Optical temperature sensing in Er3+-Yb3+ codoped CaWO4 and the laser induced heating effect on the luminescence intensity saturation

Article abstract of DOI:10.1016/j.jallcom.2017.08.007

CaWO₄ phosphors doping with different Er³⁺ and Yb³⁺ ion concentrations were synthetized through high temperature solid-state reaction technique. Excited by 980 nm laser, the influence of rare earth dopant concentration on

Full text of DOI:10.1016/j.jallcom.2017.08.007

Journal of Alloys and Compounds 726 (2017) 547e555
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: http://www.elsevier.com/locate/jalcom
Optical temperature sensing in Er3þ-Yb3þ codoped CaWO4 and the
laser induced heating effect on the luminescence intensity saturation
,
b
,
,
, d
Wei Xu a *, Ying Cui a, Yuwei Hu a, Longjiang Zheng a, Zhiguo Zhang b c, Wenwu Cao b
a School of Electrical Engineering, Yanshan University, Qinhuangdao 066004, China
b Condensed Matter Science and Technology Institute, Harbin Institute of Technology, Harbin 150001, China
c Laboratory of Sono- and Photo- Theranostic Technologies, Harbin Institute of Technology, Harbin 150001, China
d Materials Research Institute, The Pennsylvania State University, Pennsylvania 16802, USA
a r t i c l e i n f o
a b s t r a c t
CaWO4 phosphors doping with different Er3þ and Yb3þ ion concentrations were synthetized through
high temperature solid-state reaction technique. Excited by 980 nm laser, the influence of rare earth
dopant concentration on the thermometry behavior of Er3þ green upconversion emissions was discussed
via the fluorescence intensity ratio method. Superior temperature sensing ability could be achieved in
CaWO4: 0.1 mol% Er3þ-3 mol% Yb3þ, the maximum sensitivity of which is about 1.05 ꢀ 10ꢁ2 Kꢁ1 at 439 K.
Meanwhile, the pumping laser induced thermal effect on the saturation of upconversion emission in-
tensity has been preliminarily studied. The temperature of the irradiation spot on sample caused by the
exposure of excitation laser is observed to be greatly enhanced as the pumping power increased. Based
on the steady-state rate equations and the temperature dependence of luminescence decay curves,
theoretical calculations have been performed. The results confirm that the laser induced heating effect
plays an important role in the luminescence intensity saturation phenomenon.
Article history:
Received 4 May 2017
Received in revised form
31 July 2017
Accepted 1 August 2017
Available online 4 August 2017
Keywords:
Optical temperature sensing
Upconversion
Dopant concentration
Luminescence intensity saturation
© 2017 Published by Elsevier B.V.
1. Introduction
because the FIR from the thermally coupled levels (TCL) of RE ions is
independent of spectrum losses and fluctuations in the excitation
Temperature is one of the fundamental physical parameters and
the realization of accurate and reliable temperature measurement
is always a challenging task. Nowadays, much attention has been
paid to the non-contact temperature sensors based on the rare
earth (RE) ions activated luminescent materials, which present
efficient luminescence, high sensitivity, good accuracy, and easy
integration [1e3]. And the temperature could be measured through
analyzing the fluorescence intensity, band shape, lifetimes, etc.
Since the luminescence characteristics of the RE ions strongly
depend on the temperature [4e6]. This contactless and non-
invasive thermometry technique has little effect on the thermal
state of detected objects and exhibit wide applications in many
fields, such as chemistry, electromagnetism, and biomedicine,
where conventional sensors can not be used. As one of the sensing
methods, the fluorescence intensity ratio (FIR) technique is exten-
sively regarded as a promising tool for mapping temperature
intensity [6]. Thus, reasonable measurement accuracy and sensi-
tivity could be obtained with this method, and several RE ions have
already been explored for the FIR related temperature survey
[7e10]. Among these ions, Er3þ is the most well studied, which has
the TCL of 4S3/2 and 2H11/2 with proper energy separation as well as
the corresponding strong green upconversion (UC) photo-
luminescence under near-infrared (NIR) light excitation [11,12].
And owing to the large absorption cross-section around NIR region,
Yb3þ is usually utilized as the sensitizer to further enhance Er3þ UC
emissions. The NIR laser excited Er3þ-Yb3þ codoped systems have
been widely considered as promising candidates for surface tem-
perature detection and biological thermal probes with preferable
measurement accuracy and resolution [13e15]. On the other hand,
host material is also a critical element determining the sensing
performance because the ion's optical properties are ruled by the
coordination environment. Recently, tungstates have attracted
much interest due to their distinguished features, including good
thermal stability, high density, as well as low phonon threshold
energy, which enable the activators to emit stable and intense
luminescence even at high temperatures [16e18]. Meanwhile, the
* Corresponding author. School of Electrical Engineering, Yanshan University,
Qinhuangdao 066004, China.
E-mail address: college_d@163.com (W. Xu).
http://dx.doi.org/10.1016/j.jallcom.2017.08.007
0925-8388/© 2017 Published by Elsevier B.V.

548
W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
study on the thermal sensing behavior of Er3þ-Yb3þ codoped
tungstates is also increasing. Lu et al. pointed out that excellent
temperature sensing properties could be achieved based on the
stark sublevels of Er3þ-Yb3þ codoped Gd2(WO4)3 phosphor [19].
Chai et al. observed color-tunable UC emissions in Er3þ-Yb3þ
codoped ZnWO4 and a thermometry sensitivity of 0.0099 K-1 was
obtained in such compound [20]. Moreover, it was reported that the
UC luminescence efficiency of NaLa(WO4)2: Er3þ-Yb3þ is compa-
Yb3þ (x ¼ 0, 1, 3, 5) and those with x mol% Er3þ (x ¼ 0.1, 0.3, 0.5, 0.7)
- 3 mol% Yb3þ were prepared with the high temperature solid-state
method. The starting materials were CaCO3 (99.99%, Aladdin), WO3
(99.99%, Aladdin), Er2O3 (99.99%, Aladdin), and Yb2O3 (99.99%,
Aladdin). The raw powders with the designed stoichiometry were
fully mixed and grinded. Then the mixtures were put into the
corundum crucibles and sintered in a furnace for 10 h at 1300 ꢂC in
air. The obtained Er3þ-Yb3þ codoped CaWO4 phosphors were finally
pressed into a thin disk with a diameter of 1.3 cm. A powder
rable to that of b
-NaYF4: Er3þ-Yb3þ but with a much higher thermal
sensitivity than that of the fluoride [21]. Despite the significant
progress in luminescent thermometry, many applications define
additional requirement, especially the enhanced measurement
sensitivity and precision. As is well known, the doping levels of RE
ions have appreciable influence on their luminescence character-
istics, i.e., the temperature response properties of Er3þ green
emissions could be improved through designing the dopant con-
centration of Er3þ and Yb3þ ions. Thus, to further develop Er3þ
luminescence based temperature sensors, it is of technological and
scientific importance to comprehend the effect of RE content on the
thermometry behavior in Er3þ-Yb3þ codoped tungstate matrix,
which, to our best knowledge, is still lacking.
diffractometer (Rigaku DMAX2500) using Cu
Ka radiation
(
l
¼ 0.154 nm) was applied to confirm the crystal phase through X-
ray diffraction (XRD) method. The measurement angular is in the
range of 10ꢂ ꢃ 2
q
ꢃ 80ꢂ with a step size of 0.02ꢂ and speed of 12ꢂ/
min. To estimate the particle size and morphology of the crystalline
powder, Hitachi S4800 scanning electron microscopy (SEM) was
used. The samples are excited by a 980 nm continuous wave diode
laser, which is fitted on a laser diode (LD) mount and linked to a
Thorlabs ITC4005 LD/temperature controller. Zolix-SBP300 grating
spectrometer equipped with a CR131 photomultiplier is used to
record the luminescence spectra. The temperature of the samples is
controlled by the experimental system reported in our previous
work [18]. The luminescence decay curves are measured by 980 nm
laser modulated through square-wave electric current and recor-
ded by Tektronix DPO 5054 oscilloscope. To guarantee the validity
of the experimental results, the position for the optical devices
keeps unchanged during the whole experiment.
Additionally, the pumping power dependence of emission in-
tensities is known to be very essential for understanding the
excitation mechanisms in UC luminescence materials. And it is
usually assumed that the order n of the UC process, i.e. the number
of the pumping photons required to excite the emitting states, is
indicated by the slope of the luminescence intensity I vs. pumping
power P in double-logarithmic representation. In many works, it
has been observed that the dependence of the UC emission in-
tensity on pumping power decreases in slope with the strength of
excitation [22e27]. This phenomenon is considered as lumines-
cence intensity saturation (LIS). Some works have been performed
to interpret the mechanisms leading to such phenomenon, and
most believe the competition between the linear decay and the UC
process of the intermediate excited states results in the LIS [22e27].
Generally, it would be hard for the UC process to transcend the
linear decay when excited by a continuous wave laser with hun-
dreds of milliwatt. In fact, the laser energy absorbed by the RE ions
is not only used to generate the luminescence, but also introduce
heat into the sample. And the temperature for the irradiation spot
on the sample would rise as the laser power is strengthened. This
heating effect is more obvious for the phosphors because of the
poor thermal conductivity and the opacity [9,28]. The temperature
increment could subsequently cause the changing in the phonon-
involved behaviors, including the multi-phonon nonradiative
relaxation, phonon-assisted energy transfer, and phonon-assisted
energy absorbance. These processes will influence the population
of the excited levels of RE ions and might contribute to the UC LIS.
However, up to now, few works has taken the laser induced heating
(LH) effect into account when discussing the LIS for the upcon-
verted materials.
3. Results and discussion
3.1. Optical temperature sensing behavior
The XRD patterns for the pure CaWO4 and Er3þ-Yb3þ codoped
CaWO4 phosphors are shown in Fig. 1(a). To clearly exhibit the
changing in the XRD patterns, intense diffraction peaks in the range
of 15ꢂ ꢃ 2
q
ꢃ 64ꢂ were depicted. All of the diffraction peaks for the
as-prepared samples can be well indexed to the tetragonal struc-
ture of CaWO4 based on the PDF card # 77e2236, and no other
diffraction peaks were detected, indicating the dopants (Er3þ and
Yb3þ ions) are successfully incorporated into the host lattice and do
not cause significant changes to the crystal structure. Meanwhile,
stronger diffraction at 2q
values around 18.8ꢂ (101) and 28.6ꢂ (112)
obviously exhibits red-shift with the increasing of Er3þ and Yb3þ
content. The diffraction peak around 18.8ꢂ is shown in Fig. 1(b) to
clearly present the red-shift phenomenon. When RE ions are
incorporated into the lattice of CaWO4, they usually substitute Ca2þ
ions in the dodecahedral sites with the generation of cation va-
cancies for charge compensation. The substitution of Ca2þ ions
possessing larger radius (~1.12 Å) by Er3þ (~1.00 Å) and Yb3þ
(~0.98 Å) with smaller radii would result in the shrinkage of the
host lattice [29,30]. Through the SEM image (see Fig. S1 in sup-
plementary material), it can be seen that the phosphors consist of
aggregated and irregular particles, and size of distinguishable
particles is in the micron range. To understand the dopant effect of
Er3þ and Yb3þ on Er3þ green luminescence based thermometry
properties, the UC emission spectra are recorded at temperatures
ranging from 307 to 773 K. To avoid the laser thermal effect on the
below calibration, the excitation power of 980 nm LD is set as low as
50 mW. Figs. 2 and 3 present the temperature dependent green UC
emission spectra for 0.5 mol% Er3þ - x mol% Yb3þ (x ¼ 0,1, 3, 5) and x
mol% Er3þ (x ¼ 0.1, 0.3, 0.5, 0.7) - 3 mol% Yb3þ codoped samples.
Two dominant emission bands of Er3þ centered around 530 and
In this work, CaWO4 was selected as the studied object, and a
series of Er3þ-Yb3þ codoped CaWO4 phosphors were prepared. The
influence of Er3þ and Yb3þ doping concentration on Er3þ green
emissions based thermometry behavior was investigated when
excited by 980 nm laser. Meanwhile, the LH effect on the LIS phe-
nomenon was studied based on Er3þ-Yb3þ codoped CaWO4 phos-
phors. The results illustrate that enhancing Yb3þ or reducing Er3þ
dopant content can modify the luminescence thermometry ability
in Er3þ-Yb3þ codoped CaWO4. Besides, the LH effect is proven to
play an important role in the LIS phenomenon.
553 nm are attributed to the transitions of 2
H
11/2 / 4I15/2 and 4S3/
2. Experimental
/
4I15/2, respectively. And the corresponding FIR between 530
2
and 553 nm luminescence (IH/IS) at different temperatures are also
included in the figures. Obviously, the emission band positions are
The CaWO4 phosphors codoping with 0.5 mol% Er3þ - x mol%

W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
549
Fig. 1. XRD patterns for the CaWO4 phosphors and CaWO4 doping with different Er3þ and Yb3þ concentrations.
Fig. 2. Green UC emission spectra for CaWO4: 0.5 mol% Er3þ-x mol% Yb3þ (x ¼ 0, 1, 3, 5) as well as the FIR plotted as a function of temperature.
observed to scarcely change as temperature rises, but the FIR has
ꢀ
ꢁ
ꢀ
ꢁ
varied with the increase of temperature. This is mainly ascribed to
the fact that the relative population between the thermally coupled
levels follows Boltzmann type distribution. According to the theory
by Wade et al., the FIR of the emissions from TCL of Er3þ ions is
modified as [6].
I
FIR≡ H
IS
AHgHsHuH
ASgSsSuS
D
E
DE
kBT
¼
exp
ꢁ
þ B ¼ C exp
ꢁ
þ B;
kBT
(1)
where IH and IS are respectively the relative luminescence intensity

550
W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
Fig. 3. Green UC emission spectra for CaWO4: x mol% Er3þ-3 mol% Yb3þ (x ¼ 0.1, 0.3, 0.5, 0.7) as well as the FIR plotted as a function of temperature.
originated from the upper and lower TCL of Er3þ; A and g represent
the spontaneous transition rate and the degeneracy; is the
detection system response at the angular frequency E is the
energy gap between the TCL; C and B are constants, and kB and T
denote the Boltzmann constant and the absolutely temperature,
respectively. The fitted results in Figs. 2 and 3 show that the fittings
all match well with the experimental values and the fitted co-
Fig. 1(b), the increment doping concentration of Er3þ and Yb3þ
could arouse slight changing in the host lattice. And the minor
alteration in the local environment of activator would induce sig-
nificant variation in the spontaneous transition rates, consequently
modulating the AH/AS. From the results given in Figs. 2 and 3, it is
assumed that the value of AH/AS for Er3þ is enlarged with increasing
Yb3þ dopant but changed from the negative relationship with Er3þ
content in CaWO4. The changing of AH/AS with Er3þ and Yb3þ
dopants can be distinctly inferred from the room-temperature UC
emission spectra. At room temperature, the emission band
centered around 530 nm has been relatively weakened compared
with that of 553 nm when increasing Er3þ concentration, and the
opposite changing trend is observed by enhancing Yb3þ doping
content. Such changing style could be clearly exhibited by the
normalized ratios among the AH/AS values, which are estimated to
be about 1: 1.18: 1.56: 1.91 for 0.5 mol% Er3þ - x mol% Yb3þ (x ¼ 0, 1,
3, 5) codoped CaWO4 and 1: 0.89: 0.81: 0.76 for x mol% Er3þ - 3 mol
% Yb3þ (x ¼ 0.1, 0.3, 0.5, 0.7) ones. And it is worthwhile to note that
similar changing trend of AH/AS can also be found in Er3þ/Yb3þ
codoped glasses [34]. On the other hand, raising Yb3þ concentra-
tion would increase the absorption coefficient of samples around
1000 nm, which is in favor of enhancing Er3þ UC emissions and also
contributes to the increment of C value. And increasing Er3þ con-
tent most probably results in the concentration quenching caused
by cross-relaxation, which would also lead to the decrease in C
value. The variation in the C consequently provides different
measurement sensitivity. The maximum sensitivity is achieved in
0.1 mol% Er3þ- 3 mol% Yb3þ codoped CaWO4, which is about
1.05 ꢀ 10ꢁ2 Kꢁ1 at 439 K. Furthermore, the measurement error can
be estimated from the calculated sensitivity [35].
s
u; D
efficients
centration. While it is interesting to note that the pre-exponential
parameter clearly changes in different samples. For 0.5 mol% Er3þ
DE and B barely change with the variation of RE con-
-
x mol% Yb3þ codoped CaWO4, the value of C is gradually enhanced
from 10.13 to 15.93 as Yb3þ concentration is increased. In contrast,
the reduction of C from 16.94 to 11.83 is observed with the incre-
ment of Er3þ ions in x mol% Er3þ - 3mol% Yb3þ codoped ones.
Several testing cycles were performed, and a good reproducibility
of the thermometry technique was obtained (see Fig. S2 in the
supplementary material), which benefits from the excellent ther-
mal stability of the phosphors evidenced by the FIR and XRD
analysis (see Figs. S3 and S4 in the supplementary material).
As for the thermometry, the measurement sensitivity is an
important parameter reflecting the changing rate of FIR with
temperature, which is defined as [13].
ꢀ
ꢁ
dFIR AHgHsHuH
D
E
DE
kBT2
S≡
¼
exp
ꢁ
kBT
ꢀ
:
(2)
dT
ASgSsSuS
The sensitivities obtained from the samples are presented in
Fig. 4. At the same temperature, the sensitivity is evidently
enhanced with Yb3þ content, whereas decreased as Er3þ dopant is
raised. Through Eq. (2), it is known that the sensitivity is dominated
by A, g,
s
,
u
, and
DE at certain temperature. And the parameters g, s,
u
, and
DE should be identical at different samples. Nevertheless, the
D
FIRkBT2
FIRDE
D
FIR
S
spontaneous transition rates and the spectral features of the Er3þ
-
DT ¼
¼
:
(3)
Yb3þ codoped oxides are sensitive to the cationic environment and
ratio between the particular RE dopants [31e33]. And as shown in
When a signal division circuitry with the same precision is

W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
551
Fig. 4. The measurement sensitivities for Er3þ-Yb3þ codoped CaWO4 phosphors.
applied, the high sensitivity is in favor of getting small error.
Compared with other Er3þ activated thermometers given in Table 1,
excellent thermometry ability could be expected by using CaWO4:
0.1 mol% Er3þ-3 mol% Yb3þ phosphor. And it is interesting to
observe that the thermometry based on Er3þ-Yb3þ codoped tung-
saturation and produce thermal effect, influencing the thermom-
etry behavior [44,45].
3.2. Laser induced heating effect on the luminescence intensity
saturation
state possesses higher sensitivities. Especially for the MWO4: Er3þ
-
Yb3þ (M ¼ Ca2þ, Sr2þ Mg2þ, Zn2þ), as RE ions are incorporated into
the matrix lattice to substitute M ions, the crystal structure could be
efficiently distorted by such substitution due to the different ion
The optical thermometry based on the FIR is known to be an
effective tool to determine the pumping laser induced heating (LH)
effect on the sample [28,46]. In the following, the LH effect on the
LIS is to be studied based on CaWO4: Er3þ-Yb3þ since the LIS phe-
nomenon has been widely observed in Er3þ/Yb3þ codoped systems
[23e27]. And considering the complex UC mechanisms involved in
Er3þ-Yb3þ codoped systems, CaWO4 doping with small Er3þ and
Yb3þ contents (0.1 mol% Er3þ-1 mol% Yb3þ synthesized in Ref. [18])
is chosen as the studied object to simplify the subsequent discus-
sion. The UC emission spectra at different excitation powers are
recorded and the temperature of the laser spot on the sample is
estimated based on the FIR method, as shown in Fig. 5. To clearly
present the change of spectra profile with the excitation power, the
luminescence intensity has been normalized according to that of
553 nm emission. It can be seen that the laser spot temperature has
radius and valent mismatch. This behavior is particularly in favor of
2
the radiation from H11/2
/
4I15/2 hypersensitive transition and
leads to higher temperature measurement sensitivity [14].
Furthermore, through the changing tendency of AH/AS with Er3þ
and Yb3þ concentration, thermometry sensitivity and accuracy can
be further improved by modifying the dopants of Er3þ and Yb3þ in
CaWO4. Yet, careful optimization is necessary to balance the CR
quenching, signal to noise ratio, and sensitivity/accuracy issues.
Meanwhile, considering the strong dependence of multi-photon
excitation process on the pumping power, attention should also
be paid to the power adjusting during the temperature measure-
ment since overlarge excitation power might induce absorption
Table 1
Thermometry properties for the optical temperature sensors based on the green UC emissions of Er3þ and Er3þ-Yb3þ codoped compounds.
Materials
Temperature range (K)
Excitation wavelength (nm)
Maximum sensitivity (Kꢁ1
)
Reference
Na0.5Gd0.5MoO4:Er3þ-Yb3þ
Y2Ti2O7:Er3þ-Yb3þ
298e778
298e673
303e600
295e973
293e720
300e510
83e583
139e468
298e513
290e490
80e480
307e473
163e583
95e775
93e773
133e773
303e773
980
980
980
980
488
980
980
980
980
980
980
980
980
980
980
980
980
0.0086
0.0067
0.0091
0.0048
0.0066
0.0045
0.0099
0.0115
0.0059
0.0058
0.0053
0.0085
0.0093
0.0128
0.0128
0.0112
0.0105
[8]
[11]
[12]
[13]
[14]
[15]
[20]
[21]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
This Work
La2O3:Er3þ-Yb3þ
Yb2Ti2O7:Er3þ-Mo6þ
Oxyfluoride glass:Er3þ
NaYTiO4:Er3þ-Yb3þ
ZnWO4: Er3þ-Yb3þ
NaLa(WO4)2: Er3þ-Yb3þ
CaLa2ZnO5:Er3þ-Yb3þ
Na2Gd2Ti3O10: Er3þ-Yb3þ
Na0.5Bi0.5TiO3:Er3þ
GdVO4:Er3þ-Yb3þ
MgWO4:Er3þ-Yb3þ
SrWO4: Er3þ-Yb3þ
SrMoO4:Er3þ-Yb3þ
NaY(WO4)2:Er3þ-Yb3þ
CaWO4:Er3þ-Yb3þ

552
W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
Fig. 5. (a) Normalized spectra of the green UC emissions at different excitation powers and (b) the calculated local temperature for Er3þ-Yb3þ codoped CaWO4 under 980 nm laser
excitation.
been significantly enhanced as the power increases. To further
understand the LH influence on the population of the excited states,
the steady-state rate equations need to be listed.
It is known that under 980 nm laser excitation, the excited levels
of Er3þ can be populated both by energy transfer (ET) from Yb3þ to
Er3þ and its own ground state absorption (GSA) as well as the
excited state absorption (ESA). But the population through GSA and
ESA could be neglected compared with that through the ET pro-
cesses (see Fig. S5 in the supplementary material). And considering
the small doping concentration of Er3þ, the cross-relaxations be-
tween Er3þ are also neglected. Then, the rate equations describing
the excitation mechanisms can be written as follows
dN1
dt
N1
t1
¼ C0NbN0 ꢁ
¼ C1NbN1 ꢁ
ꢁ C1NbN1
(4)
(5)
dN2
dt
N2
t2
ꢁ
GN2
Fig. 6. The energy level diagrams for Yb3þ and Er3þ as well as the UC mechanism
(TP ¼ Thermal population).
dN3
dt
¼
G
N2 ꢁ N3=t3
(6)
(7)
4
temperature, it is rational to deem that ions on the F7/2 state de-
dNb εNalPP
dt
Nb
¼
ꢁ C0NbN0 ꢁ C1NbN1 ꢁ
tYb
4
cays simply nonradiatively to the S3/2 state [47]. So it can be
hcs
4
considered that the population of S3/2 is directly originated from
the ET process: Yb3þ: 2F5/2 þ Er3þ: 4I11/2/Yb3þ: 2F7/2 þ Er3þ: 4S3/2
where N0, N1, N2, N3, Na, Nb are the population densities on Er3þ
4I15/2 4I11/2 4S3/2 2H11/2, and Yb3þ 2F7/2 2F5/2 states, respectively;
t1 t2 t3
tYb are the lifetimes for Er3þ: 4I11/2, 4S3/2, 2H11/2, and Yb3þ
2F5/2 states, respectively; C0 and C1 are the ET coefficients;
is the
lP and ε represent
:
2
[47,48]. And the population of H11/2 is totally attributed to the
,
,
,
,
:
,
4
thermal population from S3/2 due to their thermally coupled
,
,
:
relationship [47]. Under steady-state excitation, Eqs. (4)e(7) give
G
4
4
2
rise to the population of S3/2 emitting level as
thermally populating rate from S3/2 to H11/2
;
the pump wavelength (980 nm) and the absorption cross section of
Yb3þ at 980 nm; P and s are the power and the square of the irra-
diation spot, respectively; h and c represent the Planck constant
and light speed, respectively. The detailed description can be con-
sulted in Fig. 6. Eqs. (5) and (6) are obtained based on the
assumption: because of the much weak luminescence from the 4F7/
a
g2
P2
þ
N2 ¼
ꢀ
(8)
ð1=t2
G
Þ=t1
where
a
¼ C1C0N0Na2ε2lP2/h2c2s2;
g
¼ C0N0þtꢁYb1 (detailed calculation
routine can be consulted in the supplementary material). In view of
of Er3þ as well as little luminescence from 2H11/2
/
4I15/2 at low
the Boltzmann type distribution between 2H11/2 and 4S3/2 levels,
G
2

W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
553
is obtained by combining Eq. (6) as
4S3/2 and 2H11/2 states have been measured, as shown in the left of
Fig. 8. The corresponding decay rates are obtained based on these
curves and fitted by Eq. (10), as shown in the right of Fig. 8. It can be
observed that the fitted curves agree well with the experimental
data. tꢁ2 1 and tꢁ3 1 at temperatures corresponding to the different
pumping powers are deduced through the fitted results. Noting the
emission intensity I is proportional to N, the slope n is calculated
based on the above theory model and presented as theoretical
value-I in Fig. 7. The theoretically calculated slopes n are observed
to agree well with the experimental results in the pumping powers
ranging from 50 to 570 mW. The decrease of n (1.10) for the
experimental data in the range of 300e570 mW implies the
occurrence of LIS, which can also be inferred via the theoretical
calculation (1.21) based on our rate equation model. As the power is
further increased (>570 mW), the n value for the experimental data
turns into a negative value due to the thermal quenching. Although
the n for the theoretical value-I keeps decreasing at powers sur-
passing 570 mW, the corresponding value is larger than that in the
experiment. This deviation is probably attributed to the neglecting
ꢀ
ꢁ
gH
DE
G
¼
exp
ꢁ
kBT
(9)
gSt3
In order to get the result in Eq. (8), it has been assumed that the
ET between the acceptor and donor do not significantly affect the
lifetime of the excited ions, implyingtꢁ11»C1Nb, i.e. the UC process
for the discussed level is neglected. Pumped by a continuous diode
laser with low power, just a few ions can be excited, implying Nx
(x¼0, a) z const. and C0N0»C1N1.
It is known that the lifetimes for the excited levels can be
described as [49].
ꢂ
ꢀ
ꢁꢃꢁp
X
X
1
t
Zw
kBT
¼
Aij þ WNR
¼
Aij þ W0 1 ꢁ exp
ꢁ
(10)
where SAij is the sum of radiative transition rates from level i to its
lower lying levels j; WNR and W0 are the multiphonon nonradiative
relaxation (MNR) rate and MNR at 0 K; ħu and p denote the phonon
energy and the number of phonons involved in the MNR, respec-
tively. The MNR for Yb3þ: 2F5/2 state is not expected due to the large
energy gap between 2F5/2 and 2F7/2. The MNR for Er3þ: 4I11/2 state is
neglected for simplifying the discussion. Meanwhile, because of the
resonant ET from Yb3þ to Er3þ when excited by 980 nm laser, C0 and
C1 are assumed as constants without lack of generality. Based on
4
of the MNR for the intermediate level I11/2 during the analysis.
Owing to the great enhancement of laser spot temperature with the
increase of the excitation powers, the MNR process for the inter-
mediate state 4I11/2 will also take place and be strengthened, which
naturally weakens the population on Er3þ 4S3/2 state and conse-
:
quently contribute to the thermal quenching for the luminescence
at high pumping powers. Here, it would be interesting to separately
understand how the MNR process influences LIS. The effect of the
MNR for Er3þ: 4I11/2 is deduced and taken into account by the ratio
Ie/It between the relative intensities corresponding to the experi-
these above assumptions, t1
constants. If the laser heating effect, i.e. the temperature dependent
parameters t2 and
in Eq. (8) are negligible, the result N2fP2 (i.e.
, a, and g in Eq. (8) can be considered as
G
mental data Ie and theoretical value-I It. The
G item in Eq. (8) is
n ¼ 2) can be obtained, as seen the ideal value in Fig. 7. However,
owing to the giant enhancement of the irradiation spot tempera-
ture on the sample with the increase of excitation power, the
thermal absorption and the MNR processes will be promoted,
temporarily discarded, and the relationship between the 553 nm
emission intensity I and the excitation power P could be written as
follow
which make the values of tꢁ2 1 t3ꢁ1 and
, G become larger and finally
a
g2
P2
1=ðt1t2
contribute to the depopulation of 4S3/2 state. Hence, the LH effect is
taken into consideration in the present work.
I ¼
ꢀ
:
(11)
Þ
It can be known from Eqs. (8)e(10) that the constants
a and g do
The calculated slopes n are shown as theoretical value-II in Fig. 7.
not influence the slope n of the luminescence intensity vs. pumping
power in double-logarithmic representation. Therefore, n can be
quantificationally calculated if tꢁ2 1 and t3ꢁ1 at temperatures corre-
sponding to the different pumping powers are given. For this pur-
pose, the temperature dependent luminescence decay curves for
It is depicted that the LIS can be obviously induced only by the
thermally strengthened MNR as the pumping power is increased.
Based on Eqs. (8)e(11) and the results shown in Fig. 5, the
relationship between the emission intensity I and the excitation
power P can be generally expressed as
Pn
f ðPÞ
If
(12)
where f(P) is a function of excitation power P, reflecting the laser
induced thermal effect, and the value of f(P) would increase as P is
raised. From the results presented in Fig. 7, some conclusions could
be achieved. When the pumping power P is relatively low, the laser
induced heating effect is not obvious. Therefore, f(P) is negligible
compared to Pn, so we can approximately get IfPn. With the in-
crease of P, f(P) is comparable to Pn, getting the n to become smaller
and leading to the LIS phenomenon. As P is further enhanced, f(P)
becomes larger than Pn, causing the decrease of I with the increase
of P and finally resulting in the luminescence thermal quenching.
4. Conclusions
CaWO4 phosphors doping with different Er3þ and Yb3þ con-
centrations were prepared, and the influence of RE dopant content
on the thermometry behavior of Er3þ green UC emissions was
discussed when excited by 980 nm cw diode laser. A maximum
sensitivity of 1.05 ꢀ 10ꢁ2 Kꢁ1 is achieved here, which is superior to
Fig. 7. The experimental and theoretical slope of the intensity for 553 nm emission vs.
the pumping power in double-logarithmic representation.

554
W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
4
2
Fig. 8. The luminescence decay curves and decay rates for the S3/2 and H11/2 states at different temperatures as well as the fitting results.
other Er3þ based sensors. And the result implies that preferable
temperature sensing properties could be realized by further raising
Yb3þ or reducing Er3þ dopants. In addition, the excitation laser
induced heating effect in Er3þ-Yb3þ codoped CaWO4 has been
studied by using the luminescence thermometry method. It is
illustrated that such heating effect plays an important role in the UC
luminescence intensity saturation and should be taken into
consideration when investigating the excitation mechanisms for
the upconverted materials.
M. Yu, Ultrahigh-sensitive optical temperature sensing based on ferroelectric
Pr3þ-doped (K0.5Na0.5)NbO3, Appl. Phys. Lett. 108 (2016), 061902e1-061902-
5.
[8] P. Du, L.H. Luo, H.K. Park, J.S. Yu, Citric-assisted sol-gel based Er3þ/Yb3þ
-
codoped Na0.5Gd0.5MoO4: a novel highly-efficient infrared-to-visible upcon-
version material for optical temperature sensors and optical heaters, Chem.
Eng. J. 306 (2016) 840e848.
[9] L. Marciniak, A. Pilch, S. Arabasz, D.Y. Jin, A. Bednarkiewicz, Heterogeneously
Nd3þ doped single nanoparticles for NIR-induced heat conversion, lumines-
cence, and thermometry, Nanoscale 9 (2017) 8288e8297.
[10] A.S. Souza, L.A.O. Nunes, I.G.N. Silva, F.A.M. Oliveira, L.L. da Luz, H.F. Brito,
M.C.F.C. Felinto, R.A.S. Ferreira, S.A. Júnior, L.D. Carlos, O.L. Malta, Highly-
sensitive Eu3þ ratiometric thermometers based on excited state absorption
with predictable calibration, Nanoscale 8 (2016) 5327e5333.
Acknowledgements
[11] B.P. Singh, A.K. Parchur, R.S. Ningthoujam, P.V. Ramakrishna, S. Singh, P. Singh,
S.B. Rai, R. Maalej, Enhanced up-conversion and temperature-sensing
behaviour of Er3þ and Yb3þ co-doped Y2Ti2O7 by incorporation of Liþ ions,
Phys. Chem. Chem. Phys. 16 (2014) 22665e22676.
This work was supported by the National Nature Science
Foundation of China (NSFC) (grant number 61505174); and the
Hebei Province Nature Science Foundation (grant number
E2016203185).
[12] R. Dey, V.K. Rai, Yb3þ sensitized Er3þ doped La2O3 phosphor in temperature
sensors and display devices, Dalton Trans. 43 (2014) 111e118.
[13] B. Dong, B.S. Cao, Y.Y. He, Z. Liu, Z.P. Li, Z.Q. Feng, Temperature sensing and
in vivo imaging by molybdenum sensitized visible upconversion lumines-
cence of rare-earth oxides, Adv. Mater. 24 (2012) 1987e1993.
Appendix A. Supplementary data
ꢀ
ꢀ
[14] S.F. Leon-Luis, U.R. Rodríguez-Mendoza, P. Haro-Gonzalez, I.R. Martín,
V. Lavín, Role of the host matrix on the thermal sensitivity of Er3þ lumines-
cence in optical temperature sensors, Sensors Actuators B. Chem. 174 (2012)
176e186.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jallcom.2017.08.007.
[15] D. He, C. Guo, S. Jiang, N. Zhang, C. Duan, M. Yin, T. Li, Optical temperature
sensing properties of Yb3þ-Er3þ co-doped NaLnTiO4 (Ln
conversion phosphors, RSC Adv. 5 (2015) 1385e1390.
¼
Gd, Y) up-
References
[16] L.Z. Zhang, H.F. Lin, G. Zhang, X. Mateos, J.M. Serres, M. Aguilo, F. Diaz,
U. Griebner, V. Petrov, Y.C. Wang, Crystal growth, optical spectroscopy and
laser action of Tm3þ-doped monoclinic magnesium tungstate, Opt. Express 25
(2017) 3682e3693.
[1] T.T. Bai, N. Gu, Micro/Nanoscale thermometry for cellular thermal sensing,
Small 12 (2016) 1590e4610.
[17] C.S. Lim, Microwave sol-gel process of KGd(WO4)2:Ho3þ/Yb3þ phosphors and
their upconversion photoluminescence properties, J. Phys. Chem. Solids 78
(2015) 65e69.
[2] S.N. Zhao, L.J. Li, X.Z. Song, M. Zhu, Z.M. Hao, X. Meng, L.L. Wu, J. Feng,
S.Y. Song, C. Wang, Lanthanide ion codoped emitters for tailoring emission
trajectory and temperature sensing, Adv. Funct. Mater. 25 (2015) 1463e1469.
[3] E.C. Ximendes, W.Q. Santos, U. Rocha, U.K. Kagola, F. Sanz-Rodríguez,
[18] W. Xu, X.Y. Gao, L.J. Zheng, P. Wang, Z.G. Zhang, W.W. Cao, Optical ther-
mometry through green upconversion emissions in Er3þ/Yb3þ-codoped
CaWO4 phosphor, Appl. Phys. Express 5 (2012), 072201e1-072201-3.
[19] H.Y. Lu, R. Meng, H.Y. Hao, Y.F. Bai, Y.C. Gao, Y.L. Song, Y.X. Wang, X.R. Zhang,
Stark sublevels of Er3þ-Yb3þ codoped Gd2(WO4)3 phosphor for enhancing the
sensitivity of a luminescent thermometer, RSC Adv. 6 (2016) S7667eS7671.
[20] X.N. Chai, J. Li, X.S. Wang, Y.X. Li, X. Yao, Color-tunable upconversion photo-
luminescence and highly performed optical temperature sensing in Er3þ/Yb3þ
co-doped ZnWO4, Opt. Express 24 (2016) 22438e22447.
ꢀ
N. Fernandez, A.D.S. Gouveia-Neto, D. Bravo, A.M. Domingo, B. del Rosa,
Unveiling in vivo subcutaneous thermal dynamics by infrared luminescent
nanothermometers, Nano Lett. 16 (2016) 1695e1703.
[4] X.F. Wang, Q. Liu, Y.Y. Bu, C.S. Liu, T. Liu, X.H. Yan, Optical temperature sensing
of rare-earth ion doped phosphors, RSC Adv. 5 (2015) 86219e86236.
ꢀ
ꢀ
[5] L. Marciniak, K. Prorok, L. Frances-Soriano, J. Perez-Prieto, A. Bednarkiewicz,
A broadening temperature sensitivity range with a core-shell YbEr@YbNd
double ratiometric optical nanothermometer, Nanoscale 8 (2016) 5037e5042.
[6] S.A. Wade, S.F. Collins, G.W. Baxter, Fluorescence intensity ratio technique for
optical fiber point temperature sensing, J. Appl. Phys. 94 (2003) 4743e4756.
[7] W. Tang, S.C. Wang, Z.L. Li, Y. Sun, L.M. Zheng, R. Zhang, B. Yang, W.W. Cao,
ꢀ
[21] C. Cascales, C.L. Paíno, E. Bazan, C. Zaldo, Ultrasmall, water dispersible,
TWEEN80 modified Yb:Er:NaGd(WO4)2 nanoparticles with record

W. Xu et al. / Journal of Alloys and Compounds 726 (2017) 547e555
555
upconversion ratiometric thermal sensitivity and their internalization by
mesenchymal stem cells, Nanotechnology 28 (2017), 185101e1-185101-11.
[22] M. Pollnau, D.R. Gamelin, S.R. Lüthi, H.U. Güdel, M.P. Hehlen, Power depen-
dence of upconversion luminescence in lanthanide and transition-metal-ion
systems, Phys. Rev. B 61 (2001) 3337e3346.
[23] J. Feenstra, I.F. Six, M.A.H. Asselbergs, R.H. van Leest, J. de Wild, A. Meijerink,
R.E.I. Schropp, A.E. Rowan, J.J. Schermer, Er3þ/Yb3þ upconverters for InGaP
solar cells under concentrated broadband illumination, Phys. Chem. Chem.
Phys. 17 (2015) 11234e11243.
[36] L. Li, C.F. Guo, S. Jiang, D.K. Agrawal, T. Li, Green up-conversion luminescence
of Yb3þ-Er3þ co-doped CaLa2ZnO5 for optically temperature sensing, RSC Adv.
4 (2014) 6391e6396.
[37] Z.Y. Zhang, C.F. Guo, H. Suo, X.Q. Zhao, N.M. Zhang, T. Li, Thermometry and up-
conversion luminescence of Yb3þ-Er3þ co-doped Na2Ln2Ti3O10 (Ln ¼ Gd, La)
phosphors, Phys. Chem. Chem. Phys. 18 (2016) 18828e18834.
[38] S.S. Wang, H. Zhou, X.X. Wang, A.L. Pan, Up-conversion luminescence and
optical temperature-sensing properties of Er3þ-doped perovskite Na0.5Bi0.5
-
TiO3 nanocrystals, J. Phys. Chem. Solids 98 (2016) 28e31.
[39] T.V. Gavrilovic, D.J. Jovanovic, V. Lojpur, M.D. Dramicanin, Multifunctional
Eu3þ- and Er3þ/Yb3þ-doped GdVO4 nanoparticles synthesized by reversemi-
celle method, Sci. Rep. 4 (2014).
[24] G.Y. Chen, T.Y. Ohulchanskyy, A. Kachynski, H. Agren, P.N. Prasad, Intense
visible and near-infrared upconversion photoluminescence in colloidal LiYF4:
Er3þ nanocrystals under excitation at 1490 nm, ACS Nano
5 (2011)
4981e4986.
[40] X.N. Chai, J. Li, Y. Zhang, X.S. Wang, Y.X. Li, X. Yao, Bright dual-mode green
emission and temperature sensing properties in Er3þ/Yb3þ co-doped MgWO4
phosphor, RSC Adv. 6 (2016) 64072e64078.
[25] F.X. Wang, F. Song, G. Zhang, Y.D. Han, Q. Li, C.G. Ming, J.G. Tian, Upconversion
and pump saturation mechanisms in Er3þ/Yb3þ codoped Y2Ti2O7 nanocrystals,
J. Appl. Phys. 115 (2014), 134310e134311-134310-7.
[41] Z. Wei, W. Zheng, Z.Y. Zhu, X.F. Guo, Upconversion of SrWO4: Er3þ/Yb3þ
:
[26] G. Zhang, F. Song, C.G. Ming, Y. Yu, C. Zhang, H.Y. Zhao, Q.R. Wang, J.D. Liu,
Photoluminescence properties and pump-saturation effect of Er3þ/Yb3þ
codoped Y2Ti2O7 nanocrystals, J. Lumin 132 (2012) 774e779.
improvement by Yb3þ codoping and temperature sensitivity for optical
temperature sensors, Chem. Phys. Lett. 651 (2016) 46e49.
[42] P. Du, L.L. Luo, J.S. Yu, Infrared-to-visible upconversion emission of Er3þ/Yb3þ
-
[27] T. Grzyb, A. Tyminski, Up-conversion luminescence of GdOF: Yb3þ, Ln3þ (Ln ¼
codoped SrMoO4 phosphors as wide-range temperature sensor, Curr. Appl.
Phys. 15 (2015) 1576e1579.
[43] P. Du, L.H. Luo, J.S. Yu, Upconversion emission, cathodoluminescence and
temperature sensing behaviors of Yb3þ ions sensitized NaY(WO4)2:Er3þ
phosphors, Ceram. Int. 42 (2016) 5635e5641.
ꢀ
Ho, Tm, Er) nanocrystals, J. Alloys Compd. 660 (2016) 235e243.
[28] A.K. Soni, V.K. Rai, Thermal and pump power effect in SrMoO4:Er3þ-Yb3þ
phosphor for thermometry and optical heating, Chem. Phys. Lett. 667 (2017)
226e232.
[29] R.D. Shannon, Acta Cryst. A 32 (1976) 751e767.
[44] X. F. Wang, Y. Wang, J. Marques-Hueso, X. H. Yan, Improving Optical tem-
perature sensing performance of Er3þ doped Y2O3 microtubes via co-doping
and controlling excitation power, Sci. Rep. doi:10.1038/s41598-017-00838-w.
[45] X.Z. Zhou, B.E. Smith, P.B. Roder, P.J. Pauzaushie, Laser refrigeration of
ytterbium-doped sodium-yttrium-fluoride nanowires, Adv. Mater. 28 (2016)
8658e8662.
[30] F. Huang, Y. Gao, J.C. Zhou, J. Xu, Y.S. Wang, Yb3þ/Er3þ codoped CaMoO4: a
promising green upconversion phosphor for optical temperature sensing,
J. Alloys Compd. 639 (2015) 325e329.
[31] G. Lakshminarayana, J.R. Qiu, M.G. Brik, G.A. Kumar, I.V. Kityk, Spectral anal-
ysis of Er3þ, Er3þ/Yb3þ- and Er3þ/Tm3þ/Yb3þ doped TeO2-ZnO-WO3-TiO2-
Na2O glasses, J. Phys. Condens. Matter 20 (2008), 375101e1-375101-8.
[32] B.R. Judd, Optical Absorption intensity of rare-earth ions, Phys. Rev. 127
(1962) 750e761.
[46] R. Dey, A. Pandey, V.K. Rai, Er3þ-Yb3þ and Eu3þ-Er3þ-Yb3þ codoped Y2O3
phosphors as optical heater, Sensors Actuators B Chem. 1901 (2014) 512e515.
[47] X.H. Xu, W.F. Zhang, D.C. Yang, W. Lu, J.B. Qiu, S.F. Yu, Phonon-assisted pop-
ulation inversion in lanthanide-doped upconversion Ba2LaF7 Nanocrystals in
glass-ceramics, Adv. Mater. 28 (2016) 8045e8050.
[33] G.S. Ofelt, Intensities of crystal spectra of rare-earth ions, J. Chem. Phys. 37
(1962) 511e520.
[34] H. Desirena, E. De la Rosa, L.A. Díaz-Torres, G.A. Kumar, Concentration effect of
Er3þ ion on the spectroscopic properties of Er3þ and Yb3þ/Er3þ co-doped
phosphate glasses, Opt. Mater. 28 (2006) 560e568.
[35] E. Maurice, G. Monnom, B. Dussardier, A. Saïssy, D.B. Ostrowsky, G.W. Baxter,
Erbium-doped silica fibers for intrinsic fiber-optic temperature sensors, Appl.
Opt. 34 (1995) 8019e8025.
[48] C.J. da Silva, M.T. de Araujo, E.A. Gouveia, A.S. Gouveia-Neto, Thermal effect on
multiphonon-assisted anti-Stokes excited upconversion fluorescence emis-
sion in Yb3þ sensitized Er3þ doped optical fiber, Appl. Phys. B 70 (2000)
185e188.
[49] M.J. Weber, Multiphonon relaxation of rare earth ions in Yttrium Orthoalu-
minate, Phys. Rev. B 8 (1973) 54e64.