JOURNAL OF
LUMINESCENCE ELSEMER
Journal of Luminescence 75 (1997) 301.-308
Optical transitions of Er3+ in lead-tellurium-germanate
glasses
Z. Pan, S.H. Morgan* Center,fbr Photonic Materials and Deoices, Physics Department. Fisk Unitwrsig, Nashoille. TN 37208, USA Received 17 July 1996; received in revised form 21 May 1997; accepted 21 May 1997 Abstract Optical-absorption, photoluminescence, and up-conversion luminescence were measured for Er3+ in lead-telluriumgermanate glasses (63-x)GeO, . xTe0,. 27PbO. lOCa0 with x = 0, 10, 20, 30, and 40. Each glass sample contains an extra 0.2 mol% Er,O,. Judd-Ofelt intensity parameters were determined and used to calculate the radiative transition rates and lifetimes of Er 3+ in glasses with different TeO, content. It has been found that the 4S3,2 + 41,,/, radiative transition rate and the infrared (800 nm) to green (547 nm) up-conversion efficiency of Er3+ increase with increasing TeO, content. Our results indicate that incorporating TeO, up to 40mol% into a lead-germanate glass network increases the luminescence efficiency due to the increased refractive index and decreased maximum phonon energy of the host glass. Keywords:
Optical
transitions;
Er 3+., Lead-tellurium-germanate
1. Introduction
Rare-earth doped lead-germanate and telluritebased glasses have been the subject of several recent investigations [l-5]. These glasses have better mechanical strength, chemical durability, and thermal stability than fluoride-based glasses [2,3]. The smaller maximum vibrational energy (75@850 cm- ‘) of germanate and tellurite glasses compared to that of silicate ( - 1150 cm- ‘) glasses, results in a smaller multiphonon decay rate of doped rare-earth ions, compared to that in silicate-based glasses [l-4]. The GeOz-based glasses exhibit better infrared (IR) transmission than SO,-based glasses due to
*Corresponding author. Tel.: + 1 615 329 8621; fax: + 1 615 329 8634; e-mail:
[email protected].
0022-23 13/97/S 17.00 r$; 1997 Elsevier Science B.V. All rights reserved PII SOO22-2313(97)001 18-X
glasses
the larger size and the heavier mass of germanium when compared to silicon. Lead-germanate basedglass systems have been previously reported as having a large glass forming region while maintaining a high infrared transmittance [6-83. A specific glass of the composition 63GeO,-27PbO-10CaO (GPC) was noted to have less OH- contamination [9], but showed a strong crystallization in the fiberpulling temperature range [2]. Tellurite glass has the lowest maximum phonon energy among common oxide glasses due to the heavier mass of tellurium atom. On the other hand, tellurite glass has a larger refractive index compared to germanate, silicate, and fluoride glasses [3]. A higher refractive index could result in a larger radiative transition rates of rare-earth ions [ 11. The PbO-TeO,-Ge02 glass was first reported as a low-loss infrared transmitting glass [lo]. We
302
2. Pan, S.H. Morgan /Journal ofLuminescence
recently reported a new lead-tellurium-germanate glass system (633x)GeO, . xTeOz .27PbO. lOCa0 [ 111. The thermal stability of the glass with respect to crystallization has significantly improved by incorporating up to 40 mol% of TeOz in the glass system. In this article, we report on the optical-absorption, photoluminescence, and up-conversion luminescence of Er3+ in lead-tellurium-germanate glasses (633x)GeO, . xTe0,. 27PbO. lOCa0 with x = 0, 10, 20, 30, and 40. Each glass sample contains an extra 0.2 mol% Er203.
2. Experiment Anhydrous oxide powders of GeO,, PbO, TeO,, Erz03, and anhydrous CaC03 powder were used to prepare glasses. The PbO and CaCO, were of common reagent grade, GeOz was of 99.9999% purity (Eagle Picher Research Lab), TeO, and Erz03 were of 99.99% purity (Alfa, Johnson Matthey). The glass compositions were in the form of (633x)Ge02. xTeOz .27PbO. 10CaO with x = 0, 10, 20, 30, and 40, respectively, and contain an extra 0.2 mol% Er203. CaO was added to reduce the OH- contamination [2]. The batch powders of 30 g were thoroughly mixed, and melted in fused silica crucibles at a temperature range of 780-l 100°C. The melts were held for 1.5 h and cast onto a copper plate and pressed by another copper plate from the top, forming a glass plate of about 2 mm thickness. The glasses were subsequently annealed at just below the glass transition temperatures for 50 min and then allowed to cool to room temperature in the furnace. Clear glasses were formed for x = 0, 10, 20, 30, and 40. These glasses appear to be of good optical quality, with no visual evidence of devitrification. However, an attempt to make glass with 50 mol% TeO, was unsuccessful because of the phase separation. The visible and near-infrared absorption spectra were measured at room temperature using a Hitachi model U-4001 spectrophotometer over a spectral range of 2.5&2000 nm. The absorption spectra of Er3+ -doped samples were measured relative to the undoped samples, and reported as the absorption coefficient of Er3+ in glasses.
75 (1997) 301-308
Table 1. Thermal properties and refractive index of (63x)GeO, xTe0,. 27PbO lOCa0 glasses with different Te02 content x. No crystallization peak was observed for x = 30 and x = 40 compositions Properties
x = 0
x=
10
x = 20
x = 30
x = 40
T, (“C) T, (“C) n (5461 A)
505 571 1.79
470 516 1.82
420 591 1.86
385
365
Note: where T,: glass transition temperature, temperature, n: refractive index.
1.91
1.95
T,: crystallization
The luminescence and up-conversion luminescence spectra were measured at room temperature using a Spex Model 1403 double grating spectrometer with a cooled RCA31034 photomultiplier and photon-counting system. The 488 nm line from a Coherent Innova 90 argon-ion laser was used for direct excitation and a Lexel cw near-infrared Ti:sapphire laser tuned to 800 nm was used for up-conversion excitation. The spectral response of the system from 28000-11300 cm-’ has been calibrated by using a standard quartz tungsten halogen (QTH) lamp. The luminescence spectra reported have been corrected for the wavelength dependence of the system. The refractive index, n, was measured using an ellipsometer, type 43603-200E, Rudolph Research. The thermal properties were analyzed by differential scanning calorimetry (DSC) on a Perkin-Elmer DSC-4 scanning calorimeter in the temperature region of 200-600°C. The thermal properties and refractive index are listed in Table 1.
3. Results and discussion 3.1. Absorption
spectra and Judd-Ofelt
analysis
The uv-visible and near-infrared absorption spectra were measured for (63-x)GeO, . xTeOz . 27PbO. lOCa0 glasses with x = 0, 10, 20, 30, and 40, each glass contains an extra 0.2 mol% Er,O,. The Er3+-doped samples were measured relative to the undoped samples, and are therefore reported as the absorption coefficient of Er3+ in glasses. Fig. 1
Z. Pan, S.H. Morgan /Journal qf‘luminescence
Wavelength
-
10000
75 (1997) 301-308
303
(nm)
_~
15000
__~A
/
20000
25000
Photon energy (cm-‘) Fig. I. Absorption spectrum 41;5.2 to the lkcl specified.
of Er3+ in a 43Ge07.
20Te0,.27PbO’
shows a typical spectrum from a sample with x = 20. The absorption edges due to the host glasses are shifted from about 350 nm for GPC glass to near 370 nm for glass with highest TeOz content. The Judd-Ofelt (J-O) approach for determining radiative and nonradiative rates of rare-earth transitions has been widely used for both crystals and glasses [12-171. In the J-O approach, the line strength for an electric-dipole transition between two .I states has been given by [15,18] &(J,
J’) =
c Q2,
((U(t)(( a’S’L’, J’)*, (1)
where Sz, are J-0 parameters, J and J’ specify the total angular momentum of initial and final states, aSL define all other quantum numbers needed to specify the states, and the ( 1)UC’)(() are double reduced matrix elements of unit tensor operators calculated in the intermediate coupling approximation. Because the matrix elements are essentially the same from host to host [12,19], we used the values calculated by Weber [15] for Er3+ in LaF3. The line strength of an absorption band can be calculated using the experimental values of the in-
IOCaO
tegrated
s
glass. Absorption
absorbance.
x(v)dv =
bands
are all from the ground
This relationship
level
is [ 12,151
8n4e2Nv m 3hc(2J + 1)
band
(2)
where r(v) is the absorption coefficient, v is the frequency (cm-‘) and v, is the mean frequency of the absorption band, h is Planck’s constant, c is the speed of light, and e is the electron charge. The J-O intensity parameters 52i then can be determined by a least-squares fitting of Eq. (1). The absorption bands measured are all dominated by electricdipole transitions except the transition 41 15;2 -+ 411312 for which magnetic-dipole transition must be considered [ 121. The oscillator strength of the magnetic-dipole transition 411 5/2 -+ 41I 3,2 was calculated using the formula given in Ref. [15]. The electric dipole oscillator strength for 41,s,2 + 41 i3,* transition was then obtained by subtracting the calculated magnetic-dipole portion from the experimental oscillator strength. A two-step least-squares fit [1] to Eq. (1) was carried out for determining J-O parameters. Results of the J-O parameters are listed in Table 2.
2. Pan, S.H. Morgan /Journal of Luminescence 75 (1997) 301-308
304
Table 2. Judd-Ofelt intensity parameters for Er3+ in (63-x) GeO, .xTeO, .27PbO lOCa0 glasses with different TeO, content x. Each glass contained an extra 0.2 mol % Erz03
Table 3. Predicted spontaneous-radiative transition rates and lifetimes of the ‘?& state of Er3+ in (63-x) GeO,.xTeO,. 27PbO. lOCa0 glass
Composition (mol %)
Wz (10-Z0cm2)
Q, (10-Zocm2)
Q.5 (10~20cmZ)
Composition (mol %)
Transition
Average frequency (cm-‘)
Wed (s- ‘)
x=0 x = 20 x = 40
5.78 * 0.30 6.16 k 0.32 6.04 f 0.29
1.46 + 0.19 1.64 _+ 0.18 1.63 + 0.21
0.94 f 0.11 0.99 f 0.09 1.02 _+ 0.11
x=0
‘%312 + 4115,* 41 1312 41 I I,2 419/z % 312 -+ 4115,z 41 1312 .+I1I,2 %,, ‘%312 -+ 411 5,2 41 13,2 4I 1*,z 419/z
18440 11850 8200 5950 18440 11850 8200 5950 18440 11850 8200 5950
1312.1 530.5 42.7 71.2 1580.4 638.9 51.7 87.5 1928.3 779.6 62.9 105.6
x = 20
The error bars given are associated with the quality of the least-squares fit and all are within 15% of the experimental values. Values of J-O parameters do not show any significant change for glasses with different TeO, content, from x = 0 to 40. These values are in agreement with that reported by Ryba-Romanowski for Er3+ in a tellurite-based glass [19], our sZz value is slightly higher, more closer to a a22value reported by Reisfelt for Er3+ in a binary germanate glass [20]. After obtaining the J-O parameters Q,, the radiative transition rates WR for all excited levels of Er3+ can be determined. The relation is [12,15,19]
w (UJ, bJ’)
=
,,;;;;
,)(x.d&d
+
hnd&ndh
(3)
where the x terms account for a local field correction for Er3 + ions and are given by xed = n(n’ + 2)2/9 for electric-dipole transitions and ,&,,d= n3 for magnetic-dipole transitions [12]. The magnetic-dipole transitions have the selection rule AS=AL=O,AJ=O, +l(butnotOc orO-+ 0) [12]. Therefore, only three transitions need to be considered. These magnetic-dipole transition rates were calculated using the formulae given in Ref. [lSJ. The radiative lifetime z of an excited state i is governed by
ri=[FWa(i,j)]-‘, where the summation is over all terminal states j. The values of the predicted spontaneous-radiative transition rates and radiative lifetimes of the
x = 40
W,, (s- ‘)
rR (ms)
0.511
0.424
0.348
state of Er3 + in (63-x)GeO, . xTe02. 27PbO. 10Ca0.0.2Er203 glasses with x = 0, 20, and 40, are listed in Table 3. According to our results, the radiative transition rates of Er3 + in (63-x)GeO, . xTe0, .27PbO. 10CaO glasses increase with increasing TeO, content x. The 4S3,2 + 411s,2 radiative transition rate has increased by about 46% from 1300 s - 1 for glass with x = 0 to 19OOs-’ for glass with x = 40. Because the values of J-O intensity parameters obtained do not show a significant change, the increased radiative transition rates are mainly from an increased local field correction factor x in the formulae, Eq. (3). The refractive index n increases from 1.79 for the glass with x = 0 to 1.95 for the glass with x = 40, which results in an enhanced local field and a larger radiative transition rate of Er3 + in glass. 4S3,2
3.2. Luminescence
and up-conversion
luminescence
The room temperature luminescence spectra of Er3+-doped (63-x)GeO, . xTe0,. 27PbO. lOCa0 glasses, excited by a 488 nm laser beam, are shown in Fig. 2, where the intensities have been corrected for the absorption at 488 nm, i.e., the luminescence intensity divided by the absorbance at 488 nm. The
of Luminescence
2. Pan, S.H. Morgan/Journal
305
75 (1997) 301- 3C.M
Wavelength (nm)
---__
500
571
667 -~
800 x=40 x=20
4s3/2 -+ 4h5/2
x=0
x20 4S3R + 4113/2
~ c41,5,2.
~___ 12500
_
15000
__L 17500
20000
Photon energy (cm-l) Fig. 2. Luminescence 488 nm and adjusted
luminescence
quantum
emitted Vq =
absorbed
spectra of Er 3+ in (67-x)Ge02. for display.
light radiation
yield, qq, is defined power power
=- r,,p rR ’
xTeOz ‘27PbO’
as [Zl]
(5)
where z,+ is the experimentally determined lifetime from the level in question and zR the radiative lifetime. The data in Fig. 3 illustrates that the luminescence quantum yield increases with increasing TeO, content in the host glass. The intensity of 4S3!z +4115,2 emission at 547 nm has increased by about twice in the glass with x = 40 compared to that in the glass with x = 0. The radiative lifetime has been calculated usof 4S3,2 -+ 41, 512 transition ing Judd-Ofelt analysis to be 0.51 ms in the glass with x = 0 and 0.35 ms in the glass with x = 40 (Table 3) the difference is due to the local field effect on Er3+ ions as discussed in the previous section. The measured lifetimes of 4S3,2 -+ 41 1s,2 transition at room temperature were about 9 us in the glass with x = 0 and about 14 us in the glass with x = 40. The quantum yield in question was then determined to be 1.8 and 4.0%, respectively.
lOCa0
glasses. The intensities
The measured governed by z exp = W,
are corrected
lifetime
+ WMP +
for the absorption
of an excited
WETI- l?
state
at
is
(6)
where WR is the total radiative rate of the state, WMP is the nonradiative rate due to multiphonon decay, and WET is the nonradiative rate due to energy transfer. In our case, the energy transfer rate between excited ions is very small because of the low Er3+ ion concentration [4,22]. The radiative rate from the 4S 3,2 level is also significantly lower than the multiphonon transition rate in question. Therefore, the measured lifetime is dominated by the multiphonon decay. The estimated multiphonon rates from measured lifetimes are 1.l x ~O’S-~ for the glass with x = 0 and 6.9 x 104s-i for the glass with x = 40. Our lifetime data therefore indicate a smaller multiphonon decay rate in the glass with a higher TeO, content. The reported nonradiative decay rates of rare-earth ions for a 3000 cm-’ gap are roughly 1.5 x lo5 s-i for germanate glass and 6 x lo4 s- ’ for tellurite glass [16]. Relaxation of rare-earth ions in glass through multiphonon processes depends on the highestenergy
306
2. Pan, S.H. Morgan /Journal of Luminescence
75 (1997) 301-30X
Wavelength (nm) 800
867
I
---
I
571
500
-
x20
l4s3/2 --f4113/2 : ‘.
I
4F~/2 -+
,’ :
/_
4l15/Z
~-12500
15000
17500
Photon energy (cm-l) Fig. 3. Up-conversion luminescence spectra of Er 3i- in (67-x) GeOz. xTe0, the absorption at 800 nm and adjusted for display.
phonons of the host glasses. Our Raman study has confirmed that the frequency of highest phonon band has decreased from 820 cm-’ for the glass with x = 0 to 750cm- ’ for the glass with x = 40 [23]. Because the rate of multiphonon decay is very sensitive to the maximum vibrational energy of the glass, a smaller multiphonon decay rate of rareearth ions is therefore expected in the glass with a higher Te02 content. According to our results, the luminescence efficiency of 4S,,, -+ 411s,2 transition in the glass with x = 40 has increased to twice that in the glass with x = 0. The mechanisms for this increase are: (i) an increased radiative transition rate due to the enhanced local field effect; (ii) a decreased multiphonon decay rate as a result of a reduced maximum phonon energy. Fig. 3 shows the room temperature up-conversion luminescence spectra of Er”+ in glasses. The intensities have been corrected for the absorption at 800 nm. This figure illustrates that the dominant emission is in the green corresponding to 4S312 + 41 1512 transition of Er3+. The green emission in the region 514-573 nm accounts for more
‘27PbO. lOCa0 glasses. The intensities are corrected
for
than 80% of the total emitted light in the spectral region studied (22 00&l 1300 cm- ‘). Fig. 3 also shows a red emission band at 662 nm and a nearinfrared band at 855 nm. These two emission bands are identified as the 4F9,2 -+ 41,5,2 and 4S3,2 + 4113,2 transitions, respectively. The energy levels in a two-step up-conversion process under 800 nm excitation are illustrated in Fig. 4. The up-conversion efficiency is defined as 1241 Pdt(Vis.)
n = P&ir.)
’
(7)
where Pemitis the emitted light power and Pabsis the absorbed incident light power. According to Fig. 4, the up-conversion efficiency increases with increasing Te02 content X. The up-conversion 4S3,z -+ 4115,2 green emission has increased by about 2.6 times in the glass with x = 40 compared to that in the glass with x = 0. Previous studies [1,25] indicated that the up-conversion efficiency is related to the population of the intermediate excited levels and the radiative quantum yield of the final emission level. In our case, 4111,2 and 4113,2 are the
Z. Pan, S.H. Morgan 1 Journal o/Lumintwence
,-
307
tellurite glasses and better thermal stability than a lead- germanate glass 63Ge02. 27PbO. 1OCaO (GPC) [1,23]. In this study, we determined the Judd-Ofelt intensity parameters and calculated the radiative transition rates of Er3+ in this new glass system. We also measured the luminescence and up-conversion luminescence. Our results indicate that the quantum yield of 4S,,2 + 41 t 5,2 transition increased by two times and the infrared (800 nm) to green (547 nm) up-conversion efficiency increased 2.6 times in the glass with x = 40 compared to that in the glass with .Y= 0. The increased luminescence efficiencies are attributed to an enhanced local held for Er3+ ions and a decreased multiphonon decay rate.
20
15
10
5.
0
75 (1997) 301 308
‘I,,,
Fig. 4. Energy-level diagram for a two-step infrared to visible up-conversion of Er 3+ in glass under 800nm excitation. The solid lines indicate radiative transition and waved lines show nonradiative decay.
intermediate levels and 4S312 is the final emission level. Our luminescence data have indicated a larger quantum yield of 4S3,z --f 41, 5,2 green emission in the glass with a higher Te02 content. In addition, larger populations in 411 1,2 and 41 13,2 levels are expected in the glass with a higher TeOz content because of a smaller multiphonon decay rate. Both of the above factors lead to a larger up-conversion efficiency as observed. In a sequentially excited up-conversion process, the up-conversion efficiency defined in Eq. (7) also depends on the infrared pump intensity [ 1,241. By using a focused infrared laser beam, a relatively large up-conversion efficiency of 1.1 x lo-’ has been obtained in a Er” ‘-doped lead-tellurium-germanate glass [ 11.
4. Conclusions We have presented the studies of the effect of modifying the glass network on the absorption and luminescence of Er3+ ions in a new glass system, in the form of (63-x)GeO* . xTe02. 27PbO. lOCa0 with .Y= 0, 10, 30, and 40. It has been reported that the lead-tellurium-germanate glasses showed a better glass forming ability than binary lead-
Acknowledgements The authors thank Dr. A. Ueda of Fisk University for conducting absorption measurements and Dr. H. Liu of University of Puerto Rico at Mayagiiez for conducting the lifetime measurements. Support for this work was provided by NASA through grant NAGW-2925.
References 111 Z. Pan. S.H. Morgan. K. Dyer. A. Ueda. H. Liu. J. Appl. Phys. 79 (I 996) 8906. 121 J. Wang. J.R. Lincoln. W.S. Brocklesby. R.S. Deol. C.J. Mackechnie, A. Pearson, A.C. Tropper. D.C. Hanna. D.N. Payne. J. Appl. Phys. 73 (1993) 8066. [31 J.S. Wang. E.M. Vogel, E. Snitzer. Opt. Mater. 3 (1994) 187. A. Loper. V. Kmg, B.H. Long, r41 Z. Pan. S.H. Morgan, W.E. Collins. J. Appl. Phys. 77 (1995) 4688. I51 Z. Pan, K. Dyer. A. Loper, S.H. Morgan, in: B.G. Potter Jr.. A. J. Bruce (Eds.). Synthesis and Application of Lanthanide-Doped Materials, The American Ceramic Society, Westerville, OH, 1996. p. 55. M.A. Piliavin. Germanate Glasses. Ch. (2) C61 A. Margaryan, Artech House. Boston, 1993. B. Piriou. and V.R. c71 S.J.L. Ribeiro. J. Dexpert-Ghys. Mastelaro. J. Non-Cryst. Solids 159 (1993) 213. 181 S. Nielsen, W.D. Lawson. and A.F. Fray. Infrared Phys. l(l961) 21. [91 A.F. Fray, S. Nielsen, Infrared Phys. I (1961) 175. A.F. Fray, Infrared Phys. [lOI S. Nielsen, W.D. Lawson, I (1961) 21.
308
[ll] [12]
2. Pan, S.H. Morgan /Journal
Z. Pan, S.H. Morgan, J. Non-Cryst. Solids 210 (1997) 130. M.D. Shinn, W.A. Sibley, M.G. Drexhage, R.N. Brown, Phys. Rev. B 27 (1983) 6635. [13] R. Reisfeld, Y. Eckstein, J. Non-Cryst. Solids 15 (1974) 125. [14] W.T. Carnall, P.R. Fields, K. Rajnak, J. Chem. Phys. 49 (1968) 4412; 4424. [is] M.J. Weber, Phys. Rev. 157 (1967) 262. [16] C.B. Layne, W.H. Lowdermilk, M.J. Weber, Phys. Rev. B 16 (1977) 10. 1171 C.B. Layne, M.J. Weber, Phys. Rev. B 16 (1977) 3259. [18] M.J. Weber, J.D. Myers, J.D. Blackburm, J. Appl. Phys. 52 (1981) 2944. [19] W. Ryba-Romanowski, J. Lumin. 46 (1990) 163.
of Luminescence
[20] [21] [22] [23]
[24] [25]
75 (1997) 301-308
R. Reisfeld, in: Structure and Bonding, vol. 13, Springer, New York, 1973, p. 136. L.A. Riseberg, M.J. Weber, in: E. Wolf (Ed.), Progress in Optics, Vol. XIV, Ch. III, Elsevier, Amsterdam, 1976. J.P. van der Ziel, L.G. Van Uitert, W.H. Grodkiewicz, R.M. Mikulyak, J. Appl. Phys. 60 (1986) 4262. Z. Pan, S.H. Morgan, A. Loper, V. Jones, A. Ueda, MRS Spring Meeting R4.3, San Francisco, 8-12 April, 1996. R.S. Quimby. M.G. Drexhage, M.J. Suscavage, Electron. Lett. 23 (1987) 32. D.C. Yeh, W.A. Sibley, I. Schneider, R.S. Afzal, 1. Aggarwal, J. Appl. Phys. 69 (1991) 1648.