Full text of DOI:10.1557/JMR.2000.0045

Journal of
MATERIALS RESEARCH
Welcome
Comments
Help
3+
Sensitizing effect of Yb on near-infrared fluorescence
4+
emission of Cr -doped calcium aluminate glasses
a)
Yong Gyu Choi and Kyong Hon Kim
Telecommunication Basic Research Laboratory, Electronics and Telecommunications Research
Institute (ETRI), Yusong P.O. Box 106, Taejon 305-600, Korea
Yong Seop Han and Jong Heo
Photonic Glasses Laboratory, Department of Materials Science and Engineering, Pohang University
of Science and Technology (POSTECH), San 31, Hyoja-dong, Nam-gu, Pohang,
Kyungbuk 790-784, Korea
(Received 6 June 1999; accepted 4 November 1999)
3+
4+
We have demonstrated that an efficient energy transfer takes place from Yb to Cr
3+
in calcium aluminate glasses. Yb improves excitation efficiency at around 980 nm,
4+
enhancing emission intensity of Cr fluorescence at 1.2–1.6 ␮m. Nonradiative energy
transfer via electric dipole–dipole interaction between ytterbium and chromium ions
3+
4+
was found to be dominant over radiative Yb → Cr energy transfer. A diffusion-
3+
4+
limited energy transfer mechanism well explains the decay behavior of Yb /Cr -
4+
codoped glasses. This codoping scheme may be applicable to other Cr -containing
crystals and glasses.
Conventional oxide glasses exhibit strong adsorption
in the 3.5–5 ␮m region, while infrared (IR) transmission
cutoff of calcium aluminate glasses is around 6 ␮m
wavelength, which is mainly attributed to low vibrational
standing associated with the processing condition and
4+
mechanism of Cr formation in glasses. For example,
4+
incorporation of Cr into fourfold coordination sites was
promoted when these glasses were melted under an inert
−1
6
4+
energy (∼ 700 cm ) of these glasses compared to that
atmosphere. However, the content of Cr ions was in-
dependent of melting atmosphere, and the relative con-
−
1
(
typically higher than 900 cm ) of the conventional
1
6+
oxide glasses. Calcium aluminate glasses have been
estimated to show the sum of scattering losses of ap-
proximately 0.04 dB/km at 1.55 ␮m. In addition, these
tent of Cr increased with increasing oxygen partial
7
pressure of the melting atmosphere.
2
4+
3
3
In T symmetry, the Cr : A → T absorption tran-
d 2 1
3
3
glasses exhibit a mechanical strength comparable to that
sition is electric dipole allowed while the A → T
2 2
transition is electric dipole forbidden. However, due to
the asymmetric phonons originated from distortion of the
3
7
of some silicate glasses. On the other hand, binary CaO–
Al O glasses have a narrow glass-forming region and a
2
3
3
3
pronounced tendency toward devitrification. However,
introduction of alkali and alkaline-earth metals to the
calcium aluminate system significantly improves the
perfect tetrahedron, the A → T transition can become
2 2
partially allowed. Therefore, strong absorption bands in
3
3
the visible region are attributed to the A → T transi-
2
1
4
glass-forming ability. On the basis of the above consid-
tion, while the weak near-infrared absorption is most
3
erations, calcium aluminate glasses can be applicable to
telecommunication uses. Renewed interests on these
glasses have risen since stable formation of the +4 oxi-
dation state of chromium dopant, which emits near-
likely arising from the T level (Fig. 1). On the other
2
hand, an excited-state absorption (ESA) conspicuously
4
+
occurs at the visible wavelength in Cr -doped crys-
9,10
4+
tals.
The same phenomenon is also evident in Cr -
5
6
infrared luminescence, has been known.
doped calcium aluminate glasses. This means that
optical pumping with wavelength longer than ∼ 900 nm,
where the absorption efficiency is low, is better to avoid
the ESA. However, a significant ground-state absorption
between 1200 and 1500 nm takes place even in a glass
doped with low concentration of Cr, i.e., 0.1 mol% Cr O
4+
Cr ion under specific fourfold crystal fields in some
oxide glasses emits 1.2–1.6 ␮m fluorescence which is
2
attributed to an intra-3d configurational transition. So
far, formation of the stable +4 oxidation state of chro-
mium has been achieved only in calcium-aluminate and
2
3
6,7
4+ 3
3
alumino-silicate glasses. There is still lack of under-
[Fig. 1(a)], which reabsorbs the Cr : T → A lumines-
2 2
cence responsible for the 1.2–1.6 ␮m fluorescence and
thereby makes a deleterious effect by lowering the cor-
responding quantum efficiency. Increase of doping con-
centration of the chromium ions to enhance the ab-
a)
Address all correspondence to this author.
278
J. Mater. Res., Vol. 15, No. 2, Feb 2000
© 2000 Materials Research Society
http://journals.cambridge.org
Downloaded: 06 Apr 2015
IP address: 150.131.192.151

Rapid Communications
3
sorption at the T level is also prohibited due to the
fluorescence in the 1.2–1.6-␮m band is about 190 nm.
2
8
probable concentration quenching effect. Thus, Cr con-
Figure 3 shows the fluorescence decay profiles of
3
+ 2
2
centration should be kept low and an efficient pump
mechanism is required for the chromium-doped crystals
and glasses for laser and amplifier applications.
In this paper, Yb ion, of which the strong absorption
peak is around 980 nm, is introduced as a sensitizer of
Yb : F → F
transition. An exponential decay
5
/2
7/2
with the 1/e fall time of 822 ␮s is evident in the 1.0 mol%
Yb O single-doped glass [Fig. 3(a)]. This measured
2
3
3+
lifetime is comparable with a radiative transition prob-
4+
Cr ion in calcium–aluminate glasses in order to provide
3+
a suitable pump band. Spectral overlap between Yb
4
+
emission and Cr absorption and the relatively high os-
4
+
cillator strength of Cr provide the theoretically ad-
3+
4+
equate backgrounds of Yb → Cr energy transfer.
1CaO–36Al O –10BaO–3ZnO (mol%) glasses
5
2
3
doped with Cr O and (or) Yb O were prepared in air by
2
3
2 3
the conventional melt-quenching method. Starting mate-
rials with purity higher than 99.9% were weighed and
mixed to yield 10-g-batch mixtures. The glass samples
were obtained by melting the mixtures in Pt crucibles at
1650 °C for 1 h and subsequent annealing at 700 °C. Up
to 1.0 mol% Yb O could be dissolved in our host com-
2
3
position without any sign of crystallization. A charge-
6+
transfer band of Cr as well as the intraconfigurational
4+
transitions of Cr are resolved in the low-intensity ab-
sorption spectrum of the Cr O -doped glasses [Fig. 1(a)].
2
3
3
+
Absorption from Yb [Fig. 1(c)] is well overlapped
4+
3
3
with the absorption of the Cr : A → T transition
2
2
[
Fig. 1(b)]. Using a 978-nm pump, fluorescence emission
enhanced in magnitude at room temperature is evident in
a glass codoped with 0.01 mol% Cr O and 1.0 mol%
Yb O (Fig. 2). The full width at half-maximum of the
FIG. 2. Fluorescence emission spectra of calcium aluminate glasses
doped with (a) 0.01 mol% Cr O only and (b) 0.01 mol% Cr O /1.0
2
3
2 3
2
3
mol% Yb O . Excitation wavelength was at 978 nm, the peak wave-
2
3
2
3
3+
length of the Yb absorption spectrum.
3
+
2
2
FIG. 3. Fluorescence decay curves of the Yb : F → F transi-
5
/2
7/2
FIG. 1. Low-intensity absorption spectra of calcium aluminate glasses
doped with (a) 0.1 mol% Cr O , (b) 0.01 mol% Cr O /1.0 mol%
tion for 1.0 mol% Yb O -doped glasses with Cr O doping concen-
trations of (a) 0, (b) 0.01, and (c) 0.1 mol%, respectively. Solid lines
were obtained from the least-squares fit to Eq. (2).
2 3 2 3
2
3
2 3
Yb O , and (c) 1.0 mol% Yb O , respectively.
2
3
2 3
J. Mater. Res., Vol. 15, No. 2, Feb 2000
279
http://journals.cambridge.org
Downloaded: 06 Apr 2015
IP address: 150.131.192.151

Rapid Communications
3+
−1
ability of Yb in this glass host (∼ 1049 ± 10 s ), which
is calculated from the absorption spectrum using the
1
1
Fuchtbauer–Ladenburg equation.3 This indicates that
+
the radiative transition rate of Yb is hardly affected by
the phonon environment. Meanwhile, when the concen-
tration of Cr O is increased from 0.01 to 0.1 mol%, the
2
3
decay profiles result in shortened decay times and clear
nonexponential decay behaviors [Figs. 3(b) and 3(c)].
Such a nonexponential decay suggests that there is an
energy transfer from ytterbium ion to another ion, most
probably chromium ion in this case.
When the activators or quenching centers are present,
a decaying behavior of the sensitizer fluorescence fol-
lows a nonexponential behavior at the beginning. At
the later stage of the decay, however, an exponen-
tial behavior starts to dominate because of either the
energy migration among sensitizer ions or ceasing of
12
donor–acceptor energy transfer. In general, both direct
sensitizer → activator energy transfer and diffusion
among sensitizer ions affect overall decay behavior of a
sensitizer. An overall deexcitation probability N(t) with
time t is expressed as follows:
FIG. 4. Dependence of ⌸(t) on ln(t) for 1.0 mol% Yb O -doped
2
3
glasses with codopant of (a) 0.1 and (b) 0.01 mol% Cr O , respec-
2
3
tively. Solid lines were plotted from the least-squares fit to the straight
line.
N(t) ס exp{−W t − ⌸(t)}
,
(1)
0
where W is a reciprocal of lifetime without any energy
0
acceptor and ⌸(t) is a magnitude of the departure from
the intrinsic exponential behavior due to the energy transfer
between sensitizer and activator ions. In a multipole–
coefficient for energy migration among donors. Then, the
critical transfer distance R can be calculated from an equa-
3/s
0
multipole interaction scheme, ⌸(t) is proportional to t ,
where the exponent s can be either 6, 8, or 10 de-
pending upon the electric-multipole character of the
6
0
tion, R ס C /W . For an isolated sensitizer–activator pair
sa
0
separated by R , the energy transfer occurs with the same
0
rate as the spontaneous deactivation in the sensitizer. The
13
sensitizer–activator interaction. Then, the exponent of
(t) can be revealed from a slope in the plot of ln{−ln
optimized R was estimated to be 4 ± 1 nm using results of
0
⌸
the least-squares fits in Figs. 3(b) and 3(c). The distance
N(t) − W t} versus ln(t). It is verified in Fig. 4 that the
3+
3+
0
is larger than that of Eu → Cr transfer in a phosphate
slope is 0.74 and 0.62 for 0.01 and 0.1 mol% Cr O3
15
2
glass.
doped glasses, respectively. The slightly higher value
Radiative energy transfer was also appreciable in the
than 0.5 for the exponent of ⌸(t), i.e., s ס 6, may suggest
3+
4+
Yb /Cr -codoped glasses, since the spectral shape of
3+
that a considerable amount of diffusion among Yb ions
intervenes in the time scale of our observation. The de-
crease in the exponent with increasing chromium con-
centration further indicates that the effect of direct energy
transfer becomes dominant as the chromium concentra-
tion increases. Therefore, both diffusion among sensi-
tizer ions and single step electric dipole–dipole transfer
from sensitizers to randomly distributed activator ions
should be considered, and the following equation derived
3+
2
2
the Yb : F → F transition was distorted accord-
5/2
7/2
4+
3+
ing to the absorption line shape of Cr around Yb
emission wavelength. On the other hand, a back energy
transfer from chromium to ytterbium seems not to be
plausible. The measured lifetimes of the infrared emis-
sion in calcium aluminate glasses are in a range of sub-
microseconds to ∼ 40 ␮s at room temperature, and the fast
transition rates mainly result from the strong nonradia-
5,16
tive multiphonon relaxation.
14
by Yokota and Tanimoto was used for the fits of ex-
perimental decay curves:
In conclusion, we have demonstrated improvement of
9
1
80-nm pump efficiency as well as enhancement of 1.2–
4+
.6 ␮m fluorescence of Cr -doped calcium aluminate
3
ր
2
1
ր
2
1
ր
2
4
3+
N͑t͒ = exp
ͭ
− W t − ⁄
3
␲
naC
t
0
sa
glasses by introduction of Yb as a sensitizer. Electric
dipole–dipole energy transfer from Yb to Cr , which
is limited by diffusion among Yb ions, is a dominant
energy transfer mechanism. This codoping scheme may
be applicable to other Cr -containing crystals and
glasses.
3+
4+
2
3
ր
4
1
+ 10.87x + 15.50x
3+
ͩ
ͪ
ͮ
.
(2)
1
+ 8.743x
4
+
Here, n is an activator concentration, C is an interac-
a
sa
−
1/3 2/3
tion parameter, and x ס DC
t
, where D is a diffusion
sa
280
J. Mater. Res., Vol. 15, No. 2, Feb 2000
http://journals.cambridge.org
Downloaded: 06 Apr 2015
IP address: 150.131.192.151

Rapid Communications
ACKNOWLEDGMENTS
5. K. Cerqua-Richardson, B. Peng, and T. Izumitani, OSA Proc.
Adv. Solid-State Lasers 13, 52 (1992).
Work at ETRI was supported by the Ministry of In-
formation and Communication of Korea. Work at
POSTECH was supported in part by the Korean Science
and Engineering Foundation (KOSEF) through the Ultra-
Fast Fiber-Optic Networks Research Center at Kwangju
Institute of Science and Technology.
6
. E. Munin, A.B. Villaverde, M. Bass, and K. Cerqua-Richardson,
J. Phys. Chem. Solids 58, 51 (1997).
7. T. Murata, M. Torisaka, H. Takebe, and K. Morinaga, J. Non-
Cryst. Solids 220, 139 (1997).
. M.J. Weber, J. Appl. Phys. 44, 4058 (1973).
. Z. Burshtein, P. Blau, Y. Kalisky, Y. Shimony, and M.R. Kokta,
IEEE J. Quantum Electron. 34, 292 (1998).
8
9
1
1
1
0. V.P. Mikhailov, N.I. Zhavoronkov, N.V. Kuleshov, A.S. Avtukh,
and V.G. Shcherbitsky, Opt. Quantum Electron. 27, 767 (1995).
1. H. Takebe, T. Murata, and K. Morinaga, J. Am. Ceram. Soc. 79,
681 (1996).
2. I.A. Bondar, A.I. Burshtein, A.V. Krutikov, L.P. Mezentseva,
V.V. Osiko, V.P. Sakun, V.A. Smirnov, and I.A. Shcherbakov,
Sov. Phys. JETP 54, 45 (1981).
REFERENCES
1. Y.G. Choi and J. Heo, J. Non-Cryst. Solids 217, 199 (1997).
2
. M.E. Lines, J.B. MacChesney, K.B. Lyons, A.J. Bruce,
A.E. Miller, and K. Nassau, J. Non-Cryst. Solids 107, 251 (1989).
13. D.L. Dexter, J. Chem. Phys. 21, 836 (1953).
14. M. Yokota and O. Tanimoto, J. Phys. Soc. Jpn. 22, 779 (1967).
15. M.J. Weber, Phys. Rev. B4, 2932 (1971).
16. X. Wu, H. Yuan, W.M. Yen, and B.G. Aitken, J. Lumin. 66&67,
285 (1996).
3
. G.Y. Onoda, Jr., and S.D. Brown, J. Am. Ceram. Soc. 53, 311
1970).
(
4
. E.V. Uhlmann, M.C. Weinberg, N.J. Kreidl, and A.A. Goktas,
J. Am. Ceram. Soc. 76, 449 (1994).
J. Mater. Res., Vol. 15, No. 2, Feb 2000
Downloaded: 06 Apr 2015
281
http://journals.cambridge.org
IP address: 150.131.192.151