Formation of Deep Electron Trap by Yb3+ Codoping

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

Functional Inorganic Materials and Devices 3+

Formation of Deep Electron Trap by Yb Codoping Leads into Super-Long Persistent Luminescence in Ce doped Yttrium Aluminum Gallium Garnet Phosphors 3+

Jumpei Ueda, Shun Miyano, and Setsuhisa Tanabe ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b02758 • Publication Date (Web): 23 May 2018 Downloaded from http://pubs.acs.org on May 23, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Formation of Deep Electron Trap by Yb3+ Codoping Leads into Super-Long Persistent Luminescence in Ce3+-doped Yttrium Aluminum Gallium Garnet Phosphors Jumpei Ueda*, Shun Miyano and Setsuhisa Tanabe

Graduate School of Human and Environmental Studies, Kyoto University, Yoshida-nihonmatsucho, Sakyo-ku, Kyoto 606-8501 Japan

KEYWORDS Persistent luminescence, Trap engineering, Ce3+, Valence state change, Garnet

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 31

ABSTRACT

The Y3Al2Ga3O12:Ce3+-Cr3+ compound is one of the brightest persistent phosphors, but its persistent luminescence (PersL) duration is not so long due to the relatively shallow Cr3+ electron trap. Comparing the vacuum referred binding energy of the electron trapping state by Cr3+ and those by lanthanide ions, we selected Yb3+ as a deeper electron trapping center.

The

Y3Al2Ga3O12:Ce3+-Yb3+ phosphors show Ce3+:5d → 4f green persistent luminescence after ceasing blue light excitation. The formation of Yb2+ was confirmed by the increased intensity of absorption at 585 nm during the charging process. This result indicates that the Yb3+ ions act as electron traps by capturing an electron. From the thermoluminescence glow curves, it was found the Yb3+ trap makes much deeper electron trap with 1.01 eV depth than the Cr3+ electron trap with 0.81 eV depth. This deeper Yb3+ trap provides much slower detrapping rate of filled electron traps than the Cr3+-codoped persistent phosphor. In addition, by preparing transparent ceramics and optimizing Ce3+ and Yb3+ concetnrations,

the Y3Al2Ga3O12:Ce3+(0.2%)-

Yb3+(0.1%) as-made transparent ceramic phosphor showed super long persistent luminescence for over 138.8 h after ceasing blue light charging.


2

Page 3 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Persistent luminescence is continuous luminescence for long duration after ceasing excitation. Until now many persistent phosphors have been reported in various compounds with different luminescence centers such as Eu2+ 1-2, Cr3+ 3-5 and so on6. On the other hand, host materials with garnet crystal structure have attracted a great deal of attentions for persistent phosphors as well as for scintillators, LED phosphors or laser materials. Some Ce3+ singly doped garnets such as (Y,Gd)3Al5-xGaxO12:Ce3+

7-9

,

Y3Sc2Al3-xGaxO12:Ce3+

10-12

,

Mg3Y2Ge3O12:Ce3+

13

and

Lu2CaMg2(Si1−xGex)3O12:Ce3+ 14 have been known to show short persistent luminescence due to limited and shallow electron traps originating from intrinsic defects. However, to improve the persistent luminescence properties, it is necessary to engineer the electron traps for persistence. In 2014, we found that Cr3+ acts as an excellent electron trap to dramatically improve the persistent luminescence intensity and duration in Ce3+-doped Y3Al5-xGaxO12 (Yttrium Aluminum Gallium Garnet, YAGG)15.

The YAGG:Ce3+-Cr3+ shows the 5d-4f green persistent

luminescence for several hours after blue-light excitation as well as white LED illumination. After the discovery of the YAGG:Ce3+-Cr3+ phosphors, we also developed transparent ceramics of YAGG:Ce3+-Cr3+ with super long persistent duration Gd3Al5-xGaxO12:Ce3+-Cr3+

17-18

16

and yellow persistent phosphors of

. In YAGG:Ce3+-Cr3+, the threshold energy for charging and the

electron trap depth formed by Cr3+ can be controlled by conduction band (CB) engineering with different Ga content, x 19. For the efficient blue light charging, it was found that the Ga content, x, in YAGG should be more than x=3.0. The blue light charging efficiency increases with increasing x because the energy gap between the lowest 5d level (5d1) and the bottom of CB becomes closer. On the other hand, the trap depth by Cr3+ electron trap is changed from 1.02 eV (x=0) to 0.41 eV (x=5.0) by increasing x in the YAGG host. In addition, with increasing x, the


3


Page 4 of 31

quantum efficiency of Ce3+:5d1-4f luminescence at room temperature decreases significantly above x=3.0

19-20

. At higher Ga content, instead of the high blue light charging efficiency, the

trap depth by Cr3+ becomes much shallower and the quantum efficiency also decreases with increasing Ga content. Thus, for instance, the YAGG:Ce3+-Cr3+ phosphors with high Ga content (x>4.0) are no longer persistent phosphors with bright and long persistent luminescence19. From the viewpoint of the blue light charging efficiency, trap depth and quantum efficiency properties, YAGG:Ce3+-Cr3+ with x=3.0 is the most balanced persistent phosphor in the series of YAGG:Ce3+-Cr3+ persistent phosphors. However, the YAGG:Ce3+-Cr3+ with x=3.0 shows a little bit shorter persistent duration while the initial persistent luminescence intensity is high. To improve the persistent luminescence duration at room temperature with keeping the blue light charging efficiency and quantum efficiency, the trap depth should be deeper in the YAGG with x ≅3 composition to avoid quick electron detrapping. For the management of electron trap depth, lanthanide (Ln) codoping can be a smart method also by taking into account the zig-zag curve of vacuum referred binding energy, VRBE, of divalent lanthanide ions (Ln2+). So far, a number of persistent phosphors codoped with Ln3+ were developed 2, 6. In these phosphors, Ln3+ works as an electron trap and the trap depth can be controlled by a type of Ln3+. This is because some Ln3+ ions can be changed into the Ln2+ states after capturing one electron and the relative energy level of Ln2+ ground state with respect to the bottom of CB is different by a type of lanthanide ions. Bos et al. reported the systematic energy shift of electron trap depth by changing Ln3+ codopants in YPO4:Ce3+ phosphors and the good agreement with the energy gap between the bottom of CB and Ln2+ ground state with the zig-zag curve in a host referred binding energy (HRBE) diagram as well as the VRBE

21-23

. Thus, for

selecting a suitable lanthanide codopant for the persistent luminescence at ambient temperature,


4

Page 5 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the VRBE diagram including the CB, valence band (VB), ground states of Ln2+ and Ln3+ becomes a strong tool. By investigating the energy gap between the bottom of CB and level of Ln2+, the trap depth can be predicted. That is, the trap depth by Cr3+ codopant corresponds to the energy gap between the bottom of CB and Cr2+. In the YAGG with x=3.0, the VRBE of Cr2+ level and the bottom of CB are located to be -2.80 eV and -1.99 eV, respectively19. Because the CB energy is not largely changed by doping with low concentration in the same host material, the deeper electron trap in the YAGG with x=3.0 host can be realized by selecting codopant whose VRBE is slightly lower than the Cr2+ level. When we check the VRBE diagram for YAGG with x=3.0 19, 24-27, the VRBE of Yb2+ is located at -3.52 eV. Therefore, we expected that Yb3+ can be a suitable deeper trap in the YAGG host with x=3.0 compared with the electron trap of Cr3+ for long persistent luminescence as shown in Figure 1 (a) and (b). In actual, it is already known that Yb3+ works well as an electron trap in some phosphors, such as the persistent phosphor of MgGeO3:Mn2+-Yb3+

28-29

and the persistent and thermoluminescence phosphor of

(Y,Gd)3Al5O12:Ce3+-Yb3+30-32, where Yb3+ act as a much deeper electron trap. In this study, we investigated the possibility of Yb3+ as a deeper electron trap than the Cr3+ electron trap in the YAGG hosts (x~3.0) and successfully developed long persistent phosphor in YAGG:Ce3+-Yb3+. Also, for improving the persistent luminescence intensity and for new unique applications, we prepared transparent ceramic phosphors of YAGG:Ce3+-Yb3+. Transparent ceramics have many advantages such as high transparency, high density, low light scattering, good crystallinity, attracting a great deal of attention as optical material hosts. In actual, the transparent ceramics were already reported and commercialized as laser host material33 and w-LED phosphor34 in garnet ceramics. Recently, we reported that the transparent ceramic persistent phosphors show better persistent luminescence performance compared with powder type persistent phosphors16,


5


18

Page 6 of 31

. This is because low light scattering enables charging inside of persistent phosphors. For the

powder of persistent phosphors, the inside of the powder compact cannot be charged due to the strong light scattering. In addition, recently, for transparent persistent phosphors with deep electron traps, the possibility to be data storage materials was reported35.

Thus, we also

investigated the persistent luminescence properties of YAGG:Ce3+-Yb3+ transparent ceramic phosphors.

2. Experimental Procedure 2.1 Synthesis of YAGG:Ce3+-Yb3+ normal ceramics Polycrystalline ceramics of Y3Al5-xGaxO12 garnet, YAGG, (x=3.0) doped with 0.5% Ce3+ and y % (y=0.05, 0.1, 0.5 and 1) Yb3+ for the dodecahedral site (A site in {A}3[B]2(C)3O12) were synthesized by solid-state reactions. To avoid concentration quenching for Ce3+ luminescence, 0.5% was selected firstly36. Chemicals of Y2O3 (99.99 %), Al2O3 (99.99 %), Ga2O3 (99.99 %), CeO2 (99.99 %) and Yb2O3 (99.9%) were used as starting materials. The powder was mixed by a ball milling system (Premium Line P-7, Fritsch) with ethanol. The obtained slurry was dried and pulverized. The dried powders were pressed at 50 MPa into 10-mm-φ × 2-mm thick pellets. The pellets were sintered at 1600 °C for 10 h in air.

2.2 Synthesis of YAGG:Ce3+-Yb3+ transparent ceramics Transparent ceramics of YAGG(x=3.0) doped with z% (z= 0.5 and 0.2) Ce3+ and 0.1% Yb3+ were prepared by solid-state reactions. Starting chemicals are the same as those used in the synthesis of conventional ceramics. The powder with stoichiometric compositions was mixed by the ball milling system with ethanol, tetraethyl orthosilicate (TEOS) and dispersant. The mixed


6

Page 7 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


powder was dried and pulverized, and then compacted to form a ceramic green body (20 mm φ × approx. 3 mm thick) under uniaxial pressing of 50 MPa. The green body was sintered at 1600°C for 10 h under a vacuum atmosphere. Some of the as-prepared samples were annealed at 1250°C for 5 h under air.

The both sides of all the obtained transparent ceramics were

polished Simultaneously by using a copper plate and diamond slurry, and the final thickness was 2.1 mm.

2.3 XRD and optical measurements The crystalline phase of ceramic samples were identified using an XRD (X-ray diffraction) equipment (Ultima IV, Rigaku) as garnet crystals (see Figure S1). For photoluminescence (PL) spectra in the range of visible to near infrared, the sample was excited by a 365nm LED and the luminescence was detected with two CCD spectrometers (USB 2000+, Ocean Optics) of different gratings with a bifurcated fiber. PL excitation (PLE) spectra were measured using a fluorescence spectrophotometer (RF-5300, Shimadzu).

For persistent luminescence (PersL)

spectra, the sample was charged by 450 nm monochromatic light by using a Xe lamp and a 450 nm bandpass filter and the persistent luminescence was detected a CCD spectrometer (QE65PRO, Ocean Optics). After the sample was charged by the 450 nm monochromatic light mentioned above, the PersL decay curve was detected by a photomultiplier tube, PMT (R3896, Hamamatsu Photonics), which was covered with a 475 nm short-cut filter and a 600 nm long-cut filter to filter out all but the Ce3+ luminescence.

The persistent luminance (mcd/m2) was

measured by using a luminance-measurement setup (BW-L1, Konica-Minolta) composed of a CCD spectrometer (Glasier X, B&W Tek Inc.), a fiber and a collimator lens. Using the obtained persistent luminance, the persistent decay curves was calibrated.

For the persistent luminance


7


Page 8 of 31

measurement of YAGG:Ce3+-Yb3+ transparent ceramics, the back side and edge of the sample was covered with aluminum foil to collect persistent luminescence. Thermoluminescence (TL) glow curves monitoring the Ce3+ luminescence were measured using the same photomultiplier and the filters as the PersL decay curve measurement. The sample was mounted on a cryostat (Helitran LT3, Advanced Research Systems) to control the temperature. The samples were excited by UV light of the Xe lamp (250 nm-400 nm) at 100 K for 10 min. and kept for another 10 min. after ceasing excitation, and then the sample temperature was increased with the heating rate of 10 K/minute up to 600 K. For the transparent ceramics, in-line transmittance spectra were measured by a UV–VIS–NIR spectrometer (UV-3600, SHIMADZU). The time course of in-ine transmittance spectra during charging process and detrapping process was measesured also by the UV-VIS-NIR spectrometer. For the charging process, the sample was illuminated by a 455 nm LED (LLS-455, OceanOptica) for different time (1, 3, 5, 10, 30 and 60 min). For the detrapping process, the spectrum was measured at different time (5 min, 60 min, 24 h and 4 days) after ceaisng blue light charging. Images of the persistent luminescence were taken by a digital camera (Sony, NEX-7) with ISO 6400, F-number 5 and different shutter speed, SS after ceasing the blue-LED charging for 5 min.

3. Results and Discussion 3.1 PL spectra of YAGG(x=3.0):Ce3+-Yb3+(y%) Figure 2 (a) shows the PL spectra excited by 365 nm and PLE spectra of 520 nm for YAGG(x=3.0):Ce3+-Yb3+ with different Yb3+ concentration.

An intense luminescence band

peaked at 510 nm was observed. This band is attributed to the transitions from the 5d1 energy level to the 4f energy levels of Ce3+. In addition to the Ce3+:5d1→4f band, sharp PL lines due to


8

Page 9 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the Yb3+:2F5/2→2F7/2 transitions were observed at around 1µm by UV excitation. With increasing Yb3+ concentration, y, the Ce3+:5d1→4f PL intensity decreases monotonically and the Yb3+:2F5/2 →2F7/2 PL intensity increases up to y=0.5% and the Yb3+ PL intensity of y=1% almost same as that of y=0.5%. From the PLE spectra of Ce3+ luminescence, two PLE bands were observed at around 440 and 350 nm in all the samples Y3Al2Ga3O12:Ce3+(0.5%)-Yb3+(y %). These PLE bands are originated from the transitions of Ce3+ ion from the ground 4f level to the 5d1 level and the second lowest 5d (5d2) level, respectively. In the PLE spectra monitoring Yb3+ luminescence between 900 and 1100 nm, similar PLE bands were observed at around 425 nm and 360 nm, which are attributed Ce3+:4f-5d1 and 5d2 (see Figure S2). These results show that the energy transfer occurs from Ce3+ to Yb3+ in the YAGG(x=3.0) host. It was suggested that the energy transfer from Ce3+ to Yb3+ is caused through the metal-to-metal-charge-transfer (MMCT) state of Ce4+-Yb2+ in the Y3Al5O12 garnet doped with Ce3+ and Yb3+, but not through the down conversion (quantum cutting) process 37-38. The energy transfer efficiency can be estimated by the lifetime of donor excited level (Ce3+:5d1).

The obtained lifetime of Ce3+:5d1 in

YAGG:Ce3+(0.5%)-Yb3+(y%) decreases with increasing Yb3+ concentration in order of 39.8 (y=0), 37.3 (y=0.05), 34.0 (y=0.1), 27.7 (y=0.3), 24.5 (y=0.5) and 18.5 ns (y=1) as shown in Figure S3. From these obtained lifetime values of Ce3+: 5d1 level, the energy transfer efficiency (ηET) from Ce3+ to Yb3+ increases with increasing Yb3+ concentration and reaches to approximately 53.5 % at Yb3+ (1%) (see Figure S4). The quantum efficiency (η) of the Ce3+:5d1-4f luminescence can be roughly estimated by η=100-

ηΕΤ (%). Thus, the observed decrease of the Ce3+:5d1→4f PL intensity with Yb3+ concentration is caused by the decrease of quantum efficiency due to the strong energy transfer.


9


Page 10 of 31

3.2 PersL spectra of YAGG(x=3.0):Ce3+-Yb3+(y%) Figure 2 (b) shows the PersL spectra of Y3Al2Ga3O12:Ce3+-Yb3+ samples with different Yb3+ concentration after 450 nm blue light charging.

All the YAGG(x=3.0):Ce3+-Yb3+ samples

exhibit the persistent luminescence of Ce3+ at around 520 nm due to the Ce3+:5d→4f transition, but of Yb3+ at 1µm not. From the observation of Ce3+ persistent luminescence and the absence of Yb3+ persistent luminescence, it is expected that Yb3+ act only as an electron trap. With increasing Yb3+ concentration, the PersL intensity of Ce3+:5d1-4f firstly increases up to y=0.1 and decreases. The lack of Yb3+ emission in persistent luminescence spectrum could be related to the absence of persistent energy transfer from Ce3+ to Yb3+ only in persistent luminescence. However, some persistent phosphor containing Yb, such as MgGeO3:Yb3+, Ca2SnO4:Yb3+ and (Ba1-xSrx)AlSi5O2N7:Yb2+-Yb3+, show NIR persistent luminescence at around 1µm due to the electron transfer or energy transfer39-41.

In order to clarify the absence of Yb3+ PersL in

YAGG:Ce3+-Yb3+, further study is needed, but it may be possibly related to the different Ce3+ Yb3+ pairs which show photoluminescence and persistent luminescence.

3.3 TL glow curves of YAGG(x=3.0):Ce3+-Yb3+(y%) Figure 2 (c) shows the TL glow curves monitoring Ce3+ luminescence of the YAGG(x=3.0):Ce3+-Yb3+ samples and the reference sample of YAGG(x=3.0):Ce3+ and YAGG(x=3.0):Ce3+-Cr3+ reported by our group 15. The Ce3+ singly doped YAGG(x=3.0) sample exhibits some weak TL glow peaks originated from the intrinsic defects at 206 and 286 K while the YAGG(x=3.0):Ce3+-Yb3+ samples exhibit only single TL glow peak from 343 to 355 K. The TL glow peak temperature of the samples with Yb3+ (0.05, 0.1 and 0.5%) are similar to each other, but the sample with Yb3+ (1%) shows slightly higher TL glow peak temperature of 355 K.


10

Page 11 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


This is probably because that the effect to the host electronic band structure by Yb3+ doping with slightly higher concentration is not negligible. The doping smaller Yb3+ ions (0.98 Å) in 8-fold coordination

42

can increase the bandgap as is observed in the composition of Y3Al5O12 and

Lu3Al5O12. When Y3+ (1.015 Å) ions in 8-fold coordination are replaced by smaller Lu3+ (0.97 Å) ions

42

, the band gap energy is known to be increased43

Compared with the

YAGG(x=3.0):Ce3+-Cr3+ sample, the TL glow peak of the YAGG(x=3.0):Ce3+-Yb3+ sample clearly shifts to higher temperatures. The peak temperatures of YAGG(x=3.0):Ce3+(0.5%)Cr3+(0.05%) and YAGG(x=3.0):Ce3+(0.5%)-Yb3+(0.05%) samples are 294 K and 346 K, respectively. These results show that Yb3+ acts as a new electron trap, which is distinguished from the traps by intrinsic defects and the electron trap by Cr3+. From the observed TL peak temperature, we estimated the trap depth using the 1st order kinetics equation 44-46 as below,

= s × exp −

(1),

where is the heating rate, is the Boltzmann constant (eV/K), is the frequency factor (s-1) and is the TL glow peak temperature (K). It is known that the 1st order kinetics equation is dominant rather than non-first-order kinetics in many compounds47. Here, we assume frequency factor to be 1 × 1013 s−1, because we already estimated it from the heating rate plot for the YAGG(x=3.0):Ce3+-Cr3+ sample19 and also because the frequency factor depends on the host material, but not on the dopants. The estimated trap depths are shown in column 3 of Table 1. The trap depth values of the YAGG(x=3.0):Ce3+(0.5%)-Yb3+(y%) with y=0.05, 0.1, 0.5 and 1 samples range within 1.00~1.04 eV, and nearly the same each other. Thus, the deeper electron trap was realized by Yb3+ codoping compared with the Cr3+ electron trap of 0.81 eV, as is predicted from the VRBE diagram.


11


Page 12 of 31

3.4 PersL decay curves of YAGG(x=3.0):Ce3+-Yb3+(y%) Figure 2 (d) shows the PersL decay curves of the YAGG(x=3.0):Ce3+-Yb3+ samples with different Yb3+ concentration after 450 nm blue light charging for 5 min.

The persistent

luminescence of all the samples is detectable by human eyes in a dark room. The PersL decay rate which corresponds to slope does not change much by Yb3+ concentration because those trap depths are almost the same. However, the PersL decay rate of Yb3+-codoping samples becomes slower compared with that of Cr3+-codoping sample. This is because the deeper electron trap of Yb3+ causes slower detrapping rate. Thus, as we expected, the PersL decay rate was successfully controlled by the deeper electron trap depth. The PersL duration on 0.32 mcd/m2 after ceasing excitation light are listed in column 4 of Table 1. With increasing Yb3+ concentration, y, the PersL intensity as well as duration increases up to y=0.1 and then decreases. The PersL intensity corresponds to the electron trap concentration, thus, it is expected that the PersL intensity will increases monotonically with increasing Yb3+ concentration. However, the optimum Yb3+ concentration for the PersL intensity was observed. The decrease of PersL intensity with higher Yb3+ concentration is probably because of the decrease of Ce3+: 5d1-4f quantum efficiency as discussed in chapter 3.1 and the increase of nonradiative recombination probability increases. From the viewpoint of long PersL duration, 0.1% of Yb3+ concentration is the best among the YAGG(x=3.0):Ce3+(0.5%)-Yb3+(y%) samples with different y. In order to improve the persistent luminescence intensity furthermore, we tried to prepare transparent ceramic persistent phosphors.

3.5 Persistent luminescence of YAGG(x=3.0):Ce3+-Yb3+ transparent ceramics


12

Page 13 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3 (a) shows the in-line transmittance spectra of the as-made and annealed YAGG(x=3.0):Ce3+(0.5%)-Yb3+(0.1%) transparent ceramics. Inset of Figure 3 (a) shows images of the as-made and annealed YAGG(x=3.0):Ce3+(0.5%)-Yb3+ (0.1%) transparent ceramics. The as-made and annealed samples show green and yellow color, respectively, and both of them possess high transparency.

The thickness of both as-made and annealed samples is 2.1 mm.

The both samples show over 50% transmittance at 1200 nm. In the transmittance spectrum of the as-made sample before charging, four absorption bands were observed at around 950, 585, 425 and 340 nm. The sharp absorption lines at around 950 nm is attributed to the Yb3+:4f-4f transitions from 2F7/2 to 2F5/2, while the broad absorption bands at around 425 nm and 340 nm are the transitions of Ce3+ from 4f to 5d1 and 5d2, respectively. The broad absorption band at 585 nm can be originated from the 4f14-4f135d1 spin allowed transition of small amount of Yb2+ which was generated by reduction of Yb3+ under vacuum sintering. On the other hand, in the transmittance spectrum of the annealed samples, the absorption band of Yb2+:4f-5d was not observed because the minor amount of Yb2+ ions were oxidized to be Yb3+. The observed small bump at 690 nm in all the transmittance spectra is due to the small change of instrument response by switching detector from InGaAs photodiode to photomultiplier tube (PMT). It is already known that the Yb2+ can be stabilized in the garnet compounds (Y3Al5O1248-51, Y3Ga5O1252, Lu3Al5O1253 and Lu3Ga5O1252) by vacuum sintering with proper charge compensators or the reduction with γ-irradiation in Yb3+-doped compounds.

In the

Y3Al5O12:Yb2+ and Y3Ga5O12:Yb2+, the first absorption band due to the spin allowed 4f14-4f135d1 transition (4f-5d) were reported to be approximately 650 and 555 nm, respectively.

The

obtained 4f-5d absorption band (585 nm) of small amount of Yb2+ in Y3Al2Ga3O12:Ce3+(0.5%)-


13


Page 14 of 31

Yb3+ (0.1%) transparent ceramics is located between the Yb2+:4f-5d absorption wavelengths of Thus, Yb2+:4f-5d absorption wavelength is blue-shifted

Y3Al5O12 and Y3Ga5O12 host.

monotonically with increasing Ga content. This tendency of Yb2+:4f-5d absorption wavelength shifting for Ga content corresponds to that of Ce3+:4f-5d1. Dorenbos summarized the Ce3+:4f-5d1 absorption wavelength in RE3Al5-xGaxO12 (RE=Gd, Y, Lu) hosts and reported the Ce3+:4f-5d1 absorption wavelength is blue shifted with increasing Ga content due to the crystal fileld. These results indicate that the absorption band at 585 nm in Y3Al2Ga3O12:Ce3+(0.5%)-Yb3+ (0.1%) transparent ceramics is originated from the Yb2+:4f-5d transition and the absorption peak wavelength of Y3Al5-xGaxO12:Yb2+ is also blue-shifted with increasing Ga content by the crystal field. In addition, Dorenbos reported that the redshift parameters for Ln3+ (D(3+,A)) is correlated with the redshift parameters for Ln2+ (D(2+,A)) by 2+, A = 0.64D3+, A − 0.233 eV 2. The redshift parameter for Ln(3+/2+) is the energy difference between the 5d energy of Ln(3+/2+) as free ion and that of Ln(3+/2+) in a compound, which includes the 5d energy sifting by centroid shift and crystal filed factors. The redshift parameters are changed by valence state of Ln, but not changed by type of lanthanide ions. The D(2+,A) and D(3+,A) can be estimated using below functions with 4f-5d absorption wavelength of Yb2+ and Ce3+ (λabsYb2+:4f-5d and λabsCe3+:4f-5d (nm)), D3+, A = 6.12 eV − D2+, A = 4.41 eV −

()*+

,-./,0123:56789 ()*+

,-./,:.3:56789

(3), (4).

The estimated D(2+,A) , D(3+, A) were listed in Table S2. Based on these equations, the Dcal(2+,A) and Yb2+:4f-5d absorption wavelength can be predicted by Ce3+:4f-5d1 absorption wavelength. However, Dorenbos et al. already reported the Dcal(2+,A) estimated from the


14

Page 15 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ce3+:4f-5d1 is about 0.6 eV smaller than the D(2+,A) estimated from the Yb2+:4f-5d in Y3Al5O12, Y3Ga5O12 garnets, and suggested that it is possibly related to the large hardness of the garnets combined with the exceptionally large crystal field splitting43. Taking into account the 0.6 eV correction, the Yb2+:4f-5d absorption wavelength in Y3Al2Ga3O12 was calculated to be 636 nm from the reported Ce3+:4f-5d absorption wavelength (435 nm) in Y3Al2Ga3O1243. The estimated Yb2+:4f-5d absorption wavelength is roughly in line with the observed absorption band. This is result also shows that the absorption band at 585 nm in Y3Al2Ga3O12:Ce3+(0.5%)-Yb3+ (0.1%) transparent ceramics is originated from the Yb2+:4f-5d transition. Figure 3 (b) and (c) shows the variation of in-line transmittance spectra of the as-made YAGG(x=3.0):Ce3+(0.5%)-Yb3+ (0.1%) transparent ceramics during charging process by 455 nm LED light and detrapping process at room temperature.

During the charging process, the

Yb2+:4f-5d absorption intensity clearly increases with increasing charging time, and the absorption intensity was almost saturated after 1 h charging. Because this absorption intensity is proportional to the Yb2+concentrations, some of the Yb3+ ions were changed into the Yb2+ state after charging. In addition, the Yb2+:4f-5d absorption intensity decreases with time after ceasing excitation as shown in Figure 3 (c).

The Yb2+ ions created by the charging process gradually

change back into the Yb3+ state by releasing the trapped electron. To visualize this changing of Yb2+ and Yb3+ concentrations by charging and detrapping processes, the integrated absorption coefficient of Yb2+:4f-5d and Yb3+:4f-4f was plotted as a function of charging time in Figure 3 (d) and as a function of time after ceasing excitation in Figure 3 (e). The Yb2+:4f-5d integrated absorption coefficient clearly increases with increasing charging time while the Yb3+:4f-4f integrated absorption coefficient shows decreasing tendency. On the other hand, for the detrapping process, Yb2+:4f-5d and Yb3+:4f-4f integrated absorption


15


coefficient decreases and slightly increases, respectively.

Page 16 of 31

The inverse correlation between

Yb2+:4f-4f and Yb3+:4f-5d absorption coefficient for the charging and detrapping processes show that Yb3+ ions act as electron traps by capturing one electron and changing into Yb2+.

In

addition, even at 4 days after ceasing blue light excitation, the 4f-5d transition of created Yb2+ was observed, which indicates that a part of electron traps by Yb3+ is still filled. Thus, these transparence ceramics shows persistent luminescence even 4 days after ceasing excitation as discussed later. The reason for the small changing of Yb3+:4f-4f integrated absorption coefficient during and after charging is because Yb3+:4f-4f absorption transition probability is lower 4f-4f transition probability than Yb2+:4f-5d transition probability 54.

In other words, because the Yb2+

ion has the parity allowed 4f-5d transition with high transition probability, the small amount of change from Yb3+ to Yb2+ can be detected easily by monitoring Yb2+ absorption intensity in transmittance spectrum during and after charging.

Thus, the YAGG:Ce3+-Yb3+ persistent

transparent ceramics is one of the special material to detect the electron trap species by the transmittance spectrum. Figure 4 (a) shows the TL glow curves of the as-made and annealed transparent ceramics of Y3Al2Ga3O12:Ce3+(0.5%)-Yb3+(0.1%) after UV charging for 10 min.

Both TL glow peak

temperatures of the as-made and annealed samples are located at approximately 339 K which is almost the same as that of opaque ceramics of the same composition. However, the width of TL glow peak of the as-made sample is much broader than that of the annealed samples. There are two possible reasons: One is the formation of additional electron trapping center by oxygen vacancies in the as-made sample due to the vacuum sintering. It was known that they act as electron traps, and as-made YAGG:Ce3+ transparent ceramics sintered under vacuum which has oxygen vacancies also showed some TL glow peaks as shown in Figure S5. Another possible


16

Page 17 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


reason is the broadening of trap depth distribution by the inhomogeneity of coordination environments around Yb3+ electron trapping center. It was known that the TL glow band was broaden by wider electron trap distribution

55-56

. So far, many researchers analyzed the electron

trap distribution in some persistent phosphors which has broad TL glow band57-60.

We also

reported that the electron trap depth distribution was broadened when the inhomogeneity increases around the Cr3+ electron trap by the substitution of the Ga ions for the Al sites in Y3Al53+ 3+ xGaxO12:Ce -Cr

persistent phosphors 19. Similar to this, the inhomogeneity can be caused by

formation of oxide vacancies by the vacuum sintering. In the both cases, its TL glow curves can become much sharper in the annealed sample because the oxide vacancies can be removed after annealing in air as below reaction, (

2; + ?@∙∙ + B) → 2;

Formation of Deep Electron Trap by Yb3+ Codoping

Recommend Documents