition of 4I11/2 → 4I13/2 and the absorption bands of Er3+ match well with the commercially available and low-cost 808 nm or 980 nm laser diode (LD)6. Among Er3+-doped materials, TeO2-based GCs have attracted great attention due to the following advantages. On one hand, GCs combine the advantages of excellent fiber-drawing ability of glass7,8 and strong crystal field of nanocrystals9,10. On the other hand, tellurite glasses possess wide transmission region (from 0.35 to 5 μm), the lowest phonon energy (~750 cm−1) among the common oxide glasses, large refractive index (~2.0), high solubility for rare earth ions and good chemical durability etc11,12,13,14. The low phonon energy is very beneficial to the 2.7 μm emission of Er3+ ions due to the narrow gap between 4I11/2 and 4I13/2 levels. The large refractive index enhances emission cross section and improves the refractive index match between nanocrystals and glass matrix.

The successful fabrication of a new cubic crystalline phase in K2O-Nb2O5-TeO2 glasses in 1995 and the investigation on structure in TeO2-PbF2-CdF2 GCs in 2002 have triggered great interests in tellurite GCs15,16. Although a few studies have been conducted on tellurite GCs, to our best knowledge, no work is reported about the 2.7 μm emission of Er3+-doped germanotellurite glasses and GCs system. Not to mention the effect of GeO2 concentration on the spectroscopic properties of Er3+-doped germanotellurite glasses. In this work, we obtained enhanced 2.7 μm emission from Er3+-doped germanotellurite glasses and GCs, and the influence of GeO2 concentration and heat-treated temperature on the spectroscopic properties of Er3+-doped germanotellurite glasses and GCs are discussed in detail.

Results and Discussion

The effect of Te/Ge ratio on the spectroscopic properties

In order to analyze the effect of GeO2 concentration on the thermal property, DSC curves of representative Te59Ge0, Te29Ge30, and Te0Ge59 samples are measured and shown in Fig. 1. We can observe that the glass transition temperature (Tg), onset crystallization temperature (Tx), and crystallization peak temperature (Tp) increase significantly with the increase of GeO2 concentration. It can be ascribed to the fact that GeO2 acts as a network former and can collect glass network and strengthen the structure of the network17. In order to evaluate the thermal stability of prepared samples, the parameter ∆T (Tx − Tg) is obtained, which is usually used to estimate the glass stability. It can be seen from Table 1 that the values of ∆T increases with the increase of GeO2 concentration. A large ∆T means the strong inhibition of nucleation and crystallization.

Figure 1
figure 1

DSC curves of Te(59−x)Gex (x = 0, 30, 59) glasses.

Table 1 The temperatures of glass transition (Tg), onset crystallization (Tx), peak crystallization (Tp), and thermal stability parameter ∆T in various glasses.

In order to investigate the evolution of glass structure with the component variation, Raman spectra are carried out on precursor glasses. Figure 2(a) shows the Raman spectra of undoped precursor glasses excited by a 532 nm laser. The peak located at 462 cm−1 can be attributed to the stretching mode of Te-O-Te symmetric bridges, the shoulder about 679 cm−1 is identified as the stretching vibration of Te-O bonds in the TeO4 trigonal bipyramid, and another peak located at 764 cm−1 is assigned to the stretching vibration of Te-O bond in TeO3+1 or TeO3 units17,18. It is known that the intensities and the positions of the Raman peaks depend on the concentration and kind of structural groups. As the GeO2 concentration increases, the peaks at 462 cm−1 and 764 cm−1 gradually shift toward higher frequencies, and at the same time the intensity of the peak at 764 cm−1 decreases. This behavior is due to the formation of GeO4 units and a part of TeO3+1 or TeO3 units are replaced by GeO4 units. One can observe that when TeO2 is absent, the peaks shifted to 536 cm−1 and 835 cm−1, which corresponds to the symmetric stretching and asymmetric stretching of Ge-O-Ge bridges in GeO4 tetrahedra, respectively19,20. Additionally, as the location of the Raman peak is relevant with the phonon energy of glass, the addition of GeO2 leads to an increase of the phonon energy of glass.

Figure 2
figure 2

(a) Raman spectra, and (b) FTIR spectra of undoped Te(62−x)Gex (x = 0~62) glasses.

To clarify the influence of GeO2 concentration on the content of OH in glass, the FTIR spectra of precursor glasses are recorded in Fig. 2(b). The bands at around 1636 cm−1 and 3439 cm−1 are belong to the bending vibration and stretching vibration of the OH group, respectively. The band at approximately 1423 cm−1 can be assigned to the stretching vibration of the group21,22. It can be easily observed that the content of OH increases gradually with the increase of GeO2 concentration. On the basis of electronegativity theory, covalency decreases with the difference of electronegativity between cation and anion ions. The values of electronegativity for Te, Ge, and O elements are about 2.1, 1.8, and 3.5, respectively23. Since the larger electronegativity difference between Ge and O indicates weaker covalency of Ge-O bond, the ionicity of Ge-O bond is stronger than that of Te-O bond. Therefore, Ge4+ will absorb more OH group, which results in the higher content of OH with the increase of GeO2 concentration. The bands in the range from 500 to 1200 cm−1 are assigned to the vibrations of glass network former, and the locations of the peaks shift to longer wavenumber with the addition of GeO2 content, which matches well with the result of Raman spectra as mentioned before.

Figure 3 shows the absorption spectra of precursor glasses and GCs in the range of 500–3200 nm. Five absorption bands at 524 nm, 653 nm, 800 nm, 978 nm and 1535 nm are found and attributed to the transitions of Er3+ ions from the ground state 4I15/2 to the excited states 2H11/2, 4F9/2, 4I9/2, 4I11/2 and 4I13/2, respectively. The shape and peak positions of each transition remain unchanged with the increase of GeO2 concentration. The inset in Fig. 3 shows the enlarged image of absorption spectra of germanotellurite glasses in the range from 2700 nm to 3200 nm. It can be clearly observed that the absorption at around 3 μm corresponding to the stretching vibrations of -OH group increases monotonically with increasing GeO2 content, which can also be explained by electronegativity theory mentioned above. Obvious absorption peaks around 980 nm indicates that the prepared samples can be efficiently pumped by a commercial 980 nm LD.

Figure 3: Absorption spectra of 3% Er3+-doped Te(59−x)Gex (x = 0~59) glasses and Te59Ge0 GCs heat-treated at different temperature for 8 h.
figure 3

The inset shows the enlarged image of absorption spectra in the range from 2700 nm to 3200 nm.

According to absorption spectra, Judd-Ofelt parameters Ωλ (λ = 2, 4, 6) can be determined by J-O theory, which have been successfully used to calculate 4 f electrons transitions of RE ions in different host materials24,25. Based on J-O theory, Ω2 is closely related to the covalency between RE ions and ligand anions, which reflects the asymmetry of local environment around the site of RE ions. The larger the Ω2 is, the stronger the covalency between RE ions and ligand anions becomes. Table 2 shows the J-O parameters of Er3+ in different samples. Based on the electronegativity theory mentioned above, the covalency of the Er-O bond increases with the increase of GeO2 concentration. As a result, the values of the parameter Ω2 show an obvious increase with increasing GeO2 content.

Table 2 J-O intensity parameters of Er3+ in different samples.

Figure 4(a) shows the MIR emission spectra of Er3+-doped germanotellurite glasses under a 980 nm LD excitation. Obvious 2.7 μm emission peaks corresponding to Er3+: 4I11/2 → 4I13/2 transition can be observed. It is found that the 2.7 μm emission intensity decreases with the increase of GeO2 concentration. It is known that the intensity of MIR emission is very sensitive to the multiphonon relaxation rate of RE ions which strongly depends on the phonon energy of their matrix. Usually, the lower the phonon energy of host matrix is, the smaller the multiphonon relaxation probability is. In addition, the OH impurity in glasses can absorb and quench 2.7 μm emission signals, which is disadvantage for MIR emission. From the previous results, both the content of OH and the phonon energy of glass increase with the increase of GeO2 concentration. As a result, the 2.7 μm emission intensity decreases with the increasing GeO2 content. Based on the absorption spectra, the absorption cross section (σabs) at 980 nm can be defined as26:

Figure 4
figure 4

(a) 2.7 μm emission spectra, and (b) Absorption cross section at 980 nm of 3% Er3+-doped Te(59−x)Gex (x = 0~59) glasses (c) 2.7 μm emission spectra of 3% Er3+-doped Te59Ge0 glass and GCs heat-treated at 400 °C–430 °C for 8 h. (d) Composition and heat-treated temperature dependence integrated intensity of 2.7 um emission peaks; (e) Absorption and emission cross sections, and (f) gain cross sections at 2.7 μm in 3% Er3+-doped Te59Ge0 GC heat-treated at 430 °C for 8 h. The inset in figure (e) shows the emission cross sections of 3% Er3+-doped Te(59−x)Gex (x = 0~59) glasses.

where OD(λ), N, and l represent the optical density, Er3+ doping concentration and sample thickness, respectively. The absorption cross section at 980 nm have been calculated and displayed in Fig. 4(b). It can be obtained that the peak absorption cross sections decrease with increasing GeO2 concentration. Higher absorption cross section is beneficial to the 2.7 μm emission. In order to estimate the 2.7 μm emission properties for potential laser applications, the emission cross sections at 2.7 μm in Er3+-doped germanotellurite glasses are calculated by the Fuchtbauer-Ladenburg equation27 and shown in the inset of Fig. 4(e). The detailed calculation processes can refer to ref. 14. We observe that the peak emission cross section decreases with the increasing GeO2 content.

Figure 5(a,b) display the NIR and UC emission spectra of Er3+-doped germanotellurite glasses, respectively. It can be found that enhanced 1.53 μm emission corresponding to Er3+: 4I13/2 → 4I15/2 transition and 532 nm, 548 nm, and 669 nm emission corresponding to Er3+: 2H11/2 → 4I15/2, 4S3/2 → 4I15/2 and 4F9/2 → 4I15/2 transitions are observed. The NIR and UC emission intensity decrease with the increase of GeO2 concentration. The inset in Fig. 5(b) shows the ratio of red to green emission peaks in Er3+-doped glasses with different GeO2 concentration and the ratio increases gradually with the increase of GeO2 concentration. This can be explained by the energy level diagram of Er3+ shown in Fig. 5(b). For the green emission, electrons at 4I15/2 level are excited to 4I11/2 level through ground state absorption (GSA) under 980 nm LD pumping, then the excited electrons at 4I11/2 level are excited further to 4F7/2 level through excited state absorption (ESA1) or energy transfer upconversion process (ETU: 4I11/2 + 4I11/2 → 4I15/2 + 4F7/2). The electrons of 4F7/2 level decay nonradiatively to the next lower 2H11/2 and 4S3/2 levels, and the 2H11/2 → 4I15/2, 4S3/2 → 4I15/2 transitions produce green emissions. The red emission is originated from the 4F9/2 → 4I15/2 transition. The excited electrons at 4I13/2 level are excited to 4F9/2 level through excited state absorption (ESA2) or the electrons at 4S3/2 level decay nonradiatively to 4F9/2 level, then the 4F9/2 → 4I15/2 transition gives rise to red emission. It can be obtained from the energy level diagram of Er3+ that the electrons of 4F7/2 and 4F9/2 levels are primarily dominated by the 4I11/2 and 4I13/2 levels, respectively. As we all know, the quite narrow energy gap between 4I11/2 and 4I13/2 levels makes the nonradiative relaxation in 4I11/2 level remarkable for hosts with large phonon energy. While due to the relatively larger energy gap between 4I13/2 and 4I15/2 levels than that between 4I11/2 and 4I13/2 levels, the influence of matrix phonon energy on the nonradiative relaxation in 4I13/2 level is relatively small. Therefore, the nonradiative relaxation rate from 4I11/2 level becomes larger than that from 4I13/2 level with the increase of GeO2 concentration. In addition, cross relaxation process (CR: 4I11/2 + 4I13/2 → 4F9/2 + 4I15/2) also makes a contribution to the red emission. As a result, the ratio of red to green emission peaks gradually grows with increasing GeO2 concentration.

Figure 5
figure 5

(a) 1.53 μm NIR and (b) UC emission spectra of 3% Er3+-doped Te(59−x)Gex (x = 0~59) glasses pumped by a 980 nm LD; (ce) Composition and heat-treated temperature v.s. integrated intensity of 1.53 um, 548 nm and 669 nm emission peaks, respectively. The inset in figure (b) shows the GeO2 concentration dependence the ratio of red to green emission peaks and the energy level diagram of Er3+ under 980 nm LD excitation.

Figure 6 reveals the fluorescence decay curves of Er3+: 4I13/2 level (Fig. 6(a)) and 4I11/2 level (Fig. 6(b)) excited by 808 and 980 nm LD as a function of GeO2 concentration. The fluorescence lifetime of a rare-earth ion in an excited state is mainly determined by the radiative decay rate, multiphonon decay rate, and energy transfer rate. Since the concentration of Er3+ ions in all of the samples is the same, the energy transfer rate of Er3+ ions should be close for all of the glass samples. Therefore, the measured lifetime of Er3+: 4I13/2 level and 4I11/2 level in glasses is mainly determined by the radiative decay rate and the multiphonon decay rate. In addition, the non-single exponential shape of the decay curves of Er3+: 4I13/2 level is probably associated to energy transfer. Table 3 shows the lifetimes of Er3+: 4I13/2 and Er3+: 4I11/2 levels for different samples. It can be seen that both the fluorescence lifetime of Er3+: 4I13/2 level and Er3+: 4I11/2 level decrease with the increase of GeO2 concentration. The content of OH and phonon energy of germanate glasses are higher than that of tellurite glasses, which can be confirmed by the FTIR and Raman spectra. Thus, the introduction of GeO2 to tellurite glass will lead to a decrease for the lifetime of Er3+: 4I13/2 level and 4I11/2 level by increasing the content of OH and phonon energy of glass, which will subsequently increase the multiphonon decay rate of Er3+ in glass samples. Therefore, it can be concluded that multiphonon decay rate has a dominant influence on the lifetime of Er3+: 4I13/2 level and 4I11/2 level for the glass samples.

Figure 6
figure 6

Fluorescence decay curves of (a) Er3+: 4I13/2 level pumped by pulsed 980 nm LD, and (b) Er3+: 4I11/2 level pumped by pulsed 808 nm LD in 3% Er3+-doped Te(59-x)Gex (x = 0~59) glasses.

Table 3 The lifetimes of Er3+: 4I13/2 and Er3+: 4I11/2 levels for different samples.

The effect of heat-treated process on the spectroscopic properties

The XRD patterns of representative Te59Ge0, Te29Ge30, and Te0Ge59 samples heat-treated at different temperature are shown in Fig. 7(a,b and c), respectively. The precursor glasses are completely amorphous with no diffraction peaks. After a heat-treated process, several diffraction peaks, a characteristic of crystalline structure, are clearly observed, which are assigned to Li2TeO3 (JCPDF: 26–1192) and BaTiO3 (JCPDF: 34–0129) for (a), Li2TeO3 (JCPDF: 26–1192), Li2GeO3 (JCPDF: 34–0659), BaTiO3 (JCPDF: 34–0129) and BaTe2O5 (JCPDF: 36–0886) for (b), Li2GeO3 (JCPDF: 34–0659) and BaTiO3 (JCPDF: 34–0129) for (c). With the increase of heat-treated temperature, the diffraction peaks become more evident and sharper, which indicates crystalline size grows gradually. What is more, the crystallization is a bulk process (Supplementary Fig. S1).

Figure 7
figure 7

(ac) XRD patterns of 3% Er3+-doped Te59Ge0, Te29Ge30, and Te0Ge59 glasses and GCs heat-treated at different temperature for 8 h. (d) the dependence of crystalline size on heat-treated temperature.

From the peak width of XRD pattern, the crystalline size of samples can be calculated on the basis of the Scherrer’s equation28:

where D is the grain size of nanocrystals, K is a dimensionless shape factor, λ is the X-ray wavelength, β is the full-width at half maximum (FWHM) of the diffraction peak and θ is the Bragg angle. The diffraction peaks at 2θ = 26.17° in (a) and (b), and 2θ = 32.35° in (c) are used to obtain the crystalline size of Li2TeO3 and Li2GeO3 nanocrystal, respectively. The calculated crystalline size dependent on heat-treated temperature is plotted in Fig. 7(d). It can be seen that the crystalline size increases monotonically with increasing heat-treated temperature.

Figure 8(a) shows the TEM image of GC-Te59Ge0 heat-treated at 400 °C for 8 h. The quasi-spherical particles dispersed in glass matrix homogeneously. The high-resolution transmission electron microscope (HRTEM) image in Fig. 8(b) displays the fine lattice structure with corresponding interplanar spacing about 0.34 nm, which can be ascribed to the (004) plane of Li2TeO3 nanocrystal. The particle size distribution histogram shown in Fig. 8(c) suggests that the particle size is distributed with diameters in the range of 8–20 nm with an average size of ~14 nm, which is in good agreement with our calculated result based on the XRD data.

Figure 8
figure 8

(a) TEM micrograph, (b) HRTEM image, and (c) size distribution histogram of 3% Er3+-doped Te59Ge0 GC heat-treated at 400 °C for 8 h.

The absorption spectra of Te59Ge0 GCs in the range of 500–3200 nm are displayed in Fig. 3. It can be seen that the baselines of the absorption curves become more precipitous toward to the shorter wavelength region with the increase of heat-treated temperature. It can be ascribed to the transmittance loss causing by the Rayleigh scatting effect, which is resulted from the growth of nanocrystals. According to absorption spectra, the J-O parameters of Er3+ in Te59Ge0 GCs are calculated and shown in Table 2. Based on J-O theory, the parameter Ω2 is sensitive to the environmental configuration symmetry of RE ions, and it decreases with the increase of heat-treated temperature, which indicates the variation of RE ions environment from glass to GCs. It can be seen from Table 2 that the value of Ω2 decreases with the increase of heat-treated temperature confirming the incorporation of Er3+ into the precipitated nanocrystals. In addition, the value of Ω6, which is proportional to the rigidity of the host, increases slightly in the transition from glass to GCs29.

Figure 4(c) shows the MIR emission spectra of Te59Ge0 glass and GCs under a 980 nm LD excitation. It is found that the 2.7 μm emission intensity enhances remarkably with the increase of heat-treated temperature. Moreover, a slight red shift of emission peak emerges in the GCs compared with the precursor glass, which is mainly due to the crystal field effect. Figure 4(d) shows the contrast of 2.7 μm emission intensity produced by Er3+-doped glasses with different heat-treated temperature and GeO2 concentrations. The 2.7 μm emission intensity enhances with the increase of heat-treated temperature under the condition of the same GeO2 concentration. According to the XRD pattern, the nanocrystals have better crystallinity at a higher temperature, inducing the enhancement the crystal field effect. Therefore, the 2.7 μm emission intensity enhances significantly with increasing crystallization temperature.

Figure 4(e) displays the absorption and emission cross sections at 2.7 μm in Er3+-doped tellurite GC heat-treated at 430 °C for 8 h. The peak absorption and emission cross sections in the present GC are 0.68 × 10−20 cm2 and 0.85 × 10−20 cm2, respectively. The obtained σem for present sample is higher than those of Er3+-doped bismuthate (0.77 × 10−20 cm2), germanate (0.70 × 10−20 cm2) glass and oxyfluoride GC (0.43 × 10−20 cm2), but smaller than the GC containing SrF2 crystals (1.20 × 10−20 cm2)30,31,32,33. Higher emission cross section is benefit for higher possibility to achieve laser output. In addition, Te59Ge0 glass still has a good transparency after heat treatment (Supplementary Fig. S2). The gain cross section, which is an important parameter to evaluate the gain performance of prepared sample, can be calculated by the following equation34:

where P is the population inversion given by the ratio between the concentration of the Er3+: 4I11/2 level and the total Er3+ concentration. The gain cross sections for different P values are shown in Fig. 4(f). It is found that the gain becomes positive when P is more than 0.4 implying that a low pumping threshold can be achieved for the Er3+: 4I11/2 → 4I13/2 laser operation.

Figure 5(c–e) show the composition and heat-treated temperature dependent integrated intensity of 1.53 μm, 548 nm and 669 nm emission peaks in Er3+-doped germanotellurite GCs. It is clear shown that all the emission intensity increase with the increasing heat-treated temperature in the case of the same GeO2 concentration. These results exhibit similar trend with MIR emission, which can be explained by the similar crystal field effect mentioned above.

Conclusions

In summary, Er3+-doped germanotellurite glasses and GCs are prepared and the structural and spectroscopic properties are investigated. From FTIR and Raman spectra, it is noted that the addition of GeO2 increases the content of OH and phonon energy of glasses. J-O intensity parameters are calculated and discussed. Moreover, the influence of GeO2 concentration and heat-treated temperature on the MIR, NIR and UC emissions are discussed in detail. Enhanced 2.7 μm emission is observed in Er3+-doped germanotellurite GCs and the tellurite GC heat-treated at 430 °C for 8 h possesses large absorption (0.68 × 10−20 cm2) and emission (0.85 × 10−20 cm2) cross sections for Er3+: 4I11/2 → 4I13/2 transition. These results imply that this GC might be a promising candidate for Er3+-doped MIR fiber lasers.

Methods

Preparation of Er3+-doped GCs

Germanotellurite glasses with molar compositions of (59-x) TeO2-x GeO2-8 TiO2-8 BaO-22 Li2O-3 Er2O3 (x = 0, 10, 20, 30, 40, 50, 59), which denoted as Te59Ge0, Te49Ge10, Te39Ge20, Te29Ge30, Te19Ge40, Te9Ge50, and Te0Ge59 correspondingly, were prepared by the conventional melt-quenching method. TeO2 (5N), GeO2 (5N), TiO2 (4N), BaCO3 (A.R.), Li2CO3 (A.R.), and Er2O3 (4N) were used as raw materials and each batch of 20 g was well-mixed and melted in a covered corundum crucible at 1100 °C for 40 min at ambient temperature. The melt was cast onto a preheated copper plate and then cooled down to room temperature. The glass samples were cut to the size of 10 mm × 10 mm and heat-treated at different temperature for 8 h to achieve GCs. Then the samples were polished to optical quality before measurements with a thickness of 1.5 mm.

Characterization

Differential scanning calorimetry (DSC) was performed in a simultaneous thermal analyzer (STA449C NETZSCH) under N2 atmosphere with heating rate at 10 °C /min in order to investigate the thermal property and determine the crystallization characteristic temperatures. XRD measurements were carried out on a X’Pert PRO X-ray diffractometer by using Cu Kα1 as a radiation source to identify the crystalline phase and estimate the nanocrystal grain size. Transmission electron microscope (TEM, 2100 F, JEOL, Japan) was performed to analyze the micrograph and microstructure of GC sample. Raman spectra were recorded in the range of 200–1100 cm−1 using a Raman spectrometer (Renishaw in Via, London, UK) with an excitation of 532 nm laser. The FTIR spectra were obtained on a Vector 33 Fourier transform infrared spectrophotometer (Bruker, Switzerland) to estimate the content of OH in glasses. The absorption spectra were measured in the range of 500–3200 nm using a Perkin-Elmer Lambda 900/UV/VIS/NIR spectrophotometer. The mid-infrared (MIR) emission spectra in the range of 2550–2850 nm were recorded by a spectrometer (zolix, Omni 5015i, Beijing, China) with a lock-in amplifier upon excitation of a 980 nm LD. The near-infrared (NIR) and upconversion (UC) emission spectra pumped by 980 nm LD were measured on a computer controlled Triax 320 type spectrofluorimeter (Jobin-Yvon Corp.). The fluorescence decay curves of Er3+: 4I11/2 and 4I13/2 levels were captured by a Tektronix TDS 3012c Digital Phosphor Oscilloscope with pulsed 808 and 980 nm LDs as excitation sources, respectively. All the measurements were carried out at room temperature.

Additional Information

How to cite this article: Kang, S. et al. Spectroscopic properties in Er3+-doped germanotellurite glasses and glass ceramics for mid-infrared laser materials. Sci. Rep. 7, 43186; doi: 10.1038/srep43186 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.