Tm3+ codoped LiNbO3 polycrystals

Tm3+ codoped LiNbO3 polycrystals

Journal of Luminescence 132 (2012) 1568–1574 Contents lists available at SciVerse ScienceDirect Journal of Luminescence journal homepage: www.elsevi...

1015KB Sizes 2 Downloads 41 Views

Journal of Luminescence 132 (2012) 1568–1574

Contents lists available at SciVerse ScienceDirect

Journal of Luminescence journal homepage: www.elsevier.com/locate/jlumin

Upconversion white-light emission in Ho3 þ /Yb3 þ /Tm3 þ codoped LiNbO3 polycrystals Lili Xing a, Rui Wang a,n, Wei Xu b, Yannan Qian a, Yanling Xu a, Chunhui Yang a, Xinrong Liu a a b

Department of Chemistry, Harbin Institute of Technology, Harbin 150001, PR China Condensed Matter Science and Technology Institute, Harbin Institute of Technology, Harbin 150001, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 July 2011 Received in revised form 19 January 2012 Accepted 30 January 2012 Available online 5 February 2012

Ho3 þ /Yb3 þ /Tm3 þ codoped LiNbO3 polycrystals exhibiting upconversion white-light under 980 nm excitation have been successfully fabricated by the high temperature solid-state reaction method. CIE coordinate of the Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystal is (0.34, 0.35), which is very close to the standard equal energy white-light illuminate (0.33, 0.33). Efficient green, red, and blue upconversion emissions have been observed. The luminescent decay dynamics are studied, and rate equations for the blue, green, and red emissions are set up to analyze the upconversion luminescence mechanism. The present results demonstrate that the competition between the linear decay and the upconversion process for the depletion of the intermediate excited states plays an important role in upconversion mechanism. The LiNbO3 with upconversion white-light will be a promising luminous material. & 2012 Elsevier B.V. All rights reserved.

Keywords: Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 Upconversion White-light

1. Introduction The development of upconversion white fluorescence materials is still a very active research field, driven by the needs for biolabel, three-dimensional solid-state multicolor display, high-density memories, photonic applications, and so on [1–5]. Therefore, the preparation of upconversion white-light has become an important task for researchers. Rare-earth (RE) ions are suitable for the upconversion process owing to their abundant energy levels, long lifetime excited states, narrow emission spectral lines and excellent chemical durability, and they can be easily populated by near-infrared radiation [6]. So far, more promising results on the upconversion white-light have been obtained by use of RE ions, such as Er3 þ / Tm3 þ /Yb3 þ , Ho3 þ /Tm3 þ /Yb3 þ , and Pr3 þ /Er3 þ /Yb3 þ codoped systems [7–13]. But the host materials are limited mainly in nano-system, glass-system, polymer-system, and so on [14–16]. However, since the functions of these host materials are always single, typically the use of LiNbO3 as host material seems particularly attractive due to its excellent piezoelectric, electrooptical, and nonlinear optical performances. Multi-function of LiNbO3 creates sufficient conditions for opening up new perspectives to the studies of integration and tiny devices. LiNbO3 has been proved to be one of the best manual materials, which exhibits high mechanical capability and excellent chemical stability. E. Cantelar et al. [17] reported the red, green, and blue light

n

Corresponding author. Tel.: þ86 15846590861. E-mail address: [email protected] (R. Wang).

0022-2313/$ - see front matter & 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jlumin.2012.01.053

simultaneous generation in a periodically poled Zn-diffused Er3 þ / Yb3 þ /LiNbO3 nonlinear channel waveguides under 850–990 nm excitation, the red and green emissions arose from energy transfers between Yb3 þ and Er3 þ ions, while the blue light was produced by quasi-phase matching processes. To the best of our knowledge, there is still no report on upconversion white-light property for LiNbO3 multi-doped system. Preferably, the upconversion technology is surprising in the fields of luminescence and the LiNbO3 crystal is one of the most promising electro-optical materials, meaning that the integration will be remarkable. In this article, Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals were successfully prepared by the high temperature solid-state reaction method. The superior upconversion white-light emissions under 980 nm excitation are demonstrated and the upconversion mechanism is analyzed. Consequently, the results provide potential for the development of white-emitting device for lighting applications.

2. Experimental The host material was LiNbO3 with the ratio of Li/Nb¼0.946, and the raw materials were Li2CO3 (4 N purity), Nb2O5 (4 N purity), Ho2O3 (4 N purity), Yb2O3 (4 N purity), and Tm2O3 (4 N purity). The powders were fully ground in an agate mortar by hand at least for 4 h. The sufficient fine and fully homogenous powders were pressed into a disk under 20 MPa. Subsequently, the samples underwent the heat treatment of 750 1C for 2 h to resolve Li2CO3 into Li2O and CO2, and then 1150 1C for 10 h to generate Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals in Al2O3 crucible.

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

1569

Table 1 Raw material compositions (mol%) of samples. Samples

Ho3 þ (mol%)

Yb3 þ (mol%)

Tm3 þ (mol%)

1# 2# 3# 4#

0.05 0.10 0.10 0.10

2.00 2.00 2.50 2.00

0.20 0.20 0.20 0.25

Our investigation chasing white-light generation commenced with the study of orthogonal experimental design on Ho3 þ / Yb3 þ /Tm3 þ /LiNbO3 polycrystals at 980 nm excitation. The details of orthogonal experimental design are shown in Supporting Information (SI). From the orthogonal experiment, LiNbO3 polycrystal with concentrations (mol%) of 0.10Ho3 þ /2.00Yb3 þ / 0.20Tm3 þ was used as a reference sample, and the concentration of Ho3 þ should be decreased while the concentration of Yb3 þ or Tm3 þ should be increased appropriately in order to realize the white-light emission under 980 nm excitation. The raw material compositions (mol%) of samples are shown in Table 1. To identify the crystallization phase, X-ray diffraction spectra of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals were measured by an XRD-6000 diffractometer using a copper Ka radiation source. The upconversion luminescence spectra were recorded by a SPEX1000M spectrometer with a photomultiplier tube under 980 nm excitation. The CIE chromaticity coordinates for the upconversion fluorescence of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals were calculated based on the 1931 CIE standard and marked in the CIE standard chromaticity diagram. All these measures were performed at room temperature.

Fig. 1. XRD patterns of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals (1#: 0.05Ho3 þ / 2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3).

3. Results and discussion 3.1. Phase and structure analysis Fig. 1 shows the X-ray diffraction patterns of Ho3 þ /Yb3 þ / Tm3 þ /LiNbO3 polycrystals at room temperature. The results show that all the diffraction peaks of the samples can be well indexed to the known phase of LiNbO3 based on the standard XRD pattern (PDF#20–0631), so the new phase does not appear in Ho3 þ /Yb3 þ / Tm3 þ /LiNbO3 polycrystals. Further information can be obtained, the doping ions (Ho3 þ /Yb3 þ /Tm3 þ ) have no influence on the structures of the samples, and the Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals are still trigonal system. So it is speculated that the doping ions may occupy the normal Li-site or Nb-site rather than interstitial sites within the lattice.

3.2. Upconversion luminescence spectra Fig. 2 shows the upconversion emission spectra of Ho3 þ /Yb3 þ / Tm3 þ /LiNbO3 polycrystals under 980 nm excitation. The blue emission has a luminescence peak at 480 nm that corresponds to Tm3 þ : 1G4-3H6 transition; the green emission band centered around 550 nm is contributed to Ho3 þ : 5S2, 5F4-5I8 transition; and the red emission has luminescence peaks at 643 nm and 665 nm that are due to Tm3 þ : 1G4-3F4 transition and Ho3 þ : 5 F5-5I8 transition. From Fig. 2, it is to be noted that the intensities of blue emissions are almost unchanged compared with green and red emissions when the concentrations of doping ions are changed. Besides, the upconversion emission spectra of all Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals under 980 nm excitation as a function of excitation power are drawn in Fig. 3. From Fig. 3, it can be seen that for all the emissions, the intensities of

Fig. 2. Upconversion emission spectra of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals under 980 nm excitation (1#: 0.05Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ / 2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3). (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

blue, green, and red emissions are enhanced with the increase of excitation power. In order to reflect the true color of luminescences, the CIE 1931 color coordinates for the upconversion emissions of Ho3 þ /Yb3 þ / Tm3 þ codoped LiNbO3 polycrystals under 980 nm excitation are calculated using the following formula: x¼

X X þ Y þZ



Y X þY þZ



Z X þY þZ

ð1Þ

where X, Y, and Z are the three tristimulus values. The tristimulus values for a color with a spectral power distribution P(l) are

1570

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

Fig. 3. Upconversion emission spectra of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals under 980 nm excitation as a function of excitation power (1#: 0.05Ho3 þ /2.00Yb3 þ / 0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3). (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

given by Z 720 PðlÞx0 ðlÞdl X¼ 400



Z

720

PðlÞy0 ðlÞdl

400



Z

720

PðlÞz0 ðlÞdl 400

ð2Þ where l is the wavelength of the equivalent monochromatic light, and x0 (l), y0 (l), z0 (l) are three color-matching functions [18]. The color coordinates (x, y) of multicolor upconversion emissions in the four samples (1#, 2#, 3#, 4#) are (0.36, 0.27), (0.35, 0.31), (0.34, 0.35), and (0.35, 0.32), respectively, which are indicated in Fig. 4. The color coordinates of 2#, 3#, and 4# samples have a good match with the standard point of equal energy white-light (0.33, 0.33) under 980 nm excitation, and the 3# sample has the best intensity and color purity. And it is obvious that the whitelight emission can also be obtained even though the red, green, and blue emissions have unequal intensities which can be verified by Ref. [19]. This performance provides the promise to achieve attractive upconversion white-light and gives rise to the enthusiasm in developing upconversion white-light illuminations for electro-optical devices. 3.3. Upconversion mechanism investigations The schematics of the populating and upconversion luminescence processes for the blue, green, and red emissions are drawn in Fig. 5. In Ho3 þ /Yb3 þ /Tm3 þ codoped systems, Yb3 þ ions act as sensitizers to absorb laser photons, the excited Yb3 þ ions in the 5 F5/2 state may transfer their excitation energy to Ho3 þ and Tm3 þ ions. First, the Ho3 þ ions in the ground state are excited to the 5I6 state via ET of neighboring Yb3 þ . Subsequent nonradiative relaxations of 5I6-5I7 populate the 5I7 level. In the second-step excitation, the same laser pumps the excited state Ho3 þ ions from 5I6 state to 5F4,5S2 states via ET, or from the 5I7 to 5F5 state via phonon-assisted ET. Besides, the 5F5 state can also be populated by the nonradiative decay of 5S2, 5F4 states. The Ho3 þ ions

Fig. 4. Color coordinates of the multicolor upconversion emissions for Ho3 þ /Yb3 þ / Tm3 þ /LiNbO3 polycrystals (1#: 0.05Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ / 2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3). ‘‘W’’ means the standard point of equal white-light (x¼ 0.33, y¼ 0.33). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

populated 5F4,5S2 states most relax radiatively to the ground state 5 I8 level, which cause green emissions. On the other hand, the red emissions of Ho3 þ ions are produced by radiative relaxations

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

from the 5F5 state to the ground state 5I8. The Tm3 þ ions in the ground state are excited to the 3H5 state by ET of neighboring Yb3 þ , and 3F4 state is populated by the nonradiative decay of 3H5 state. And then the Tm3 þ ions are excited to 3F2/3 state via phonon-assisted ET, and relaxed to 3H4 state by nonradiative transition. Finally, the excited state Tm3 þ ions are pumped from 3 H4 state to 1G4 state via phonon-assisted ET. The Tm3 þ ions populated 1G4 level relax radiatively to the ground state 3H6 level and the intermediate state 3F4 level, which generate blue emissions and red emissions, respectively. Besides, from the orthogonal experiment (SI), it is concluded that the ET process from Tm3 þ ions to Ho3 þ ions exists in the upconversion process. It is known that at low excitation power densities, for nearly an ‘‘unsaturated’’ upconversion process, the upconversion emission intensity (If) depends on the excitation power (P) according

Fig. 5. Energy level diagram and white-light upconversion mechanism for Ho3 þ / Yb3 þ /Tm3 þ /LiNbO3. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

1571

to the power law IfpPn, where n is the number of photons needed to produce the fluorescence which can be determined from the slope of the emission intensity versus the laser excitation power in a log–log plot [20]. However, at higher power densities, the fluorescence intensity becomes independent of the excitation power density [21,22]. Some previous studies indicated that the power dependences of upconversion emissions became linear (n ¼1) at high excitation densities due to the ‘‘saturation’’ of the upconversion processes [23]. To analyze the white-light upconversion mechanism in Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystal, the dependences of upconversion emission intensity on excitation power for all Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 samples are investigated under 980 nm excitation (Fig. 6). As shown in Fig. 6, the slopes of blue, green, and red emissions are 1.63, 1.03, and 0.99 for sample 1#, 1.67, 1.21, and 1.13 for sample 2#, 1.18, 0.94, and 0.93 for sample 3#, and 1.32, 1.10, and 1.05 for sample 4#. Besides, it is to be noted that the slopes of blue emissions for all the samples are unusually small at low excitation power densities, which are 1.00, 0.78, 0.59, and 0.58 for sample 1#, 2#, 3#, and 4#, respectively. The typical red and green emissions are two-photon processes and the blue emission is a three-photon process at 980 nm excitation. It can be seen that the slopes largely deviate from the typical values. This can be attributed to the competition between the linear decay and the upconversion processes for the depletion of the intermediate excited states, which is described by M. Pollnau et al. [23] theoretically, the so-called ‘‘saturation’’ phenomenon. So the luminescence temporal decays are measured and analyzed. In the measurement of luminescence decay dynamics, the continuous wave from 980 nm laser diode was tuned into pulsing by a signal generator. And the luminescence decay curves were measured by a digital phosphor oscilloscope (Tektronix DPO 4140). Fig. 7 shows the luminescence decay dynamics of 3F4-3H6 (Tm3 þ ), 5I7-5I8 (Ho3 þ ), and 5I6-5I8 (Ho3 þ ) in sample 3# at room temperature. It is clear that for all the emissions the intensities decay exponentially. The exponential lifetime constants in the sample 3# are deduced to be 3.3 ms for the 3F4-3H6 of Tm3 þ ions (l ¼ 1800 nm), 4.7 ms for the

Fig. 6. Dependences of upconversion emission intensity on excitation power for all Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals (1#: 0.05Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3). (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

1572

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

Fig. 7. Decay dynamics of the 3F4 level of Tm3 þ ions (a), 5I7 level of Ho3 þ ions (b), and 5I6 level of Ho3 þ ions (c) in sample 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3.

5

I7-5I8 of Ho3 þ ions (l ¼2000 nm), and 0.46 ms for the 5I6-5I8 of Ho3 þ ions (l ¼1150 nm). In order to understand the luminescence dynamics better, the rate equations under steady-state excitation are set up based on the model drawn in Fig. 5. The rate equations describing the excitation mechanisms in Yb3 þ –Tm3 þ system are as follows: N0 ¼ const:

ð3Þ

dN 1 ¼ WN b N0 WNb N 1 R1 N1 ¼ 0 dt

ð4Þ

dN 2 ¼ WN b N1 WNb N 2 R2 N2 ¼ 0 dt

ð5Þ

dN 3 ¼ WN b N2 R3 N 3 ¼ 0 dt

ð6Þ

where Nb, W denote the population density of excited state Yb3 þ at any time, the ET rate from Yb3 þ to Tm3 þ , respectively. Ni (i¼ 0–3) represents the population of Tm3 þ on level i. Ri (i¼ 1–3) is the total decay rate of Tm3 þ on level i. Based on the Eq. (3)–(6) and the upconversion decay dynamic of 3F4-3H6 of Tm3 þ ions, the lifetime of 3F4 (Tm3 þ ) level is so long (t0 ¼3.3 ms) that the linear decay term in Eq. (4) can be neglected (Rpt  1 ), then N1 EN0pP0, N2pP1, and N3pP2. So we can conclude that the slopes of blue and red emissions for Tm3 þ are close to 2. The rate equations describing the excitation mechanisms in Yb3 þ –Ho3 þ system are Eq. (7)–(11). N00 ¼ const:

ð7Þ

dN 01 ¼ b21 N02 W 0 N b N01 R01 N 01 ¼ 0 dt

ð8Þ

dN 02 ¼ W 0 Nb N 00 W 0 Nb N02 R02 N02 ¼ 0 dt

ð9Þ

dN 03 ¼ W 0 N b N01 þ b43 N04 R03 N 03 ¼ 0 dt

ð10Þ

dN 04 ¼ W 0 N b N02 R04 N 04 ¼ 0 dt

ð11Þ

where W0 , and bij denote, respectively, the ET rate from Yb3 þ to Ho3 þ , and the nonradiative relaxation rate from level i to j. Ni0 (i¼0–4) represents the population of Ho3 þ on level i. Ri0 (i¼ 1–4) is the total decay rate of Ho3 þ on level i. For the generation of green emission, it has only one upconversion path as mentioned above. Based on the Eqs. (7), (9), and (11) and the upconversion decay dynamic of the 5I6-5I8 of Ho3 þ ions, the lifetime of 5I6 (Ho3 þ ) level is 0.46 ms, so the linear decay term or the upconversion term in Eq. (9) cannot be neglected solely. If the linear decay (5I6-5I8) is the dominant depletion mechanism of level N20 , we can neglect the upconversion term in Eq. (9). It follows from Eqs. (7) and (9) that N20 pP1, and from Eq. (11) that N40 pP2. So it can be concluded that the slopes of green emissions for Ho3 þ approach to 2, corresponding to one limit. In contrast, if the upconversion process (5I6-5F4,5S2) is dominant, we can neglect the linear decay term in Eq. (9), resulting in N20 pP0 and N40 pP1. So it can be obtained that the slopes of green emissions for Ho3 þ approach to 1, corresponding to the other limit. The competition between linear decay and upconversion may give rise to a consequence that the slope of green emission for Ho3 þ is between the two limits (1 and 2). The generation of red emission has two pathways which have been mentioned above ((1) and (2) in Fig. 5). For the pathway (1), based on the Eqs. (7), (9), (10), and (11) and the upconversion decay dynamic of the 5I6-5I8 of Ho3 þ ions; similarly, we can have a conclusion that the slope of red emission for Ho3 þ is between 1 and 2 approximately. Besides, from Eq. (10), the W0 NbN10 can be neglected in pathway (1), so we find that N04 =N 03 ¼ R03 =b43 ¼ const:pP 0 . For the pathway (2), based on the Eqs. (7)–(10) and the upconversion decay dynamics of the 5I7-5I8

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

1573

That is because the probability of receiving energy for Ho3 þ ions will be increased with the increase in Ho3 þ content. And it is to be noted that the green, red emissions have also been reduced and the blue emission has been found invariable with the increasing of Tm3 þ content. It is speculated that decrease in green, red emissions are dominated by the lower probability of receiving energy for Ho3 þ ions with the increase of Tm3 þ content. And the invariant blue emission can be explained by the ET process from Tm3 þ to Ho3 þ : 3H4(Tm3þ )þ 5I8(Ho3 þ )-3F4(Tm3þ )þ 5 I6(Ho3 þ ). With the increasing of Tm3þ , the probability of ET will be increased greatly. It is also conjectured that the ET process from Tm3 þ to Ho3þ : 3H4(Tm3þ )þ 5I8(Ho3þ )-3F4(Tm3þ )þ 5I6(Ho3þ ) may be responsible for the unusual small slopes of Tm3þ at low excitation power densities, which is similar to the photon avalanche process.

4. Conclusion

Fig. 8. Intensity ratios R (g/r) of red emission to green emission as the increase of excitation power for all Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals (1#: 0.05Ho3 þ / 2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 2#: 0.10Ho3 þ /2.00Yb3 þ /0.20Tm3 þ /LiNbO3; 3#: 0.10Ho3 þ /2.50Yb3 þ /0.20Tm3 þ /LiNbO3; 4#: 0.10Ho3 þ /2.00Yb3 þ /0.25Tm3 þ /LiNbO3). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

of Ho3 þ ions, the lifetime of 5I7 (Ho3 þ ) level is so long (t0 ¼4.7 ms) that the linear decay term (R10 N10 ) in Eq. (8) can be neglected, the WNbN2 in Eq. (9) and the b43N40 in Eq. (10) can also be neglected in this case, resulting in N10 pP0 and N20 pN30 pP1. So we can conclude that the slope of red emission for Ho3 þ is about 1. Besides, it can be deduced that N 04 =N 03 ¼ R03 ðR01 þ W 0 N b Þ= b21 R04 pP1 in pathway (2). So it is proved that the slope of red emission for Ho3 þ is similar to 1. It is reasonable to attribute the lower slope of red emission than green emission to the pathway (2). Through above analysis, the competition between the linear decay and the upconversion process for the depletion of the intermediate excited states results in the deviation of slope from the typical value. The long lifetimes of 3F4(Tm3 þ ) level, 5 I7(Ho3 þ )level, and 5I6(Ho3 þ ) level in Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystal system are the foundation to the lower slopes compared with the typical values for the green, red, and blue emissions, which is very benefit to improve the efficiency of upconversion luminescence. The intensity ratios R(g/r) of the green emission to the red emission for all the samples are obtained at different excitation power (Fig. 8). As shown in Fig. 8, the values of R(g/r) approaches a constant, the excitation power has little influence on R(g/r). N 04 R03 0 Based on the analyses above, that is, ¼ const:pP in 0 ¼ b N 43 3 N0 R0 ðR0 þ W 0 N Þ pathway (1) and N04 ¼ 3 b1 R0 b pP 1 in pathway (2) for red 21 4 3 emission of Ho3 þ , it is speculated that the red emission is dominated by pathway (1). So we can have a conclusion that the red emission of Ho3 þ mainly originates from the nonradiative decay of green emission. Meanwhile, the analysis coincides with the experimental results, namely the intensity of green emission is stronger than the red emission as shown in Fig. 2. Moreover, on the basis of upconversion mechanisms, the upconversion emission spectra of Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 (Fig. 2) are discussed in detail. It is clear that the green, red emissions have been increased and the blue emission is nearly the same with the increase in Yb3 þ content. Upon increasing the concentration of Yb3 þ , more energy will be absorbed and transferred, besides, the photon count for blue emission is larger than red or green emission, so the intensities of green and red emissions are increased obviously compared to blue emission. When the Ho3 þ content is increased, the red and green emissions are greatly enhanced and blue emission is almost unchanged.

Ho3 þ /Yb3 þ /Tm3 þ /LiNbO3 polycrystals have been successfully prepared by the high temperature solid-state reaction method. Bright white-light is generated through frequency upconversion under 980 nm excitation at room temperature. CIE coordinate of the 0.1Ho3 þ /2.5Yb3 þ /0.2Tm3 þ /LiNbO3 is (0.34, 0.35), which is very close to the standard equal energy white-light illuminate (0.33, 0.33). The red, green, and blue emissions can be ascribed to Ho3 þ : 5F5-5I8, Tm3 þ : 1G4-3F4; Ho3 þ : 4S2, 5F4-5I8, and Tm3 þ : 1 G4-3H6 transitions, respectively. Efficient blue, green, and red upconversion emissions have been observed. The luminescent decay dynamics show that the decay time constants of the 3F4 level of Tm3 þ ions (l ¼1800 nm), 5I7 level of Ho3 þ ions (l ¼2000 nm), and 5I6 level of Ho3 þ ions (l ¼1150 nm) are 3.3 ms, 4.7 ms, and 0.46 ms, respectively. Rate equations are also set up to analyze the upconversion emission dynamics. The researches on upconversion mechanisms indicate that the competition between the linear decay and the upconversion process for the depletion of the intermediate excited states plays an important role in upconversion mechanism, and the red emission of Ho3 þ mainly originates from the nonradiative decay of green emission. Such upconversion white-light in Ho3 þ /Yb3 þ /Tm3 þ / LiNbO3 makes it an excellent candidate application in lighting device and instrument integration.

Acknowledgment We gratefully acknowledge financial support by the NSF of China (10732100), and the NSF of Heilongjiang Province (B200903). The authors also thank Prof. ZhiGuo Zhang from Department of Physics of Harbin Institute of Technology for their kindest help in the tests of upconversion spectra.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jlumin.2012.01.053.

References ¨ ¨ [1] S. Heer, K. Kompe, H.U. Gudel, et al., Adv. Mater. 16 (2004) 2102. [2] K. Teshima, S.H. Lee, N. Shikine, et al., Cryst. Growth Des. (2001). A–E. [3] H.T. Amorim, M.V.D. Vermelho, A.S. Gouveia-Neto, et al., J. Solid State Chem. 171 (2003) 278. [4] D.Q. Chen, Y.S. Wang, Y.L. Yu, et al., J. Solid State Chem. 181 (2008) 2763. [5] G. Glaspell, J. Anderson, J.R. Wilkins, et al., J. Phys. Chem. C. 112 (2008) 11527. [6] D.Q. Chen, Y.S. Wang, K.L. Zheng, et al., Appl. Phys. Lett. 91 (2007) 251903. [7] M. Ma, L.W. Yang, G.Z. Ren, et al., J. Lumin. 131 (2011) 1482.

1574

[8] [9] [10] [11] [12] [13] [14] [15] [16]

L. Xing et al. / Journal of Luminescence 132 (2012) 1568–1574

N.K. Giri, D.K. Rai, S.B. Rai, Appl. Phys. B 91 (2008) 437. V. Jambunathan, X. Mateos, M.C. Pujol, et al., J. Lumin. 131 (2011) 2212. Y. Dwivedi, Anita Rai, S.B. Rai, J. Appl. Phys. 104 (2008) 043509. F. Lahoz, I.R. Martin, J.M. Calvilla-Quintero, Appl. Phys. Lett. 86 (2005) 051106. F. Wang, R.R. Deng, J. Wang, et al., Nat. Mater. 10 (2011) 968. G.Z. Ren, S.J. Zeng, J.H. Hao, J. Phys. Chem. C. 115 (2011) 20141. ¨ X.H. Ma, H. Zhou, et al., J. Phys. Chem. C. 112 (2008) 15071. W.C. Lu, N.Q. Wang, X. Zhao, C.M. Li, et al., J. Lumin. 130 (2010) 1044. A. Rizzo, N. Solin, L.J. Lindgren, et al., Nano Lett. 10 (2010) 2225.

[17] E. Cantelar, G.A. Torchia, J.A. Sanz-Garcia, et al., Appl. Phys. Lett. 83 (2003) 2991. [18] H. Gong, D.L. Yang, X. Zhao, et al., Opt. Mater. 32 (2010) 554. [19] G.Y. Chen, Y. Liu, Y.G. Zhang, et al., Appl. Phys. Lett. 91 (2007) 133103. [20] X.H. Pan, J.D. Yu, S. Yoda, et al., J. Alloys Compd. 509 (2011) 7504. [21] D. Li, B. Dong, X. Bai, et al., J. Phys. Chem. C. 114 (2010) 8219. [22] M. Pollnau, J. Alloys Compd. 341 (2002) 51. ¨ [23] M. Pollnau, D.R. Gamelin, S.R. Luthi, et al., Phys. Rev. B. 61 (2000) 3337.