Journal of Luminescence 182, February 2017, Pages 200–207 http://dx.doi.org/10.1016/j.jlumin.2016.10.033 journal homepage: http://www.sciencedirect.com/science/journal/00222313/182/supp/C
Influence of codoping on the luminescence properties of YAG:Dy for high temperature phosphor thermometry Ellen Hertle a, c, Liudmyla Chepyga
b, c
, Miroslaw Batentschuk , Lars Zigan b
a, c
a
Lehrstuhl für Technische Thermodynamik (LTT), Friedrich-Alexander-Universität Erlangen-Nürnberg, Am Weichselgarten 8, 91058 Erlangen, Germany b Institute of Materials for Electronics and Energy, Friedrich-Alexander-Universität Erlangen-Nürnberg, Martensstraße 7, 91058 Erlangen, Germany c Erlangen Graduate School in Advanced Optical Technologies (SAOT), Friedrich-AlexanderUniversität Erlangen-Nürnberg, Paul-Gordan-Str. 6, 91052 Erlangen, Germany *Corresponding author:
[email protected] Corresponding author at: Lehrstuhl für Technische Thermodynamik, Friedrich-Alexander-Universität Erlangen-Nürnberg, Am Weichselgarten 8, D-91058 Erlangen, Germany. Tel.: +49 9131 8529770; fax: +49 9131 8529901.
© 2016 Elsevier Ltd. All rights reserved. One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modifications of the content of this paper are prohibited. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/
In uence of codoping on the luminescence properties of YAG:Dy for high temperature phosphor thermometry Ellen Hertlea,c , Liudmyla Chepygab,c , Miroslaw Batentschukb , Lars Zigana,c,∗ a Lehrstuhl für Technische Thermodynamik (LTT), Friedrich-Alexander-Universität Erlangen-Nürnberg, Am Weichselgarten 8, 91058 Erlangen, Germany
b Institute of Materials for Electronics and Energy, Friedrich-Alexander-Universität Erlangen-Nürnberg, Martensstraÿe 7, 91058 Erlangen, Germany
c Erlangen Graduate School in Advanced Optical Technologies (SAOT), Friedrich-Alexander-Universität Erlangen-Nürnberg, Paul-Gordan-Str. 6, 91052 Erlangen, Germany
Abstract
The eects of codoping on the temperature-dependent luminescence of YAG:Dy were investigated for application in high temperature thermometry. YAG:Dy is a well-known thermographic phosphor suitable for high temperature measurements. However, its decay time is too long for studying temperatures in fast transient technical processes. Codoping is a known method for increasing the signal intensity and for decreasing luminescence decay times. Five co-doped samples with three dierent sensitizers, namely Tm, Tb and Pr were synthesized by solid-state method. Changes in spectral emission behavior and temperature dependent decay time are presented up to 1600 K. The intensity ratio used for temperature measurements was similar for all samples, and decay time decreased for all codopings. Codoping by Tm decreased signal intensity considerably, thus YAG:Dy,Tm is not suitable for high temperature measurements. However, codoping YAG:Dy with Tb and especially Pr resulted in improved luminescent characteristics, and these phosphors are promising for high temperature phosphor thermometry.
Keywords: Thermographic phosphors, High temperature thermometry, Codoping ∗
Lars Zigan
Email address:
[email protected] (Lars Zigan )
Preprint submitted to Journal of Luminescence
October 19, 2016
1. Introduction
Thermographic phosphor thermometry has become an established technique for remote and non-intrusive temperature measurements. The phosphors employed usually consist of a ceramic powder host matrix doped with a rare-earth 5
or transition metal activator. They are means to provide temporally and spatially resolved temperature elds for both surface temperature measurements [1] as well as for thermometry of gaseous ows [2, 3, 4]. Depending on the technical application studied, phosphors with dierent characteristics such as temperature sensitivity range, decay time, luminescence
10
intensity, emission wavelength or cross sensitivities are needed. Investigation of the thermographic properties for various phosphors were presented by Allison and Gillies [4] and Aldén et al. [5]. An overview over the decay time characteristics of dierent phosphor materials is given in [1]. An extensive summary on phosphor applications for dierent temperature ranges can be found in [6, 7].
15
The phosphorescent materials mainly used for high temperature thermometry are presented in Table 1. The widest temperature range is covered by YAG:Dy and YAG:Tm showing temperature dependent exposure times of 1-103 µs or 0.1-102 µs, respectively. The luminescence properties of thermographic phosphors strongly depend
20
on the interaction between host lattice and dopant ions. As a rule, rare earth Table 1: Survey of main phosphors employed for high temperature measurements.
Phosphor
Temperature range
Decay time
References
YAG:Dy
300-1700 K
1 · 103 - 1 µs
[8, 9, 10, 11]
YAG:Pr
300-1100 K
190 µs
YAG:Tm
300-1700 K
1 · 102 - 0.1 µs
[15, 16]
Y2 O3 :Eu
290-1300 K
1 · 10 - 1 µs
[17, 18]
GdAlO3 (GAP):Cr
300-1350 K
1 · 104 - 1 µs
[19]
BaMg2 Al10 O17 (BAM):Eu
300-1300 K
2 - 0.01 µs
2
3
[12, 13, 14]
[20, 21]
activators have a weak coupling of the 4f electrons with the lattice. This results in narrow emission lines and higher thermal quenching temperatures, which is advantageous in comparison to transition metals [22]. The YAG crystal structure is body-centered cubic, with the rare earth dopant substituting 3 Y at the 25
dodecahedral 3 Y-sites in the crystal lattice [23]. In general, increasing temperature in uences the spectral behavior of rare earths. Thermalization eects occur when two energy levels lie relatively close together, as for example the 4 F9/2 and the 4 I15/2 energy levels of YAG:Dy. At higher temperatures the population of higher energy levels rises, changing the
30
intensity distribution of the spectrum corresponding to Boltzmann's law. The radiative transition decay time is reduced due to quenching, vibrational relaxation and internal conversions. Additionally, broadening of linewidth becomes more pronounced caused by lattice vibrations [24]. The Dieke diagram can help to understand the emission behavior of the
35
activator ions. An excerpt for the four studied rare earths, namely Dy3+ , Pr3+ , Tm3+ and Tb3+ is given in Figure 1. Y3 Al5 O12 (YAG) has proven to be an e cient host for high temperature thermometry, having a high melting point of 2213 K [26]. Moreover, many lanthanide activators show higher temperature sensitivity ranges in YAG as in
40
other host matrices [1]. YAG:Dy is widely used in combustion application for thermometry showing emission up to 1700◦ C from both 4 F9/2 and 4 F7/2 level [16, 27], see Fig. 1. Its emission characteristics are not aected by combustion environments [28]. YAG:Dy has the additional advantage of emitting in the blue region, thus being less aected by black body radiation at high temperatures
45
[21]. Based on the work of Heyes [22], who developed a method to design phosphors for high temperature, at the moment YAG:Dy is the phosphor with the highest capability for thermometry at high temperatures. However, YAG:Dy has a long emission decay time of several 100 microseconds at room temperature. This introduces measuring inaccuracies at short
50
exposure times of the detector due to low signals. At long exposure times black body radiation at high temperatures cannot be satisfactorily suppressed [21]. 3
40
35
1
I6
1
D2
30
355 nm
Energy / 10³ cm-1
5
25
5
20 1
D2
1
G4
D3
4
F7/2
4
I15/2 F9/2
4
D4
7
G4
15
10
5/2 3
7/2
H5
9/2 11/2
5 3
0
4
6
3
13/2
H5
H4
Pr
7
F6
Tb
6
H15/2
Dy
3
H6
Tm
Figure 1: The Dieke energy diagram for the rare earths Pr3+ , Tb3+ , Dy3+ and Tm3+ following [25]. The green shading between levels notes the largest energy gap for the rare earths.
4
In addition, high-temperature thermometry above 1000 K is still limited because of signal loss due to thermal quenching. It is possible to simultaneously reduce quenching and improve luminescence intensity by substituting the 55
tetrahedral site Al3+ -O2- in YAG:Dy with B3+ -N3- [29, 30]. Enhanced Dy3+ emission by a factor of 2.5 to 4 has been observed for Gd3 Al5 O12 stabilized with Lu3+ compared to the simple YAG:Dy [31]. Another option to improve luminescence properties is to use a second sensitizer in addition to the original activator. Codoping YAG:Dy with Er3+ [8, 27]
60
or Bi3+ [32] led to an increase in signal intensity, the latter enhancing luminescence by about seven times at room temperature. Enhanced luminescence has also been observed for codoping Y2 Zr2 O7 :Dy3+ with Li+ [33]. Using alternative codoping combinations can also increase luminescence intensity while decreasing the decay time simultaneously. Of special interest for
65
high temperature applications are other rare earth sensitizer such as Tm3+ , Tb3+ and Pr3+ . These lanthanide ions have similar excitation wavelengths, are temperature sensitive and both Tm3+ , and Pr3+ have shorter decay times than Dy3+ [12, 15]. Energy transfer processes between Dy3+ and Tb3+ were studied in a YF3
70
matrix [34], for gadolinium oxide hosts [35] and Y2 O3 nanophosphors [36]. Likewise the eect of codoping Dy3+ ions with Tb3+ ions has been studied for various glasses [37, 38, 39, 40], nding non-radiative energy resonance transfers due to dipole-dipole interactions. In this context shorter decay times of about 25% of the original time for the Dy3+ luminescence were presented [34, 37]. Temper-
75
ature dependence from 30 K to 300 K as well as decreasing decay times with higher Tb concentrations in Dy3+ -Tb3+ codopings have been observed for a CaMoO4 host matrix [41]. Energy transfer between Tm3+ to Dy3+ has been observed for tellurite [42] and selenide [43] glasses. YSZ co-doped by (Dy3+ + Tm3+ ) was used for study-
80
ing luminescence intensity of thermal barrier coatings, focusing on porosity and phase composition [44]. No information on codoping eects for Dy3+ and Pr3+ ions can be found in 5
the literature. However, singly-doped YAG:Pr has a short luminescent lifetime while oering good temperature measurement capabilities up to 1100 K [12]. 85
This paper concentrates on the in uence of Tm3+ , Tb3+ and Pr3+ codopings on the luminescent behavior of YAG:Dy. Dierent activator concentrations are presented and their in uence on the absolute signal intensity, the decay time and the spectral behavior are studied.
2. Materials and method
90
2.1. Sample preparation and characterization Y(3-x) Al5 O12 :x% Dy3+ and Y(3-x-y) Al5 O12 :x % Dy3+ , y% RE3+ phosphors were fabricated by conventional high temperature solid-state method, with RE being Pr3+ , Tb3+ and Tm3+ respectively. In the following sections the expression of Y(3-x) Al5 O12 :x% Dy3+ and Y(3-x-y) Al5 O12 :x % Dy3+ , y% RE3+ are
95
abbreviated as YAG:Dy(x%),RE(y%), where x is the dopant concentration in mol % and y is the sensitizer concentration in mol %, respectively. In order to decrease sintering temperature and to improve the phosphor e ciency B2 O3 was chosen as a ux. Stoichiometric amounts of Y2 O3 ,Al2 O3 , Pr2 O3 , Tb2 O3 , Tm2 O3 , B2 O3 (all, 99.99%, Alfa Aesar) Dy2 O3 (99.9%, Reacton) were used as
100
starting materials in the present work. The starting powders were mixed and ground in an agate mortar by hand. Subsequently, they were red in an alumina crucible at 1673 K for 7 h in air to produce the nal samples. After ring, samples were cooled to room temperature in the furnace and were ground again to powder for subsequent use. The crystal structure of the phosphors was exam-
105
ined by X-ray diraction carried out ex situ with a PANalytical X'Pert powder diractometer using standard BraggBrentano geometry with ltered Cu-Kα radiation source (λ = 1.5406 A) and an X'Celerator solid-state stripe detector at room temperature. Morphology and elemental analysis of the phosphors were performed with a scanning electron microscope (SEM) (JEOL) equipped
110
with a eld emission gun operated at 15 kV. The photoluminescence (PL) and photoluminescence excitation (PLE) spectra of powders at room temperature 6
Phosphor Sample
Dichroic Mirror
LP 355
BS PD2
Lens Set-Up PD1 Bandpass
filters
Oven (300-1700K) HR 355 Nd:YAG-Laser
Figure 2: Schematic view of the set-up used for phosphor emission decay characterisation. For spectral analysis a spectrometer replaced the two photo diodes.
were measured with a spectro uorometer (Jasco FP-8500). Optical measurements were conducted under identical conditions for all samples with a spectral resolution of 5 nm on both the excitation and emission side. 115
2.2. Experimental set-up for high temperature measurements A high temperature oven (Supertherm HT 18/10) with optical access was used for spectral emission and decay time investigations. The phosphor samples were placed into corundum ceramic sample carriers and reproducibly positioned inside the furnace. To excite luminescence, the third harmonic of a
120
pulsed Nd:YAG laser (Quantel Q-smart 450) with a pulse duration of 6 ns and a repetition rate of 10 Hz was used. A low average pulse energy of 8 mJ was selected to avoid even slight in uences of the laser pulse energy on the decay or spectral behavior, which has been observed for some phosphors such as BAM:Eu J [45]. This laser energy resulted in a laser uence of about 0.03 cm 2 , which is far
125
J below the the critical limit of 1.5 cm 2 observed for gaseous ows [3]. Figure 2
shows a schematic view of the experimental set-up. A 355 nm dichroic mirror guided the laser beam inside the furnace while allowing the transmission of the luminescence emission. On the detection side a spherical lens collected the luminescence signal, while a second spherical lens 130
focused the signal on the detector. Additionally, a 355 nm longpass edge lter 7
was positioned between the spherical lenses to block the laser light ahead of the detector. The luminescence spectrum was monitored using a Ocean Optics Flame-S spectrometer with a wavelength range of 400-700 nm. The spectrometer exhibits 135
a grating of 1200 lines/mm, a 16 bit 2048 Pixel CCD detector and a slit width of 25 µm. The spectrometer was triggered, and the integration time was set to 2 ms in order to avoid strong interference of black-body radiation. The decay time was determined using two high speed photo diodes (Thorlabs, DET 210). Using a single-edge imaging- at dichroic beam splitter (484 nm edge
140
BrightLine) and two bandpass lters of 494±10 nm and 465±15 nm allowed to evaluate the decay time of two dierent wavelength regions simultaneously. The photo current was read out temporally resolved by an oscilloscope (PicoScope 6, real-time sampling rate of 200 µs/div). The temperature in the furnace was varied between 300 K and 1600 K. A
145
type B thermocouple (shielded to avoid radiation eects, 0.75% inaccuracy) was used to monitor the temperature in the proximity of the phosphor sample. For each experimental condition 150 single shots were recorded for the spectral data and 32 single shots were taken for decay time measurements.
3. Data evaluation
150
3.1. Phase identi cation and morphology X-ray diraction was carried out to prove the successful synthesis of the different codopings (see Fig. 3). The measurements revealed well-de ned Bragg re ections of pure crystalline powder samples. This indicates that the six samples are well-crystallized in a phase of cubic crystalline structure, without appre-
155
ciable changes due to the dierent Dy3+ and Dy3+ , RE3+ (Tm3+ , Tb3+ , Pr3+ ) doping concentration. These patterns have been indexed to the YAG garnet structure. In YAG, Y is eightfold coordinated in a distorted cubic D2 local site; this site is traditionally labelled as the dodecahedral site because the distorted cube produces a polyhedron with 12 triangular faces. Al is in an octahedral 8
YBO
Tm(1%) 3
Intensity (a.u )
Tb(1%)
10
30
40
(640)
(642)
(444)
50
(721)
(631)
(532)
(521)
(440)
(431)
(422)
(400)
(420)
20
(321)
(220)
(211)
Pr (1%)
60
70
2-Theta / degree
Figure 3: XRD patterns of YAG:Dy(2%), YAG:Dy(2%),Tm(1%), YAG:Dy(2%),Tb(1%) and YAG:Dy(2%),Pr(1%) 160
environment and in a tetrahedral environment. Due to ionic size considerations in the YAG lattice, the RE3+ ions are expected to predominantly enter into the distorted dodecahedral sites by replacing the Y3+ ions and to be coordinated to eight O2 - ions [46]. There were no additional or missing re ections in the X-ray diraction patterns which could imply the choice of a dierent unit cell or the
165
presence of unknown by-products (Fig. 3). Furthermore, results revealed that some of the samples contained a minor secondary yttrium orthoborate phase YBO3 (3%) as side-product. The presence of YBO3 phase can be explained by solubility limit of the ux B2 O3 in Y3 Al5 O12 . A SEM study was performed to investigate the surface morphology and crys-
170
tallite sizes of the six synthesized phosphors fabricated by conventional high temperature solid-state method at 1673 K. Fig. 4 exemplary shows SEM images for YAG:Dy(2%) and YAG:Dy(2%),Tm(1%). The size of the particles varies from a few to several micrometers and the particles are conglomerated due to the ux and high temperature used during synthesis. When compar-
175
ing the dierent powders the overall morphological structure is similar for all
9
YAG:Dy(2%),Tm(1%)
YAG:Dy(2%) YAG:Dy(2%)
10
YAG:Dy(2%)
1
10
m
m
YAG:Dy(2%),Tm(1%)
m
1
m
Figure 4: Particle morphology of YAG:Dy(2%) (left) and YAG:Dy(2%),Tm(1%) (right)
synthesized samples.
3.2. Spectral investigation The luminescence emission of thermographic phosphors can be used to determine temperature by using either the intensity ratio (IR) or the decay time 180
method. Emission spectra were evaluated in a temperature range up to 1600 K. From 300 K to 700 K spectral data was taken for every 100 K, above that measurements were conducted in 50 K increments. The intensity ratio was calculated from discrete peak values of two emission bands, a temperature dependent band around 458.5 nm due to the 4 I15/2 →
185
6
H15/2 transition and a second emission band centering at 484 nm correspond-
ing to the 4 F9/2 → 6 H15/2 transition [1]. For each of these transitions YAG:Dy shows several sharp emission lines, see also Fig. 6. At room temperature the emission bands originating from the considered levels are nearly overlapping and the 4 I15/2 emission is very weak. Increasing temperature leads to a signif190
icant increase in emission intensity, which indicates the 4 I15/2 level population thermal growth. In order to observe the strongest ratio, in the following the peaks at 458.5 nm and 497 nm were compared, with the latter emission peak 10
also belonging to the 4 F9/2 → 6 H15/2 transition. This intensity ratio is often used for thermometry [5, 15] and shows a more temperature sensitive behav195
ior in contrast to a calculation based on broader emission bands [8], although these are used in practical two-dimensional thermometry application. In this work only the temperature dependent behavior of the intensity ratio which is calculated from peak values is analyzed for dierent co-doped thermographic phosphors. This procedure is reasonable for the assessment of the general lu-
200
minescence characteristics of the dierent samples. Additionally, the spectral emission was studied in the wavelength range up to 700 nm as YAG:Dy shows another characteristic emission band at 571 nm due to the 4 F9/2 → 6 H13/2 transition [1].
3.3. Decay time evaluation 205
The decay curves for the 4 F9/2 to 6 H15/2 and 4 I15/2 to 6 H15/2 transition of singly doped Dy3+ have a multi-exponential character. Thus for the evaluation of the temperature dependent decay time, the decay curves have been tted to the following bi-exponential function [47], which eectively models the decay behavior of YAG:Dy:
I(t) = A1 exp(−t/τ1 ) + A2 exp(−t/τ2 ), 210
(1)
where τ1 and τ2 are the short and long decay decay times, respectively and
A1 and A2 their respective weight factors. Based on the equation above, the average decay time τ is calculated from [47]
τ=
A1 τ12 + A2 τ22 . A1 τ1 + A2 τ2
(2)
4. Results and discussion
4.1. Spectral analysis 215
Figure 5 shows the excitation spectra for YAG:Dy(2%), YAG:Dy(2%),Tm(1%), YAG:Dy(2%),Tb(1%) and YAG:Dy(2%),Pr(1%), monitoring only the 583 nm emission of Dy3+ . Sharp excitation peaks from 290 nm to 400 nm correspond 11
5/2
, P
YAG:Dy(2%)
11/2
6
YAG:Dy(2%),Tm(1%)
4
I
YAG:Dy(2%),Tb(1%)
H
15/2
YAG:Dy(2%),Pr(1%)
,
15/2
6
H
15/2
4
I
13/2
4
F
7/2
6
15/2
M 4 6
H
15/2
15/2
H 4000
6
F
6
6
H
H
15/2
6
15/2
H
4
15/2
G 4
9/2
4
I
2000 11/2
PL intensity (a.u.)
4
M
17/2
6
, P
7/2
6000
0 250
300
350
400
450
500
W avelength / nm
Figure 5: Excitation spectra of YAG:Dy(2%), YAG:Dy(2%),Tm(1%), YAG:Dy(2%),Tb(1%) and YAG:Dy(2%),Pr(1%) for 583 nm emission. Only the energy levels of Dysprosium are shown.
to the transitions from the ground state to the excited states of the 4f con gurations [48, 49]. All phosphors show the strongest band at 352 nm, which is 220
close to the third harmonic of Nd:YAG lasers. Compared to the singly doped YAG:Dy(2%), PLE intensity increased for Tb3+ and Pr3+ codopings, whereas we have decreased emission for Tm3+ codoping. The Pr3+ codoping showed a broad absorption band between 250 nm and 300 nm related to the 3 H4 → 4 F5D transition of Pr3+ [50, 51]. For the Tb3+ codoping a very pronounced peak ap-
225
peared at 324 nm, which is the most intense spin-forbidden 4f → 5d transition of terbium [52]. The emission spectra at room temperature for an excitation wavelength of 355 nm are given in Fig. 6 for YAG:Dy(2%), YAG:Dy(2%),Tm(1%), YAG:Dy(2%),Tb(1%) and YAG:Dy(2%),Pr(1%). Only the emission in the blue region due to the 4 F9/2
230
→ 6 H15/2 transition is shown, which is relevant for temperature measurements. The spectra are normalized to the 497 nm peak of YAG:Dy(2%). The overall spectral emission characteristics were similar for all samples. In fact, codoping by Pr3+ led to distinctly stronger luminescence intensities compared to the
12
2.5
YAG:Dy(2%) YAG:Dy(2%),Tm(1%)
PL intensity (a.u.)
2.0
YAG:Dy(2%),Tb(1%) YAG:Dy(2%),Pr(1%)
1.5
1.0
0.5
0.0 450
460
470
480
490
500
510
W avelength / nm
Figure 6: Emission spectra due to the 4 F9/2 → 6 H15/2 transition of YAG:Dy(2%), YAG:Dy(2%),Tm(1%), YAG:Dy(2%),Tb(1%) and YAG:Dy(2%),Pr(1%) for 355 nm excitation. Spectra are normalized to the 497 nm peak of YAG:Dy(2%).
reference phosphor YAG:Dy(2%). Codoping with Tb3+ slightly increased the 235
signal intensity, whereas codoping by Tm3+ resulted in reduced signal intensity. The normalized emission spectrum of YAG:Dy(2%) for the measured temperature range is given in Fig. 7. It shows the emission centering at 484 nm due to the 4 F9/2 → 6 H15/2 transition as well as a strongly temperature dependent
240
peak around 458.5 nm due to the 4 I15/2 transition (see Fig. 7). With increasing temperature the ratio between the emission in the blue at 484 nm and the 4 F9/2
→ 6 H13/2 transition at 571 nm became slightly more pronounced towards the yellow. The absolute emission intensities of YAG:Dy(2%) over the measured temper245
ature range showed the expected drop in signal intensity for higher temperatures (see Fig. 8). At 1400 K the signal of the whole emission band around 484 nm decreased by about 50% compared to the signal intensities at 300 K for the 4
F9/2 → 6 H15/2 transition used for intensity ratio measurements. For both the
4
F9/2 → 6 H15/2 at 484 nm as well as the 4 F9/2 →6 H13/2 transition at 571 nm a 13
2.5
300 K
Normalized signal intensity / -
4
F
6 9/2
H
600 K 13/2
900 K
2.0
1200 K 1500 K
1.5
4
1.0 4
I
15/2
6
H
F
6 9/2
H
15/2
15/2
0.5 4
F
6 9/2
H
11/2
0.0 400
450
500
550
600
650
700
W avelength / nm
Figure 7: Normalized spectra for the reference phosphor YAG:Dy(2%). Spectra are normalized to the peak at 497 nm. 250
broadening of the lines with increasing temperature was visible. A comparison of the absolute signal intensities of all tested samples between 430 nm and 530 nm is presented in Fig. 9. For all phosphor samples the signal intensities started dropping around 900 K to 1000 K due to thermal quenching and other non-radiative processes. Both samples with Tm3+ codoping showed
255
decreased signal intensities by up to 40% as compared to YAG:Dy(2%). As expected, a higher amount of Dy3+ led to higher signal intensities in accordance with [15], however, the absolute emission of YAG:Dy(3%),Tm(1%) was still below that of YAG:Dy(2%). Co-doping with Tb3+ did not yield any improvements at room temperature, but higher signal intensities were measured above 900 K
260
as compared to YAG:Dy(2%). The Pr3+ codopings showed strongly increased signal intensities up to 1000 K. This was followed by a decline in absolute emission, so that after 1200 K the dierences in signal intensity of YAG:Dy(2%) and YAG:Dy(2%),Pr(1%) were only marginal. The intensity ratio for the dierent codopings is given in Figure 10. All
265
samples showed a similar, increasing trend with temperature. Regarding the
14
Absolute Intensities [430-530nm] (a.u)
Figure 8: Absolute signal intensities for the reference phosphor YAG:Dy(2%) from 300 K to 1500 K
YAG:Dy(2%) YAG:Dy(2%),Pr(0.5%)
80
YAG:Dy(2%),Pr(1%) YAG:Dy(2%),Tb(1%) YAG:Dy(2%),Tm(1%) YAG:Dy(3%),Tm(1%)
60
40
20
0 400
600
800
1000
1200
1400
1600
Temperature /K
Figure 9: Absolute signal intensities for the 4 F9/2 → 6 H15/2 and 4 I15/2 → 6 H15/2 transition of dierent phosphor samples from 300 K to 1600 K
15
0.8
Intensity ratio [458.5 nm / 497 nm]
YAG:Dy(2%) YAG:Dy(2%),Pr(0.5%) YAG:Dy(2%),Pr(1%) 0.6
YAG:Dy(2%),Tb(1%) YAG:Dy(2%),Tm(1%) YAG:Dy(3%),Tm(1%)
0.4
0.2
0.0 200
400
600
800
1000
1200
1400
1600
Temperature / K
Figure 10: Temperature calibration curve for the dierent codoped samples. The intensity ratio is calculated based on the ratio between the 4 I15/2 → 6 H15/2 and 4 F9/2 → 6 H15/2 transition.
Tm3+ codopings, the increase of the intensity ratio occurred already at lower temperatures. In contrast, the curves for the Tb3+ and Pr3+ codopings lie slightly below the reference YAG:Dy.
YAG:Dy(2%),Tm(1%) and YAG:Dy(3%),Tm(1%) 270
The main emission peaks of Tm3+ lie at 460 nm (1 D2 → 3 F4 ) and 480 nm (7 G4 → 3 H6 ) [1].
Regarding the spectra of the codoped sample the tem-
perature dependent emission peak at 458.5 nm was visible at lower temperatures compared to YAG:Dy(2%), while the peak at 484 nm of Dy3+ was less pronounced (see Fig. 11). This suggests a possible energy transfer from 275
Dy3+ to Tm3+ at energies around 2.55 eV. The normalized emission spectra of YAG:Dy(3%),Tm(1%) exhibited almost no dierence compared to the YAG:Dy(2%),Tm(1%) sample (not shown here). It is noteworthy that up to 1000 K the absolute luminescence signal levels were higher for YAG:Dy(3%),Tm(1%) compared to YAG:Dy(2%),Tm(1%) by about fty percent, see Fig. 9. This can
280
be attributed to the growing probability of energy transfer processes leading to 16
2.5
Normalized signal intensity / -
300 K 600 K 900 K
2.0
1200 K 1500 K
1.5 480 nm
460 nm 1.0
0.5
0.0 400
450
500
550
600
650
700
W avelength / nm
Figure 11: Normalized spectra for YAG:Dy(2%),Tm(1%) at temperatures from 300 K to 1500 K. Spectra are normalized to the peak at 497 nm.
luminescence for higher dopant concentrations. This eect, however, is limited by an increasing probability of a non-radiative energy transfer between excited dopants at too high dopant concentrations known as concentration quenching [4]. Above 1000 K signal levels of both Tm3+ -doped samples converged, which 285
indicates an increase in non-radiative cross-relaxation between two elements and concentration quenching. Dierences in the decay times of the 4 F9/2 ("497 nm") and 4 I15/2 ("458 nm") transition appeared for the YAG:Dy samples codoped with Tm3+ . Comparing the YAG:Dy(2%),Tm(1%) to the YAG:Dy(2%), the 4 F9/2 transition showed
290
similar decay times. However, higher doping levels of Dy3+ in YAG:Dy(3%),Tm(1%) led to lower decay times of the 4 F9/2 transition by about 15%. Similar observations regarding shorter decay times with increasing Dy3+ concentrations were made by Jovicic et al. [15]. In contrast, for the 4 I15/2 transition codoping YAG:Dy with Tm3+ resulted in a decrease in decay time by 20% for both sam-
295
ples (see Figure 12). However, comparing the absolute signal intensities of the Tm3+ codoping to the YAG:Dy(2%) sample, the overall signal intensity was up
17
0.8
0.7
Decay time / ms
0.6
0.5
0.4
YAG:Dy(2%) [458nm]
0.3
YAG:Dy(2%) [484nm] YAG:Dy(2%),Tm(1%) [458nm]
0.2
YAG:Dy(2%),Tm(1%) [484nm] YAG:Dy(3%),Tm(1%) [458nm]
0.1
YAG:Dy(3%),Tm(1%) [484nm] 0.0 200
400
600
800
1000
1200
1400
1600
1800
Temperature / K
Figure 12: Temperature dependent decay time characteristics of YAG:Dy(2%),Tm(1%) and YAG:Dy(3%),Tm(1%) compared to YAG:Dy(2%)
to 40% lower. This suggests a possible energy transfer from Dy3+ to Tm3+ with subsequent energy quenching. Certainly it makes the phosphor uninteresting for further use in high temperature thermography. 300
YAG:Dy(2%),Tb(1%) Single doped YAG:Tb has photoluminescence peaks at 490 nm, 544 nm, 585 nm and 619 nm assigned to the 5 D4 → 7 Fj (j=6,5,4,3) transition of the terbium ion [53, 1]. Comparing the emission spectrum of YAG:Dy(2%),Tb(1%) to that of the uncodoped YAG:Dy sample, an additional peak at 544 nm appeared
305
which corresponds to the 5 D4 → 7 F5 transition of terbium (see Figure 13). The emission line was most prominent at room temperature, but decreased with increasing temperatures. Small changes in the spectral behavior were observable around 490 nm compared to YAG:Dy. Neither the transition at 490 nm nor at 544 nm (due to energy transfers) did signi cantly change the intensity ratio
310
and thus, the temperature sensitivity of the phosphor, see Fig. 10. Absolute integrated signal intensities of YAG:Dy(2%),Tb(1%) were comparable to that of the singly doped sample in a wide temperature range, see Fig. 9. 18
2.5
Normalized signal intensity / -
585 nm
300 K 600 K 900 K
2.0
1200 K 1500 K
1.5 490 nm
544 nm 1.0
619 nm
0.5
0.0 400
450
500
550
600
650
700
W avelength / nm
Figure 13: Normalized spectrafor YAG:Dy(2%),Tb(1%) at temperatures from 300 K to 1500 K. Spectra are normalized to the peak at 497 nm.
Evaluating the decay time of the codoped samples with regard to the reference sample of YAG:Dy(2%), YAG:Dy(2%),Tb(1%) showed a 10-15% lower 315
decay time for 4 I15/2 emission (Fig. 14). This is probably due to energy transfer from Dy3+ to Tb3+ . The luminescence decay time of the 4 F9/2 transition had a similar duration at room temperature compared to the singly doped sample but started to decrease earlier at 1000 K. Thus, the terbium codoping showed similar spectral intensities despite lower decay times in comparison to the reference
320
sample.
YAG:Dy(2%),Pr(1%) and YAG:Dy(2%),Pr(0.5%) The emission spectra of YAG:Dy(2%),Pr(1%) are given in Figure 15. The 1
D2 → 1 G4 transition of Pr3+ resulting in an emission peak at 607 nm was
visible for the YAG:Dy(2%),Pr(1%) and became less signi cant for the lower 325
Pr3+ concentration (not shown here). At room temperature both Pr-codopings had a higher signal intensity than the singly doped sample, see Fig. 9. Absolute intensities increased by about 30% for both the 1% Pr and the 0.5% Pr sample. However, absolute signal intensities dropped more strongly for both codopings 19
0.8
0.7
Decay time / ms
0.6
0.5
0.4
0.3
YAG:Dy(2%) [458nm]
0.2
YAG:Dy(2%) [484nm] YAG:Dy(2%),Tb(1%) [458nm]
0.1
YAG:Dy(2%),Tb(1%) [484nm] 0.0 200
400
600
800
1000
1200
1400
1600
1800
Temperature / K
Figure 14: Temperature dependent decay time characteristics of YAG:Dy(2%),Tb(1%) compared to YAG:Dy(2%)
compared to YAG:Dy at higher temperatures. Thus above 1200 K the dier330
ence in luminescence intensity between YAG:Dy(2%),Pr(1%) and YAG:Dy(2%) became marginal. Codoping with Pr3+ led to a decay time decrease by 15% compared to the single doped YAG:Dy(2%), see Figure 16. Regarding the overall temperature behavior similar curve shapes became apparent for both the codoped phosphor
335
as well as the single doped reference sample. Small dierences could be observed between the two Pr3+ codopings of 1% and 0.5%, with marginally lower decay times for the YAG:Dy(2%),Pr(1%). Based on these measurements, codoping YAG:Dy with Tm3+ is not ex-
340
pedient due to the strong decrease in signal intensity. Conversely, codopings with Pr3+ and Tb3+ showed promising luminescent characteristics, both having higher signal intensities at decreased overall decay times. Out of these two lanthanides the Pr3+ -codoping had the stronger eect.
20
2.5
Normalized signal intensity / -
300 K 600 K 900 K
2.0
1200 K 1500 K
1.5 485 nm
1.0
607 nm
0.5
0.0 400
450
500
550
600
650
700
W avelength / nm
Figure 15: Normalized spectra for YAG:Dy(2%),Pr(1%) at temperatures from 300 K to 1500 K. Spectra are normalized to the peak at 497 nm.
0.8
0.7
Decay time / ms
0.6
0.5
0.4
YAG:Dy(2%) [458nm]
0.3
YAG:Dy(2%) [484nm] YAG:Dy(2%),Pr(0.5%) [458nm]
0.2
YAG:Dy(2%),Pr(0.5%) [484nm] YAG:Dy(2%),Pr(1%) [458nm]
0.1
YAG:Dy(2%),Pr(1%) [484nm] 0.0 200
400
600
800
1000
1200
1400
1600
1800
Temperature / K
Figure 16: Temperature dependent decay time characteristics of YAG:Dy(2%),Pr(1%) and YAG:Dy(2%),Pr(0.5%) compared to YAG:Dy(2%)
21
5. Conclusion and outlook
345
The eects of codoping on the temperature-dependent luminescence of YAG:Dy were investigated regarding the use of these materials in high temperature thermometry. Six luminescent ceramic µm powder materials co-doped with Tm3+ , Tb3+ and Pr3+ were synthesized and characterized. YAG:Dy was chosen as a reference since this phosphor has up to date the highest temperature mea-
350
surement capabilities. Spectral emission behavior and temperature dependent decay time were presented up to 1600 K. Codoping by Tm3+ decreased signal intensity considerably, thus this combination is not suitable for phosphor thermometry. In contrast, the intensity ratio remained almost unchanged and the absolute signal intensity increased by up to 30% for codopings with Pr3+
355
and Tb3+ , while decay times decreased by 10 to 20 %. Thus, for thermometry codoping YAG:Dy(2%) with Pr3+ and Tb3+ enhanced the luminescent characteristics considerably, especially for detection at shorter exposure times. This work clearly indicated that it is possible to improve high temperature phosphor materials through codoping. Additional studies on this topic are necessary to
360
further reduce decay time and increase absolute signal intensities at high temperatures. The addition of other sensitizers or modi cation of the host matrix could lead to improved signals at high temperatures, which is the focus of future work.
Acknowledgements
365
The authors gratefully acknowledge nancial support of the DFG (Deutsche Forschungsgemeinschaft). We would like to thank Dr. Mykhailo Sytnyk for SEM measurements and Prof. Stefan Will for helpful discussions.
References
[1] M. Chambers, D. Clarke, Annual Review of Materials Research 39 (2009) 370
325359.
22
[2] B. Fond, C. Abram, A. L. Heyes, A. M. Kempf, F. Beyrau, Optics Express 20 (2012) 2211822133. [3] J. Linden, N. Takada, B. Johansson, M. Richter, M. Alden, Applied Physics B 96 (2009) 237240. doi:10.1007/s00340-009-3608-7. 375
[4] S. W. Allison, G. T. Gillies, Review of Scienti c Instruments 68 (1997) 26152650. doi:10.1063/1.1148174. [5] M. Aldén, A. Omrane, M. Richter, G. Särner, Progress in Energy and Combustion Science 37 (2011) 422461. doi:10.1016/j.pecs.2010.07.001. [6] J. Brübach, C. P itsch, A. Dreizler, B. Atakan, Progress in Energy and
380
Combustion Science 39 (2013) 3760. doi:10.1016/j.pecs.2012.06.001. [7] J. Brübach, T. Kissel, M. Frotscher, M. Euler, B. Albert, A. Dreizler, Journal of Luminescence 131 (2011) 559564. doi:10.1016/j.jlumin.2010.10.
017. [8] G. Jovicic, L. Zigan, S. Will, A. Leipertz, Measurement Science and Tech385
nology 26 (2014) 015204. doi:10.1088/0957-0233/26/1/015204. [9] A. Heyes, S. Seefeldt, J. Feist, Optics & Laser Technology 38 (2006) 257 265. doi:10.1016/j.optlastec.2005.06.012. [10] S. Skinner, J. Feist, I. Brooks, S. Seefeldt, A. Heyes, Sensors and Actuators B: Chemical 136 (2009) 5259. doi:10.1016/j.snb.2008.10.070.
390
[11] A. Hashemi, A. Vetter, G. Jovicic, M. Batentschuk, C. J. Brabec, Measurement Science and Technology 26 (2015) 17. doi:10.1088/0957-0233/26/
7/075202. [12] J. Jordan, D. A. Rothamer, Applied Physics B 110 (2012) 285291. doi:10.
1007/s00340-012-5274-4. 395
[13] D. Witkowski, D. A. Rothamer, A. Lynch, V. Belovich, Measurements (2013). 23
[14] N. J. Neal, J. Jordan, D. Rothamer, SAE International Journal of Engines 6 (2013) 300318. doi:10.4271/2013-01-0562. [15] G. Jovicic, L. Zigan, S. Will, A. Leipertz, Zeitschrift für Physikalische 400
Chemie 229 (2015) 977997. doi:10.1515/zpch-2014-0603. [16] M. Cates, S. Allison, S. Jaiswal, D. Beshears, YAG: Dy and YAG: Tm Fluorescence Above 1400 C, United States. Department of Energy, 2003. [17] A. H. Khalid, K. Kontis, H.-Z. Behtash, Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering 224
405
(2010) 745755. doi:10.1243/09544100jaero560. [18] N. Ishiwada, T. Ueda, T. Yokomori, Luminescence 26 (2010) 381389. doi:10.1002/bio.1237. [19] J. I. Eldridge, T. J. Bencic, D. Zhu, M. D. Cuy, D. E. Wolfe, S. W. Allison, D. L. Beshears, T. P. Jenkins, B. Heeg, R. P. Howard, et al., Implementa-
410
tion of an Ultra-Bright Thermographic Phosphor for Gas Turbine Engine Temperature Measurements, Technical Report, NASA Technical Report, 2014. [20] B. Fond, C. Abram, F. Beyrau, Applied Physics B 118 (2015) 393399. doi:10.1007/s00340-014-5997-5.
415
[21] G. Särner, M. Richter, M. Aldén, Optics Letters 33 (2008) 13271329. [22] A. Heyes, Journal of Luminescence 129 (2009) 20042009. doi:10.1016/j.
jlumin.2009.03.041. [23] S. Geller, Zeitschrift für Kristallographie-Crystalline Materials 125 (1967) 147. 420
[24] G. Blasse, B. Grabmaier, Luminescent materials, Springer Verlag Berlin, 1994. [25] G. H. Dieke, H. M. Crosswhite, Applied Optics 2 (1963) 675686. 24
[26] J. Caslavsky, D. Viechnicki, Journal of Materials Science 15 (1980) 1709 1718. 425
[27] A. Jaber, L. Zigan, A. Sakhrieh, A. Leipertz, Measurement Science and Technology 24 (2013) 075202. doi:10.1088/0957-0233/24/7/075202. [28] M. Yu, G. Särner, C. C. M. Luijten, M. Richter, M. Aldén, R. S. G. Baert, L. P. H. de Goey, Measurement Science and Technology 21 (2010) 037002. [29] W. Y. Kwong, A. Steinberg, Y. H. Chin, Optics Letters 39 (2014) 6166
430
6169. doi:10.1364/ol.39.006166. [30] L. M. Chepyga, G. Jovicic, A. Vetter, A. Osvet, C. J. Brabec, M.
Batentschuk,
Applied
Physics
B
122
(2016).
doi:10.1007/
s00340-016-6487-8. [31] J. Li, J.-G. Li, S. Liu, X. Li, X. Sun, Y. Sakka, Journal of Materials Chem435
istry C 1 (2013) 76147622. doi:10.1039/c3tc31413h. [32] Z. Mu, Y. Hu, L. Chen, X. Wang, Journal of Luminescence 131 (2011) 16871691. doi:10.1016/j.jlumin.2011.03.072. [33] F. Gu, S. F. Wang, M. K. Lü, G. J. Zhou, D. Xu, D. R. Yuan, Langmuir 20 (2004) 35283531. doi:10.1021/la049874f.
440
[34] L. Beauzamy, B. Moine, P. Gredin, Journal of Luminescence 127 (2007) 568574. doi:10.1016/j.jlumin.2007.03.007. [35] V. Bedekar, D. P. Dutta, M. Mohapatra, S. V. Godbole, R. Ghildiyal, A. K. Tyagi, Nanotechnology 20 (2009) 125707. doi:10.1088/0957-4484/20/12/
125707. 445
[36] S. Som, S. K. Sharma, T. Shripathi, Journal of Fluorescence 23 (2013) 439450. doi:10.1007/s10895-013-1160-7. [37] X. Wan, Y.-Q. Lin, S.-L. Tie, J.-Y. Shen, Journal of Non-Crystalline Solids 357 (2011) 34243429. doi:10.1016/j.jnoncrysol.2011.06.011. 25
[38] H. Lin, E. Y.-B. Pun, X. Wang, X. Liu, Journal of Alloys and Compounds 450
390 (2005) 197201. doi:10.1016/j.jallcom.2004.07.068. [39] K. Sobha, K. Rao, Journal of physics and chemistry of solids 57 (1996) 12631267. [40] X.-Y. Sun, M. Gu, S.-M. Huang, X.-L. Liu, B. Liu, C. Ni, Physica B: Condensed Matter 404 (2009) 111114. doi:10.1016/j.physb.2008.10.
455
039. [41] X. Li, M. Dong, F. Hu, Y. Qin, L. Zhao, X. Wei, Y. Chen, C. Duan, M. Yin, Ceramics International 42 (2016) 60946099. doi:10.1016/j.ceramint.
2015.12.167. [42] T. Sasikala, L. Rama Moorthy, A. Mohan Babu, T. Srinivasa Rao, Journal 460
of Solid State Chemistry 203 (2013) 5559. doi:10.1016/j.jssc.2013.03.
058. [43] Z. Yang, L. Luo, W. Chen, Journal of the Optical Society of America B 23 (2006) 358362. [44] W. Wang, J. Wei, H. Hong, F. Xuan, Y. Shan, Journal of Alloys and 465
Compounds 584 (2014) 136141. doi:10.1016/j.jallcom.2013.08.046. [45] J. Brübach, J. Feist, A. Dreizler, Measurement Science and Technology 19 (2008) 559564. [46] A. B. Muñoz-García, E. Artacho, L. Seijo, Physical Review B 80 (2009) 014105. doi:10.1103/physrevb.80.014105.
470
[47] J. R. Lakowicz, Principles of uorescence spectroscopy, Springer Science & Business Media, 2013. [48] L. Wu, Y. Zhang, M. Gui, P. Lu, L. Zhao, S. Tian, Y. Kong, J. Xu, Journal of Materials Chemistry 22 (2012) 6463. doi:10.1039/c2jm15506k.
26
[49] M. Klimczak, M. Malinowski, J. Sarnecki, R. Piramidowicz, Journal of Lu475
minescence 129 (2009) 18691873. doi:10.1016/j.jlumin.2009.04.073. [50] M. Wisniewska, A. J. Wojtowicz, T. Lukasiewicz, Z. Frukacz, Z. Galazka, M. Malinowski, Radio- and VUV - excited luminescence of YAP:Cr, YAP:Pr and YAG:Pr, Technical Report, DTIC Document, 2001. [51] S. Zhou, Z. Fu, J. Zhang, S. Zhang, Journal of Luminescence 118 (2006)
480
179185. doi:10.1016/j.jlumin.2005.08.011. [52] P. Dorenbos, Journal of Luminescence 91 (2000) 91106. doi:10.1016/
S0022-2313(00)00197-6. [53] H. Aizawa, T. Katsumata, S. Komuro, T. Morikawa, H. Ishizawa, E. Toba, Sensors and Actuators A: Physical 126 (2006) 7882. doi:10.1016/j.sna. 485
2005.09.021.
27
