- Email: [email protected]
Ratiometric luminescence thermometry with different combinations of emissions from Eu3+ doped Gd2Ti2O7 nanoparticles
Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Contents lists available at ScienceDirect
Journal of Luminescence journal homepage: www.elsevier.com/locate/jlumin
Ratiometric luminescence thermometry with different combinations of emissions from Eu3 þ doped Gd2Ti2O7 nanoparticles Vesna Lojpur, Sanja Ćulubrk, Miroslav D. Dramićanin n University of Belgrade, Vinča Institute of Nuclear Sciences, P.O. Box 522, 11001 Belgrade, Serbia
art ic l e i nf o
a b s t r a c t
Article history: Received 28 October 2014 Received in revised form 22 December 2014 Accepted 9 January 2015
Herein, Eu3 þ doped Gd2Ti2O7 nanoparticles were tested for application in ratiometric luminescence thermometry. It is shown that two combinations of emissions: one that uses two emissions of Eu3 þ ions and one that uses one emission of Eu3 þ ions and trap emission of Gd2Ti2O7 provide thermometry over the 303–423 K temperature range with relative sensitivities between 0.14% K 1 and 0.95% K 1. Thermometry based on two Eu3 þ emissions from 5D0 to 5D1 levels has a higher relative sensitivity, but lower absolute sensitivity than thermometry based on one Eu3 þ emission and trap emission of Gd2Ti2O7. The tested material is prepared by Pechini-type polymerized complex route and is composed of agglomerated nanoparticles of 30–50 nm in size with pure-phase cubic structure (space group Fd3m) as evidenced from electron microscopy and X-ray diffraction measurements. & 2015 Elsevier B.V. All rights reserved.
Keywords: Thermometry Luminescence Pyrochlores Rare earths Nanoparticles
1. Introduction Luminescence thermometry is most commonly performed through relative emission intensity measurements (ratiometric intensity measurements; fluorescence intensity ratio—FIR) [1–5]. This type of measurements is insensitive to fluctuations of excitation light or other changes in measurement conditions, and more importantly, it is self-referencing (i.e., these types of measurements do not have to refer to any temperature standard) [5]. Ideally, one of the used emissions should be independent of temperature (internal reference) and, then, a calibration between the ratios of emissions is indicative of temperature. When used with rare earth ion-doped materials, this technique involves the use of emissions that originate from two closely spaced, “thermally coupled” excited energy levels of the rare earth ions. The relative population of these levels follows a Boltzmanntype distribution and it is dependent on the temperature and the energy difference between levels (energy gap) [2,6]. The main mechanism behind “thermally coupled” energy levels is thermalization: when two energy levels of the RE activator are closely separated by a difference of 2000 cm 1 or less, the upper level will not fluoresce at low temperatures since electrons do not have enough energy to bridge the energy gap. As the temperature increases, the upper level becomes populated and hence the emission from this level gradually increases in intensity at the expense of the lower level population.
n
Corresponding author. E-mail address: [email protected] (M.D. Dramićanin).
Recently, the new concept of ratiometric luminescence thermometry that exploits trap emission of the host as a self-referencing standard and the emission of the activator ions as temperature indicator has been demonstrated using TiO2 nanopowders doped with Eu3 þ ions [7] and Sm3 þ ions [8], and Zn2SiO4 doped with Mn2 þ ions [9]. This concept provided high relative sensitivities and the possibility to perform luminescence thermometry with activator ions which have only single intense emission band (as in the case of Mn2 þ :Zn2SiO4 [9]). Here, we aimed to compare performance of two above mentioned concepts of ratiometric luminescence thermometry with Eu3 þ doped Gd2Ti2O7 nanoparticles. Gd2Ti2O7 exhibit notable luminescence after incorporation of rare earth ions [10–16]. In this cubic-type host rare-earth dopants replace Gd3 þ ions in crystallographic sites with a strict center of symmetry (D3d) [16]. The emission spectrum of Eu3 þ doped Gd2Ti2O7 consists of emissions from Eu3 þ spin-forbidden f–f electronic transitions [15] and trap emission of Gd2Ti2O7 induced by the presence of oxygen vacancies [17]. Therefore, both ratiometric intensity concepts are feasible with this material.
2. Experimental 2.1. Material synthesis procedure Gd2Ti2O7:5 at% Eu3 þ nanopowder was prepared by Pechinitype polymerized complex route, described in detail in our previous work [15]. In brief, this synthesis method is based on polyesterification between citric acid (CA) and ethylene glycol. For
http://dx.doi.org/10.1016/j.jlumin.2015.01.027 0022-2313/& 2015 Elsevier B.V. All rights reserved.
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
synthesis of Gd2Ti2O7:5 at% Eu3 þ nanopowder, titanium (IV)isopropoxide, gadolinium(III)-nitrate, citric acid and ethylene glycol were mixed in 1:1:5:20 M ratio. First, titanium (IV)-isopropoxide (Alfa Aesar, 97%) was dissolved in ethylene glycol (Lach-Ner, 99%) under constant magnetic stirring. Then citric acid (Kemika, 99.5%) was added to the solutions and stirred until complete dissolution was achieved. After that, appropriate amounts of Gd2O3 (Alfa Aesar, 99.9%) and Eu2O3 (Alfa Aesar, 99.9%) were dissolved in hot, concentrated nitric acid, evaporated to dryness and joined with titanium (IV)-isopropoxide/EG/CA mixture. Mixtures were stirred for 1 h at 60 1C until they became transparent and further heated at 130 1C for a few hours to promote polymerization. Black, amorphous resins were fired at 350 1C for 30 min. At the end, pure phase of Gd2Ti2O7:5 at% Eu3 þ nanopowder was obtained after annealing at 880 1C for 4 h and naturally furnace cooling to the room temperature.
Fig. 1. Experimental setup for temperature dependent photoluminescence measurement.
2.2. Instruments and measurements Phase composition of the sample was checked with X-ray powder diffraction measurement performed on a Rigaku Smartlab diffractometer. Diffraction data were collected in the 2θ range from 101 to 901, counting 0.71/min in 0.02 steps. The morphological features and chemical purity were investigated by means of scanning electron microscopy (SEM JEOL, JSM-6610LV) with INCA energy dispersive X-ray analysis. For this purpose the powder sample was dispersed on Cu holder and sputtered with Au. Transmission electron microscopy (TEM) is conducted using JOEL-JEM 2100 instrument equipped with LaB6 cathode and operated with 200 kV. For TEM measurements powders were used without any additional preparation. Photoluminescence spectra were collected using a Fluorolog-3 Model FL3-221 spectrofluorometer system (Horiba Jobin-Yvon), over the temperature range from 293 to 433 K. The photoluminescence measurements were performed under continuous excitation from a 450 W Xenon lamp at a wavelength of 393 nm. The samples were placed in a custommade temperature controlled furnace, and emission spectra were collected via an optical fiber bundle. The temperature of the samples was controlled within the accuracy of 70.5 1C by a temperature control system utilizing proportional-integral-derivative feedback loop equipped with T-type thermocouple for temperature monitoring. Experimental setup shown in Fig. 1, consists of the following parts: 1. Lamp, 2. Detector, 3. Monochromator, 4. Optical fiber, 5. Lens, 6. Sample, 7. Oven, 8. T-regulator, 9. Controller and 10. PC.
3. Results and discussion Pyrochlores with formula A2B2O7 crystallize into a face-centered cubic lattice with space group Fd-3m, No 227. Unit cell contains eight molecules (Z¼8) and four crystallographically nonequivalent sites. The A cation (usually of ionic radius 1 Å) is eight-coordinated and located within scalenohedra (distorted cubes) that contain six equally spaced O atoms at a slightly shorter distance from the central cation. The smaller B cation (ionic radius 0.6 Å) is sixcoordinated and located within a trigonal antiprism with all six anions at equal distances from the central cation. In the pyrochlore structure, D3d symmetry of the metal cation point group requires that these two polyhedra are neither octahedra nor cubes (but trigonal antiprisms and scalenohedra), even though many authors do refer to them as octahedral and cubic coordination polyhedra [15]. XRD patterns of Gd2Ti2O7:5 at% Eu3 þ powder are presented in the Fig. 2. The main diffraction peaks are indexed according to the ICCD card No. 01-074-9640. There were no peaks related to any other phases indicating sample without impurities. High intensity is a consequence of high crystallinity of the powder.
Fig. 2. XRD diffraction patterns for Gd2Ti2O7:5 at% Eu3 þ powder indexed according to the ICCD card No. 01-074-9640.
Morphology of the sample obtained through Pechini-type polymerized complex route was showed in Fig. 3. The micrograph provided by SEM analysis, Fig. 3a, indicated that Gd2Ti2O7:5 at% Eu3þ consists of large chunks with a size of several tens of microns. The elements distribution determined by the energy dispersive X-ray mapping analysis, examined chemical homogeneity of the powder as shown in Fig. 3b. The elemental maps of Gd, TiO and Eu confirmed uniform distribution of all elements throughout the material. The EDX spectrum showed in Fig. 3c revealed the presence of Eu3 þ ions and the purity of rare earth doped Gd2Ti2O7. Further magnification was followed by TEM measurements. The TEM micrograph is presented in Fig. 3d and revealed that chunks are entirely composed of nanoparticles. It is evident that particles are agglomerated and have irregular, round and rectangular shapes. Size of nanoparticles is mostly around 30–50 nm. Luminescence spectra of Gd2Ti2O7:5 at% Eu3 þ powders measured as a function of temperature (293–423 K) are shown in the Fig. 4. Two distinct spectral regions can be observed: high energy broad band spectral region (400–550 nm) that belongs to the trap emission of the Gd2Ti2O7 host and the low energy spectral region that belongs to emission of Eu3 þ ions. Emission of dopant is composed of distinctive bands peaking at 589 nm, 597 nm (5D0-7F1) 611 nm and 627 nm (5D0-7F2) that arise from spin-forbidden electron transition. In addition, small intensity emission from 5D1-7F1 transition can be observed at about 535 nm. It is clearly seen from the Fig. 4that dopant emission is extremely sensitive to temperature, showing a rapidly decreasing intensity with the temperature increase while changes of trap emission are minimal with the change of temperature in the measured range. Small changes in trap emission intensities are a consequence of slight changes in optical properties of the host with temperature such as band gap energy, reflectivity and changes of
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
3
Fig. 3. (a) SEM image of Gd2Ti2O7:5 at% Eu3 þ powder (b) EDX mapping of elements distribution inside the particles (c) EDX spectrum (d) TEM micrograph of Gd2Ti2O7:5 at% Eu3 þ powder.
emissions from the 5D1 level (I1) and 5D0 level (I0) to the ground level is described by the following equation: FIR1 ¼
Fig. 4. Photoluminescence spectra of 5.0 at% Eu3 þ doped Gd2Ti2O7 powder over the temperature range of 293–563 K (temperature increment was 20 K).
experimental parameters during the measurements. On the other hand, the sharp decline of dopant emission is due to the non-radiative relaxation processes that increase with increasing of temperature. Changes of emission spectral shape with temperature provide measures suitable for luminescence thermometry. The use of intensity ratio between two emission lines (FIR) in the emission spectrum is a self-referencing thermometry method that gained a significant attention in recent times [1–6]. In the present case two combinations of emissions can be considered. First combination involves the use of emissions that originate from two closely spaced, “thermally coupled” excited energy levels of the activator ion (Eu3 þ ). The relative population of these levels follows a Boltzmann type distribution and it is dependent on the temperature and the energy difference between levels (energy gap). In the materials doped with Eu3 þ , temperature can be determined using the FIR technique by observing emissions from 5D1 to 5D0 energy levels that are separated 1700 cm 1 (depending on the host material). In this case, the fluorescence intensity ratio (FIR) of
I 1 g 1 A1 hν1 ΔE10 ΔE10 ¼ B1 exp ; ¼ exp I 0 g 0 A0 hν0 kT kT
ð1Þ
where g1 and g0 are the degeneracies, A1 and A0 are the spontaneous emission rates, and ν1 and ν0 are the frequencies of the transitions from 5D1 level to 5D0, respectively; h stands for Planck's constant, andΔE10 for the energy difference between 5D1 level and 5D0 level. At low temperatures there is no emission from the high-energy level, since the population of this level vanishes as the temperature approaches 0 K, and, therefore, the FIR has zero value at 0 K, as described by Eq. (1). However, in fluorescence measurements, instruments frequently detect small signals, even when there is no emission (originating from detector dark currents, etc.). Therefore, the ratio of emissions will have a finite value, A1, at low temperatures, and the FIR1 of two Eu3 þ emissions can be described as follows: ΔE10 : ð2Þ FIR1 ¼ A1 þB1 exp kT A second combination involves one emission of the activator ion and the host material trap emission. The temperature dependence of the ratio of these emissions (FIR2) can be described by the Arrhenius-type Mott equation (frequently referred to as the energy-gap law): ΔE ð3Þ FIR2 ¼ A2 þB2 exp kT where A2 and B2 are constants, ΔE is the energy gap between 5D0 level and 7F6 level. It should be kept in mind that energy differences (ΔE10 and ΔE) in Boltzmann and Mott equations have different physical meanings. Both FIR's combinations are tested for luminescence thermometry in 303–423 K temperature range, as shown in the Fig. 5 (the peak
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
4
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Fig. 5. Temperature dependence of FIR's: (a) of two Eu3 þ emissions and (b) of Eu3 þ emission and host defect state emission. Experimental data are shown with the symbols; the solid lines represent fits using Eqs. (2) and (3).
Fig. 6. Absolute sensitivity vs temperature for (a) FIR of two Eu3 þ emissions and (b) FIR of one Eu3 þ emission and the host defect emission.
Fig. 7. Relative sensitivity vs temperature for (a) FIR of two Eu3 þ emissions and (b) FIR of one Eu3 þ emission and the host defect emission.
intensities were calculated after subtraction of the background signal). In the case of Boltzmann type thermometry, Fig. 5a, the following values are obtained after fitting of experimental data with Eq. (2): A1 ¼ 0.048, B1 ¼ 35.53, and ΔE10 ¼1718 cm 1. In the case of thermometry that uses ratio of Eu3þ and host defect emission, Fig. 5b, the following values are obtained after fitting of experimental data with Eq. (3): A2 ¼2.168, B2 ¼321.1, and ΔE¼1656.2 cm 1. The important parameters for the applicability of the any sensing method are absolute and relative sensitivities of the measurements. The rate at which the fluorescence intensity ratio changes in temperature is known as the absolute sensitivity of luminescence
thermometry, Sa, which is given by the following equation: dFIR : Sa ¼ dT The relative sensor sensitivity, Sr can be found from: 1 dFIR ; Sr ¼ 100% FIR dT
ð4Þ
ð5Þ
and is in fact the normalized absolute sensor sensitivity with respect to the FIR. Fig. 6a and b shows absolute sensitivity of thermometry performed from FIR of two Eu3 þ emissions and from
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
FIR of one Eu3 þ emission and the host defect emission, respectively. In the same way Fig. 7 presents relative sensitivity values of two FIR combinations. Both absolute sensitivities show a similar trend of change in temperature. The smallest values are found at room temperature and absolute sensitivities become larger with temperature increase. In the case of FIR of two Eu3 þ emissions value of absolute sensitivity is in the range from 2.77 10 4 K 1 to 14.26 10 4 K 1. Values of absolute sensitivity of FIR of one Eu3 þ emission and the host defect emission are one order of magnitude larger than in the previous case, and are in the range from 3.23 10 3 K 1 to 15.3 10 3 K 1. The higher values are achieved in the latter case because of higher FIR values. On the other hand, the relative sensitivity of FIR of two Eu3 þ emissions (from 0.47% K 1 to 0.95% K 1) is slightly larger than the relative sensitivity of FIR of one Eu3 þ emission and the host defect emission (from 0.14% K 1 to 0.46% K 1). So, when comparing applicability of two FIR approaches (different combinations of emission ratios) one can note that conventional Boltzmann-type approach provides higher relative sensitivity of thermometry than an approach that exploits one Eu3 þ emission and the host defect emission. On the other hand, host defect state emission is of higher intensity than emission from 5D1-7F1 transition, therefore it can be recorded more easily. 4. Conclusion The emission spectrum of Eu3 þ doped Gd2Ti2O7 nanoparticles consists of emissions from 5D0 to 5D1 levels of Eu3 þ ions and trap emission of Gd2Ti2O7. Emission of Eu3 þ ions is quenched when temperature of the material is increased, and the ratio of emissions from 5D0 to 5D1 level changes in temperature following Boltzmann-type trend. Therefore, Eu3 þ emissions in Gd2Ti2O7 nanoparticles can be used for ratiometric luminescence thermometry in the 303–423 K temperature range with relative sensitivity changing from 0.47% K 1 to 0.95% K 1. Over the same temperature range the trap emission of Gd2Ti2O7 is fairly sensitive to temperature changes and, as a result, it can serve as an excellent internal standard for ratiometric luminescence thermometry. FIR of trap emission and emission from 5D0 level of Eu3 þ ions provides thermometry that varies from 0.14% K 1 to 0.46% K 1 over the 303–423 K temperature
5
range. These relative sensitivity values are slightly smaller compared to those found using two Eu3þ emissions. Then again, absolute sensitivity of thermometry is higher for FIR that uses trap emission than for FIR of two Eu3þ emissions, and trap emission is of stronger intensity than emission from 5D1 Eu3 þ level. To conclude, Eu3þ doped Gd2Ti2O7 nanoparticles can be used for ratiometric luminescence thermometry with two conceptually different combinations of emission lines with relatively high values of the relative sensitivities of measurements (see for comparison the comprehensive list of relative sensor sensitivity values presented by Brites et al. [1]).
Acknowledgment This work is supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Grant no. 45020) and the APV Provincial Secretariat for Science and Technological Development of the Republic of Serbia, through Project no. 114-451-4787. References [1] C.D.S. Brites, P.P. Lima, N.J.O. Silva, A. Millan, V.S. Amaral, F. Palacio, L.D. Carlos, Nanoscale 4 (2012) 4799. [2] D. Jaque, F. Vetrone, Nanoscale 4 (2012) 4301. [3] D.X. Wang, O.S. Wolfbeis, R.J. Meier, Chem. Soc. Rev. 42 (2013) 7834. [4] H.S. Peng, S.H. Huang, O.S. Wolfbeis, J. Nanopart. Res. 12 (2010) 2729. [5] A.H. Khalid, K. Kontis, Sensors 8 (2008) 5673. [6] V.M. Lojpur, M.G. Nikolić, M.D. Dramićanin, J. Appl. Phys. 115 (2014) 203106. [7] M.G. Nikolić, Ž. Antić, S. Ćulubrk, J.M. Nedeljković, M.D. Dramićanin, Sens. Actuators B: Chem. 201 (2014) 46. [8] M.D. Dramićanin, Ž. Antić, S. Ćulubrk, S.P. Ahrenkiel, J. Nedeljković, Nanotechnology 25 (48) (2014) 485501. [9] V. Lojpur, M.G. Nikolić, D. Jovanović, M. Medić, Ž. Antić, M.D. Dramićanin, Appl. Phys. Lett. 103 (2013) 141912. [10] K.M. Lin, C.C. Lin, C.Y. Hsiao, Y.Y. Li, J. Lumin. 127 (2007) 561. [11] Y. Zhang, L. Ding, X. Pang, W. Zhang, J. Rare Earths 27 (2009) 900. [12] Y. Zhang, C. Jia, Z. Su, W. Zhang, J. Alloy. Compd. 479 (2009) 381. [13] M.L. Pang, J. Lin, J. Fu, Z.Y. Cheng, Mater. Res. Bull. 39 (2004) 1607. [14] C.C. Ting, Y.C. Chien, W.F. Sung, ECS J. Solid State Sci. Technol. 2 (2013) 105. [15] S. Ćulubrk, Ž. Antić, M. Marinović-Cincović, P.S. Ahrenkiel, M.D. Dramićanin, Opt. Mater. 37 (2014) 598. [16] P.A. Tanner, Chem. Soc. Rev. 42 (2013) 5090. [17] N.D. Abazović, M.I. Čomor, M.D. Dramićanin, D.J. Jovanović, S.P. Ahrenkiel J.M. Nedeljković, J. Phys. Chem. B 110 (2006) 25366.
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
Contents lists available at ScienceDirect
Journal of Luminescence journal homepage: www.elsevier.com/locate/jlumin
Ratiometric luminescence thermometry with different combinations of emissions from Eu3 þ doped Gd2Ti2O7 nanoparticles Vesna Lojpur, Sanja Ćulubrk, Miroslav D. Dramićanin n University of Belgrade, Vinča Institute of Nuclear Sciences, P.O. Box 522, 11001 Belgrade, Serbia
art ic l e i nf o
a b s t r a c t
Article history: Received 28 October 2014 Received in revised form 22 December 2014 Accepted 9 January 2015
Herein, Eu3 þ doped Gd2Ti2O7 nanoparticles were tested for application in ratiometric luminescence thermometry. It is shown that two combinations of emissions: one that uses two emissions of Eu3 þ ions and one that uses one emission of Eu3 þ ions and trap emission of Gd2Ti2O7 provide thermometry over the 303–423 K temperature range with relative sensitivities between 0.14% K 1 and 0.95% K 1. Thermometry based on two Eu3 þ emissions from 5D0 to 5D1 levels has a higher relative sensitivity, but lower absolute sensitivity than thermometry based on one Eu3 þ emission and trap emission of Gd2Ti2O7. The tested material is prepared by Pechini-type polymerized complex route and is composed of agglomerated nanoparticles of 30–50 nm in size with pure-phase cubic structure (space group Fd3m) as evidenced from electron microscopy and X-ray diffraction measurements. & 2015 Elsevier B.V. All rights reserved.
Keywords: Thermometry Luminescence Pyrochlores Rare earths Nanoparticles
1. Introduction Luminescence thermometry is most commonly performed through relative emission intensity measurements (ratiometric intensity measurements; fluorescence intensity ratio—FIR) [1–5]. This type of measurements is insensitive to fluctuations of excitation light or other changes in measurement conditions, and more importantly, it is self-referencing (i.e., these types of measurements do not have to refer to any temperature standard) [5]. Ideally, one of the used emissions should be independent of temperature (internal reference) and, then, a calibration between the ratios of emissions is indicative of temperature. When used with rare earth ion-doped materials, this technique involves the use of emissions that originate from two closely spaced, “thermally coupled” excited energy levels of the rare earth ions. The relative population of these levels follows a Boltzmanntype distribution and it is dependent on the temperature and the energy difference between levels (energy gap) [2,6]. The main mechanism behind “thermally coupled” energy levels is thermalization: when two energy levels of the RE activator are closely separated by a difference of 2000 cm 1 or less, the upper level will not fluoresce at low temperatures since electrons do not have enough energy to bridge the energy gap. As the temperature increases, the upper level becomes populated and hence the emission from this level gradually increases in intensity at the expense of the lower level population.
n
Corresponding author. E-mail address: [email protected] (M.D. Dramićanin).
Recently, the new concept of ratiometric luminescence thermometry that exploits trap emission of the host as a self-referencing standard and the emission of the activator ions as temperature indicator has been demonstrated using TiO2 nanopowders doped with Eu3 þ ions [7] and Sm3 þ ions [8], and Zn2SiO4 doped with Mn2 þ ions [9]. This concept provided high relative sensitivities and the possibility to perform luminescence thermometry with activator ions which have only single intense emission band (as in the case of Mn2 þ :Zn2SiO4 [9]). Here, we aimed to compare performance of two above mentioned concepts of ratiometric luminescence thermometry with Eu3 þ doped Gd2Ti2O7 nanoparticles. Gd2Ti2O7 exhibit notable luminescence after incorporation of rare earth ions [10–16]. In this cubic-type host rare-earth dopants replace Gd3 þ ions in crystallographic sites with a strict center of symmetry (D3d) [16]. The emission spectrum of Eu3 þ doped Gd2Ti2O7 consists of emissions from Eu3 þ spin-forbidden f–f electronic transitions [15] and trap emission of Gd2Ti2O7 induced by the presence of oxygen vacancies [17]. Therefore, both ratiometric intensity concepts are feasible with this material.
2. Experimental 2.1. Material synthesis procedure Gd2Ti2O7:5 at% Eu3 þ nanopowder was prepared by Pechinitype polymerized complex route, described in detail in our previous work [15]. In brief, this synthesis method is based on polyesterification between citric acid (CA) and ethylene glycol. For
http://dx.doi.org/10.1016/j.jlumin.2015.01.027 0022-2313/& 2015 Elsevier B.V. All rights reserved.
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
synthesis of Gd2Ti2O7:5 at% Eu3 þ nanopowder, titanium (IV)isopropoxide, gadolinium(III)-nitrate, citric acid and ethylene glycol were mixed in 1:1:5:20 M ratio. First, titanium (IV)-isopropoxide (Alfa Aesar, 97%) was dissolved in ethylene glycol (Lach-Ner, 99%) under constant magnetic stirring. Then citric acid (Kemika, 99.5%) was added to the solutions and stirred until complete dissolution was achieved. After that, appropriate amounts of Gd2O3 (Alfa Aesar, 99.9%) and Eu2O3 (Alfa Aesar, 99.9%) were dissolved in hot, concentrated nitric acid, evaporated to dryness and joined with titanium (IV)-isopropoxide/EG/CA mixture. Mixtures were stirred for 1 h at 60 1C until they became transparent and further heated at 130 1C for a few hours to promote polymerization. Black, amorphous resins were fired at 350 1C for 30 min. At the end, pure phase of Gd2Ti2O7:5 at% Eu3 þ nanopowder was obtained after annealing at 880 1C for 4 h and naturally furnace cooling to the room temperature.
Fig. 1. Experimental setup for temperature dependent photoluminescence measurement.
2.2. Instruments and measurements Phase composition of the sample was checked with X-ray powder diffraction measurement performed on a Rigaku Smartlab diffractometer. Diffraction data were collected in the 2θ range from 101 to 901, counting 0.71/min in 0.02 steps. The morphological features and chemical purity were investigated by means of scanning electron microscopy (SEM JEOL, JSM-6610LV) with INCA energy dispersive X-ray analysis. For this purpose the powder sample was dispersed on Cu holder and sputtered with Au. Transmission electron microscopy (TEM) is conducted using JOEL-JEM 2100 instrument equipped with LaB6 cathode and operated with 200 kV. For TEM measurements powders were used without any additional preparation. Photoluminescence spectra were collected using a Fluorolog-3 Model FL3-221 spectrofluorometer system (Horiba Jobin-Yvon), over the temperature range from 293 to 433 K. The photoluminescence measurements were performed under continuous excitation from a 450 W Xenon lamp at a wavelength of 393 nm. The samples were placed in a custommade temperature controlled furnace, and emission spectra were collected via an optical fiber bundle. The temperature of the samples was controlled within the accuracy of 70.5 1C by a temperature control system utilizing proportional-integral-derivative feedback loop equipped with T-type thermocouple for temperature monitoring. Experimental setup shown in Fig. 1, consists of the following parts: 1. Lamp, 2. Detector, 3. Monochromator, 4. Optical fiber, 5. Lens, 6. Sample, 7. Oven, 8. T-regulator, 9. Controller and 10. PC.
3. Results and discussion Pyrochlores with formula A2B2O7 crystallize into a face-centered cubic lattice with space group Fd-3m, No 227. Unit cell contains eight molecules (Z¼8) and four crystallographically nonequivalent sites. The A cation (usually of ionic radius 1 Å) is eight-coordinated and located within scalenohedra (distorted cubes) that contain six equally spaced O atoms at a slightly shorter distance from the central cation. The smaller B cation (ionic radius 0.6 Å) is sixcoordinated and located within a trigonal antiprism with all six anions at equal distances from the central cation. In the pyrochlore structure, D3d symmetry of the metal cation point group requires that these two polyhedra are neither octahedra nor cubes (but trigonal antiprisms and scalenohedra), even though many authors do refer to them as octahedral and cubic coordination polyhedra [15]. XRD patterns of Gd2Ti2O7:5 at% Eu3 þ powder are presented in the Fig. 2. The main diffraction peaks are indexed according to the ICCD card No. 01-074-9640. There were no peaks related to any other phases indicating sample without impurities. High intensity is a consequence of high crystallinity of the powder.
Fig. 2. XRD diffraction patterns for Gd2Ti2O7:5 at% Eu3 þ powder indexed according to the ICCD card No. 01-074-9640.
Morphology of the sample obtained through Pechini-type polymerized complex route was showed in Fig. 3. The micrograph provided by SEM analysis, Fig. 3a, indicated that Gd2Ti2O7:5 at% Eu3þ consists of large chunks with a size of several tens of microns. The elements distribution determined by the energy dispersive X-ray mapping analysis, examined chemical homogeneity of the powder as shown in Fig. 3b. The elemental maps of Gd, TiO and Eu confirmed uniform distribution of all elements throughout the material. The EDX spectrum showed in Fig. 3c revealed the presence of Eu3 þ ions and the purity of rare earth doped Gd2Ti2O7. Further magnification was followed by TEM measurements. The TEM micrograph is presented in Fig. 3d and revealed that chunks are entirely composed of nanoparticles. It is evident that particles are agglomerated and have irregular, round and rectangular shapes. Size of nanoparticles is mostly around 30–50 nm. Luminescence spectra of Gd2Ti2O7:5 at% Eu3 þ powders measured as a function of temperature (293–423 K) are shown in the Fig. 4. Two distinct spectral regions can be observed: high energy broad band spectral region (400–550 nm) that belongs to the trap emission of the Gd2Ti2O7 host and the low energy spectral region that belongs to emission of Eu3 þ ions. Emission of dopant is composed of distinctive bands peaking at 589 nm, 597 nm (5D0-7F1) 611 nm and 627 nm (5D0-7F2) that arise from spin-forbidden electron transition. In addition, small intensity emission from 5D1-7F1 transition can be observed at about 535 nm. It is clearly seen from the Fig. 4that dopant emission is extremely sensitive to temperature, showing a rapidly decreasing intensity with the temperature increase while changes of trap emission are minimal with the change of temperature in the measured range. Small changes in trap emission intensities are a consequence of slight changes in optical properties of the host with temperature such as band gap energy, reflectivity and changes of
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
3
Fig. 3. (a) SEM image of Gd2Ti2O7:5 at% Eu3 þ powder (b) EDX mapping of elements distribution inside the particles (c) EDX spectrum (d) TEM micrograph of Gd2Ti2O7:5 at% Eu3 þ powder.
emissions from the 5D1 level (I1) and 5D0 level (I0) to the ground level is described by the following equation: FIR1 ¼
Fig. 4. Photoluminescence spectra of 5.0 at% Eu3 þ doped Gd2Ti2O7 powder over the temperature range of 293–563 K (temperature increment was 20 K).
experimental parameters during the measurements. On the other hand, the sharp decline of dopant emission is due to the non-radiative relaxation processes that increase with increasing of temperature. Changes of emission spectral shape with temperature provide measures suitable for luminescence thermometry. The use of intensity ratio between two emission lines (FIR) in the emission spectrum is a self-referencing thermometry method that gained a significant attention in recent times [1–6]. In the present case two combinations of emissions can be considered. First combination involves the use of emissions that originate from two closely spaced, “thermally coupled” excited energy levels of the activator ion (Eu3 þ ). The relative population of these levels follows a Boltzmann type distribution and it is dependent on the temperature and the energy difference between levels (energy gap). In the materials doped with Eu3 þ , temperature can be determined using the FIR technique by observing emissions from 5D1 to 5D0 energy levels that are separated 1700 cm 1 (depending on the host material). In this case, the fluorescence intensity ratio (FIR) of
I 1 g 1 A1 hν1 ΔE10 ΔE10 ¼ B1 exp ; ¼ exp I 0 g 0 A0 hν0 kT kT
ð1Þ
where g1 and g0 are the degeneracies, A1 and A0 are the spontaneous emission rates, and ν1 and ν0 are the frequencies of the transitions from 5D1 level to 5D0, respectively; h stands for Planck's constant, andΔE10 for the energy difference between 5D1 level and 5D0 level. At low temperatures there is no emission from the high-energy level, since the population of this level vanishes as the temperature approaches 0 K, and, therefore, the FIR has zero value at 0 K, as described by Eq. (1). However, in fluorescence measurements, instruments frequently detect small signals, even when there is no emission (originating from detector dark currents, etc.). Therefore, the ratio of emissions will have a finite value, A1, at low temperatures, and the FIR1 of two Eu3 þ emissions can be described as follows: ΔE10 : ð2Þ FIR1 ¼ A1 þB1 exp kT A second combination involves one emission of the activator ion and the host material trap emission. The temperature dependence of the ratio of these emissions (FIR2) can be described by the Arrhenius-type Mott equation (frequently referred to as the energy-gap law): ΔE ð3Þ FIR2 ¼ A2 þB2 exp kT where A2 and B2 are constants, ΔE is the energy gap between 5D0 level and 7F6 level. It should be kept in mind that energy differences (ΔE10 and ΔE) in Boltzmann and Mott equations have different physical meanings. Both FIR's combinations are tested for luminescence thermometry in 303–423 K temperature range, as shown in the Fig. 5 (the peak
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
4
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Fig. 5. Temperature dependence of FIR's: (a) of two Eu3 þ emissions and (b) of Eu3 þ emission and host defect state emission. Experimental data are shown with the symbols; the solid lines represent fits using Eqs. (2) and (3).
Fig. 6. Absolute sensitivity vs temperature for (a) FIR of two Eu3 þ emissions and (b) FIR of one Eu3 þ emission and the host defect emission.
Fig. 7. Relative sensitivity vs temperature for (a) FIR of two Eu3 þ emissions and (b) FIR of one Eu3 þ emission and the host defect emission.
intensities were calculated after subtraction of the background signal). In the case of Boltzmann type thermometry, Fig. 5a, the following values are obtained after fitting of experimental data with Eq. (2): A1 ¼ 0.048, B1 ¼ 35.53, and ΔE10 ¼1718 cm 1. In the case of thermometry that uses ratio of Eu3þ and host defect emission, Fig. 5b, the following values are obtained after fitting of experimental data with Eq. (3): A2 ¼2.168, B2 ¼321.1, and ΔE¼1656.2 cm 1. The important parameters for the applicability of the any sensing method are absolute and relative sensitivities of the measurements. The rate at which the fluorescence intensity ratio changes in temperature is known as the absolute sensitivity of luminescence
thermometry, Sa, which is given by the following equation: dFIR : Sa ¼ dT The relative sensor sensitivity, Sr can be found from: 1 dFIR ; Sr ¼ 100% FIR dT
ð4Þ
ð5Þ
and is in fact the normalized absolute sensor sensitivity with respect to the FIR. Fig. 6a and b shows absolute sensitivity of thermometry performed from FIR of two Eu3 þ emissions and from
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i
V. Lojpur et al. / Journal of Luminescence ∎ (∎∎∎∎) ∎∎∎–∎∎∎
FIR of one Eu3 þ emission and the host defect emission, respectively. In the same way Fig. 7 presents relative sensitivity values of two FIR combinations. Both absolute sensitivities show a similar trend of change in temperature. The smallest values are found at room temperature and absolute sensitivities become larger with temperature increase. In the case of FIR of two Eu3 þ emissions value of absolute sensitivity is in the range from 2.77 10 4 K 1 to 14.26 10 4 K 1. Values of absolute sensitivity of FIR of one Eu3 þ emission and the host defect emission are one order of magnitude larger than in the previous case, and are in the range from 3.23 10 3 K 1 to 15.3 10 3 K 1. The higher values are achieved in the latter case because of higher FIR values. On the other hand, the relative sensitivity of FIR of two Eu3 þ emissions (from 0.47% K 1 to 0.95% K 1) is slightly larger than the relative sensitivity of FIR of one Eu3 þ emission and the host defect emission (from 0.14% K 1 to 0.46% K 1). So, when comparing applicability of two FIR approaches (different combinations of emission ratios) one can note that conventional Boltzmann-type approach provides higher relative sensitivity of thermometry than an approach that exploits one Eu3 þ emission and the host defect emission. On the other hand, host defect state emission is of higher intensity than emission from 5D1-7F1 transition, therefore it can be recorded more easily. 4. Conclusion The emission spectrum of Eu3 þ doped Gd2Ti2O7 nanoparticles consists of emissions from 5D0 to 5D1 levels of Eu3 þ ions and trap emission of Gd2Ti2O7. Emission of Eu3 þ ions is quenched when temperature of the material is increased, and the ratio of emissions from 5D0 to 5D1 level changes in temperature following Boltzmann-type trend. Therefore, Eu3 þ emissions in Gd2Ti2O7 nanoparticles can be used for ratiometric luminescence thermometry in the 303–423 K temperature range with relative sensitivity changing from 0.47% K 1 to 0.95% K 1. Over the same temperature range the trap emission of Gd2Ti2O7 is fairly sensitive to temperature changes and, as a result, it can serve as an excellent internal standard for ratiometric luminescence thermometry. FIR of trap emission and emission from 5D0 level of Eu3 þ ions provides thermometry that varies from 0.14% K 1 to 0.46% K 1 over the 303–423 K temperature
5
range. These relative sensitivity values are slightly smaller compared to those found using two Eu3þ emissions. Then again, absolute sensitivity of thermometry is higher for FIR that uses trap emission than for FIR of two Eu3þ emissions, and trap emission is of stronger intensity than emission from 5D1 Eu3 þ level. To conclude, Eu3þ doped Gd2Ti2O7 nanoparticles can be used for ratiometric luminescence thermometry with two conceptually different combinations of emission lines with relatively high values of the relative sensitivities of measurements (see for comparison the comprehensive list of relative sensor sensitivity values presented by Brites et al. [1]).
Acknowledgment This work is supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Grant no. 45020) and the APV Provincial Secretariat for Science and Technological Development of the Republic of Serbia, through Project no. 114-451-4787. References [1] C.D.S. Brites, P.P. Lima, N.J.O. Silva, A. Millan, V.S. Amaral, F. Palacio, L.D. Carlos, Nanoscale 4 (2012) 4799. [2] D. Jaque, F. Vetrone, Nanoscale 4 (2012) 4301. [3] D.X. Wang, O.S. Wolfbeis, R.J. Meier, Chem. Soc. Rev. 42 (2013) 7834. [4] H.S. Peng, S.H. Huang, O.S. Wolfbeis, J. Nanopart. Res. 12 (2010) 2729. [5] A.H. Khalid, K. Kontis, Sensors 8 (2008) 5673. [6] V.M. Lojpur, M.G. Nikolić, M.D. Dramićanin, J. Appl. Phys. 115 (2014) 203106. [7] M.G. Nikolić, Ž. Antić, S. Ćulubrk, J.M. Nedeljković, M.D. Dramićanin, Sens. Actuators B: Chem. 201 (2014) 46. [8] M.D. Dramićanin, Ž. Antić, S. Ćulubrk, S.P. Ahrenkiel, J. Nedeljković, Nanotechnology 25 (48) (2014) 485501. [9] V. Lojpur, M.G. Nikolić, D. Jovanović, M. Medić, Ž. Antić, M.D. Dramićanin, Appl. Phys. Lett. 103 (2013) 141912. [10] K.M. Lin, C.C. Lin, C.Y. Hsiao, Y.Y. Li, J. Lumin. 127 (2007) 561. [11] Y. Zhang, L. Ding, X. Pang, W. Zhang, J. Rare Earths 27 (2009) 900. [12] Y. Zhang, C. Jia, Z. Su, W. Zhang, J. Alloy. Compd. 479 (2009) 381. [13] M.L. Pang, J. Lin, J. Fu, Z.Y. Cheng, Mater. Res. Bull. 39 (2004) 1607. [14] C.C. Ting, Y.C. Chien, W.F. Sung, ECS J. Solid State Sci. Technol. 2 (2013) 105. [15] S. Ćulubrk, Ž. Antić, M. Marinović-Cincović, P.S. Ahrenkiel, M.D. Dramićanin, Opt. Mater. 37 (2014) 598. [16] P.A. Tanner, Chem. Soc. Rev. 42 (2013) 5090. [17] N.D. Abazović, M.I. Čomor, M.D. Dramićanin, D.J. Jovanović, S.P. Ahrenkiel J.M. Nedeljković, J. Phys. Chem. B 110 (2006) 25366.
Please cite this article as: V. Lojpur, et al., J. Lumin. (2015), http://dx.doi.org/10.1016/j.jlumin.2015.01.027i