Optical Properties of the TeO2-PbO-Li2O-Nd2O3 Glass System
*M.R.Sahar, A.Kassim, M.S.Rohani and R. Arifin
Advanced Optical Material Research Group, Physics Dept., Faculty of Science
Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor Darul Takzim
e-mail*: m-rahim@dfiz2.fs.utm.my
________________________________________________________________________
Abstract
Nd-doped Tellurite glasses of the (80-x)TeO2-10PbO-10Li2O-xNd2O3 system, where x = 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0
mol% have successfully been synthesized by melt-quenching techniques. Their absorption spectra had been
characterized in the ranges of 400-900 nm while the luminescence spectra are measured in the range of 400 to 1500
nm. The experimental,
f exp . and
calculated,
f cal
oscillator strength results are found to vary from 1.1566 x 10-6 to
0.8041 x 10-6 and from 1.0257 x 10-6 to 0.7715 x 10-6 with respect to the Nd3+content respectively. Meanwhile, the
Judd-Ofelt parameters,  2 varies from 3.472 x 1020cm2 to 2.576 x 1020cm2,
x 1020cm2 and
 6 varies
 4 varies from 2.110 x 1020cm2 to 2.108
from 2.424 x 1020cm2 to 2.392 x 1020cm2 with respect to Nd3+ content. From the
luminescence spectra it is observed that the emission occurs at 485nm, 560nm, 605nm, 880nm, 1062nm and 1340nm,
which is due to 4G9/2 → 4I9/2, 2G9/2 → 4I11/2 , 2H11/2 → 4I9/2 , 4F3/2 → 4I9/2 , 4F3/2 → 4I11/2 and 4F3/2 → 4I13/2 transitions
respectively.
Keywords: Tellurite glass, Sellmeier equation, oscillator strength, Judd-Ofelts parameters
________________________________________________________________________
PACS Code: 78.55 Qr
1. Introduction
Glasses containing heavy networks formers doped with rare earth ions have attracted a
considerable interest due to their low vibrational frequencies which cause inefficient non-radiative
relaxation [1]. Incorporation of rare earth into various glass oxides has been a key to the development of
many optical devices such as infrared lasers, IR-visible upconverters, fibre and waveguide amplifiers for
optical transmission network [2]. Presently, only rare earth ions are known to perform the lasing action in
glassy hosts [3]. Among these materials, particular attention has recently been devoted to the study of
tellurite glasses. Tellurite is well known to be good not only in mechanical strength and chemical durability
but also stable against atmospheric moisture, low melting temperature and good in corrosion resistance. It
also possesses excellent IR transmission [4] and high optical quality with high refractive index, [5, 6].
In addition, tellurite glass which possess lower phonon energies has been proved to be the most
stable hosts for obtaining efficient luminescence in rare earth compare to other oxide glasses [7]. In the past
several year, extensive works have been carried out for the Nd3+ doped tellurite glass since it has emerges
as an excellent lasing ion at 1063nm with a transition of 4F3/2 → 4I11/2 [8-10]. In this work, a systematic
study on the optical transition and the quantum efficiency of Nd-doped tellurite glass will be reported and
discussed with respect to the composition.
2. Experimental
The tellurite glass of (88-x)TeO2-10PbO-2Li2O-xNd2O3 system is prepared by melt-quenching
technique. Batches of 20g are prepared from certified reagent grades of TeO2 (99.95% purity), Li2CO3
(97%), PbO (98% purity) and Nd 2O3 (99.995%). The chemicals are firstly mixed thoroughly in a platinum
crucible before being heated at 1000 oC for half an hour. After the batch is completely melted, the melts
was cast onto the preheated stainless steel plate followed by annealing at 300 oC for 5 hours before allowed
to cool down to room temperature. The glass is then cut and polished at the thickness of about 2.0mm.
Electronic absorption spectra are determined at room temperature by using a Perkin Elmer UV
Spectroscopy in the range of 400 – 900 nm. The luminescence spectra are also obtained at room
temperature by using Nanosecond Luminescence Spectroscopy System, Model NT340/1 Ekspla excited at
585nm using the tunable Nd: YAG laser system NT342. The signal is monitored by monochromator
SP2300 equipped with photomultiplier in the photon counting mode and recorded under data acquisition
unit (DAQ).
3. Theory
The experimental oscillator strengths f exp . of the absorption spectral transitions of Nd 3+ doped in
tellurite glasses are calculated using the relation [14,15];
f exp =
where
2.303mc 2
E ( )d
e 2 Nd 
…………………………………………..(1)
N is the number density of rare earth ions, e is the charge of the electron, m is mass of the
electron,
c is the velocity of light, d is the sample thickness and E ( ) is the optical density. Meanwhile,
the calculated oscillator strengths,
f cal. of transition from the ground state to excited state is given by:
f cal. =
8 2 mc (n 2  2) 2
9n
3h (2 J  1)
  ( J U
t
t
 ' J ' ) 2 ………….
(2)
where
n is the refractive index of the host, h the Plank constant, 2J  1 is the degeneracy of the ground
state,
 t are the Judd-Ofelt parameters and the
  ( J U
t
t
 ' J ' ) 2 represents the doubly reduce
matrix elements which are host independence parameters. Three Ωt parameters where λ = 2, 4, 6 were
obtained by fitting using a least-squares method to the electric dipole contributions of the experimental
oscillator strength of the observed transitions to the calculated ones [1]. In this case, the magnetic dipole
contributions are subtracted from the observed oscillator strengths by using the procedure described by
Serqueira et. al [18].
A root- mean square error which gives a measure of the uncertainty resulting from the calculated and
experimental oscillator strength is given by:
 rms =
( f
cal .
 f exp . ) 2
……………………………(3)
q p
where q is the number of transition and p is the number of parameters.
The spontaneous emission probability or the radiative transition probability,
excited state (  J ) to a particular final state
AR (J :  ' J ' ) 
where,
64 4 v 3
3h(2 J  1)

( ' J ' ) is given by [17];
n ( n 2  2) 2
S ed  n 3 S md
9
S ed is the electric dipole line strength as given by;
S ed  e 2   t (J U t  ' J ' ) ………………………….(5)
and
AR for a transition from the
S md is the magnetic-dipole line strength as given by:
 …………..(4)
S md 
where
e2h2
(J L  2S  ' J ' ) 2 ……………….(6)
2
2 2
16 m c
(J L  2S  ' J ' ) are the reduced matrix elements of the operator L  2S . The
magnetic-dipole transitions have the selection rule
S  L  0, J  0, 1(but not 0  or
0  0 [16].
Since the fluorescent level relaxation generally involves transitions to all low-lying levels, therefore the
total radiative probability
AT may be written as;
AT (J )   ' J ' AR (J ,  ' J ' ) ………………………..(7)
Consequently, the fluorescent branching ratio
(  R ) which gives the most probable transition to occur is
given by;
 R ( J ,  ' J ' ) 
and the radiative lifetime
R
A(J ,  ' J ' )
AT (J )
…………………..…(8)
is given by,
 R (J )  [ AT (J ) 1 ] ………………………………….(9)
The quantum efficiency (  ) is calculated from the radiative lifetime  R obtained from Judd-Ofelts theory
and the measured experimental lifetime

T
R
T
as given by:
……………………………………… (10)
4. Results
4.1 Absorption spectra.
An absorption spectrum for (88-x)TeO2-10PbO-2Li2O-xNd2O3 glass system with wavelength
ranging from 400-900 nm at room temperature is shown in Figure 1. From Figure 1, it can be seen clearly
that all samples exhibit the absorption peaks around 11415 cm-1, 12422 cm-1, 13351 cm-1, 14663 cm-1,
15949 cm-1, 17094 cm-1, 18975 cm-1, 19455 cm-1, 21645 cm-1 and 23202 cm-1 which correspond to the
transitions from ground state of 4I9/2 to the excited state of 4F3/2, 4F5/2, 4F7/2, 4F9/2, 2H11/2, 4G5/2, 4G7/2, 4G9/2,
4
G11/2 and 2P1/2 respectively.
The nominal composition and the results of the experimental, f exp . and calculated,
f cal. oscillator
strength are given in Table 1. It should be noted that the experimental oscillator strength of absorption
bands of Nd3+ in these glasses are determined with the known values of Nd 3+ concentration, sample
thickness, peak position and peak areas by using Equation (1). As can be seen, the experimental and
calculated oscillator strength results are found varies from 1.157 x 10-6 to 0.804 x 10-6 and from 2.527 x 106
to 2.243 x 10-6 respectively with respect to the mol percentages of Nd 3+ for the transition 4I9/2 to 4G5/2. The
least square fitting procedure is then applied to determine the J-O intensity parameters Ωt and the result is
shown in Table 2. From Table 2 it can be seen that the J-O intensity parameters Ω2, Ω4 and Ω6 varies with
respect to mol% of Nd3+. For Ω2, the value varies from 3.4715 x 1020 cm2 to 2.5760 x 1020 cm2. For Ω4, the
value varies from 2.110 x 1020 cm2 to 2.108 x 1020 cm2 whereas for Ω6 the result was in a range of 2.424 x
1020 cm2 to 2.392 x 1020 cm2 respectively. Using these values, the electric dipole line strength, S ed can be
obtained by Equation (5). But since the magnetic contribution in an electronic transition within the 4fn
ground configuration of the Nd3+ ions is very small [9,10] then the magnetic dipole line strength, S md can
be neglected. Then the radiative transition probability, AR which represent the transition relaxation from
4
I9/2 to 4G5/2 may be obtained from Equation (4) and the values are inserted in Table 2.
4.2 Luminescence Spectroscopy
Figure 2 shows the luminescence spectrum of the (88-x)TeO2-10PbO-2Li2O-xNd2O3 glass system
under excitation at 585 nm. From Figure 2 it can be seen that there are seven significant emission peaks
centered at around 450nm, 485nm, 560nm, 605nm, 880nm, 1062nm and 1340nm which correspond to the
transition of 4G11/2 → 4I11/2, 4G11/2 → 4I13/2, 4G11/2 → 4I15/2, 4F3/2 → 4I9/2, 4F3/2 → 4I11/2 and 4F3/2 →4I13/2
respectively. The emission spectrum shown in Figure 2 represented the rest of the Nd3+ doped tellurite
glasses emission spectra. For the near infrared spectral region, three significant spontaneous transitions
probability peaks are observed with a transition of 4F3/2 → 4I11/2 is found to be dominant than the other two.
It also shows that there is no significant change in emission spectral peaks as the Nd3+ incorporated into the
glass system with different concentration.
Meanwhile, the branching ratio,  R which, correspond to the respective emission can be
calculated from Equation (8) and the values are inserted in Table 3. The results show that the branching
ratio,
 is maximum at 4F3/2 → 4I9/2 transition. Consequently, the calculated lifetime,  cal . and the
expected quantum efficiency,  can be obtained fro Equation (9) and Equation (10) respectively. These
values are also tabulated in Table 3. The quantum efficiency,
 which gives the ratio of total radiative
decay rate to total decay rate are found to be 78% for the visible emission and 70% for the near infrared
emission. The decrement of the experimental lifetimes to calculated lifetimes for emission spectra is mostly
due to the non-radiative losses factor.
4. Discussion
4.1 Absorption spectra and Judd-Ofelt Analysis
From Figure 1, it is observed that all glasses show great similarity in their general absorption shape with ten
significant absorption peaks which are assigned to the transitions from the 4I9/2 ground state to their
respective excited state. The positions of the absorption bands are similar to those observed for other Nd 3+
doped glasses [2]. Meanwhile, the values of the intensity parameters appear to be in the range usually
observed for oxide glasses [1]. Figure 3 shows the measured and calculated oscillator line strength of the
glass plotted against the Nd3+ content for the transition of 4I9/2 to 4G5/2. As can be seen from Figure 3, both
of the line strength decreases with respect to Nd 3+ content. This is due to the fact that an increasing of Nd3+
content will increases the number density of Nd3+ ions which is inversely proportionate to the measured
oscillator line strength, f exp . .Furthermore, an addition of Nd3+ may decreases the amount of the covalent
bond and the glass rigidity as indicated by the decreasing values of
 2 and  6 respectively. Jorgensen and
Reisfeld [17] in their work noted that the Ω2 parameter is the indicative of the amount of covalent bonding
while the Ω6 parameter is related to the rigidity of the host. Serquiera et. al. [18] in their work also found
that the decreases of Ω2 with respect to Nd content are due to the existence of the less centrosymmetrical
ion site and therefore covalent bond with the ligands. A similar result has also been observed by Rao et. al
[10] which shows that the Ω2 parameter reveals the dependence of the covalency between Nd ions and
ligand anions which is due to the asymmetry of the local environment around the Nd 3+ site. From Table 2,
it can be seen that the glass containing 0.5 mol% of Nd3+ (S11) posseses the highest Ω2 value which
indicate the highest covalent character [19]. Meanwhile, the small rms deviation,  rms obtained as in Table
1 indicate a good fit between the experimental and calculated oscillator strengths [20].
It also found out that the transition 4G5/2 is the most dominant spectral absorption intensity for all
samples. A similar trend has also been observed by other workers [14,15]. These sharp adsorption bands
are due to the inhomogeneously broadened f  f transitions that are essentially of electric-dipole nature
without considering the magnetic dipole line strength which can be neglected [9-10,21].
4.2 Luminescence Spectroscopy
From the luminescence spectra (Figure 2), seven emission transitions which are centered nearly at
450, 485, 560, 605, 880, 1062 and 1340 nm are detected. The Nd-YAG excitation at 585nm populates the
excited state of 4G5/2 which through the excitation state absorption (ESA) the electron are re-excited to the
high-lying metastable state from where some ions relax to 4G11/2 and 4F3/2 level through the multiphonon
non-radiative emission respectively. The transition from 4G11/2 to lower lying state therefore emitting the
blue emission (450nm), green emission (485nm), yellow emission (560nm) and orange emission (605nm)
whereas the transition from 4F3/2 to lower lying state emitting the near infrared emission (880nm, 1062nm,
1340nm). It is obvious that the 485nm green emission peak is dominant than all other emission peaks in the
visible spectral region as also been observed by other workers [11,22].
The branching ratio,  R , the calculated
 cal and measured  exp . lifetimes, together with radiative
quantum efficiency,  for the tellurite glass under the 4F3/2 →4I11/2 transitions are tabulated in Table 4.
From Table 4, it can be seen that the branching ratio varies from 32.170 to 32.560 depends on the content
of Nd3+. This result indicates that the amount of rare earth in the glass would play a major role in the
radiative transition probability. A similar results has also been reported elsewhere [24] thus supported the
argument. It also observed that the calculated lifetimes increases from 3.360ms to 3.314ms while the
measured lifetimes are found to decrease from 2.644ms to 2.085ms. Since the addition of Nd3+ into the
glass would increase the value of
a decrease in
 exp is
 4 and  6 , thus the rigidity of the network, then the increase in  cal or
very much expected [24]. Meanwhile, it was reported that the radiative lifetime is
inversely proportional to
 4 and  6 [25]. Figure 4 shows the quantum efficiency,  plotted against the
Nd3+ content for tellurite glasses. As can be seen from Figure 4, the quantum efficiency decreases with
respect to Nd3+ content. This indicates that the probable transitions levels are very much affected by the
increasing of non-radiative decay which is probably due to the intensified of cross-relaxation among the
Nd3+ ions [21]. The decreasing could also be due to the reduction of the interatomic distance of Nd3+ ions
that will cause the cross-relaxation [13].
5. Conclusion
Nd3+-doped tellurite glasses of the system TeO2-PbO-Li2O has successfully been synthesized by
using conventional method. Through the absorption spectra, the Judd-Ofelt parameters are found to vary in
the range usually found for the oxide glass. The experimental and the calculated oscillator strength are
found to be in the order of 10-6, increases as the Nd3+ content increases. Meanwhile, the spectral emission
peaks are found to center at 485 nm, 560nm, 605nm, 880nm, 1062nm and 1340nm respectively. The
quantum efficiency up to more than 75% may be obtained and the value is observed to be decreases as the
Nd3+ content increases.
5. Acknowledgement
The author would like to thanks Universiti Teknologi Malaysia and MOSTI for the research grant under
Vot 74532.
6. References
[1] – Raffaella Rolli, Karl Gatterer, Mario Wachtler, Marco Bettinelli, Adolfo Speghini, David Ajo, J.
Spectrochimica Acta Part A, (2001), 2009-2017.
[2] – Upendra Kumar K., Prathyusha V.A., Babu P., Jayasankar C.K., Joshi A.S.,
Speghini A., Bettinelli M., J. Spectrochimica Acta Part A. (2006)
[3] - M.R.Sahar, K.Sulhadi, M.S.Rohani, J.Non-Cryst.Solids, 354, (2008), 1179-1181.
[4] – M.R.Sahar, A.K.Jehbu, M.M.Karim, J. Non-Cryst. Solids, 213&214, (1997),164-167
[5] – Lin H., Tanabe S., Lin L., Yang D.L., Liu K., Wong W.H., Yu J.Y., Pun E.Y.B., J. Physics Letters A
358, (2006) 474-477.
[6]- Wang J.S., Vogel E.M., Snitzer E., Jackel J.L., Da Silva V.L., Silberberg Y., J. Non-Cryst.Solids 178,
(1994), 109.
[7] – Nazabal V., Todoroki S., Nukui A., Matsumoto T., Suehara S., Hondo T., Araki T., Inuoe S., Rivero
C., Cardinal T., J.Non-Cryst. Solids 325, (2003), 85-102.
[8] – Ajit Kumar G., Biju P.R., Venugopal C., Unnikrishnan N.V., J. Non-Cryst. Solids 221, (1997),47-58.
[9] – Cases R., Chamarro M.A., Alcala R., Rodriguez V.D. (1991), J. Luminescence 48 & 49, (1991),
509-514.
[10] – Rao T.V.R., Reddy R.R., Nazeer Ahammed Y., Parandamaiah M., Sooraj Hussain N., Buddhuhu S.,
Purandar K., J. Infrared Physics & Technology 41, (2000), 247-258.
[11] – Jaba N., Kanoun A. Mejri H., Maaref H., Brenier A., J.Physics: (2000).
[12] – Sidebottom D.l., Hruschka M.A., Potter B.G., Brow R.K., J.Non-Cryst. Solids 222, (1997),282-289
[13]- Yunlong Yu, Daqin Chen, En Ma, Yuansheng Wang, Zhongjian Hu, J.Spectrochimica Acta Part
A. (2006)
[14] – Judd, B.R., Phys. Rev. 127 (1962), 750
[15] – Ofelts, G.S, J.Chem.Physics 37(1962), 511
[16] – Pan, Z., Morgan, S.H., J. Luminescence 75 (1997), 301-308
[17] – Jorgensen C.K., Reisfeld R., J.Less Common Mat. 93, (1983),107
[18] – Serqueira, E.O., Dantas N.O., Monte A.F.G., Bell M.J.V. J.Non-Cryst. Solids 352, (2006),
3628-3632.
[19]- Subbalakshmi,P., Veeriah, N., J. Phys. Chem. Solids 64, (2003), 1027-1035
[20] – Moorthy, L.R., Jayasimhadri, M., Saleem, S.A., Murthy, D.V.R., J.Non-Cryst. Solids
353 (2007)1392-1396.
[21] – Hamit K., Huseyin C., Natali C., Alphan S., Gonul O., J.Luminescence 128 (2008),1501-1506.
[22] – Amorim H.T., Vermelho M.V.D., Gouveia-Neto, A.S., Cassanjes F.C., Ribeiro S.J.L., Messadeq Y.,
(2002) J.Alloys Comp. 346, (2002), 282-284.
[23] – G.A.Kumar, De La Rosa Cruz, K.Ueda, A.Martinez, O.Barbosa Garcia, Opt.Mat. 22(2003),201.
[24] – S.Mohan, K.S.Thind, G.Sharma, Brazilian Journal of Physics, 37, No.4, (2007)1306-1312.