# Fileset

[cmb_apss.pdf](https://mdr.nims.go.jp/filesets/e2d62201-d00b-463e-a65e-964faf2450da/download)

## Creator

[TODOROKI, Shin-ichi](https://orcid.org/0000-0003-3986-1900), INOUE, Satoru

## Rights



## Other metadata

[Combinatorial Fluorescence Lifetime Measuring System for Developing Er-Doped Transparent Glass Ceramics](https://mdr.nims.go.jp/datasets/6bd16158-2c55-45e3-bb5e-8bc37868bdbe)

## Fulltext

Combinatorial Fluorescence Lifetime MeasuringSystem for Developing Er-Doped Transparent GlassCeramicsS. Todoroki∗ and S. InoueAdvanced Materials Laboratory, National Institute for Materials Science,Namiki 1-1, Tsukuba, Ibaraki 305-0044, JAPANAbstractFluorescence lifetime of Er3+ was measured for F-doped tellurite glasses with parallel heattreatment under a temperature gradient atmosphere in order to find the annealing condi-tion to make transparent glass ceramics in which Er3+ ions are located in the precipitatedcrystals. The preparation and annealing of the samples were performed in a vertical tem-perature gradient furnace, where molten glass was sucked into a pre-heated Pyrex glasstube. The annealing temperature range is between 350◦C and 800◦C. After the annealingtreatment, time-resolved fluorescence emission of Er3+ (1.55um; excitation light source is977nm) were measured sequentially along the tube. The lifetime of the emission was about2.6 msec for as-prepared glass. We have found that the lifetime increased to 5.2 msec whenthe glass was annealed at 470◦C for 5 min and 550◦C for 5 min successively, although itstransparency was lost. This increase implies that the Er3+ ions are embedded in fluorine-rich phase to bring about reduced non-radiative emission. We are now continuing to findthe condition to get transparent glass ceramics.Key words: tellurite glass, erbium, fluorescence lifetimePACS:06.60.-c, 07.20.Hy, 07.60.-j, 78.55.-mPreprint submitted to Elsevier Science 7 September 20081 IntroductionTransparent glass ceramics, especially nano-crystallite-dispersed glass materialsplay an important role in optical and/or photonics devices because they combinemerits of glass and crystal materials[1]. One of the promising works was doneby Wang and Ohwaki[2]. They made a aluminosilica oxyfluoride glass sample inwhich Er3+-doped PbF2 nano-crystallites are precipitated. After the heat treatmentfor precipitation, the sample showed 100 times larger green upconversion fluores-cence of Er3+. This is due to a significant decrease of non-radiative decay of Er3+,that is, reduced vibration energy of surrounding ligands; oxygen to fluorine.In order to determine the recipe of such kind of materials, the following parametersshould be optimized. (1) glass composition in which Er3+ ions are incorporated inprecipitated non-oxide phase, (2) the first annealing temperature and time for nucle-ation, and (3) the second annealing temperature and time for crystal growth. There-fore, development of new transparent glass ceramics needs considerable experi-ments and/or empirical sense, and only few recipes have been found until now[1].Such investigations would be accelerated efficiently if combinatorial methodologyis applied.Recently, the authors have developed a combinatorial evaluation system for ther-mal stability of glass materials with low-softening temperature[3,4]. In this system,about-40-cm-long glass sample can be prepared quickly and successively annealed∗ Corresponding author.Email address:TODOROKI.Shin-ichi@nims.go.jp (S. Todoroki).URL: http://www.geocotoes.com/Tokyo/1406/ (S. Todoroki).2under a temperature gradient of from 350◦C to 800◦C(see Fig. 1(a)). Thus, thissystem is appropriate for surveying annealing conditions.In this study, Er3+-doped tellurium oxyfluride glasses are annealed and the lifetimeof Er3+ fluorescence of 1.5µm is evaluated by a newly constructed equipment. Tel-lurite glass system is chosen because of the following reasons. (1) Its melting tem-perature is so low (∼ 800◦C) that our combinatorial system can treat with the melt.(2) Tellurite glass is known to show attractive properties such as non-linear opticaleffect[5], acousto-optics effect[6] and ultra-wideband Raman amplification in fiberform[7]. (3) The authors have succeeded to make a new optical coupling structurewhere a pair of optical fibers are spliced via tellurite glass melt recently[8], whichis expected to realize new optical devices.2 ExperimentalThe glass composition in this study is 75TeO2-22MgO-2YF3-1ErF3(in mol%), whichis chosen because of the following reasons. (1) TeO2-MgO system gives a largeglass forming region[9]. (2) ionic radii of Y and Er are nearly the same. (3) TeO2-YO3/2 system gives no glass forming region[9]. A mixture of reagent grade powermaterial of each components are melted at about 800◦C in a Pt crucible placedat the bottom of the vertical temperature gradient furnace shown in Fig. 1(a). APyrex glass capillary (75cm× 8mm OD× 1.5mm ID) was inserted into the fur-nace from the top so as to be heated under a temperature gradient of800 ∼ 350◦Cin 50cm-span. The temperature profile inside the capillary along its length had beendetermined in advance by inserting a thermo-couple into a empty capillary.The inserted capillary was held in the furnace for 10 min in order to reach thermally3equilibrium state. Then, the glass melt was sucked into the capillary by a vacuumpump, which was connected to the top of the capillary, to make a sample library.The melt stopped rising within 1 second because its temperature decreases and itsviscosity increases during the run of about 300mm. Then the library was movedto an appropriate positions in the furnace to be annealed at different temperaturein parallel. After the annealing treatment, the capillary is pulled out completely.Three samples which are annealed differently were prepared(see Fig. 1(b)); (0) noannealing, (1) annealed for 5min, and (2) another successive annealing at a differentposition for 5min.The sample library is mounted in a fluorescence lifetime measuring equipmentshown in Fig. 2. Its light source is a CW Ti:Sapphire laser of 977nm (Spectra-Physics, 3900) pumped by LD-pumped YAG laser (SpectraPhysics, Millenia Vs).Its detector is a multi-channel spectrometer (Soma Optics, S-2700) which can storetime-resolved spectra in every millisecond. These three are optically connectedthrough multi-mode optical fibers. For fluorescence lifetime measurement, the CWlight is chopped by a mechanical shutter and fluorescence decay curve for 1.5µmband of Er3+ (4I13/2 →4I15/2) is stored. This measurement is performed sequen-tially along the sample library in every 1mm. The whole system is controlled by areal-time operating system (ART-Linux) on a personal computer.3 ResultsFigure 3 shows the annealing condition of the sample libraries, that is, every 1mmsegment in the libraries is plotted as a function of annealing temperatures, the 1stannealing temperature,T1 and the 2nd,T2. For example, the annealing conditionsfor the segment A2 in the Sample (2) and A1 in the Sample (1) are(T1, T2) =4(470, 550) andT1 = 470, respectively. Sample (0) is also plotted for reference. A0and A1 are located at the same position in the library. Consequently, the relationamong these segments is represented as follows.(melt)↓A01st anneal−−−−−−−−−−−−−−−−−→T1 = 470 for 5min ↓A12nd anneal−−−−−−−−−−−−−−−−−→T2 = 550 for 5min ↓A2(1)As the heat treatment proceeded, crystallization occurred and the portion of trans-parent segments in the library decreased. Appearance of each segment is judged byeye and plotted in Fig. 3 by the following way. Completely transparent segment isplotted as©, white segment as·, the rest, i.e. not white but not completely trans-parent, as•. For convenience, the third one is called as “opaque” from here. Theappearance of the segments in Sample (0) varies along the length because theirinitial temperatures in quenching are different each other.Fluorescence spectra of Er3+ for the segment of A0, A1 and A2 (see Fig. 3) isshown in the left part of Fig. 4. The shapes of the spectra forA0 andA1 are almostthe same, while that for A2 is different from the others. Fluorescence lifetime ofeach segment is calculated from the stored decay curves of 1.533nm fluorescenceon the assumption that the curve is expressed by a single exponential(see the rightpart of Fig. 4). Again the lifetime ofA0 andA1 are nearly the same, while that of A2is two times larger than the others. Figure 5 shows the calculated lifetime plottedalong the library or as a function of the 1st annealing temperature,T1. An increaseof lifetime is observed after the 2nd heat treatment. For the the segments in Sample(2) at aroundT1 = 470 (A2 in Fig. 3), not only lifetime but also the shape of theirfluorescence spectra was also changed(see Fig. 4).54 DiscussionAlthough the lifetime does not vary after the 1st heat treatment (compare Sample(0) and (1) in Fig. 3), an increase is observed after the 2nd one. So let us examinethe segments shown in Eq. (1) and the following(see Fig. 3).(melt)↓B01st anneal−−−−−−−−−−−−−−−−−→T1 = 550 for 5min ↓B1(2)Among these segments, only A2 shows the largest lifetime value of about 5.2 msec.The rest shows nearly the same value of about 2.6 msec. Thus, the increase inlifetime occurred in the 2nd heat treatment ofT2 = 550. On the contrary, the 1stannealing ofT1 = 550 did not bring about any increase inB1. Thus, the cause ofthis increase is considered to be due to a phenomena during the 1st annealing ofT1 = 470.As for the appearance of the segments, the annealing of 550◦C made A2 white butB1 remain transparent. The white color must be due to the precipitated crystalscontaining TeO2, for it is the main component of this glass. Thus, it is natural toconsider that the heat treatment at 470◦C on this glass composition caused nucle-ation and the further annealing at 550◦C promoted crystal growth on the generatednuclei.Therefore, the change in lifetime can be explained in the viewpoint of coordinationenvironment around Er3+. In the as-prepared segments, Sample (0), it is reason-able to assume that Er3+ ions are coordinated mainly by oxygen anions. The situ-ation is not changed after the 1st heat treatment because both the lifetime and theshape of fluorescence spectra are not changed. After the 2nd annealing at around550◦C, Er3+ ions are not incorporated by the precipitated oxide crystals and stay in6fluorine-rich residual phase.When Er3+ ions in the residual phase are coordinated by fluorine anions, electronsin the excited state of Er3+ stays longer, in other words, the non-radiative decayrate is smaller compared with that in oxide matrix, because the vibration energy ofsurrounding fluorine is smaller than that of oxygen. The change in spectral shapeshown in Fig. 4 also support the change in coordination around Er3+.Among the samples in this study, the segments showing increased lifetime simulta-neously lost their transparency. This combinatorial evaluation system is, however,proved to be a powerful tool to develop transparent glass ceramics because onlythree samples, with one preparation process and one measurement per each, canprovide rich information described above. We are now continuing to find the con-dition to prepare transparent glass ceramics.5 ConclusionCombinatorial evaluation system for Er3+-doped glass materials is developed. Itincludes a parallel annealing furnace for sample preparation and a sequential mea-surement equipment for fluorescence lifetime. Er3+-doped tellurium magnesiumoxyfluoride glass samples are prepared and annealed at various temperatures. Then,the lifetime of 4I13/2 →4I15/2 fluorescence (∼1.5µm) is evaluated. The lifetimeshowed a twofold increase when the glass is annealed at 470◦C for 5min and suc-cessively at 550◦C for 5min. This is considered to be due to a change in cordinationaround Er3+ ions, that is, oxygen ligands are replaced by fluorine anions during thegrowth of oxide crystals in the 2nd heat treatment. This system is useful to surveyconditions for making transparent glass ceramics.7References[1] P. A. Tick, Ultra-transparent glass-ceramics, in: Proc. 6th Int. Symp. on New Glass,1997, pp. 99–120.[2] Y. Wang, J. Ohwaki, New transparent vitroceramics codoped with Er3+ and Yb3+ forefficient frequency upconversion, Appl. Phys. Lett. 63 (24) (1993) 3268–3270.[3] S. Todoroki, S. Inoue, T. Matsumoto, Combinatorial evaluation system for thermalproperties of glass materials using a vertical furnace with temperature gradient, Appl.Surface Sci. 189 (3–4) (2002) 241–244.[4] S. Todoroki, T. Matsumoto, S. Inoue, Rapid and quantitative determination ofcrystallization tendency of zinc tellurite glass melt by using temperature-gradientfurnace, in: I. Takeuchi, J. M. Newsam, L. T. Wille, H. Koinuma, E. J. Amis(Eds.), Combinatorial and Artificial Intelligence Methods in Material Science, Vol.700 of Material Research Society Symposium Proceedings, Material Research Society,Pennsylvania, USA, 2002, pp. 209–214.[5] K. Tanaka, K. Kashima, K. Hirao, N. Soga, A. Mito, H. Nasu, Second harmonicgeneration in poled tellurite glasses, Jpn. J. Appl. Phys. 32 (6B) (1993) L843–L845.[6] T. Yano, A. Fukumoto, A. Watanabe, Tellurite glass: A new acousto-optic material, J.Appl. Phys. 42 (10) (1971) 3674–3676.[7] A. Mori, H. Masuda, K. Shikano, K. Oikawa, K. Kato, M. Shimizu, Ultra-widebandtellurite-based raman fibre amplifier, Electron. Letters 37 (24) (2001) 1422–1443.[8] S. Todoroki, A. Nukui, S. Inoue, Formation of optical coupling structure between twoends of silica glass optical fibers by inserting tellurite glass melt, J. Ceram. Soc. Jpn.110 (5) (2002) 476–478.[9] M. Imaoka, Glass Handbook, Asakura Shoten, Tokyo JAPAN, 1975, Ch. 8.3, pp. 880–903, (in Japanese).8Figure CaptionsFig. 1 (a: left) The structure of vertical temperature gradient furnace used in thisstudy and (b: right) an illustration showing 2-step annealing procedure(see text).Fig. 2 The fluorescence lifetime measuring equipment used in this study(see text).Fig. 3 Annealing condition of the sample libraries and appearance of the annealedglass segments inside. (0) with no annealing treatment for reference. (1) with 1stannealing for 5min only. (2) with two successive heat treatments, each for 5min.Thick line plotted by©: completely transparent segment, thin line by· : white, andmedium line by•: opaque(see text). An and Bn indicate specific glass segments.These notations are also used in Fig. 4 and Fig. 5.Fig. 4 (Left) Fluorescence spectra of Er3+ and (right) decay curves of 1533nmfluorescence for the glass segments of A0, A1 and A2(see Fig. 3). For the decaycurve, triangle points are not used for lifetime calculation.Fig. 5 Lifetime of 1.533nm fluorescence of Er3+ plotted along the library or as afunction of the 1st annealing temperature,T1.9Vacuum PumpGlass Melt5-zone furnaceGlass tubeLoading system 2Loading system 1TimeTemp. .... T1.... T2(0) (1) (2)Fig. 1. (a: left) The structure of vertical temperature gradient furnace used in this study and(b: right) an illustration showing 2-step annealing procedure(see text).Light SourceSpectrometerSampleMulti-ch.ART-LinuxMoving MechanismFig. 2. The fluorescence lifetime measuring equipment used in this study(see text).10Fig. 3. Annealing condition of the sample libraries and appearance of the annealed glasssegments inside. (0) with no annealing treatment for reference. (1) with 1st annealing for5min only. (2) with two successive heat treatments, each for 5min. Thick line plotted by©:completely transparent segment, thin line by· : white, and medium line by•: opaque(seetext). An and Bn indicate specific glass segments. These notations are also used in Fig. 4and Fig. 5.11Fig. 4. (Left) Fluorescence spectra of Er3+ and (right) decay curves of 1533nm fluorescencefor the glass segments of A0, A1 and A2(see Fig. 3). For the decay curve, triangle pointsare not used for lifetime calculation.12Fig. 5. Lifetime of 1.533nm fluorescence of Er3+ plotted along the library or as a functionof the 1st annealing temperature,T1.13