Physica B 413 (2013) 5558

Contents lists available at SciVerse ScienceDirect

Physica B
journal homepage: www.elsevier.com/locate/physb

Violet and near-ultraviolet upconversion luminescence in La3Ga5.5Ta0.5O14

codoped with Er3 and Yb3
S. Georgescu a, A.M. Voiculescu a, C. Matei a,b, A.G. Stefan a,b, O. Toma a,n
a
National Institute for Laser, Plasma and Radiation Physics, 409 Atomistilor Street, Magurele, Ilfov, 077125, Romania
b
University of Bucharest, Faculty of Physics, 405 Atomistilor Street, 077125, Magurele-Ilfov, Romania

a r t i c l e i n f o abstract

Article history: Violet (410 nm, 2H9/2-4I15/2) and near-ultraviolet (380 nm, 4G11/2-4I15/2) upconversion luminescence,
Received 4 October 2012 were obtained for the rst time in Er:Yb:La3Ga5.5Ta0.5O14 ceramic sample, synthesized by solid-state
Received in revised form reaction, by pumping at 973 nm. Yb3 Er3 energy transfer processes accounting for population of the
11 December 2012
emitting levels are discussed. The lifetime of the emitting level 2H9/2, responsible for violet
Accepted 22 December 2012
luminescence, was measured and its quantum efciency was estimated.
Available online 4 January 2013
& 2013 Elsevier B.V. All rights reserved.
Keywords:
Upconversion
Luminescence
Rare earth
Langatate
Er3
Yb3

1. Introduction The langatate crystal (La3Ga5.5Ta0.5O14LGT) doped with erbium

The possibility to obtain phosphors that emit ultraviolet (UV)
radiation for pumping in infrared (IR), may lead to important
applications in biology, medicine and treatment of wastewater. Given
the small penetration depth of the UV radiation in biological tissues
or in turbid media [1], the phosphors emitting in UV and pumped in
IR can produce useful UV radiation deep inside the material.
Generation of upconversion luminescence (UCL) from visible
to UV proved to be of interest in the improvement of the
photocatalytic activity of some materials [25]. Generally, for
pumping with visible radiation, two-photon processes are neces-
sary to produce UV emission. Generation of UV radiation for IR
pumping involves n-photon processes with nZ3, whose ef-
ciency is much lower. Nevertheless, the superior penetration
depth for IR radiation in turbid media could compensate the
inferior IR to UV upconversion efciency in rapport with visible to
UV one, when a larger penetration depth is required.
When excited in IR, besides the green and red luminescence,
the materials doped with erbium and ytterbium emit in blue,
violet and ultraviolet. Under IR pumping, upconversion UV emis-
sion in systems doped with erbium have been obtained by at least
three-photon processes [613]; higher order processes as ve-
photon [13] or even six-photon UC uorescence have also been
reported [14].
shows a high quantum efciency for the (2H11/2, 4S3/2) level. According
to the data published by Ryba-Romanowski et al. [15], this quantum
efciency for a LGT crystal doped with 0.57 at% erbium is 72%,
comparable with that for Er:LiYF4 (70%, using the radiative lifetime
(571 ms, taking into account the thermalization of 2H11/2 and 4S3/2)
from Ref. [16] and the measured lifetime (400 ms) given in Ref. [17]).
In a recent paper [18], we analyzed the green ((2H11/2, 4S3/2)-
4
I15/2) and red (4F9/2-4I15/2) luminescence, excited in infrared
with a diode laser, of Er3 in LGT doped with 1 at% erbium and
3 at% ytterbium. A rather high quantum efciency (58%) of the
(2H11/2, 4S3/2) emitting level was obtained.
In the present paper, we extend the study of IR-pumped,
upconversion luminescence in LGT:Er:Yb, by examining the violet
(2H9/2-4I15/2, at  408 nm) and near-ultraviolet (4G11/2-4I15/2, at
 380 nm) emissions. The mechanisms leading to the population of
the emitting levels are also investigated and the levels lifetimes and
quantum efciency are discussed. Besides, weak bluegreen (4G11/
4 4 4 4
2- I13/2 at 508 nm) and blue ( F7/2- I15/2 at 488 nm and ( F3/2,
4
F5/2)-4I15/2, at 450 nm) luminescence was observed. To our best
knowledge, violet, near-ultraviolet and blue UCL in compounds from
the langasite family doped with Yb3 and Er3 has not been
reported.
2. Experimental

n
Corresponding author. Tel.: 40 214574550x2133; fax: 40 214 574489. The LGT powder doped with erbium and ytterbium (1 at% and
E-mail address: octavian.toma@inpr.ro (O. Toma). 3 at%, respectively) was synthesized by solid-state reaction.

0921-4526/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.physb.2012.12.045
56 S. Georgescu et al. / Physica B 413 (2013) 5558
Stoichiometric quantities of high purity oxides (La2O3, Ga2O3,
Er2O3, Yb2O3, Ta2O5) were mixed in an agate mortar, pressed with
a hydraulic press at 2.5 MPa and then annealed in air at 1350 1C
for 35 h. As a result of the thermal treatment, a solid, ceramic
sample was obtained. The sample was cut, polished and washed
in an ultrasonic bath to remove the abrasive particles. The
luminescence of the LGT:Yb:Er solid sample was excited in IR,
at 973 nm, with various pump powers, using the laser diode
DioMod980/30/400. The experimental set-up for luminescence
measurements contains a Horiba Jobin-Yvon monochromator
(model 1000 MP), a S-20 photomultiplier and the SR830 lock-in
amplier on line with a computer. The pump beam was modu-
lated using the mechanical chopper (model SR450) in order to
allow the improvement of the selectivity of the experimental
setup by a proper choice of the phase of the lock-in. The same
ceramics sample was used for diffuse reectance and decay
measurements. The experimental setup for diffuse reectance
measurements was presented elsewhere [19]. For decay measure-
ments, the luminescence was excited with the third harmonics of
a pulsed Nd:YAG laser. All the measurements were performed at
room temperature.
3. Results and discussion
The phase purity of LGT:Er:Yb was analyzed previously [18] by
X ray diffraction. Besides the diffraction lines of LGT phase, some
other lines, but with very low intensities, were observed. The
extralines belong to perovskite LaGaO3 and b-Ga2O3 phases.
The ceramic sample used in experiments is shown in Fig. 1.
This sample was cut from the sintered pellet. In contrast with
langatate single crystals grown in air, which show a rather strong
reddish color due to the color centers (Ref. [20] and references
therein), our ceramic sample, though annealed in air for a rather
long time, does not show the reddish color. Only the surface of the
sample shows this coloration. This coloration, observed also in the
nanocrystals from the langasite family, whose intensity increases
with the annealing temperature, screens the absorbed or emitted
radiation, mainly in the short wavelength domain [21]. In our
case, the reduction of the reddish coloration permits an easier
observation of the violet and near-UV emitted luminescence.
The diffuse reectance spectrum obtained for the ceramic
sample is given in Fig. 2. Though the cut ceramic sample looks
white, the diffuse reectance baseline is slightly tilted towards
short wavelength domain (mostly for l o400 nm), denoting that
some color centers are still present.
Fig. 1. LGT:Er(1 at%):Yb(3 at%) ceramic sample. Left: the cut and polished surface; right: the surface of the pellet, as obtained after the thermal treatment in air.
The green and red luminescence of LGT:Er:Yb for IR pumping
was presented elsewhere [18]. Here, we present the luminescence
spectrum in the 375512.5 nm domain, for pump at 973 nm
(Fig. 3). This domain corresponds to the following Er3 transitions:
4
G11/2-4I15/2, 2H9/2-4I15/2, (4F3/2, 4F5/2)-4I15/2, and 4G11/2-
4
I13/2, partially superposed with 4F7/2-4I15/2. For comparison
purposes, we introduced in the inset the luminescence spectrum
of LGT:Er:Yb in the spectral range 350 nm700 nm. Due to the
much higher intensity of the green luminescence ((2H11/2, 4S3/2)-
4
I15/2; see inset), the tail of the 2H11/2-4I15/2 luminescence band
is superposed on the weaker 4G11/2-4I13/2 and 4F7/2-4I15/2 lines
(Fig. 3, black line; R output of the lock-in amplier). In our
sample, the (2H11/2, 4S3/2) level has a rather long lifetime:
160 ms [18]. Due to the narrow energy gaps between 4G11/2, and
2
H9/2 ( 1400 cm  1) between 4F5/2 and 4F7/2 (  1500 cm  1), as
well as between 4F7/2 and 2H11/2 (  1100 cm  1), we expect much
shorter lifetimes for these levels. Therefore, we can separate the
spectra corresponding to 4G11/2-4I13/2 and 4F7/2-4I15/2 transi-
tions from the green tail by a proper choice of the lock-in phase.
We have chosen the phase of the lock-in to eliminate the green
((2H11/2, 4S3/2)-4I15/2) luminescence (Y output of lock-in ampli-
er). The result is shown in Fig. 3 with gray lines. Due to the very
low intensity of the diffraction lines corresponding to the para-
sitic phases (see Fig. 1 in Ref. [18]) we consider that the dopant
ions included in these phases do not contribute signicantly to
the sample luminescence.
For the elucidation of the population mechanisms of the
various emitting levels, we investigated the dependence of their
luminescence intensities on the pump power. For unsaturated
Fig. 2. Diffuse reectance spectrum of LGT:Er:Yb. The absorption transitions are
labeled.
 S. Georgescu et al. / Physica B 413 (2013) 5558 57

Fig. 3. Near-UV, violet and blue upconversion luminescence of Er3 in LGT:Er:Yb. Black lines: R output of the lock-in amplier; gray lines: the spectra of 4G11/2-4I13/2 and
4
F7/2-4I15/2 are separated from the tail of the 2H11/2-4I15/2 line by a proper choice of the phase of the lock-in amplier (Y output). The luminescence spectrum extended
from near-UV to red is presented in the inset. The spectra are not corrected for the spectral responsivity of the experimental apparatus.

of 800 cm  1, are thermalized. An increase of the local tempera-

ture favors the population of 2H11/2. Therefore, since the local
temperature of the sample increases monotonously with the
pump power, a larger slope for the luminescence corresponding
to the 2H11/2-4I15/2 transition is expected. In fact, as Fig. 4 shows,
the slope for this transition is 1.95 while, for 4S3/2-4I15/2, the
slope is 1.76. This property (increase of the uorescence intensity
ratio FIR of the 2H11/2-4I15/2 and 4S3/2-4I15/2 with tem-
perature) is used to measure temperature [25].
It can be shown that, in the three-level approximation (2H11/2,
4
S3/2 and 4I15/2), the temperature increase of the sample, Dt, due
to the laser irradiation, when the incident power increases from
P1 to P2, is related to the difference between the slopes of 2H11/2
and 4S3/2 in the double logarithmic plot, Dn, by the approximate
relation:

kB T 21

Dt  Dnln P 2 =P 1 1
DE
where T1 is the absolute temperature of the sample corresponding
Fig. 4. Double logarithmic plot of luminescence intensity vs. incident laser power. to the incident power P1 and
kB is the Boltzmann constant. In our
The difference between the slopes of 4S3/2 and 2H11/2 (which are thermalized) is case, Dn 0.19 and ln P 2 =P 1  0.9 (see Fig. 4). For T1 300 K and
due to the temperature increase as the result of the laser irradiation. For
DE800 cm  1, it results Dt E13 1C. This temperature variation is
discussion, the dependence of the total green luminescence is also shown.
relatively small and does not inuence signicantly the results of
the luminescence measurements.
upconversion processes, the luminescence intensity is propor- The main energy transfer (ET) upconversion processes Yb3 -
tional with the nth power of the incident pump power, where n is Er3 are summarized below and shown in Fig. 5.
the number of pump photons necessary to populate the emitting
level. In a double logarithmic plot, luminescence intensity vs.  ET1: 2F5/2 (Yb3 ) 4I15/2 (Er3 )-2F7/2 (Yb3 ) 4I11/2 (Er3 );
incident pump intensity, n is the slope of this dependence. Such  ET2: 2F5/2 (Yb3 ) 4I11/2 (Er3 )-2F7/2 (Yb3 ) 4F7/2
dependence, for the violet and near-UV luminescence, is given in multiphonon
(Er3 ) ! (2H11/2, 4S3/2) (Er3 );
Fig. 4. The intensity of the other luminescence bands in Fig. 3 is 2
 ET3: F5/2 (Yb3 ) 4I13/2 (Er3 )-2F7/2 (Yb3 ) 4F9/2
too low to be represented reliably in the double logarithmic plot. 3
(Er ) phonons;
In order to understand the processes which populate the levels 2
 ET4: F5/2 (Yb3 ) 4F9/2 (Er3 )-2F7/2 (Yb3 ) 2H9/2
responsible for violet (2H9/2) and UV (4G11/2) luminescence, the
(Er3 ) phonons;
dependence of the intensity of green (2H11/2-4I15/2 and 4S3/2-
4
 ET5: 2F5/2 (Yb3 ) (2H11/2, 4S3/2) (Er3 )-2F7/2 (Yb3 ) 2K15/2,
I15/2) and red (4F9/2-4I15/2) on the pump intensity is also shown. 4 multiphonon
The slope in the double logarithmic plot of the green lumines- G9/2 (Er3 ) ! 4G11/2 (Er3 ).
cence in Fig. 4 is 1.83, i.e. a two-photon process. The slope for red
(4F9/2-4I15/2) luminescence is 1.57 (also a two-photon process), The slope for 4G11/2 luminescence is 2.73 while for 2H9/2 is
due to the saturation of the population of the intermediate levels 2.47; for both levels, three-photon processes are necessary. An
(4I13/2, in this case) [2224]. examination of the energy level schemes of Er3 and Yb3 (Fig. 5)
The local temperature increase, as a result of the incident corroborated with the values of the slopes, suggests that the level
4
pump power, is evidenced in the separate loglog plots of the two G11/2 is populated by ET5. Indeed, the slope of 4G11/2 (2.73)
green luminescence bands (2H11/2-4I15/2 and 4S3/2-4I15/2). differs from the slope of thermalized levels (2H11/2, 4S3/2) which
At room temperature, these levels, being separated by a gap DE is 1.83 with 0.9 ( E1), i.e. one supplementary photon is
58 S. Georgescu et al. / Physica B 413 (2013) 5558
Fig. 5. Energy level schemes of Yb3 and Er3 and the main energy transfer
processes from Yb3 to Er3 represented by curved dashed arrows. The dotted
arrows indicate the multiphonon transitions.
necessary to promote the system from (2H11/2, 4S3/2) in 4G11/2.
Similarly, the slope of 4F9/2 is 1.57 which differs from the slope of
2
H9/2 (2.47) also by 0.9 ( E1). Therefore, the level 2H9/2 is
populated by energy transfer from 4F9/2 (ET4). Since the gap
between the Er3 levels 4G11/2 and 2H9/2 is rather low ( 1400 cm  1),
we expect that the multiphonon transition from 4G11/2 could have a
contribution to the population of 2H9/2 level [26].
The luminescence intensity of the blue emitting levels (4F3/2,
4
F5/2) is too low to have reliable data about its intensity depen-
dence on the pump intensity. According to the energy level
scheme of Er3 (Fig. 5), they are populated by multiphonon
transitions from 2H9/2. The 4F7/2 level of Er3 is populated by
the two-photon process 2F5/2 (Yb3 ) 4I11/2 (Er3 ) -2F7/2
(Yb3 ) 4F7/2 (Er3 ).
The third harmonics (355 nm, 10 ns) of the Nd:YAG laser was
used in order to measure the lifetimes of the higher emitting
levels of Er3 in LGT. The energy of the 355 nm photon corre-
sponds approximately to the Er3 transition 4I15/2-2G7/2 (not
shown in Fig. 5). By rapid multiphonon transitions, the Er3 levels
4 Commun. 148 (2008) 96.
G11/2 and 2H9/2 are populated.
Under UV pumping, the crystals from the langasite family
(langasite La3Ga5SiO14, langanite La3Ga5.5Nb0.5O14, langatate)
show a rather strong luminescence in the UV-visible domain
[27]. Due to the superposition of the Er3 luminescence with host
luminescence under 355 nm excitation, the measurement errors
are rather high. The uorescence lifetime of the energy level of
interest 2H9/2 was estimated from the experimental decay
curve: 1 ms. This level has the most intense luminescence in the
violetnear-UV spectral range; the luminescence of the other
levels (see Fig. 3) is much less intense and could not be separated
from the background luminescence of the host material.
The radiative lifetimes of the Er3 levels between 4I13/2 and
2 O. Zaharko, E.N. Domoroschina, A.B. Dubovsky, Cryst. Res. Technol. 47 (2012)
H11/2 in LGT were calculated in Ref. [15] using the JuddOfelt
theory [2830]. Using the JuddOfelt parameters given in Ref.
[15] and the refraction indexes from Ref. [31], we calculated the
radiative lifetime of level 2H9/2 (132 ms). Using the uorescence
lifetime previously estimated, the quantum efciency of 2H9/2
could be calculated: 0.76%. This rather low value is mainly due to
the small energy gap under this energy level (  1700 cm  1) and
the relatively high energy of the phonons characteristic to oxide
hosts. For comparison, we calculate the quantum efciency of
level 2H9/2 in a crystal with phonons of lower energy, LiYF4:Er. The
radiative lifetime of this is 440 ms [16] and the measured one is
10 ms [17]. The resulting quantum efciency is 2.3%.
4. Conclusions
LGT:Yb(3 at%):Er(1 at%) ceramic sample with good phase pur-
ity and low concentration of color centers was obtained by solid-
state reaction in air. Blue, violet and near-ultraviolet lumines-
cence was obtained by pumping in IR, at 973 nm. Due to the
interference of the strong luminescence of the LGT host under
355 nm pumping, only the lifetime of the Er3 level 2H9/2,
responsible for the violet luminescence was measured and the
quantum efciency of this level was estimated.
During the IR pumping, the temperature of the irradiated
sample increases. An estimation of the temperature excursion
was made from the difference between the slopes, in the double
logarithmic plot, of 2H11/2 and 4S3/2 luminescence intensity func-
tion of the incident pump power.
The quantum efciency of the violet emitting level is rather
low (o1%) due to the relatively high energy of the phonons
characteristic to oxide hosts.
Acknowledgment
This work was supported by the Romanian National Research
Council (CNCS) in the frame of the Projects IDEI 82/06.10.2011
and PD 51/05.10.2011.
References
[1] A. Vogel, V. Venugopalan, Chem. Rev. 103 (2003) 577.
[2] Z. Zhang, W. Wang, W. Yin, M. Shang, L. Wang, S. Sun, Appl. Catal. B 101
(2010) 68.
[3] J. Gao, X. Luan, J. Wang, B. Wang, K. Li, Y. Li, P. Kang, G. Han, Desalination 268
(2011) 68.
[4] T. Li, S. Liu, H. Zhang, E. Wang, L. Song, P. Wang, J. Mater. Sci. 46 (2011) 2882.
[5] L. Yin, J. Gao, J. Wang, X. Luan, P. Kang, Y. Li, K. Li, X. Zhang, Res. Chem.
Intermed. 38 (2012) 523.
[6] D.N. Patel, R.B. Reddy, Appl. Opt. 37 (1998) 7805.
[7] H. Liang, G. Chen, H. Liu, Z. Zhang, J. Lumin. 129 (2009) 197.
[8] G. Wang, W. Qin, J. Zhang, J. Zhang, Y. Wang, C. Cao, L. Wang, G. Wei, P. Zhu,
R. Kim, Opt. Mater. 31 (2008) 296.
[9] G.Y. Chen, H.C. Liu, H.J. Liang, G. Somesfalean, Z.G. Zhang, Solid State
[10] C. Joshi, K. Kumar, S.B. Rai, J. Appl. Phys. 105 (2009) 123103.
[11] G. Wang, W. Qin, L. Wang, G. Wei, P. Zhu, D. Zhang, F. Ding, J Rare Earths
27 (2009) 394.
[12] J. Zhang, L. Zhang, S. Lv, W. Qin, J. Nanosci. Nanotechnol. 10 (2010) 1974.
[13] K. Zheng, L. Wang, D. Zhang, D. Zhao, W. Qin, Opt. Exp. 18 (2010) 2934.
[14] K. Zheng, D. Zhao, D. Zhang, N. Liu, W. Qin, J. Fluorine Chem. 132 (2011) 5.
[15] W. Ryba-Romanowski, G. Dominiak-Dzik, P. Solarz, Mol. Phys. 101 (2003)
1067.
[16] C. Li, Y. Guyot, C. Linares, R. Moncorge, M.F. Joubert, OSA Proc.Adv. Solid-State
Lasers 15 (1993) 91.
[17] M. Pollnau, W. Luthy, H.P. Weber, J. Appl. Phys. 77 (1995) 6128.
[18] S. Georgescu, O. Toma, A.M. Voiculescu, C. Matei, R. Birjega, L. Petrescu, Phys.
B 407 (2012) 1124.
[19] S. Georgescu, A.M. Voiculescu, C. Matei, O. Toma, L. Gheorghe, A. Achim, Rom.
Reports Phys. 62 (2010) 128.
[20] G.M. Kuzmicheva, I.A. Kaurova, V.B. Rybakov, S.S. Khasanov, A. Cousson,
131.
[21] S. Georgescu, A.M. Voiculescu, O. Toma, S. Nastase, C. Matei, M. Osiac, J. Alloys
Compd. 507 (2010) 470.
[22] R.A. Hewes, J.F. Sarver, Phys. Rev. 182 (1969) 427.
[23] M. Pollnau, D.R. Gamelin, S.R. Luthi, H.U.Gudel, Phys. Rev. B 61 (2000) 3337.
[24] G. Chen, G. Somesfalean, Y. Liu, Z. Zhang, Q. Sun, F. Wang, Phys. Rev. B
75 (2007) 195204.
[25] S.A. Wade, S.F. Collins, G.W. Baxter, J. Appl. Phys. 94 (2003) 4743.
[26] R. Yan, Y. Li, Adv. Funct. Mater. 15 (2005) 763.
[27] Y. Hu, F.W. Wang, H.L. Lin, Mater. Chem. Phys. 107 (2008) 82.
[28] B.R. Judd, Phys. Rev. 127 (1962) 750.
[29] G.S. Ofelt, J. Chem. Phys. 37 (1962) 511.
[30] R.D. Peacock, Struct. Bonding 22 (1975) 83.
[31] J. Stade, L. Bohaty, M. Hengst, R.B. Heimann, Cryst. Res. Technol. 37 (2002)
1113.