Luminescence dynamics in Tb(3+)-doped CaWO(4) and CaMoO(4) crystals.

Abstract

Single crystals of CaWO(4) and CaMoO(4) doped with Tb(3+) have been grown by the flux growth method. Their luminescence properties have been investigated in the 10-600 K temperature range under different experimental conditions. In spite of very similar spectra at low temperature upon excitation at 365 nm, the crystals show a very different behavior as the temperature is raised or the excitation wavelength is changed. These differences have been accounted for on the basis of models that take into consideration the position of the energy levels of the rare earth relative to the bandgap of the host material.

Full text

Page 1

1
Luminescence Dynamics in Tb3þ-Doped CaWO4and CaMoO4Crystals
2
Enrico Cavalli,*,†Philippe Boutinaud,‡Rachid Mahiou,§Marco Bettinelli,^and Pieter Dorenbos||
3
†Dipartimento di Chimica Generale ed Inorganica, Chimica Analitica, Chimica Fisica, Universit? a di Parma,
Parma, Italy,‡Laboratoire des Mat? eriaux Inorganiques, Clermont Universit? e, ENSCCF, BP10448, F-63000
Clermont-Ferrand,France,§CNRS,UMR6002,F-63177Clermont-Ferrand,France,^LaboratoryofSolidState
Chemistry, DB, Universit? a di Verona, and INSTM, UdR Verona, Verona, Italy, and
Sciences, Delft University of Technology, Delft, The Netherlands
4
5
6
||Faculty of Applied
7
8
Received December 9, 2009
9
Single crystals of CaWO4and CaMoO4doped with Tb3þhave been grown by the flux growth method. Their lumine-
scencepropertieshavebeeninvestigatedinthe10-600Ktemperaturerangeunderdifferentexperimentalconditions.
Inspiteofverysimilarspectraatlowtemperatureuponexcitationat365nm,thecrystalsshowaverydifferentbehavior
as the temperature is raised or the excitation wavelength is changed. These differences have been accounted for on
thebasisofmodelsthattakeintoconsiderationthepositionoftheenergylevelsoftherareearthrelativetothebandgap
of the host material.
10
11
12
13
14
15
1. Introduction
There is an increasing interest in Tb3þ-doped tungstates
and molybdates as new active media for phosphor
technology.1-3In this perspective more effort is presently
being devoted to the development of suitable synthetic
strategies4,5than to the analysis of the processes governing
the luminescence dynamics. Hence, the optimization of the
phosphor performances is usually achieved by means of
comparative tests that, even if effective from a practical
viewpoint, do not always provide significant information
on the luminescence dynamics. We think, however, that this
aspect is important and deserves a thorough investigation.6
ThepresentpaperisfocusedonCaWO4:Tb3þandCaMoO4:
Tb3þ, two materials already known for their interesting
luminescence properties.4,7,8We have grown single crystals
of CaWO4and CaMoO4doped with Tb3þand measured
their luminescence spectra and emission decay profiles in
different experimental conditions at temperatures ranging
from 10 to 600 K. The experimental results evidence signifi-
cant differences in the spectroscopic behavior of the two
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
compounds. The data have been analyzed in the framework
of models accounting for the effects of the host lattice on
the excited-state dynamics of rare earth dopants such as Pr3þ
or Tb3þ.
36
37
38
39
2. Experimental Section
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
2.1. Crystal Growth and Properties. CaWO4 (CWO) and
CaMoO4 (CMO) single crystals nominally doped with 0.5
mol%Tb3þweregrownbythefluxgrowthmethodusingNa2WO4
or Na2MoO4as a solvent in the 1350-600 ?C temperature
range.9Analytical grade CaO (98%, Carlo Erba), Na2CO3
(99%, Aldrich), WO3 (99%, Aldrich) or MoO3 (99.5%,
Aldrich), and Tb4O7(99.9%, Aldrich) were used as starting
materials. The crystals have a tetragonal scheelite (CaWO4)
structure with space group I41/a and unit cell parameters a =
5.243A˚andc=11.376A˚forCWO10anda=5.226A˚andc=
11.430A˚forCMO.11TheTb3þionsoccupytheCa2þsiteswith
8-fold oxygen coordination (distorted dodecahedron, actual
pointgroupS4).Thedifferenceinchargeiscompensatedbythe
accommodationofNaþions,presentinthegrowthmixture,or
by the formation of cationic vacancies. Whatever the case,
it implies perturbations of the crystal field around the active
ions and then some inhomogenous broadening of the spectral
features. The electronic structures of the host matrices have
been extensively investigated12-15and are still the subject of
*Corresponding author. E-mail: enrico.cavalli@unipr.it.
(1) Liao, J.; Qiu, B.; Wen, H.; You, W. Opt. Mater. 2009, 31, 1513–1516.
(2) Liao,J.;Qiu,B.;Wen,H.;Chen,J.;You,W.Mater.Res.Bull.2009,4,
1863–1866.
(3) Wang,Z.;Liang,H.;Wang,Q.;Luo,L.;Gong,M.Mater.Sci.Eng.,B
2009, 164, 120–123.
(4) Li, G.; Wang, Z.; Quan, Z.; Li, C.; Lin, J. Cryst. Growth Des. 2007, 7,
1797–1802.
(5) Hou, Z.;Li, C.;Yang, J.; Lian, H.; Yang, P.; Chai, R.; Cheng, Z.;Lin,
J. J. Mater. Chem. 2009, 19, 2737–2746.
(6) Cavalli, E.; Boutinaud, P.; Cucchietti, T.; Bettinelli, M. Opt. Mater.
2009, 31, 470–473.
(7) Van Uitert, L. G.; Soden, R. R. J. Chem. Phys. 1960, 32, 1161–1164.
(8) Zhang, Z.-J.; Chen, H.-H.; Yang, X.-X.; Zhao, J.-T. Mater. Sci. Eng.,
B 2007, 145, 34–40.
(9) Cavalli, E.; Bovero, E.; Belletti, A. J. Phys. Condens. Matter 2002, 14,
5221–5228.
(10) Zalkin, A.; Templeton, D. H. J. Chem. Phys. 1964, 40, 501–504.
(11) G€ urmen, E.; Daniels, E.; King, J. S. J. Chem. Phys. 1971, 55, 1093–
1097.
(12) Zhang, Y.; Holzwarth, N. A. W.; Williams, R. T. Phys. Rev. B 1998,
57, 12738–12750.
(13) Fujita, M.; Itoh, M.; Takagi, S.; Shimizu, T.; Fujita, N. Phys. Status
Solidi B 2006, 243, 1898–1907.
(14) Groenink, J. A.; Hakfoort, C.; Blasse, G. Phys. Status Solidi A 1979,
54, 329–336.
Inorganic Chemistry | 3b2 | ver.9 | 10/5/010 | 12:47 | Msc: ic-2009-02445c | TEID: deb00 | BATID: 00000 | Pages: 5.99
Inorg. Chem. XXXX, XXX, 000–000
DOI: 10.1021/ic902445c
A
pubs.acs.org/IC
rXXXX American Chemical Society

Page 2

59
60
61
62
63
discussion. On the basis of the most recent developments, we
located the energy values of the fundamental excitation (FE)
(corresponding to the lowest excitation peak) at 4.2. eV in the
CMO and 5.2 in the CWO case, and those of the conduction
bands at 4.5 and 5.6 eV, respectively.
2.2. Spectroscopic Measurements. The 10 K emission spectra
were recorded using a spectroscopic system consisting of a
450 W Xe lamp fitted with a 0.22 Spex Minimate monochro-
matorassourceanda1.26mSpexmonochromatorwithaRCA
C31034 photomultiplier to analyze and detect the output radia-
tion. The crystals were mounted onto the coldfinger of a
He-cryocooler (Air Products Displex DE-202). The 300-600 K
spectra were measured using a Triax 550 monochromator
equipped with a nitrogen-cooled CCD camera and a R928
Hamamatsu photomultiplier (Jobin-Yvon Symphony system).
The excitation light was selected from a xenon lamp using a
Triax 180 monochromator. The samples were mounted on a
homemade copper holder heated by a thermocoax wire con-
nected to a Thermolyne regulator. The decay profiles were
measured in the 10-300 K temperature range upon 355 nm
laser excitation using a pulsed Nd:YAG laser (Quanta System
model SYL 202); the emission was isolated by means of a
Hilger-Watts model D330 double monochromator and de-
tected with a Hamamatsu R943-022 photomultiplier connected
to a LeCroy WS422 transient digitizer.
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
3. Emission Measurements
3.1. CWO:Tb3þ.Theenergylevelcompositionandthe
low-temperature spectra of CWO:Tb3þhave already
85
86
87
been thoroughly investigated in the past.16-18The lumi-
nescence spectra measured at 77 and 300 K (Figure 1
upon360nmexcitation(incorrespondencewiththeTb3þ
absorption) are in agreement with previous literature.
Theyhavesimilarintensityandarecomposedofboth5D3
and5D4emission features having, as expected, a signifi-
cant broadness (full width at half-maximum, fwhm, on
the order of 75-80 cm-1at low temperature) as a con-
sequence of the doping mechanisms. The wavelength
dependence of the emission is shown in Figure 1b. It
can be noted that excitation at wavelengths shorter than
320 nm (see the excitation spectrum reported in the inset
of the figure) induces only a weak
overlapping the broad tungstate emission and a more
intense5D4emission. The overlap integral between tung-
state emission and Tb3þabsorption is nonzero for both
5D3(ground-stateabsorptionat382nm)and5D4(487nm)
states. The fact that the energy transfer process results in
thepreferentialpopulationofthe5D4levelallowsinferring
that another mechanism is involved in the depletion of the
5D3level.Inordertoexplorethisaspect,wehavemeasured
the temperature evolution of the emission in the 300-
600 K range upon excitation at 360 nm. The results,
reported in Figure 2, show that the intensity of the5D4
emission slightly increases with the temperature, whereas
thatofthe5D3rapidlydecreasesabove400K,asevidenced
in the inset of the figure.
The decay profiles of both5D3and5D4emissions have
beenmeasuredinthe10-298Krangeupon355nmpulsed
laser excitation. The5D3curves are reported in Figure 3
They can be reproduced by the Inokuti-Hirayama model
for energy transfer in the absence of migration:19
88F1
a)
89
90
91
92
93
94
95
96
97
98
99
5D3luminescence
100
101
102
103
104
105
106
107
108
109
110F2
111
112
113
114
115
116F3
a.
117
118
φðtÞ ¼ A exp -t
τ-R
t
τ
? ?3=s
"#
ð1Þ
Figure 1. (a)77and298KemissionspectraofCaWO4:Tb3þ.λexc.:360nm.
(b) Room-temperature emission of CaWO4:Tb3þmeasured upon different
excitation wavelengths. The excitation spectrum is shown in the inset.
Figure 2. Emission spectra of CaWO4:Tb3þ, measured in the 300-600
K temperature range. The temperature behaviors of the integrated
intensities of the 440 and 550 nm bands are reported in the inset. The
redlinerepresentsthe fit bymeansoftheStruckand Fongermodel(eq4)
of the5D3intensity profile.
(15) Treadway, M. J.; Powell, R. C. J. Chem. Phys. 1974, 61, 4003–4011.
(16) Wortman, D. E. Phys. Rev. 1968, 175, 488–498.
(17) Leavitt,R.P.;Morrison,C.A.;Wortman,D.E.J.Chem.Phys.1974,
61, 1250–1251.
(18) Page, A. G.; Godbole, S. V.; Sastry, M. D. J. Phys. Chem. Solids
1989, 50, 571–575.
(19) Inokuti, M.; Hirayama, F. J. Chem. Phys. 1965, 43, 1978–1989.
B
Inorganic Chemistry, Vol. XXX, No. XX, XXXX
Cavalli et al.

Page 3

119
where φ(t) is the emission intensity after pulsed excitation,
Aistheintensityoftheemissionatt=0,τisthelifetimeof
the isolated donor, R is a parameter containing the energy
transfer probability, and s = 6 for dipole-dipole (D-D), 8
for dipole-quadrupole (D-Q), and 10 for quadrupole-
quadrupole (Q-Q) interaction. We have fitted the decay
curves by means of eq 1 by considering a D-D process and
A, τ, and R as adjustable parameters. The τ value obtained
from the fit is 425 μs at 10 K and gradually decreases to
360-370 μs as the temperature increases up to 100 K; then
it remains constant (see inset of Figure 3a). This small
change canbe tentatively accountedfor by the variation of
the thermal population of the Stark levels of the emitting
5D3state. The parameter R provides information on the
probability of the energy transfer process:
120
121
122
123
124
125
126
127
128
129
130
131
132
133
R ¼4
3πΓ 1-3
s
??
NaR03
ð2Þ
134
whereΓisthegammafunction,Natheconcentrationofthe
acceptor expressed in ions3cm-3, and R0is the critical
distance. The value of R deduced from the 10 K decay
profile is 0.85, and that of the critical distance is about
12 A˚, compatible with the average distance between the
active ions (15 A˚) evaluated by considering a statistical
135
136
137
138
139
140
distributionofthedopantsinsidethehostlattice.Thisvalue
istypicalforcross-relaxationprocessesinvolvingTb3þions
insolids.20-22The5D4decaycurvesareshowninFigure3b.
They can be fitted to a difference of two exponentials,
allowing the evaluation of both the rise time and the decay
timeoftheemittinglevel.Therisingcomponentvariesfrom
140to180μs.Thesevaluesaresignificantlyshorterthanthe
5D3decay times of the isolated donor deduced from the
Inokuti-Hirayama model. However they are consistent
with the 1/e decay times calculated from the profiles of
Figure3a,indicatingthatfeedingoccursfromthe5D3state.
The5D4decay times range from 615 to 550 μs (see inset of
Figure 3b), indicating that probably the5D4emission is
scarcely affected by nonradiative processes.
3.2. CMO:Tb3þ.WehavesummarizedinFigure4
results of emission measurements in the 10-298 K range.
The 10Kspectrumexcitedat365nmisverysimilartothe
spectra of CWO:Tb3þ(see Figure 1) and can be commen-
ted on accordingly. Nevertheless, its temperature depen-
dence is very different: a strong decrease of the lumine-
scence intensity occurs in the 10-100 K range (see inset
141
142
143
144
145
146
147
148
149
150
151
152
153
154F4
athe
155
156
157
158
159
160
Figure 3. (a) Decay profiles of the
measured as a function of the temperature upon 355 nm excitation
(emission wavelength: 410 nm). In the inset the temperature dependence
of the decay times has been reported. (b) Decay profiles of the
emission of CaWO4:Tb3þ, measured as a function of the temperature
upon 355 nm excitation (emission wavelength: 540 nm). In the inset the
temperature dependence of both decay and rise times has been reported.
5D3 emission of CaWO4:Tb3þ,
5D4
Figure 4. (a) Low-temperature emission spectra of CaMoO4:Tb3þ.
Inset: temperature evolution of the5D3f7F4and5D4f7F5manifolds in
the10-105Krange.(b)Room-temperatureemissionspectraofCaMoO4:
Tb3þ, measured upon different excitation wavelengths. Excitation spec-
trum is shown in the inset.
(20) Park, J. Y.; Jung, H. C.; Rama Raju, G. S.; Moon, B. K.; Kim, J. H.
J. Lumin. 2010, 130, 478–482.
(21) Bodenschatz, N.; Wannemacher, R.; Heber, J.; Mateika, D. J.
Lumin. 1991, 47, 159–167.
(22) Tachihante, M.; Zambon, D.; Arbus, A.; Zahir, M.; Sadel, A.;
Cousseins, J. C. Mater. Res. Bull. 1993, 28, 605–613.
Article
Inorganic Chemistry, Vol. XXX, No. XX, XXXX C

Page 4

161
of Figure 4a for some representative features), and at
300 K practically only the5D4emission is present. This
behavior cannot be ascribed to multiphonon relaxation
(MPR) for at least two reasons: first because it has not
been observed in the case discussed above despite the fact
that calcium tungstate and molybdate have very similar
vibrational properties23and, second, because the gap
between the5D3and the next lower lying5D4level, on
the order of 5500 cm-1, requires more than six high-
energy phonons (about 880 cm-1) to be bridged. We
therefore conclude on the existence of a host-related
quenching channel for the5D3emission in CMO:Tb3þ.
In the 10 K spectrum measured upon excitation at
300 nm, i.e., in correspondence with the host absorption,
the5D4luminescence overlaps the orange broad band
molybdateemissionandnofeaturesoriginatingfrom5D3
are present. Since there is no overlap between the host
emission and the5D3absorption band (at about 380 nm),
we assume that the energy transfer from the host popu-
lates only the5D4level. The room-temperature emission
spectra measured for different excitation wavelengths are
shown in Figure 4b. The
absent, except for a very weak feature being present only
in the 375 nm excited spectrum. For excitation wave-
lengths shorter than 375 nm a weak host emission is still
observable.Theexcitationspectrum,shownintheinsetof
Figure4b,presentstheTb3þ4f-4ffeatures(350-390nm),
the host-related broad band (below 320 nm) and a shoul-
der in the 330 nm region. The origin of this band will be
discussedinthefollowing.Themeasurementsinthehigh-
temperature regime (Figure 5
quenching of the5D4emission in the 400-600 K range.
The integrated intensity of the
reported in the inset of the figure as a function of the
temperature.
The decay profiles of both5D3and5D4emissions have
been measured as a function of the temperature. The5D3
curves are reported in Figure 6
can be reproduced by means of eq 1 for a D-D process,
yieldingτvaluesinthe270-280μsrange.Thevalueofthe
R parameter at 10 K is 1.32, and that of the critical
distance obtained from eq 2 is about 14 A˚, close to the
average distance between the doping ions (15 A˚) evalu-
ated by statistical considerations. These values are con-
sistent with those obtained for CWO:Tb3þ, indicating
that in the low-temperature limit the emission dyna-
mics of the two systems are similar and affected by the
energytransferprocess.Asthetemperatureexceeds70K,
the decays become faster and faster and the Inokuti-
Hirayama fit becomes worse. The temperature depen-
denceoftheevaluateddecaytimesisplottedintheinsetof
Figure 6a. The decay profiles of the5D4luminescence are
plottedinFigure6b.Intheirlongtimetailtheyarealways
single exponential with decay times ranging from 460 to
520 μs, indicating that this emission is not significantly
affected by nonradiative processes up to room tempera-
ture. The short time region shows a more complex
behavior: up to 70 K the curves are nearly coincident
and present a rising component with a time constant on
theorderof70-75μs.Increasingfurtherthetemperature
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
5D3luminescence is nearly
183
184
185
186
187
188
189
190
191F5
) evidence a total thermal
192
193
5D4f7F5transition is
194
195
196
197
198F6
a. Up to 70 K these curves
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
leads to a progressive shortening of the rise time, which
fallstozeroatabout200K.Thisallowstheobservationof
a short-lived component ascribed to the molybdate emis-
sion (the 355 nm excitation being in correspondence with
a residual host absorption). This component is certainly
present also in the low-temperature curves and affects
their initial profiles. This accounts for the discrepancy
222
223
224
225
226
227
Figure 5. Emission spectra of CaMoO4:Tb3þ, measured in the 300-
600 K temperature range. The temperature behavior of the integrated
intensityofthe540nmbandisreportedintheinset.Theredlinerepresents
the fit by means of the Struck and Fonger model (eq 4).
Figure 6. (a) Decay profiles of the
measured as a function of the temperature upon 355 nm excitation
(emission wavelength: 410 nm). In the inset the temperature dependence
of the decay times is reported. (b) Decay profiles of the5D4emission of
CaMoO4:Tb3þ, measured as a function of the temperature upon 355 nm
excitation (emission wavelength: 540 nm). In the inset the temperature
dependence of both decay and rise times is reported.
5D3emission of CaMoO4:Tb3þ,
(23) Basiev, T. T.; Sobol, A. A.; Voronko, Y. K.; Zverev, P. G. Opt.
Mater. 2000, 15, 205–216.
D
Inorganic Chemistry, Vol. XXX, No. XX, XXXX
Cavalli et al.

Page 5

228
observed between the low-temperature values of the rise
time and of the5D3decay time. In contrast to the case of
CWO:Tb3þ, the particular temperature dependence of
the decay curves of CMO:Tb3þin the short time range
clearly demonstrates a change in the relaxation mecha-
nism of the5D3level between 70 and 200 K. This point
will be discussed in the next section.
229
230
231
232
233
234
235
3. Discussion
CWO:Tb3þand CMO:Tb3þcrystals exhibit very similar
emission spectra at low temperature, upon Tb3þexcitation.
This indicates that thecoordination symmetriesand thenthe
crystalfields(CF) aroundtheactiveionsarenearlyidentical,
as expected on the basis of their structural properties. How-
ever,thespectrameasureduponhostexcitationaswellasthe
temperature behavior of both emitted intensities and decay
profilesarestronglyhostdependentandcannotbeaccounted
for by considering only the properties of the isolated centers.
In the past 10 years Dorenbos has been collecting data and
models specifically aimed at constructing and predicting the
location of lanthanide impurity electronic levels relative to
those of the host compound; see, for example, ref 24. The
parameter set required to construct level schemes is still
occasionally being refined and improved.25The most recent
parameter set can be found in ref 26, where also schemes
arereportedontherelativepositionoftheenergylevelsofthe
rare earth trivalent (and divalent) ions with respect to the
bandgapofanumberofhostlattices.Wethinkthatthiscould
be asuitablestarting model inorderto interprettheemission
properties of the Tb3þ-doped crystal. This scheme can be
compiled on the basis of the following considerations. In the
presence of oxidizing ions such as d0closed-shell transition
metals(Mnþ=Ti4þ,V5þ,Nb5þ,Ta5þ,Mo6þ,W6þ)itispos-
sible to observe the photoinduced redox process
Tb3þþMnþf Tb4þþMðn-1Þþ
concomitant with the formation of an intervalence charge
transfer (IVCT) state that actively participates in the fluore-
scencedynamicsofthesystem.Thisprocessusuallygivesrise
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
to a band in the excitation spectrum.27In the investigated
materialshowever,theassignmentoftheIVCTbandsisnota
simple task, since the excitation spectra of the host emissions
are complicated by the presence of features ascribed to
defectivecenters.14,15It is, however, possibleto estimatewith
good approximation the energy of an IVCT transition by
means of the empirical equation proposed by Boutinaud
et al.:28
265
266
267
268
269
270
271
IVCTðLn3þ,cm-1Þ ¼ 58800-49800
χoptðMnþÞ
dminðLn3þ-MnþÞ
!
ð3Þ
272
where χopt(Mnþ) is the optical electronegativity of the d0
transition metal ion and dmin(Ln3þ-Mnþ) its shortest dis-
tancefrom thelanthanide (Pr3þorTb3þ) ion.Thecalculated
values, about 330 nm for CMO and 310 nm for CWO, are in
correspondence with shoulders observed on the low-energy
sideoftheexperimentalexcitationbands(insetsofFigures1b
and 4b) and can be considered reasonable. The IVCT can be
considered an electronic transition from the ground state of
therareearthtothetransitionmetalconstitutingthelattice,29
and its energy provides information on the location of the
Tb3þlevels relative to the host bands. The method to con-
struct level schemes proposed by Dorenbos26allows then
extending this information to all trivalent lanthanide ions.
The obtained schemes, shown in Figure 7
stateenergies(withtheexceptionofPr3þandTb3þ,forwhich
also the excited levels are shown), are of general utility. The
procedure applied to construct these schemes is extensively
describedinref26.HereEFEisreferredtoastheenergyofthe
fundamentalexcitation, i.e.,themaximumofthehostexcita-
tion band. We proposed in ref 26 that the bottom of the
conduction band (CB) is located at about 1.08 times EFE
abovethetopofthevalenceband(VB)andthattheenergyof
the electron after IVCT is at 1.04 times EFEabove the top of
theVB.Allenergylevelsrepresentedintheschemesshouldbe
viewedaselectron-donatinglevels.Theenergyonthevertical
axis is then the difference between the energy required to
remove an electron from the level involved and the energy
necessary to remove an electron from the top of the valence
273
274
275
276
277
278
279
280
281
282
283
284
285F7
for the ground-
286
287
288
289
290
291
292
293
294
295
296
297
298
299
Figure 7. Position of the ground-state energy along the Ln3þseries in CaWO4(a) and CaMoO4(b).
(24) Dorenbos, P. J. Phys.: Condens. Matter 2000, 15, 8417–8434.
(25) Dorenbos, P. J. Alloys Compd. 2009, 488, 568–573.
(26) Dorenbos, P.; Krumpel, A. H.; van der Kolk, E.; Boutinaud, P.;
Bettinelli, M.; Cavalli, E. Opt. Mater., doi:10.1016/j.optmat.2010.02.021.
(27) Boutinaud, P.; Putaj, P.; Mahiou, R.; Cavalli, E.; Speghini, A.;
Bettinelli, M. Spectrosc. Lett. 2007, 40, 209–220.
(28) Boutinaud, P.; Cavalli, E.; Bettinelli, M. J. Phys.: Condens. Matter,
2007, 19, 386230 (1-11).
(29) Krumpel,A.;vanderKolk,E.;Dorenbos,P.;Boutinaud,P.;Cavalli,
E.; Bettinelli, M. Mater. Sci. Eng., B 2008, 147, 114–120.
Article
Inorganic Chemistry, Vol. XXX, No. XX, XXXX E

Page 6

300
band. The two models can be improved on the basis of the
present experimental results relative to the Tb3þion. The
temperature behavior of the5D3emission of CWO:Tb3þ
(Figure 2) can be analyzed by means of the Struck and
Fonger model30for a thermally induced crossover to a
Franck-Condon shifted state:
?
whereAiscloseto107andEistheactivationenergyfromthe
4fnstate to its crossover with the quenching state (here the
IVCT one). The fit of the profile reported in the inset of
Figure 2 provides a value of about 5500 cm-1for the
activation energy. This energy has been used to compile the
diagram shown in Figure 8 a. This scheme allows under-
standing the excitation wavelength dependence of the emis-
sion reported in Figure 1b: excitation in the 300-250 nm
range results in fact in the population of the IVCT state that
preferentially relaxes to the5D4, almost bypassing the5D3
state. In the case of CMO:Tb3þthere is no high-temperature
5D3emission,whereasthetemperaturedependenceofthe5D4
level emission in the 400-600 K range is reasonably repro-
ducedusing eq4,yielding a value of about 5700 cm-1forthe
activationenergy.Onthisbasiswehavecompiledthescheme
of Figure 8b, which allows understanding the tempera-
ture behavior observed in Figure 5. The relative position of
the5D3and IVCT potential curves is in fact compatible with
thermalization occurring even at low temperature, resulting
in the concomitant5D3depopulation and5D4population
through the IVCT state.
Let us now take into consideration the decay curve
measurements. The5D3and5D4decay profiles of CWO:
Tb3þevidence an energy transfer process in which the5D3
level decays nonradiatively to5D4. In agreement with pre-
viousliterature,31weconcludethatthisprocessconsistsofthe
cross-relaxation mechanism:
301
302
303
304
305
IðTÞ
I0
¼
1þA
E
kT
???-1
ð4Þ
306
307
308
309
310
311 F8
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
5D3,7F0f5D4,7F6
333
representedbymeansofdottedarrowsinFigure8a.Theweak
temperaturedependenceoftheresultingdecaytimesaccounts
for the nearly (but not completely) resonant character of the
334
335
336
process, the involved energy gaps being about 5800 cm-1
(5D3-5D4) and5750cm-1(7F0-7F6),asreportedforTb3þin
LaF3.32By considering the thermal behavior of the
emission, which is practically constant up to 400 K (see inset
of Figure 2), we can conclude that in the 150-400 K
temperature range this temperature-independent quenching
channel is prevalent, at the present dopant concentration,
upon other quenching processesoccurring through the IVCT
level. For CMO:Tb3þthe situation is rather different: in the
low-temperature regime the energy transfer process takes
place as in the previous case. For temperatures above 70-
80K;however,theshapesofboth5D3and5D4profileschange
significantly, and this behavior cannot be accounted for by
invoking phonon-mediated effects. According to Figure 8b,
we are in the presence of a progressive change in the5D3
depleting mechanism involving the thermal population of the
IVCTstateanditsradiationlessdecaybycrossovertothe5D4
level. This mechanism is consistent with the progressive
shortening of the5D3decay times and of the related rise
times in the emission profile of5D4. In this case the crossover
to the IVCT state appears to be more important than cross-
relaxation processes.
337
338
5D3
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
4. Conclusions
359
The optical spectra of CaWO4 and CaMoO4 crystals
doped with Tb3þ, measured in different experimental condi-
tions,evidencesignificantdifferencesthatcannotbeascribed
to the isolated centers. The observed phenomena have been
accounted for in the framework of a unitary model that
involves the formation of an IVCT state participating in the
excited-state dynamics of the systems. This has implied the
compilation of a “host lattice plus rare earth dopant” energy
levelschemethathasbeendemonstratedtobeeffectiveinthe
analysisoftheexperimentalobservations.Thisapproachwill
be soon extended to other systems in order to confirm its
validity.
360
361
362
363
364
365
366
367
368
369
370
371
Acknowledgment. The Italian authors acknowledge the
MIUR (ItalianMinistryfor the University andthe Scien-
tific Research) for the financial support in the course of
the project PRIN 2007.
372
373
374
Figure 8. EnergylevelschemeandfluorescencedynamicsinCaWO4:Tb3þ(a)andCaMoO4:Tb3þ(b).Thesolidlinesrepresenttheabsorption(excitation)
and radiative emission processes; the dotted lines describe the nonradiative processes.
(30) Struck, C. W.; Fonger, W. H. J. Appl. Phys. 1971, 42, 4515–4516.
(31) May, P. S.; Sommer, K. D. J. Phys. Chem. A 1997, 101, 9571–9577.
(32) Carnall, W. T.; Crosswhite, H.; Crosswhite, H. M. Energy level
structure and transition probabilities of the trivalent lanthanides in LaF3;
Argonne National Laboratories: Argonne, IL, 1977.
F
Inorganic Chemistry, Vol. XXX, No. XX, XXXX
Cavalli et al.