Er³⁺/Ho³⁺-Codoped Fluorotellurite Glasses for 2.7 µm Fiber Laser Materials

Abstract

This work reports the enhanced emission at 2.7 µm in Er³⁺/Ho³⁺-codoped fluorotellurite glass upon a conventional 980 nm laser diode. The significantly reduced green upconversion and 1.5 µm emission intensity in Er³⁺/Ho³⁺-codoped samples are observed. The results suggest that the Er³⁺: ⁴I_13/2 state can be efficiently depopulated via energy transfer from Er³⁺ to Ho³⁺ and the detailed energy transfer mechanisms are discussed qualitatively. The energy transfer efficiency from Er³⁺: ⁴I_13/2 to Ho³⁺: ⁵I₇ is calculated to be as high as 67.33%. The calculated emission cross-section in Er³⁺/Ho³⁺-codoped fluorotellurite glass is 1.82 × 10⁻²⁰ cm². This suggests that Er³⁺/Ho³⁺-codoped fluorotellurite glass is a potential material for 2.7 µm fiber laser.

1. Introduction

Owing to the increased interest in mid-infrared laser fiber (2–5 µm) used in laser surgery and remote chemical sensing fields, considerable researches have been performed to searching for new materials to use as hosts for mid-infrared laser hosts especially for Er³⁺ 2.7 µm [1,2]. Among many alternatives, fluoride fibers have emerged as natural candidates for mid-infrared laser materials because of their low phonon energy which decreases the rate of phonon-assisted nonradiative transitions [3,4]. However, fluoride fibers suffer from poor thermal stability and require complex fabrication route. Usually, the mid-infrared emission of Er³⁺ can hardly be observed in oxide glasses owing to the large phonon energy. However, it is well known that oxide glasses are more chemically and thermally stable. Among all the oxide glasses, tellurite glasses emerge as good candidates for mid-infrared fiber laser materials because of their lowest phonon energy (760 cm⁻¹) among all the oxide glasses with large refractive index and a broad transmission window (0.4–6 μm) [5,6,7].

Er³⁺ is an ideal luminescent center for 2.7 µm mid-infrared emission corresponding to the ⁴I_11/2→⁴I_13/2 transition. However, Er³⁺ suffers from self-terminating of the ⁴I_11/2 level because of the much shorter lifetime of the emitting level (⁴I_11/2) as compared to the terminal laser level (⁴I_13/2). Fortunately, codoping with other ions such as Pr³⁺, Nd³⁺, Tm³⁺ and Ho³⁺ have been proved to be feasible alternatives to enhance the 2.7 µm emission [8,9,10,11]. The strong OH⁻ absorption around 3 µm is another important fact to obtain efficient Er³⁺ 2.7 µm emission. As is reported before [12], the addition of fluoride in the tellurite glasses was proved to be an effective way to reduce OH⁻¹ groups and increase the radiative transition probabilities of 2.7 µm emission. Therefore, we prepare the Er³⁺/Ho³⁺-codoped fluorotellurite glass and evaluate the spectroscopic parameters based on the absorption spectra using the Judd-Ofelt theory. The detailed energy transfer processes based on the measured upconversion, near-infrared and mid-infrared fluorescence spectra are also discussed.

2. Experimental Section

The investigated fluorotellurite glasses in this study have the following molar compositions: 85TeO₂-10PbF₂-5LaF₃-1ErF₃-xHoF₃(x = 0, 2), hereafter named TF glass. The samples were prepared using high-purity of powder. Well-mixed, 25 g batches of the samples were placed in an aluminum crucible and melted at 900 °C for 30 min. Then the melts were cast on a preheated steel plate and annealed for 3 h. at a temperature 10 °C below the Tg before they were naturally cooled to room temperature. The annealed samples were polished with a thickness of 1 mm for the optical property measurements.

The absorption spectra were recorded by a Perkin-Elmer Lambda 900 UV/VIS/NIR spectrophotometer in the wavelength range of 400–1700 nm. The fluorescence spectra were measured with an Edinburg FLSP920 type spectrometer upon excitation at 980 nm. The fluorescence lifetime was measured with the instrument FLSP920 (Edinburgh instruments Ltd., UK). All the measurements were carried out at room temperature.

3. Results and Discussion

3.1. Absorption Spectra and Judd-Ofelt Analyses

Figure 1 shows the absorption spectra of Er³⁺ singly and Er³⁺/Ho³⁺-codoped TF glasses. All the intrinsic absorption transitions of Er³⁺ and Ho³⁺ in the region from 300 to 1700 nm are retained and labeled in Figure 1. The strong absorption around 980 nm of the Er³⁺/Ho³⁺-codoped sample indicates that this glass can be excited efficiently by a 980 nm laser diode (LD). It can be seen that Er³⁺: ⁴F_9/2, Ho³⁺: ⁵F₅ and Er³⁺: ⁴S_3/2, Ho³⁺: (⁵S₂ + ⁵F₄) are very close, which show that the energy transfer processes in Er³⁺/Ho3+-codoped TF glasses are expected to be efficient.

Figure 1. Absorption spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped samples.

Figure 1. Absorption spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped samples.

The Judd-Ofelt theory [13,14] has been commonly applied to determine the important spectroscopic and laser parameters of rare earth doped glasses. Judd-Ofelt theory has been described in other literature in detail [15]. Basically, the Judd-Ofelt analyses were applied using the experimental oscillator strengths of the absorption bands obtained from absorption spectra. Judd-Ofelt intensity parameters and oscillator strengths provide indirect information of the symmetry and bonding of rare-earth polyhedra within the matrix. Experimental (f_mea) and theoretical (f_cal) oscillator strength for representative transitions of Er in TF glass are tabulated in Table 1. Then the Judd-Ofelt intensity parameters, Ω_λ (λ = 2, 4, 6), can be derived using a least-square fitting approach and are shown in Table 2. Generally, Ω_λ is closely related to the structure change of the sites of rare earth ligand and the basicity of the glass network. It is hypersensitive to the change of composition of host materials. As is shown in Table 2, the calculated Ω₂ of Er³⁺ and Ho³⁺ in the present glass is lower than that other oxide glasses since the addition of fluoride in the tellurite glass can reduce the covalency and ligand of the glass matrix. The Ω_λ parameters follow the trend Ω₂ > Ω₄ > Ω₆ in present TF glass. It should be mentioned that the root-mean-square is 2.98 × 10⁻⁶ for Er³⁺/Ho³⁺-codoped TF glass. The larger value of the fitting is due to the overlap of energy levels of Er³⁺ and Ho³⁺ we select to calculate the intensity parameters Ω_λ.

Table 1. Measured and calculated oscillator strength of Er³⁺ in TF glass.

Absorption	Wavelength (nm)	Oscillator strength (10−6)
		Measured	calculated
4I15/2→4I11/2	975	0.752	0.882
4I15/2→4I9/2	799	0.344	0.496
4I15/2→2H11/2	521	7.448	9.269
4I15/2→4F7/2	488	3.784	3.139
4I15/2→2G11/2	378	20.337	16.597

Table 1. Measured and calculated oscillator strength of Er³⁺ in TF glass.

Absorption	Wavelength (nm)	Oscillator strength (10−6)
		Measured	calculated
4I15/2→4I11/2	975	0.752	0.882
4I15/2→4I9/2	799	0.344	0.496
4I15/2→2H11/2	521	7.448	9.269
4I15/2→4F7/2	488	3.784	3.139
4I15/2→2G11/2	378	20.337	16.597

Table 2. .Judd–Ofelt intensity parameters of Er³⁺ and Ho³⁺ in various glasses.

	Ωt(10−20 cm2)	TF	Fluoride	Phosphate	Germanate	Silicate
Er3+	Ω2	3.83	2.98	6.65	5.81	4.23
	Ω4	2.00	1.40	1.52	0.85	1.04
	Ω6	1.37	1.04	1.11	0.28	0.61
	Ref.	This work	[16]	[17]	[17]	[17]
Ho3+	Ω2	3.80	1.86	5.60	6.66	5.84
	Ω4	3.28	1.90	2.72	6.06	2.38
	Ω6	1.10	1.32	1.87	2.26	1.75
	Ref.	This work	[18]	[19]	[20]	[21]

Table 2. .Judd–Ofelt intensity parameters of Er³⁺ and Ho³⁺ in various glasses.

	Ωt(10−20 cm2)	TF	Fluoride	Phosphate	Germanate	Silicate
Er3+	Ω2	3.83	2.98	6.65	5.81	4.23
	Ω4	2.00	1.40	1.52	0.85	1.04
	Ω6	1.37	1.04	1.11	0.28	0.61
	Ref.	This work	[16]	[17]	[17]	[17]
Ho3+	Ω2	3.80	1.86	5.60	6.66	5.84
	Ω4	3.28	1.90	2.72	6.06	2.38
	Ω6	1.10	1.32	1.87	2.26	1.75
	Ref.	This work	[18]	[19]	[20]	[21]

Table 3 represents the radiative transition probabilities (A_r), branching ratios (β) and radiative lifetime (τ_r) of certain levels of Er³⁺ ions which are calculated using the above obtained Judd-Ofelt intensity parameters. It is shown that the radiative probabilities A_r of Er³⁺: ⁴I_11/2→⁴I_13/2 transition is 53.26 s⁻¹ for Er³⁺/Ho³⁺-codoped samples.

Table 3. The radiative transition probability of electric dipolar transitions (A_ed), radiative transition probability of magnetic dipolar transitions (A_md), branching ratio (β) and radiative lifetime (τ_rad) of Er³⁺/Ho³⁺-codoped glasses.

Transitions	Aed (s−1)	Amd (s−1)	Er3+/Ho3+	τ (ms)
			β (%)
4I13/2→4I15/2	331.21	79.48	100.00	3.02
4I11/2→4I15/2	317.20	0	83.47	2.63
4I13/2	62.80	22.51	16.53
4I9/2→4I15/2	319.89	0	74.63	2.33
4I13/2	104.17	0	24.30
4I11/2	4.55	4.55	1.06
4F9/2→4I15/2	3,488.20	0	91.48	0.26
4I13/2	175.72	0	4.61
4I11/2	144.22	0	3.78
4I9/2	4.92	0	0.13
4S3/2→4I15/2	2,627.95	0	62.97	0.24
4I13/2	1,329.76	0	31.86
4I11/2	82.77	0	1.98
4I9/2	132.94	0	3.19
2H11/2→4I15/2	11,737.54	0	-
4F7/2→4I15/2	6,764.75	0	-
4F5/2→4I15/2	1,638.04	0	-
2H9/2→4I15/2	3,161.88	0	40.52	0.13
4I13/2	3,480.09	0	44.60
4I11/2	1,042.72	0	13.36
4I9/2	41.44	0	0.53
4F9/2	77.59	0	0.99

Table 3. The radiative transition probability of electric dipolar transitions (A_ed), radiative transition probability of magnetic dipolar transitions (A_md), branching ratio (β) and radiative lifetime (τ_rad) of Er³⁺/Ho³⁺-codoped glasses.

Transitions	Aed (s−1)	Amd (s−1)	Er3+/Ho3+	τ (ms)
			β (%)
4I13/2→4I15/2	331.21	79.48	100.00	3.02
4I11/2→4I15/2	317.20	0	83.47	2.63
4I13/2	62.80	22.51	16.53
4I9/2→4I15/2	319.89	0	74.63	2.33
4I13/2	104.17	0	24.30
4I11/2	4.55	4.55	1.06
4F9/2→4I15/2	3,488.20	0	91.48	0.26
4I13/2	175.72	0	4.61
4I11/2	144.22	0	3.78
4I9/2	4.92	0	0.13
4S3/2→4I15/2	2,627.95	0	62.97	0.24
4I13/2	1,329.76	0	31.86
4I11/2	82.77	0	1.98
4I9/2	132.94	0	3.19
2H11/2→4I15/2	11,737.54	0	-
4F7/2→4I15/2	6,764.75	0	-
4F5/2→4I15/2	1,638.04	0	-
2H9/2→4I15/2	3,161.88	0	40.52	0.13
4I13/2	3,480.09	0	44.60
4I11/2	1,042.72	0	13.36
4I9/2	41.44	0	0.53
4F9/2	77.59	0	0.99

3.2. Fluorescence Spectra Analyses and Energy Transfer Mechanisms

Figure 2 presents the mid-infrared and near-infrared emission spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped TF glasses under excitation at 980 nm. The emission at 2.7 µm corresponds to the transition of the Er³⁺: ⁴I_11/2→⁴I_13/2. The emissions at 1.53 and 2.05 µm come from the transition of Er³⁺: ⁴I_13/2→⁴I_15/2 and Ho³⁺: ⁵I₇→⁵I₈, respectively. 2.7 µm emission can be observed in both kinds of samples. However, the intensity of the 2.7 µm increases with the addition of the HoF₃, which demonstrates the effective sensitization of Ho³⁺ ions.

Figure 2. Near-infrared and Mid-infrared fluorescence spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped fluorotellurite glasses.

Figure 2. Near-infrared and Mid-infrared fluorescence spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped fluorotellurite glasses.

From the experimental phenomenon and theoretical data, possible mechanisms [22] for the emission bands are discussed based on the simplified energy levels of Er³⁺ and Ho³⁺ presented in Figure 3. Ions on the Er³⁺: ⁴I_15/2 state are excited to the ⁴I_11/2 state by ground state absorption (GSA) when the sample is pumped by 980 nm LD. The involved energy transfer mechanisms processes based on the ⁴I_11/2 level are as follows: excited state absorption (ESA1), Er³⁺: ⁴I_11/2 + a photon→Er³⁺: ⁴F_7/2; ETU1, Er³⁺: ⁴I_11/2 + ⁴I_11/2→⁴I_15/2 + ⁴F_7/2; ETU3, ⁴I_11/2(Er³⁺) + ⁵I₆(Ho³⁺)→⁴I_15/2(Er³⁺) + ⁵F₄(Ho³⁺); ⁴I_11/2→⁴I_13/2 transition with 2.7 µm emission; ET1(a nonresonant process), from the Er³⁺: ⁴I_11/2 level to the Ho³⁺: ⁵I₆ level, energy excess (1470 cm^—1) is given to the matrix; non-radiatively relaxation to the Er³⁺: ⁴I_13/2 level.

The Er³⁺: ⁴I_13/2 level is populated owing to the nonradiative relaxation from the upper ⁴I_11/2 level. There exist four main energy transfer processes for the Er³⁺: ⁴I_13/2 level in present glass as follows: excited state absorption (ESA2), Er³⁺: ⁴I_13/2 + a photon→Er³⁺: ⁴I_9/2; ETU2 [23], Er³⁺: ⁴I_13/2 + ⁴I_13/2→⁴I_15/2 + ⁴F_9/2; ETU4, ⁴I_13/2 (Er³⁺) + ⁵I₆(Ho³⁺)→⁴I_15/2(Er³⁺) + ⁵F₅(Ho³⁺) ; ⁴I_13/2→⁴I_15/2 transition with 1.5 µm emission; ET2(a nonresonant process), from the Er³⁺: ⁴I_13/2 level to the Ho³⁺: ⁵I₇ level, energy excess (1398 cm⁻¹) is given to the matrix. After ET2 process, 2.05 µm emission can be observed due to the Ho³⁺: ⁵I₇→⁵I₈. The ESA2 process populates the Er³⁺: ⁴F_9/2 level which relaxes radiatively to the ground state with red emission around 660 nm and non-radiatively to the next lower Er³⁺: ⁴I_9/2 level. The energy stored in the Er³⁺: ⁴I_9/2 can partly be non-radiatively decayed to the Er³⁺: ⁴I_11/2 level, which is beneficial to the 2.7 µm emission.

Figure 3. Energy level schemes of Er³⁺ and Ho³⁺ and energy transfer sketch map between Er³⁺ and Ho³⁺.

Figure 3. Energy level schemes of Er³⁺ and Ho³⁺ and energy transfer sketch map between Er³⁺ and Ho³⁺.

The visible emission spectra for Er³⁺ doped and Er³⁺/Ho³⁺-codoped glasses upon 980 nm excitation are shown in Figure 4. Three visible emission peaks centered at 525, 550 and 660 nm are observed. As discussed above, the stored energy in the ⁴F_7/2 level after ESA1 process decays non-radiatively to the lower levels ²H_11/2 and ⁴S_3/2. Then the Er³⁺: (²H_11/2 + ⁴S_3/2)→⁴I_15/2 transitions bring green emissions (525 and 548 nm). Meanwhile, the excited Ho³⁺ ions at (⁵S₂ + ⁵F₄) and ⁵F₅ levels also generate 550 and 660 nm emissions, respectively. It is noted that the fluorescence intensity of red emissions becomes stronger when codoped with Ho³⁺ while the green emission becomes weaker. Because the energy gaps of the ET1 and ET2 process are relatively small, so both processes can happen efficiently. It is expected that the ESA1 and ESA2 processes will be reduced while the sample is codoped with Ho³⁺, so part of green and red emission can be attributed to the Ho³⁺ upconversion emissions. Since the energy gap of the ET2 process is smaller than that of ET1 process, the energy transfer efficiency of ET2 is expected to be higher than that of ET1, this is also proved from the extremely decreased 1.5 µm emission in the Er³⁺/Ho³⁺-codoped samples, consequently ions on the ⁵I₇ level are much more than that on the ⁵I₆ level. Then the ⁵F₅ level that is populated through ETU4 generates stronger 660 nm emission while the ⁵S₂ (⁵F₄) levels that are populated through ETU3 generate weaker 550 nm emission.

3.3. Fluorescence Spectra Analyses and Energy Transfer Mechanisms

The energy transfer efficiency has been estimated from the measured lifetime of the 1.53 µm emission of Er³⁺ singly and Er³⁺/Ho³⁺-codoped samples by the following formula [24]:

(1)

where τ_EH and τ_E are the Er³⁺ lifetime monitored at 1.53 µm with and without Ho³⁺ ions, respectively. The lifetime decay curves of the Er³⁺: ⁴I_13/2 level with and without Ho³⁺ in TF glasses are measured and shown in Figure 5. The values for lifetime of Er³⁺: ⁴I_13/2 level for Er³⁺ and Er³⁺/Ho³⁺-codoped samples are 5.25 ms and 1.71 ms, respectively. The decrease of lifetime of the Er³⁺: ⁴I_13/2 state indicates the existence of ET2 process. In addition, the energy transfer efficiency from the Er³⁺: ⁴I_13/2 to Ho³⁺: ⁵I₇ level is calculated to be 67.33%.

Figure 4. Upconversion emission spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped glass samples.

Figure 4. Upconversion emission spectra of Er³⁺ and Er³⁺/Ho³⁺-codoped glass samples.

Figure 5. Fluorescence decay curves of Er³⁺: ⁴I_13/2 level in Er³⁺ singly and Er³⁺/Ho³⁺-codoped samples.

Figure 5. Fluorescence decay curves of Er³⁺: ⁴I_13/2 level in Er³⁺ singly and Er³⁺/Ho³⁺-codoped samples.

3.4. Cross-Sections Analyses

The emission cross section is an important factor for evaluating the emissive ability of luminescent center. The absorption and emission cross sections could be calculated from Füchtbauer–Ladenburg equation [25] and McCumber theory [26]:

(2)

(3)

where λ is the wavelength, A_rad is the spontaneous transition probability, I(λ) is the fluorescence spectra intensity, n and c represent the refractive index and the speed of light, Z_L and Z_U are partition functions of the lower and upper manifolds, respectively. The maximum absorption (σ_abs) and emission cross section (σ_em) (both peaking at 2708 nm) in Figure 6 are 1.54 × 10⁻²⁰ cm² and 1.82 × 10⁻²⁰ cm², respectively, which is larger than the values reported in Ref. [27,28,29]. Hence, Er³⁺/Ho³⁺-codoped TF glass with promising properties has potential applications for 2.7 µm laser material.

Figure 6. Absorption and emission cross sections of Er³⁺: ⁴I_11/2→⁴I_13/2 in TF glasses.

Figure 6. Absorption and emission cross sections of Er³⁺: ⁴I_11/2→⁴I_13/2 in TF glasses.

4. Conclusions

Enhanced 2.7 µm emission was obtained in Er³⁺/Ho³⁺-codoped fluorotellurite glass. This suggests that Ho³⁺ can be a feasible approach to obtain efficient Er³⁺ 2.7 µm emission pumped by common 980 nm LD in fluorotellurite glass for practical applications. It was also found that the green upconversion and 1.5 µm emissions extremely decreased in the Er³⁺/Ho³⁺-codoped glass. The energy transfer mechanisms between Er³⁺ and Ho³⁺ were discussed in detail and the energy transfer efficiency was calculated to be 67.33%. Larger absorption and emission cross sections of Er³⁺: ⁴I_11/2→⁴I_13/2 in TF glasses were obtained which were 1.54 × 10⁻²⁰ cm² and 1.82 × 10⁻²⁰ cm², respectively. These results suggest that Er³⁺/Ho³⁺-codoped fluorotellurite glass has potential applications for 2.7 µm fiber laser materials.