Comparative investigation on the spectroscopic properties of Pr3+-doped boro-phosphate, boro-germo-silicate and tellurite glasses

Spectrochimica Acta Part A 93 (2012) 223–227

Contents lists available at SciVerse ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Comparative investigation on the spectroscopic properties of Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses Liaolin Zhang, Guoping Dong, Mingying Peng, Jianrong Qiu ∗ State Key Laboratory of Luminescent Materials and Devices, and Institute of Optical Communication Materials, South China University of Technology, Wushan Road 381, Guangzhou 510640, China

a r t i c l e

i n f o

Article history: Received 9 December 2011 Received in revised form 16 February 2012 Accepted 22 February 2012 Keywords: Phonon energy Multi-phonon relaxation Pr3+ -doped glasses

a b s t r a c t We report on the spectroscopic properties of Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses. The stimulated absorption and emission cross sections were estimated. Only one emission at 596 nm and 605 nm is observed in Pr3+ -doped boro-phosphate and boro-germo-silicate glasses, respectively, while three emissions at 605 nm, 612 nm and 645 nm are observed in Pr3+ -doped tellurite glass when excited at 467 nm. The fluorescence lifetime at 600 nm in Pr3+ -doped boro-phosphate, boro-germosilicate and tellurite glasses is 137 ␮s, 73 ␮s and 51 ␮s, respectively. The emissions from Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses show different decay behaviors and can be well explained by multiphonon relaxation theory. © 2012 Elsevier B.V. All rights reserved.

1. Introduction In the past decades, Pr3+ -doped glasses have attracted much attention due to their diverse fluorescence covering visible and near-infrared wavelength regions. They have been applied for the solid state lasers, fiber lasers and fiber amplifiers [1,2]. Recently, more efforts are devoted to the visible fluorescence property of Pr3+ -doped glasses for the potential applications of Pr3+ -doped glass fiber laser in the medical treatment, optical date storage, and display etc. In the early years, due to the lack of pump source in the UV and blue regions, visible fiber lasers were focused on the upconversion glass fiber laser. Blue laser was achieved based on Pr3+ -doped ZBLAN fiber [3]. However, due to the intrinsic characteristic of upconversion process, much energy was lost during the pump process. Besides, high pump power was required to gain the laser output, and the slope efficiency of upconversion fiber laser was low. All above drawbacks of upconversion fiber laser limited its practical applications. More recently, the development in GaN-related laser diodes (LDs) has enabled the light sources from ultraviolet (UV) to blue wavelength range available on the market in watt levels, which in turn advances the rapid development of visible fiber lasers [4–8]. In 2009, Okamoto et al. reported the visible to NIR tunable fiber laser based on Pr3+ -doped ZBLAN fluoride glass fiber by using 448 nm GaN multimode laser diode (LD) as a pump source [4]. Subsequently, Fujimoto et al. and Nakanishi

∗ Corresponding author. E-mail address: [email protected] (J. Qiu). 1386-1425/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.saa.2012.02.076

et al. demonstrated red laser oscillation from Pr3+ -doped fluoride glass fibers pumped by 442 nm GaN LDs [6]. However, the output power and slope efficiency of Pr3+ -doped fluoride glass fiber lasers are not enough for the practical applications. Much more efforts are required to improve the slope efficiency and the output power of the fiber laser. In addition, the fluoride glass fibers are intrinsically unstable when exposed to the environment. Therefore, it is necessary to exploit oxide glass fibers replacing the fluoride glass fibers to obtain high output power of the visible laser. The emission of Pr3+ ions is known to be strongly dependent on the host materials [9–12], especially in the visible wavelength region due to the radiative transitions of 1 D2 and 3 P0 energy levels. Thus, it is necessary to study the spectroscopic properties of Pr3+ -doped glasses in various glass matrices. Balda et al. reported the visible emission from Pr3+ -doped lead germanate glass [9], and fluorophosphate glass [12]. The cross relaxation and multiphonon relaxation processes of 1 D2 and 3 P0 energy levels in the Pr3+ -doped glasses were also discussed by Gyu Choi et al. [13,14]. However, to the best of our knowledge, there have been no comparative investigations on the fluorescence properties of Pr3+ -doped oxide glasses e.g. boro-phosphate, boro-germo-silicate and tellurite etc., while these oxide glasses are regarded as the desirable host materials for the visible fiber laser due to their excellent mechanical, thermal and optical properties. Herein, in this paper, spectroscopic properties of Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses were examined and compared. Time-resolved luminescence spectra of Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses were also measured to clarify the multiphonon relaxation processes in above glasses. Pr3+ -doped tellurite glass,

224

L. Zhang et al. / Spectrochimica Acta Part A 93 (2012) 223–227

due to its larger absorption cross section and emission cross section, diverse wavelength emissions and lower multiphonon relaxation rate, would be more desirable material for visible fiber laser.

Tellurite

The sample glasses were prepared by the conventional melting method with the compositions (in mol%): 60P2 O5 –4B2 O3 –7Al2 O3 –10K2 O–17.9BaO–0.1Pr2 O3 , 55SiO2 –8B2 O3 –5Al2 O3 –14Li2 O–2Na2 O–10GeO2 –5.74Y2 O3 – 0.1Pr2 O3 , 75TeO2 –20ZnO–5Na2 O–0.1Pr2 O3 . Analytical reagents P2 O5 , H3 BO3 , Al(OH)3 , K2 CO3 , Li2 CO3 , Na2 CO3 , ZnO, BaCO3 and 4 N Pr6 O11 , GeO2 were used as starting materials. 30 g mixture of the raw materials was weighed according to the above nominal composition and homogeneously grounded and mixed, and then transferred to an alumina crucible. The melting temperature was 1250 ◦ C, 1300 ◦ C and 850 ◦ C for the boro-phosphate, boro-germo-silicate, and tellurite glasses, respectively. And the samples were named as P, S and T glass, respectively. All the samples were melted at corresponding temperature for 30 min, and casted into the preheated stainless steel plate at 300 ◦ C. The samples were then annealed at corresponding glass transition temperature for 3 h, and cut into suitable size. Two big faces of each sample were polished for optical measurements. Absorption spectra were measured with Lambda 900 spectrophotometer. Refractive indexes of the samples were measured by a WYV-V prism refractive photometer. Glass densities were obtained according to the Archimedes method. Static excitation and emission spectra, time-resolved luminescence spectra, fluorescence lifetime were measured with Edinburgh FLS920 Fluorospectrophotometer with double xenon lamp houses, one was 450 W for steady state measurement, another was 60 W microsecond flashlamp for dynamic spectra recording. All the measurements were performed at room temperature. 3. Results and discussion Fig. 1 shows absorption spectra of P, S and T glasses. The absorption bands at 444 nm, 467 nm, 480 nm, 590 nm can be assigned to the transitions of 3 H4 → 3 Pj(j = 2, 1, 0) , and 3 H4 → 1 D2 respectively. No shift is observed for the peak position of absorption bands in P, S and T glasses. The stimulated absorption cross section can be obtained according to the equation [15]: a () =

2.303 × OD() Nl

Intensity (a.u.)

2. Experimental

3H 4

Silicate

3P 1

3H 4

3H 4

3P 2

3P 0

Phosphate 420

440

460

480

500

520

Wavelength (nm) Fig. 2. Excitation spectra of 0.1 mol% Pr3+ -doped P(a), S(b) and T(c) glasses monitored emission at 600 nm.

where N is the ions concentration of the sample glass, l is the thickness of the sample and OD() is the optical density. The stimulated absorption cross sections around 444 nm, 467 nm, and 480 nm of P, S and T glasses are listed in Table 1. The absorption cross sections at 444 nm and 467 nm of T glass are larger than that of P and S glasses. However, the absorption cross section at 480 nm of S glass is the largest, while that of P glass is the smallest. The lager stimulated absorption cross section at 444 nm and 467 nm of P glass indicates that P glass can be more efficiently pumped using GaN LD at 444 nm and 467 nm, while S glass can be efficiently pumped at 480 nm. Fig. 2 shows the excitation spectra of P, S and T glasses by monitoring emission at 600 nm. Three main excitation bands corresponding to the 3 H4 → 3 Pj(j = 2, 1, 0) transitions can be found in the excitation spectra of Pr3+ -doped glasses, which are in a good agreement with the absorption spectra. Due to the Stark split of Pr3+ energy level, the excitation bands around 467 nm are split into three bands. Fig. 3 shows the emission spectra of P, S and T glasses. Only one emission band at 596 nm and 605 nm can be observed in the

(1)

Tellurite

(a)

(b)

Intensity (a.u.)

Absorption intensity (a.u.)

Silicate

3

P2 3

P1 3 P0

1

D2

Phosphate 450

500

550

600

650

700

Wavelength (nm) Fig. 1. Absorption spectra of 0.1 mol% Pr3+ -doped P, S and T glasses.

Tellurite

600 630 660 690

Phosphate

Tellurite Silicate

400

Phosphate

Silicate 500

550

600

650

700

Wavelength (nm) Fig. 3. Emission spectra of 0.1 mol% Pr3+ -doped P(a), S(b) and T(c) glasses excited at 467 nm (Inset shows the emission spectra of 0.1 mol% Pr3+ -doped P(a), S(b) and T(c) glasses excited at 590 nm).

L. Zhang et al. / Spectrochimica Acta Part A 93 (2012) 223–227

225

Table 1 The stimulated absorption and emission cross sections for the 0.1 mol% Pr3+ -doped phosphate, silicate and tellurite glasses. Glass systems

Parameters Stimulate absorption cross section (×10−20 cm2 )

Phosphate Silicate Tellurite

 ab (444 nm)

 ab (467 nm)

 ab (480 nm)

 em (600 nm)

 em (645 nm)

6.11 3.91 14.17

2.39 0.52 5.35

2.98 5.35 4.66

2.32 2.88 6.11

– – 0.14

emission spectra of P and S glasses excited at 467 nm, respectively. However, three emission bands at 605 nm, 612 nm and 645 nm are observed in the emission spectrum of T glass. The emission at 645 nm of T glass can be assigned to the 3 P0 → 3 F2 transition as Fig. 4 shows. It is interesting that the emission at 645 nm is the strongest among the three emission bands in T glass, while there is no emission band at 645 nm in P and S glasses. When excited at 590 nm, only an emission band at about 600 nm is observed in P, S and T glasses (the inset of Fig. 3). The emission band shows a slight red shift in the order of P, S and T glasses, and the variation tendency is in an agreement with that of the emission band at 600 nm when excited at 467 nm. It can be confirmed that the emission around 600 nm is dominated by the 1 D2 → 3 H4 transition in P and S glasses when excited at 467 nm, since the emission spectra excited at 467 nm is same with the emission spectra when excited to 1 D2 energy level. The emission spectrum of T glass excited at 467 nm is complicated. However, there is no emission band at 612 nm was found when excited to 1 D2 energy level, therefore, we proposed that the emission at 612 nm of T glass is dominated by the transition of 3 P0 → 3 H6 , and the emission band at 605 nm excited at 467 nm is dominated by the transition 1 D2 → 3 H4 . It is well known that 1 D2 energy of Pr3+ is more sensitive to Pr3+ concentration than 3 P [9,12,16,17], and Pr3+ concentration of T glass is lager than that 0 of P and S glass, results in the cross relaxation rate of 1 D2 in T glass is lager than that of P and S glasses, therefore, the apparent emission at 612 nm from 3 P0 energy was observed in T glass, while the apparent emission at 605 nm were observed from 1 D2 energy level in P and S glasses. The stimulated emission cross section can be calculated by the equation [18]: 4 Aed 8cn2 eff

(2)

where  is the average emission wavelength, Aed is the transition probability and eff is the effective emission line-width calculated by the equation:



eff =

I()d Imax ()

(3)

The calculated stimulated emission cross sections of P, S and T glasses are also listed in Table 1. The stimulated emission cross section of T glass is much larger than that of P and S glasses. It can be deduced that T glass is more suitable for the visible fiber laser material due to its larger absorption and emission cross sections, and diverse wavelength emissions. Fig. 5 shows the luminescence decay curves of P(a), S(b) and T(c) glasses monitored emissions at 600 nm using a 467 nm microsecond lamp as excitation source. The luminescence decay curves of P, S and T glasses can be fitted by the single exponential decay. The fluorescence lifetime of P, S and T glasses is 137 ␮s, 73 ␮s and 51 ␮s, respectively. The measured lifetime can be regarded as the lifetime of 1 D2 energy level of Pr3+ -doped glasses, since the emission at 600 nm is dominated by the 1 D2 → 3 H4 transitions of Pr3+ in glasses. Multiphonon relaxation process associated with 3 P0 and 1 D2 energy levels may affect the emission spectra of P, S and T glasses. It is well known that the multiphonon relaxation process results from the interaction between the electronic levels of the rare-earth ions and vibrations of the host materials. Thus, the host materials play an important role for the multiphonon relaxation processes. The multiphonon relaxation rate against the energy gap and maximum phonon energy of host materials can be described by the following equation [19]: WMPR = ˇel exp[−˛(E − 2hωmax )]

25 3

P P P

3 2 3 1 0

20

(4)

where ˛ and ˇel are constants, which can be obtained from the previous works [19–23], E is the energy gap to the next lower level and hωmax is the maximum phonon energy of the host

1

G4 F 3 4 F 3 3 F 3 2 H6 3 H5 3 H4 3

5

0 Fig. 4. The energy level diagram of Pr3+ ions.

Intensity (a.u.)

D2

645nm

600-630nm

600-630nm

590nm

10

488nm

15

467nm

1

448nm

-3 -1 10 cm )

em =

Stimulate emission cross section (×10−24 cm2 )

(a) (b) (c) 0

20

40

60

80

100

120

140

Time ( s) Fig. 5. Luminescence decay curves of 0.1 mol% Pr3+ -doped P(a), S(b) and T(c) glasses.

226

L. Zhang et al. / Spectrochimica Acta Part A 93 (2012) 223–227

10 s 20 s 40 s 60 s 80 s 100 s

(a)

Intensity (a.u.)

10 s 20 s 40 s 60 s 80 s 100 s

580

600

620

640

660 550

575

(b)

600

625

650

(c)

10 s 20 s 40 s 60 s 80 s 100 s

550

600

650

700

Wavelength (nm) Fig. 6. Time-resolved luminescence spectra of 0.1 mol% Pr3+ -doped P(a), S(b) and T(c) glasses excited at 467 nm, the delay time is coded in the figure.

materials. The multiphonon relaxation rate involving 3 P0 energy level can be calculated according to Eq. (4), producing 352,440 s−1 , 199,850 s−1 and 8130 s−1 for P, S and T glasses, respectively. The multiphonon relaxation rate of 1 D2 energy level is also calculated, and is 1.58 × 10−2 s−1 , 0.86 × 10−2 s−1 and 0.079 × 10−2 s−1 for P, S and T glasses, respectively. The multiphonon relaxation process from 1 D2 energy level can be ignored when compared with the multiphonon relaxation process from 3 P0 energy level, since the former multiphonon relaxation rate are several orders of magnitude smaller than the later multiphonon relaxation rate. The depopulation rate from 3 P0 to its next lower energy level 1 D2 of Pr3+ -doped P and T glasses is 40 times larger than that of Pr3+ -doped T glass. Therefore, higher phonon energy of phosphate glass and silicate glasses rapidly depopulate the 3 P0 level, and feed efficiently to 1 D energy level through multiphonon relaxation processes com2 pared with T glass. The delay time of spectrum measurement of P and S glasses are too short to recode the emission transition from 3 P , thus, there is no emission band at 645 nm excited at 467 nm. 0 Though the delay time is same among the three measurements, the depopulation rate of 3 P0 in T glass is much slower than that of P and S glasses, only a part of phonons depopulate from 3 P0 within the delay time, thus emission at 645 nm was detected in static luminescence measurement of T glass. The fluorescence lifetime of 1 D2 energy level is determined by the multiphonon relaxation rate and cross relaxation rate associated with 1 D2 . However, the multiphonon relaxation rate of 1 D2 can be neglected according to above discussion. Therefore, the cross relaxation processes associated with 1 D2 dominate the fluorescent lifetime. It is well known that the ion concentration plays an important role for the cross relaxation. The Pr3+ ions concentration in P, S and T glasses is calculated to be 0.14 × 1027 cm−3 , 0.215 × 1027 cm−3 and 0.239 × 1027 cm−3 , respectively. Furthermore, rare-earth ions can be homogeneously distributed in P glass than S and T glasses due to the special structure of P glass [24], and the ions concentration of P glass is smaller than that of S and T glass. It largely reduces the cross relaxation probability between the rare-earth ions. Thus, the fluorescence lifetime of 1 D2 of Pr3+ in P glass is much longer than that of Pr3+ in S and T glasses. In addition, the Pr3+ ions concentration in T glass is slightly higher than that of S glass, thus, the fluorescence lifetime of 1 D2 of Pr3+ in T glass is slightly shorter than that of S glass due to cross relaxation. Fig. 6 shows the time-resolved luminescence spectra of P(a), S(b) and T(c) glasses. At beginning, two emission bands at 600 nm and 645 nm are found in the time-resolved luminescence spectra of P and S glasses. The emission band at 645 nm disappeared when the

delay time reached 20 ␮s. It is due to the rapidly depopulation of through multiphonon relaxation process as discussed above. Three emission bands at 605 nm, 612 nm and 645 nm are observed in the time-resolved luminescence spectra of Pr3+ -doped T glass at beginning. The intensity of emission at 605 nm from 1 D2 is much weaker than the emissions at 612 nm and 645 nm from 3 P0 level at staring time, since the multiphonon relaxation rate from 3 P0 of Pr3+ -doped T glass is not enough to populate to 1 D2 energy level. However, when the delay time reaches 80 ␮s, a part of photons have been transferred from 3 P0 to 1 D2 energy level through multiphonon relaxation process, resulting in the emission from 1 D2 at 605 nm which is comparable with the emissions at 612 nm and 645 nm from 3 P0 energy level. When the delay time is beyond 100 ␮s, the emission of Pr3+ -doped T glass would be dominated by transition of 1 D2 → 3 H4 since most photons may have populated to 1 D2 energy level through multiphonon relaxation process. When the delay time exceeds 20 ␮s, the time-resolved luminescence spectra of P, S and T glasses coincide well with the steady emission spectra. The different decay behavior of the emissions in Pr3+ -doped glasses is mainly due to the multiphonon relaxation processes. 3P 0

4. Conclusions The spectroscopic properties of Pr3+ -doped boro-phosphate, boro-germo-silicate and tellurite glasses were studied. The absorption cross section at 442 nm and 467 nm of Pr3+ -doped tellurite is larger than that of Pr3+ -doped boro-phosphate and boro-germosilicate glasses. When excited at 467 nm, only one emission at 596 nm and 605 nm is observed in Pr3+ -doped boro-phosphate and boro-germo-silicate glasses, respectively, while three emissions at 605 nm, 612 nm and 645 nm are observed in Pr3+ -doped tellurite glass. The fluorescence lifetime at 600 nm in Pr3+ -doped borophosphate, boro-germo-silicate and tellurite glasses is 137 ␮s, 73 ␮s and 51 ␮s, respectively. The emissions from Pr3+ -doped borophosphate, boro-germo-silicate and tellurite glasses show different decay behaviors and can be explained by multiphonon relaxation theory. The present investigation is of value for designing Pr3+ doped glasses as visible fiber laser materials. Acknowledgments This work was financially supported by the National Natural Science Foundation of China (Grant Nos. 51072054, 51072060, 51132004), Fundamental Research Funds for the Central

L. Zhang et al. / Spectrochimica Acta Part A 93 (2012) 223–227

Universities (Grant No. 2011ZZ0001, 2011ZB0001 and 2011ZP0002), Natural Science Foundation of Guangdong Province (Grant Nos. 1045106410104887, S2011030001349), China Postdoctoral Science Special Foundation (Grant No. 201104350), National Basic Research Program of China (973 program) (2011CB808102), Chinese Program for New Century Excellent Talents in University (Grant No. NCET-11-0518) and Fok Ying Tong Education Foundation (Grant No. 132004). References [1] S. Carter, D. Szebesta, S. Davey, R. Wyatt, M. Brierley, P. France, Electron. Lett. 27 (1991) 628–629. [2] Y. Miyajima, T. Sugawa, Y. Fukasaku, Electron. Lett. 27 (1991) 1706–1707. [3] D. Baney, G. Rankin, K.W. Chang, Opt. Lett. 21 (1996) 1372–1374. [4] H. Okamoto, K. Kasuga, I. Hara, Y. Kubota, Opt. Express 17 (2009) 20227–20232. [5] Y. Fujimoto, O. Ishii, M. Yamazaki, Electron. Lett. 45 (2009) 1301–1302. [6] Y. Fujimoto, O. Ishii, M. Yamazaki, Electron. Lett. 46 (2010) 1285–1286. [7] T. Yamashita, Y. Ohisi, Jpn. J. Appl. Phys. 46 (2007) L991–L993. [8] M. Farries, P. Morkel, J. Townsend, Electron. Lett. 24 (1988) 709–711.

227

[9] R. Balda, J. Fernandez, A. de Pablos, J. Fdez-Navarro, J. Phys.: Condens. Matter 11 (1999) 7411–7421. [10] Y. Sharma, S. Tandon, S. Surana, Mater. Sci. Eng. B: Solid 77 (2000) 167–171. [11] Z. Mazurak, S. Bodyi, R. Lisiecki, J. Gabry´s-Pisarska, M. Czaja, Opt. Mater. 32 (2010) 547–553. [12] R. Balda, J. Fernandez, I. Saez de Ocariz, J. Adam, A. Mendioroz, E. Montoya, Opt. Mater. 13 (1999) 159–165. [13] Y. Gyu Choi, B. Hom Kim, J. Heo, Chem. Phys. Lett. 368 (2003) 625–629. [14] Y. Gyu Choi, J. Hyun Baik, J. Heo, Chem. Phys. Lett. 405 (2005) 436–440. [15] D. McCumber, Phys. Rev. 134-2A (1964) A299–A306. [16] H. Chen, R. Lian, M. Yin, L. Lou, W. Zhang, S. Xia, J. Krupa, J. Phys.: Condens. Matter 13 (2001) 1151–1158. [17] R. Balda, J. Fernandez, I. Saez de Ocariz, M. Voda, A. Garcia, Phys. Rev. 59 (1999) 9972–9980. [18] M. Inokuti, F. Hirayama, J. Chem. Phys. 43 (1965) 1978–1989. [19] J.M.F. van Dijk, M.F.H. Schuurmans, J. Chem. Phys. 78 (1983) 5317–5323. [20] C. Layne, W. Lowdermilk, M. Weber, Phys. Rev. B 16 (1977) 10–21. [21] G. Ajithkumar, P. Gupta, G. Jose, N. Unnikrishnan, J. Non-Cryst. Solids 275 (2000) 93–106. [22] R. Reisfeld, L. Boehm, Chem. Phys. Lett. 49 (1977) 251–254. [23] M. Hehlen, N. Cockroft, T. Gosnell, A. Bruce, Phys. Rev. B 56 (1997) 9302–9318. [24] S. Jiang, M. Myers, N. Peyghambarian, J. Non-Cryst. Solids 239 (1998) 143–148.