Functionalized Inorganic Fluorides (Synthesis, Characterization & Properties of Nanostructured Solids) || Preparation and Properties of Rare-Earth-Containing Oxide Fluoride Glasses

18 Preparation and Properties of Rare-earth-Containing Oxide Fluoride Glasses Susumu Yonezawa, Jae-ho Kim and Masayuki Takashima Graduate School of Engineering, University of Fukui, Bunkyo 3–9–1, Fukui 910–8507, Japan 18.1 Introduction Oxide fluoride glasses contain two different anions that have different valence electrons and different degrees of polarization. It is interesting to compare properties of the oxide fluoride glass with those of the oxide or fluoride glass. No reports have described a glass that can contain a rare-earth fluoride with such high contents. The properties of oxide fluoride glasses have not yet been summarized systematically, as they have been for oxide glasses or fluoride glasses. It is interesting to study preparation processes and character- istics of these glasses to develop new functional materials. We have reported the prepara- tion and properties of oxide fluoride glasses containing rare-earth elements [1–4]. The oxide fluoride glasses are anticipated as a new optical material. Every rare-earth element has unique optical properties because of its arrangement of electrons in the 4f orbital. It is important to find matrices into which these rare earth elements can be doped and that can exhibit their performance to develop new optical and magnetic materials. During the preparation process [5] of some materials of solid electrolyte containing rare earth elements, one type of glass was obtained by chance. Based on quantitative analyses using EPMA, this glass contained all the starting elements (O, F, and two species of rare Functionalized Inorganic Fluorides: Synthesis, Characterization & Properties of Nanostructured Solids Edited by Alain Tressaud © 2010 John Wiley & Sons Ltd. ISBN: 978-0-470-74050-7 earth elements). In addition, Si and Al, which were not contained in starting materials, were detected in the glass. The binary rare-earth oxide fluoride solid electrode was prepared using a solid state reaction between a rare earth oxide (Ln2O3) and a rare earth fluoride (Ln’F3) at a temperature higher than 1000 �C. During the mixing process, an agate ball mill was used to mix the starting materials. During calcination, the mixture was heated to over 1000 �C after it was packed into an aluminium tube. The Si and Al in the glass had to be supplied from these materials as glass network formers. This glass contained more than 70wt% of rare earth elements, as inferred from the results of the analysis using EPMA (fundamental parameter method (FP method)). Actually, the FP method is the quasi-quantitative analysis method using the instrument’s standard library for the peak intensities of the elements. The resultant glass was a new oxide fluoride glass containing large amounts of rare-earth elements. Several reports have described preparation of glasses containing rare earth elements for use as optical or magneto-optical materials [6–10]. 18.2 Preparation and Basic Characteristics of Oxide Fluoride Glasses Containing LnF3 This section presents a description of the preparation process and basic characterization of oxide fluoride glasses containing CeF3 and NdF3 (light rare earth), TbF3 (middle rare earth), and HoF3 (heavy rare earth). Based on those results, preparation methods were extended to other rare-earth elements in the lanthanide series. Properties of those glasses were compared and summarized. 18.2.1 Preparation of Oxide Fluoride Glasses Containing LnF3 The glass in the NdF3-SiO2-Al2O3 system has been obtained once. However, the reprodu- cibility of the synthesis was not confirmed. Detailed analyses of the glass revealed that AlF3 was contained in the product in the NdF3-SiO2-Al2O3 system, meaning that some part of NdF3 was hydrolysed to form HF; this HF has reacted with Al2O3 to form AlF3. Controlling the content of AlF3 is apparently an important factor to prepare the oxide fluoride glass reproducibly. However, the higher temperature for melting the sample causes large variation of the AlF3 content in the product because AlF3 sublimes at a temperature that is remarkably higher than 700 �C. In addition, rare-earth trifluorides readily undergo pyrohydrolysis to form their oxides at temperatures higher than 1000 �C [11]. The network forming oxide GeO2 was chosen because it has the lowest melting point among SiO2 (1730 �C), Al2O3 (2045 �C), and GeO2 (938 �C) in this study. In addition, various fluorides were tested as glass network modifiers and additives to lower the melting point. Consequently, the oxide fluoride glasses containing light rare earth (NdF3) were obtainable reproducibly in the system of NdF3-AlF3-GeO2. The glass was analysed using the ZAF method with EPMA, which showed clearly that the obtained glass was incapable of maintaining its nominal composition. The Al content, especially, tended to decrease from the starting mixture ratio. The Al contents in the glass never became greater than 34mol%, even if the starting material contained more than 34mol% Al. The Nd content 546 Functionalized Inorganic Fluorides reached 63mol% as the maximum value. Glasses with a rare earth content higher than 30mol% have not been reported to date. These oxide fluoride glasses with a high content of rare earths are anticipated as new functional materials. Figure 18.1 presents a phase diagram of TbF3-AlF3-GeO2. The maximum TbF3 content was 50mol% in the glass. Reportedly, rare earth trifluorides hydrolysed at a temperature higher than 700 �C [11,12]; a lower melting temperature is preferred in order to avoid composition changes during high-temperature processing. Figure 18.2 shows that the results of quantitative analyses obtained by EPMA measurement of TbF3-AlF3-GeO2 glasses indicated that the cationic compositions in the glass deviated from that of the starting ratio. Furthermore, the composition change in the glass increased concomitantly with increasing AlF3 content. This system contained no glass with 50mol% TbF3, even though the glass was obtained from a starting mixture containing 50mol% TbF3. The amount of AlF3 must be optimized to realize good reproducibility because the excess of AlF3 drastically alters the glass composition from the starting ratio. The BaF2 was added to starting materials instead of AlF3 because BaF2 sublimates only slightly, but hydrolyses at melting temperatures of 1200 �C. Figure 18.3 presents a phase diagram of the NdF3-BaF2-GeO2 system. This system had a large glass formation compo- sition area. The NdF3 content in the glass reached 40mol% as a nominal composition. Figure 18.4 presents the cationic composition in the glasses of NdF3-BaF2-GeO2 system resulting from quantitative analyses using EPMA. The glass obtained in this system maintained the starting cationic compositions better than the NdF3-AlF3-GeO2 system did. Apparently, by adding BaF2, the oxide fluoride glasses were obtained reproducibly and the nominal composition was maintained in the glass product. Overall, BaF2 was a better component to prepare the oxide fluoride glass containing NdF3 than AlF3 was. TbF3 / mol% AlF3 / mol% GeO2 / mol% Figure 18.1 Phase diagram of TbF3-AlF3-GeO2 system shown by nominal composition. Closed and open circles respectively represent crystal and glass phases. Reproduced from [2] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 547 Based on results of the glass preparation in the ternary system glass, preparation of 40LnF3-20BaF2-40GeO2 (Ln; La–Nd, Sm–Lu) glasses by melting at 1200 �C for 1.5 h was attempted. Only in cases of NdF3, SmF3, EuF3, and GdF3 (middle rare-earth fluorides) were glasses obtained. Al / mol% Tb / mol% Ge / mol% Figure 18.2 Cationic composition in the glass of TbF3-AlF3-GeO2 system glasses measured using EPMA. Reproduced from [2] by permission of Elsevier BaF2 / mol% NdF3 / mol% GeO2 / mol% Figure 18.3 Phase diagram of NdF3-BaF2-GeO2 system shown by nominal composition. Closed and open circles respectively represent crystal and glass phases. Reproduced from [2] by permission of Elsevier 548 Functionalized Inorganic Fluorides Choosing the system containing HoF3, a glass-formation process with a heavy rare-earth trifluoride was investigated. Several chemicals were tested for use as additive components to obtain oxide fluoride glasses at lower melting temperatures and with lower weight loss during melting. Finally, results show that both BaF2 and AlF3 are necessary to obtain, reproducibly, the glasses containing HoF3; in addition, accurate controls of their contents are needed. Figure 18.5 portrays the glass-forming condition relative to the quantity of AlF3 and melting temperature. The oxide fluoride glasses were obtainable reproducibly by adding more than 6mol% AlF3; excessive AlF3 changed the glass composition, as described above. The AlF3 content would be better maintained at ca. 10mol% to obtain glass reproducibly without a composition change through the melting process. It was possible to prepare the glass at 1175 �C in case of GeO2 system, although about 1300 �C was needed in the case of the SiO2 system. Figure 18.6 portrays a phase diagram of the HoF3-BaF2-AlF3-GeO2 system in which the AlF3 content was fixed at 10mol%. This diagram shows that this glass system has a wide composition range for glass formation. An oxide fluoride glass was prepared with maximum content of HoF3 of 50mol%. Figure 18.7 presents results of quantitative analyses of HoF3-BaF2-AlF3-GeO2 glasses using EPMA (ZAF method). Consequently, only slight deviation of the cationic composition in the glass product from the nominal composition of the starting mixture was recognized. Regarding 10HoF3-10BaF2- 10AlF3-70GeO2 and the 50HoF3-10BaF2-10AlF3-30GeO2 glasses, the respective analy- tical cationic ratios in the product glasses were Ho: Ba: Al: Ge¼ 6.1: 9.0: 11.6: 73.2 and 41.4: 13.6: 8.8: 36.1. These values approximately agreed with the nominal composition. Finally, the glasses were obtained reproducibly in the LnF3-BaF2-10AlF3-GeO2 system for Ln ¼ La – Lu (except for Pm). Ba / mol% Nd / mol% Ge / mol% Figure 18.4 Cationic composition in the glass of NdF3-BaF2-GeO2 system measured using EPMA. Reproduced from [2] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 549 0 5 10 15 20 25 1200 1190 1180 1170 1160 1150 1140 x / mol% Te m pe ra tu re / ° C Figure 18.5 Classification of the products for 10HoF3-(30-x)BaF2-xAlF3-60GeO2 systems at various melting temperatures. * and • respectively signify glass and crystalline phases. Reproduced from [2] by permission of Elsevier BaF2 + 10AlF3 / mol% HoF3 / mol% GeO2 / mol% Figure 18.6 Phase diagram of HoF3-BaF2-10AlF3-GeO2 shown by nominal composition. The AlF3 content was fixed at 10mol%. Closed and open circles respectively denote crystal and glass phases. Reproduced from [2] by permission of Elsevier 550 Functionalized Inorganic Fluorides Regarding the CeF3-BaF2-10AlF3-GeO2 and CeF3-BaF2-10AlF3-SiO2 systems, the col- our of the glasses prepared in the manner described above (in an argon atmosphere) was brown. This might result from the mixed valency of Ce3þ and Ce4þ because of the decomposition of CeF3 during the glass preparation process. To prepare pale yellow glasses in the CeF3-BaF2-AlF3-SiO2 system, the melting process was conducted in a CO atmo- sphere. At the composition, around 10CeF3-20BaF2-10AlF3-60SiO2 or 20CeF3-10BaF2- 10AlF3-60SiO2, the glass was reproducibly prepared by heating at 1300 �C for 90min. Optimizing the conditions such as heating temperature, holding time and heating rate, and controlling the hydrolysis of rare earth fluoride, the composition range to prepare the glass was extended slightly. Brown glass tended to be obtained in cases of higher CeF3 contents, even in CO. It was possible to produce a glass containing 40mol% of CeF3 as a maximum. During the glass-preparation process, the sample weight decreased about 18wt%. Fluorine must be lost completely if this weight loss occurred because of the hydrolysis only. However, it was confirmed byX-ray fluorescent spectroscopy (XFS) measurement that 43 and 22% of fluorine in the starting mixture remained respectively in the products prepared in CO and Ar atmospheres. Therefore, the hydrolysis of CeF3 was controlled in a CO atmosphere although CeF3 was hydrolysed during both preparation processes in CO and Ar atmospheres. The hydrolysis of LnF3 proceeds to produce Ln2O3 via LnOxFy [11]. For cerium, the final product of hydrolysis is CeO2, which might be derived from Ce2O3 and/or CeOxFy: the hydrolysis of CeF3 includes several reaction processes. For the glasses examined here, the reaction from CeOxFy to CeO2 is apparently controlled in a CO atmosphere. For melting in an Ar atmosphere, the oxidative decomposition of CeOxFy (reaction with H2O and O2 to produce HF) proceeds and Ce4þ is generated in the sample (x value increases). The total reaction from CeF3 to CeO2 might be characterized by the following equation: 2CeF3 þ Ba + 10Al / mol% Ho / mol% Ge / mol% Figure 18.7 Cationic composition in the glass of HoF3-BaF2- AlF3-GeO2 system measured using EPMA. Reproduced from [2] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 551 3H2O þ 1/2O2 ¼ 2CeO2 þ 6HF. The O2 that is ubiquitous in the atmosphere might be an oxidizer in this case. Furthermore, CO might greatly reduce the partial pressure of O2 in the atmosphere by CO þ 1/2O2 ¼ CO2 equilibrium. 18.2.2 Density and Refractive Index Densities and refractive indexes of NdF3-Al2O3-SiO2, NdF3-AlF3-GeO2, NdF3-BaF2- GeO2, TbF3-BaF2-AlF3-GeO2, HoF3-BaF2-AlF3-GeO2, and 10LnF3-20BaF2- 10AlF3- 60GeO2 system glasses are presented in Figure 18.8. In that figure, (–) and (þ) respectively correspond to data of typical oxide and fluoride glasses taken from the database, INTERGLAD [13]. The oxide glasses in this case include quartz glass [14], borosilicate glass (COVER 18-18; Iwaki Glass Co. Ltd.), PbO-WO3-P2O5-CdO-TiO2 glass [15], and various oxide glasses in the literature found in the INTERGLAD database [13]. In addition, ZBLAN [16–21] glass and others found in the database were chosen as fluoride glasses. Curves (A) and (B) shown in Figure 18.8 were derived respectively by fitting the equation (n ¼ C1exp (C2d) where C1 and C2 are constants) to data of oxide glasses and fluoride glasses. These empirical curves emphasize characteristics of the oxide and the fluoride glasses. Plots of the oxide fluoride glasses are between curves (A) and (B). Especially, TbF3-BaF2-10AlF3-GeO2, HoF3-BaF2-10AlF3-GeO2, and LnF3-BaF2-10AlF3-GeO2 (Ln; Y–Nd, Sm–Lu) glasses used in this study exhibited a constant refractive index irrespective of density. The relation between density and refractive index is generally given as the Lorentz–Lorenz equation [15]. n2 � 1 n2 þ 2V¼ 4pN� 3 ¼RL (18:1) Therein, n, V, N, �, and RL respectively represent the refractive index, molecular volume, Avogadro’s number, polarizability, and molecular refraction. The Gladstone–Dale equa- tion is derived from Equation (18.1) when n is close to 1, as shown below [15]. n� 1¼ R v ¼ 2pN� v ¼ 2pN� M d¼RGd (18:2) In those equations,M, d, and RG respectively denote the molecular weight, density, and the Gladstone–Dale constant. Generally, this equation is used to express the relation between density and the refractive index of a glass. The RG value varies with glass composition. The RG of the fluoride, the oxide, and the oxide fluoride glasses were, respec- tively, 0.9 � 10�4 – 1.4 � 10�4, 1.8 � 10�4 – 2.9 � 10�4, and 1.0 � 10�4 – 1.8 � 10�4. Data of the NdF3-Al2O3-SiO2 system closely approximated the curve (A) corresponding to oxide glasses in Figure 18.8. The glass network of NdF3-Al2O3-SiO2 was suggested to be an oxide such as Al2O3 and SiO2. However, in the case of NdF3-AlF3-GeO2 or NdF3-BaF2- GeO2, systems in which the glass matrix consists of both an oxide and a fluoride, the plots of the refractive index against the density deviated from curve (A) to curve (B), 552 Functionalized Inorganic Fluorides corresponding to that of the fluoride glasses. For LnF3-BaF2-AlF3-GeO2 glasses, the plots were located between curves (A) and (B). Data of glasses that consisted of 70mol% oxide and 30mol% fluoride (10HoF3-10BaF2-10AlF3-70GeO2) closely approximated curve (A) for oxide glasses. Data of glasses consisting of 30mol% oxide and 70mol% fluoride (50HoF3-10BaF2-10AlF3-30GeO2) closely approximated curve (B) for the fluoride glass area. Consequently, the oxide fluoride glasses were located at the intermediate area in the relation between the refractive index and the density. No data for the simple oxide or fluoride glass was reported in this region previously. 18.2.3 Glass Transition Temperature Respective glass transition temperatures of the HoF3-BaF2-10AlF3-GeO2 glasses, TbF3-BaF2-AlF3-GeO2 glasses, TbF3-BaF2-10AlF3-SiO2 glasses, and 10LnF3- 20BaF2-10AlF3-60GeO2 glasses are presented in Tables 1–4. The glass transition Density / × 103 kg m–3 2 3 4 5 6 7 8 R ef ra ct ive in de x / – 2.2 2.0 1.8 1.6 1.4 1.2 Figure 18.8 Relation between density and the refractive index for various glasses. Reproduced from [2] by permission of Elsevier. � : oxide glass* • : LnF3-BaF2-AlF3-GeO2 (Ln; Y–Nd, Sm–Lu)þ : fluoride glass* * : TbF3-BaF2-AlF3-GeO2 ~ : NdF3-Al2O3-SiO2 & : HoF3-BaF2-AlF3-GeO2 n : NdF3-AlF3-GeO2 & : TbF3-BaF2-AlF3-SiO2 ^ : NdF3-BaF2-GeO2 (* : data from INTERGRAD[14]) Rare-earth-Containing Oxide Fluoride Glasses 553 temperatures of BaF2-free glasses were higher than those of quaternary systems such as LnF3-BaF2-AlF3-GeO2 glasses. Glasses containing a larger amount of BaF2 as a net- work modifier tend to have lower glass transition temperatures because divalent ions such as Ba2þ cut the glass network. Fluoride glasses such as ZBLAN [16–21] exhibit a glass transition temperature around 300–400 �C. Oxide fluoride glasses obtained in this study must be more thermally stable than the fluoride glasses. Table 18.1 Glass transition temperatures of HoF3-BaF2-AlF3-GeO2 glasses. Reproduced from [2] by permission of Elsevier HoF3:BaF2:AlF3:GeO2 Tg / oC 10:10:10:70 592.0 20:10:10:60 599.8 30:10:10:50 614.4 40:10:10:40 608.3 50:10:10:30 572.3 10:20:10:60 599.9 20:20:10:50 600.9 30:20:10:40 585.7 40:20:10:30 564.5 50:20:10:20 541.8 10:30:10:50 576.8 20:30:10:40 561.9 Table 18.2 Glass transition temperatures of TbF3-BaF2-AlF3-GeO2 glasses. Reproduced from [2] by permission of Elsevier TbF3:BaF2:AlF3:GeO2 Tg / oC 10:10:10:70 606.7 20:10:10:60 631.1 30:10:10:50 619.9 40:10:10:40 628.8 50:10:10:30 618.2 10:20:10:60 578.4 20:20:10:50 603.4 30:20:10:40 593.4 40:20:10:30 574.0 10:30:10:50 593.0 20:30:10:40 583.4 30:30:10:30 571.9 30:0:10:60 675.2 40:0:10:50 669.1 50:0:10:40 639.9 30:0:20:50 664 40:0:20:40 610.8 30:0:30:40 605.1 40:0:30:30 622.2 554 Functionalized Inorganic Fluorides 18.3 Optical and Magnetic Properties of LnF3-BaF2-AlF3-GeO2 (SiO2) Glasses 18.3.1 Optical Properties of HoF3-BaF2-AlF3-GeO2 Glasses Some reports have described studies of oxide fluoride, oxyfluoride, and fluorophosphate glasses [22–28]. Oxide fluoride glasses have been researched as host materials for opti- cally active ions because they have low phonon energies that correspond to oxide glasses, and high chemical and mechanical stabilities related to fluoride glasses. Although oxide and fluoride ions have similar ionic radii, the ratio of oxide and fluoride ions in the glass must alter the coordination structure that affects the elements’ functionality because of their different valences. For example, binary rare earth metal oxide fluorides such as Nd2Eu2O3F6 have an ordered ionic configuration that engenders higher electric Table 18.3 Glass transition temperatures of TbF3-BaF2-AlF3- SiO2 glasses. Reproduced from [2] by permission of Elsevier TbF3-BaF2-AlF3-SiO2 Tg / �C 30-10-10-50 666.5 40-10-10-40 633.1 50-10-10-30 590.0 10-20-10-60 657.6 20-20-10-60 608.3 30-20-10-40 588.7 Table 18.4 Glass transition temperatures of LnF3-BaF2-AlF3-GeO2 (Ln; Y-Nd, Sm-Lu) glasses. Reproduced from [2] by permission of Elsevier Composition Tg / oC 10YF3-20BaF2-10AlF3-60GeO2 608.4 10LaF3 -10BaF2-10AlF3 -70GeO2 630.3 10CeF3-20BaF2-10AlF3-60GeO2 585.3 10PrF3-20BaF2-10AlF3-60GeO2 582.2 10NdF3-20BaF2-10AlF3-60GeO2 590.1 10SmF3-20BaF2-10AlF3-60GeO2 589.6 10EuF3-20BaF2-10AlF3-60GeO2 586.7 10GdF3-20BaF2-10AlF3-60GeO2 592.8 10TbF3-20BaF2-10AlF3-60GeO2 578.4 10DyF3-20BaF2-10AlF3-60GeO2 599.3 10HoF3-20BaF2-10AlF3-60GeO2 599.9 10ErF3-20BaF2-10AlF3-60GeO2 609.8 10TmF3-20BaF2-10AlF3-60GeO2 579.5 10YbF3-20BaF2-10AlF3-60GeO2 583.4 10LuF3-20BaF2-10AlF3-60GeO2 581.0 Rare-earth-Containing Oxide Fluoride Glasses 555 conductivity than that of YSZ-11 [29]. Oxide fluoride materials containing multication species can exhibit unique properties because of their ordered-disordered ionic configura- tion. In this section, optical properties related to contents of LnF3 were reviewed using HoF3 as a probe. The results of absorption spectra measurements in the ultraviolet–visible region showed that the HoF3 contents affected the peak pattern. In that case, the Judd–Ofelt theory [30], used along with the results of the absorption spectra and the refractive indices measure- ments, can be useful to obtain information about the glass structure. Some reports have described dependence of the spectra on the glass matrix species using calculation of Judd– Ofelt intensity parameters (Ol parameters) for Ho 3þ in different host lattices [12,23, 31– 38]. The refractive indices were measured at 488, 540 and 641 nm to calculate the Ol parameters. They were used to determine the relation between the refractive index (n(l)) and the wavelength (l) by least squares fitting to the Sellmeier’s dispersion equation [38] n2ðlÞ¼1þ Sl 2 l2 � l02 (18:3) where S and lo are constants. The respective values of S and lo obtained for 10HoF3- 20BaF2-10AlF3-60GeO2 glass were 300 and 1.14. Using Equation (18.3), the refractive index was recalculated at the specific wavelength. On the other hand, no dependence of the refractive index on the wavelength was observed under the condition in this study for 50HoF3-20BaF2-10AlF3-20GeO2 glass. Therefore, n(l) was assumed to be constant, as 1.57. According to the method described in references [12, 23, 30–35, 39–42], the experimental oscillator strengths, fexp, of the aJ -> bJ’ transition at the transition mean wavelength l are presented in Table 18.5 and Table 18.6. From these values, the intensity parameters in the Judd–Ofelt theory were obtained as O2 ¼ 1.96 � 10�20, O4 ¼ 0.64 � 10�20, and O6 ¼ 0.11 � 10�20 cm2 for 10HoF3- 20BaF2-10AlF3-60GeO2 glasses and O2 ¼ 0.05 � 10�20, O4 ¼ 0.10 � 10�2 and O6 ¼ 0.02 � 10�20 cm2 for 50HoF3-20BaF2-10AlF3-20GeO2 glasses. The calculated oscillator strength, fcal, is also summarized in Table 18.6. The respective rms deviations of fexp and Table 18.5 Measured and calculated oscillator strength for Ho3þ ions in 10HoF3-20BaF2- 10AlF3-60GeO2 glass. Reproduced from [2] by permission of Elsevier 5I8! Wavelength (nm) Oscillator strength f ( x 10 �6) fexp fcal Df ( x 10 �6) 5F5 645.9 1.78 1.81 �0.030 5S2 542.6 0.388 0.634 �0.246 5F4 539.0 2.03 1.79 0.240 5F3 485.1 0.534 0.752 �0.218 5F2 474.6 0.438 0.300 0.138 3K8 468.4 0.504 0.526 �0.022 5G5 417.8 0.947 1.67 �0.723 5G4 384.4 0.587 0.187 0.400 3K7 379.3 0.282 0.0855 0.197 556 Functionalized Inorganic Fluorides fcal were d ¼ 3.9 � 10�7 and d ¼ 1.1 � 10�7 for 10HoF3-20BaF2-10AlF3-60GeO2 and 50HoF3-20BaF2-10AlF3-20GeO2 glasses. Some empirical correlations of the intensity parameter and the local structure of the lanthanide ions have been stated in the literature [12,23, 30–35]. In general, O2 increases with the asymmetry of the local structure and the degree of covalency of the lanthanide–ligand bonds, whereasO6 decreases with the degree of covalency. For glass matrices prepared in this study, no clear conclusions are discernible from Ol parameters because of their large uncertainty, especially for O2, but the small value of O2 together with very small O6 for our glasses compared to that for LaAlO3 crystals containing Ho3þ [42] were inferred to result from the lower covalency of the holmium–ligand bonds: the environment around Ho3þ in the glasses prepared in this study could be strongly ionic. This tendency is apparently truer for 50HoF3-20BaF2-10AlF3- 20GeO2 glasses than for 10HoF3-20BaF2- 10AlF3-60GeO2 glass. The ratios of O2/O6 were, respectively, 20 and 2.5 for 10HoF3-20BaF2-10AlF3-60GeO2 and 50HoF3-20BaF2- 10AlF3- 20GeO2 glasses. These ratios imply that the local structure in the 50HoF3- 20BaF2-10AlF3-20GeO2 glasses is more symmetric than that in 10HoF3-20BaF2- 10AlF3-60GeO2 glass. This fact is reflected in the change in the intensity ratio in fluores- cence spectra with the Ho3þ contents. It can be applied to control the optical properties of this glass system. 18.3.2 Optical Properties of CeF3-BaF2-AlF3-SiO2 Glasses The glasses in this system were brown when they were produced, even in inert gas (Ar). The brown colour is attributable to the presence of both Ce3þ and Ce4þ ions that have different energy levels in the glass. The energy transition that takes place between Ce3þ and Ce4þ causes the brown colour. This mixed-valence state of cerium ion resulted from hydrolysis and oxidation at high temperature. It was first found as a result of this study that Table 18.6 Measured and calculated oscillator strength for Ho3þ ions in 50HoF3-20BaF2- 10AlF3-20GeO2 glass. Reproduced from [2] by permission of Elsevier 5I8! Wavelength (nm) Oscillator strength f ( x 10 �6) fexp fcal Df ( x 10 �6) 5F5 645.9 0.300 0.278 0.022 5S2 542.6 0.117 0.0494 0.068 5F4 539.0 0.190 0.276 �0.086 5F3 485.1 0.125 0.0602 �0.065 5F2 474.6 0.0782 0.0478 �0.030 3K8 468.4 0.0275 0.0331 0.006 5G5 417.8 0.189 0.0869 �0.102 5G4 385.4 0.0151 0.0411 0.026 3K7 382.3 0.0366 0.00861 �0.028 3F2 357.5 0.258 0.00136 �0.257 3F4 334.3 0.0635 0.00323 �0.060 Rare-earth-Containing Oxide Fluoride Glasses 557 colourless or light yellow CeF3-BaF2-AlF3-SiO2 glass was obtained instead of brown glass when it was produced in a CO atmosphere. Oxidization of cerium ion is controlled in this case. The amount of Ce3þ in the glass increased and a very low content of Ce4þ was achieved. These glasses show blue emission from Ce3þ under UV irradiation (365 nm). Although previous reports describe that the addition of carbon powder during melting is effective to avoid oxidation of Ce3þ to Ce4þ in the case of oxide glasses, the decomposi- tion and/or hydrolysis of CeF3 was not avoided in the case of the oxide fluoride glasses. In addition, the Pt or Pt/Au container was badly damaged when using a H2 atmosphere. In this work, glasses having different characteristics can be prepared under an Ar or CO atmosphere. The brown Ce3þ - Ce4þ oxide fluoride glass produced in Ar atmosphere emitted no fluorescence, although the glass produced in CO gas exhibited blue emission under UV irradiation (365 nm). Two broad peaks are observed at 320 nm and 348 nm in the excitation spectra for the glass produced in CO gas. The peak at 320 nm was larger than that at 348 nm. Therefore, the wavelength for the excitation of the emission spectra was deter- mined as 320 nm for 10CeF3-20BaF2-10AlF3-60SiO2 glasses. Figure 18.9 depicts profiles of the emission spectra excited at 310, 320, 330, 357, and 366 nm. Radiation absorption and emission of Ce3þ occur because of the transition of an electron from the 4f orbital into the 5d orbital [43]. Generally, the profile in the emission spectrum corresponding to the 4f– 5d transition is broader than in that corresponding to the 4f–4f transition. Figure 18.9 shows that the peak position for excitation at 320 nm was located at a shorter wavelength than that at 357 nm excitation. To investigate these differences in the emission profiles, the peak was deconvoluted using a Gaussian function. Figure 18.10 depicts the result of the peak analysis of the emission profile excited at 320 nm. The results for all profiles portrayed in Figure 18.9 are presented in Table 18.7. Every emission peak consisted of three peaks that have a peak position at about 410 (peak 1), 445 (peak 2), and 490 nm (peak 3). Matsui et al. [44] also reported the existence of three peaks from peak analysis of the Ce3þ emission spectrum in Y2SiO5:Ce 3þ crystal. The difference between peak 1 and peak 2 might correspond to the split (2000 cm�1) of 2F5/2 and 2F7/2 by spin–orbit interac- tion [45–48]. Peak 3 corresponds to the presence of Ce3þ with CN¼7, whereas peak 1 and peak 2 correspond to that with CN¼ 6 reported in the literature. Therefore, Ce3þ of two kinds might be located in different environments in the considered glass. Peak analysis results show that the intensity of peak 3 tends to be stronger when the excitation wave- length is elongated. The quantum efficiency for peak 3 might depend on the environment around the Ce3þ. Figure 18.11 depicts the energy calculated from each wavelength of peak 1 and peak 2. The energy differences between peak 1 and peak 2 are approximately 1600 cm�1; they became small as compared to the energy difference of 2F5/2, and 2F7/2. As reported by Matsui et al., the presence of two kinds of Ce3þ cations in Y2SiO5:Ce 3þ seems to cause the change in an energy difference between the 2F5/2 and 2F7/2 levels [44]. The results depicted in Figure 18.5 might be consistent with that inference. Variation must be present in the environment around Ce3þ in the glasses here. Figures 18.12–18.15 present XPS spectra of F1s, O1s, Ce3d, and Ce4d in glasses prepared in an Ar and a CO atmospheres. The F1s peak in the glass prepared in a CO atmosphere was more intense than that in the glass prepared in an Ar atmosphere, as presented in Figure 18.6, and the peak position shifted to a lower energy than that of CeF3. Therefore, less fluoride ion exists in the glass prepared in an Ar atmosphere than in the 558 Functionalized Inorganic Fluorides glass prepared in a CO atmosphere. This difference is consistent with the result of quantitative analysis of fluoride ion by XFS measurement, as described previously. The oxidative decomposition of Ce3þ containing intermediate (CeOxFy where 2xþ y¼ 3) such as CeOF generated by hydrolysis of CeF3 was apparently controlled in a CO atmosphere. The fluoride ion attracts the electron cloud more strongly than the oxide ion. Therefore, the electron cloud seems to approach to fluoride ion in CeOxFy (2xþ y¼ 3) to a greater degree than in CeF3, which causes the smaller binding energy of F1s electron of the oxide fluoride glass than that of CeF3. A single peak existed in XPS spectra of O1s in the glasses prepared 370 420 470 Wavelength / nm In te ns ity / a rb . u n its 520 570 357 nm λEx = 366 nm 330 nm 320 nm 310 nm Figure 18.9 Emission spectra of 10CeF3-20BaF2-10AlF3-60SiO2 glass prepared in CO for several excitation wavelengths. Reproduced from [1] by permission of Elsevier 350 400 450 Wavelength / nm peak 1 peak 3 peak 2 In te ns ity / ar b. u n its 500 550 600 Figure 18.10 Decomposition of the emission spectrum of 10CeF3-20BaF2-10AlF3- 60SiO2 glass prepared in CO under the excitation wavelength of 320 nm. The solid line shows the observed spectrum. Dashed lines show calculated spectra. Reproduced from [1] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 559 here, although a double peak was observed in that in CeO2. The peak position for the glasses prepared in a CO or an Ar atmosphere was higher than that for the quartz glass. The reason is that the electron density around O2- might be lowered because the electron cloud is with- drawn toward F� via cerium ion. The oxide and fluoride ions in the glass have different electronic states, respectively, from those in simple oxide and fluoride. To elucidate the state of the valence of cerium in the glasses, Ce3d and Ce4d spectra are depicted in Figures 18.14 and 18.15. As references for Ce3þ and Ce4þ, wemeasured CeF3 and CeO2, respectively. The peak profiles for Ce3d and Ce4d electrons in the glasses prepared in a CO and an Ar atmosphere were similar. However, differences were apparent in their peak positions. Figures 18.14 and 18.15 show that the peaks of the glass prepared in a CO atmosphere located at lower binding energy (about 0.3 eV) compared to those in an Ar atmosphere. The peak position and/or profile inXPS spectra depend on the valence state. Themean valence of cerium ion increases when the oxidative decomposition of CeOxFy proceeds. The ratio of Ce3þ/Ce4þ of the glasses prepared in a CO atmosphere was larger than in glasses prepared in an Ar atmosphere. Therefore, the peaks in the XPS spectra of the glasses prepared in a CO atmosphere have to locate at lower binding energy than the glasses prepared in an Ar atmosphere. In fact, this difference was observed in Figures 18.14 and 18.15. In Ce3d spectra, the peak at 919 eV that appeared in the spectrum of CeO2 was not observed in that of the glasses (Figure 8.14). Furthermore, the peak pattern of the glasses in Figure 8.14 differed completely from that of CeF3 [49–51]. Figure 18.9 portrays the Ce4d spectra of both glasses. The peak near 105 eV corresponds to Si2p. The peaks at 109 (not identified yet), 123 (4d5/2), and 126 (4d3/2) eV that appeared in the spectrum of CeO2 were not observed in spectra of the glasses (Figure 8.14). The peak patterns of the glasses in Figure 8.15 differed completely from that of CeF3. The electronic state of cerium ion must be distinctive for the glass obtained in this study. In other words, the presence of both fluoride and oxide ions might impart a unique electronic state to cerium. 300 320 340 Excitation wavelength /nm peak 2: 5d→4f(2F7/2) peak 1: 5d→4f(2F5/2) En er gy le ve l / cm –1 21 500 22 000 22 500 23 000 23 500 24 000 24 500 25 000 360 Figure 18.11 Energy differences of peak 1 and peak 2 for several excitation wavelengths of 10CeF3-20BaF2-10AlF3-60SiO2 glass prepared in CO. Reproduced from [1] by permission of Elsevier 560 Functionalized Inorganic Fluorides T ab le 1 8 .7 A n al ys is o f em is si o n sp ec tr a o f 1 0 C eF 3 -2 0 B aF 2 -1 0 A lF 3 -6 0 Si O 2 gl as s u si n g a G au ss ia n fu n ct io n fo r se ve ra l ex ci ta ti o n w av el en gt h s. R ep ro d u ce d fr o m [1 ] b y p er m is si o n o f El se vi er P ea k 1 P ea k 2 P ea k 3 l E x A m p li tu d e C en te r FH W M A m p li tu d e C en te r FH W M A m p li tu d e C en te r FH W M W av el en gt h n m W av el en gt h n m W av el en gt h n m 3 1 0 2 7 .9 4 4 1 0 .3 4 –0 .7 8 4 9 .1 8 6 1 .3 3 4 4 6 .0 3 –3 .2 9 7 7 .3 0 2 6 .9 3 4 9 9 .0 0 –1 9 .0 8 1 2 2 .9 8 3 2 0 2 8 .9 2 4 1 0 .9 3 –0 .6 7 4 9 .5 7 6 2 .6 6 4 4 7 .1 2 –3 .5 7 7 9 .9 7 2 5 .0 7 5 0 0 .5 1 –1 9 .0 1 1 3 0 .6 0 3 3 0 2 5 3 9 4 1 1 .6 1 –0 .7 8 4 8 .1 2 6 0 .2 4 4 4 5 .7 1 –3 .9 8 7 7 .0 1 2 7 .6 5 4 9 5 .0 7 –2 3 .7 4 1 2 1 .4 9 3 5 7 1 6 7 8 4 1 3 .9 0 –0 .2 7 4 2 .5 0 6 0 .3 2 4 4 2 .8 6 –3 .4 1 7 6 .1 9 3 3 .0 5 4 8 6 .5 5 –1 7 .3 7 1 2 6 .2 2 3 6 6 1 2 .3 2 4 1 8 .6 8 –0 .2 7 3 2 .4 9 5 2 .7 7 4 4 5 .2 6 –1 .3 5 7 3 .3 4 4 3 .4 0 4 7 3 .7 5 –2 .7 5 1 3 8 .9 6 695 690 CeF3 glass (CO) glass (Ar) Binding energy / eV In te ns ity / ar b. u n its 685 Figure 18.12 F1s spectra of CeF3 and 10CeF3-20BaF2-10AlF3-60SiO2 glasses prepared in CO and Ar atmospheres. Reproduced from [1] by permission of Elsevier 538 533 Binding energy / eV In te ns ity / ar b. u n its 528 CeO2 Quartz glass glass (CO) glass (Ar) Figure 18.13 O1s spectra of CeO2, quartz glass and 10CeF3-20BaF2-10AlF3-60SiO2 glasses prepared in CO and Ar atmospheres. Reproduced from [1] by permission of Elsevier 562 Functionalized Inorganic Fluorides 130 125 120 115 Binding energy / eV CeF3 CeO3 glass (CO) glass (Ar) 4d 3/2 In te ns ity / a rb . u n its 110 105 100 4d 5/2 Figure 18.15 Ce4d spectra of CeF3, CeO2 and 10CeF3-20BaF2-10AlF3-60SiO2 glasses prepared in Ar and CO atmospheres. Reproduced from [1] by permission of Elsevier 925 915 905 Binding energy / eV CeF3 CeO3 glass (CO) glass (Ar) 919 eV In te ns ity / a rb . u n its 895 885 875 Figure 18.14 Ce3d spectra of CeF3, CeO2 and 10CeF3-20BaF2-10AlF3-60SiO2 glasses prepared in Ar and CO atmospheres. Reproduced from [1] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 563 18.3.3 Optical Properties of the Glasses Co-doped with TbF3 and SmF3 Reportedly, some rare earth co-doped systems exhibit enhancement of fluorescence intensity [52–54]. Oxide fluoride glasses that can contain rare earth fluorides of more than 50mol% have been reported above [55, 56]. It is interesting to study the emission properties of such glasses in which two rare earth ions are near one another. This section presents a description of some unique optical properties of the glass containing a large amount of TbF3 doped with a small amount of SmF3 were described. Figure 18.16 presents the relation between I600/540 (the ratio of the peak intensity at 540 nm originated from Tb3þ to that at 600 nm originated from Sm3þ) and the contents of SmF3. In fact, I600/540 increased concomitantly with increasing content of SmF3 (< 2 wt%), whereas I600/540 decreased with the contents of SmF3 (> 2 wt%). Some reports have described that concentration quenching of Sm3þ fluorescence occurred around 1–2mol% [57,58]. The concentration quenching of Sm3þ fluorescence occurs when the content of SmF3 is more than 2 wt% in Figure 18.16. The intensities of fluorescence at 600 nm of Sm3þ in 20TbF3-20BaF2 – 10AlF3 – 50GeO2 /mol%þ 2wt% SmF3 were measured as 1.2 � 104 cps, whereas that in 20TbF3–20BaF2–10AlF3–50GeO2 /mol%þ 5 wt% SmF3 was 5.3 � 103 cps. This fact reflects that Sm3þ is dispersed identically in the 20TbF3 – 20BaF2 – 10AlF3 – 50GeO2 glass without a phase-separated or otherwise clustered situation. Figure 8.17 presents the relation between I540/600 and the contents of TbF3. Results showed that I540/600 was proportional to x 2.05 by least-squares fitting, where x was in xTbF3-20BaF2-10AlF3-(70-x)GeO2 (mol%). The distance of Tb 3þ�Tb3þ in the glass (D) is inversely proportional to the third power of the Tb3þ concentration, so I540/600 is almost perfectly inversely proportional to the sixth power of D in this case. The theoretical calculation of Forster–Dexter [59, 60] indicated that the resonant energy transfer prob- ability is inversely proportional to the sixth power of the distance between two centres if the two centres belong to a dipolar transition. Therefore, the change in the relation between I540/600 and the contents of TbF3 presented in Figure 18.17 are explainable mainly using 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0 x / wt% In te ns ity ra tio I 6 00 /5 40 Figure 18.16 Relation between the intensity ratio (I540/600) and contents of SmF3 in 20TbF3 – 20BaF2 – 10AlF3 – 50GeO2 /mol%þ x / wt% SmF3 glasses. Reproduced from [2] by permission of Elsevier 564 Functionalized Inorganic Fluorides concentration quenching of the Tb3þ fluorescence [61]. The TbF3 content must be more than 20mol%toobtainaglass thatexhibitsan intenseorangeemission.Thisphenomenonisaunique property for this LnF3 - BaF2 - AlF3 - GeO2 system that can contain TbF3more than 50mol%. The fluorescence of 20TbF3 – 20BaF2 – 10AlF3 – 50GeO2 þ 0.05wt%SmF3 glass exhibited temperature dependence. The intensities of the peak at 540 nm originated from Tb3þ and the peak at 600 nm that originated from Sm3þ are mutually equivalent at room temperature. After heating to 573K or cooling to 77K, the intensity of the fluorescence that originated from Sm3þ decreased. The emission colour changed from orange to green through yellow. The glass phase was stable at temperatures of 77–573K because 673K is much less than the glass transition temperature (876K) of 20TbF3 – 20BaF2 – 10AlF3 – 50GeO2þ 0.05wt%SmF3 glass. The emission colour changed reversibly according to the temperature. Figure 18.18 portrays the relation between the intensity ratio of I600/540 and temperature. The change in the emission colour from orange to green was recognized when I600/540 was less than 0.7: I600/540 decreased gradually by heating to 673K. That emission colour changed from orange to green through yellow. In the case of cooling, I600/540 changed drastically and the emission colour changed to green at around 77K. Relating the fluorescence peak to the energy transition J! J’, where J and J’ correspond to initial and final states respectively, the electrons in the ground state are excited and the popula- tion of J’ becomes higher with increasing temperature. The electrons at J therefore barely transfer to J’. The intensity of the emission was lowered. The Tb3þ has a larger energy gap separating J and J’ than Sm3þ does. Therefore, the temperature affected the emission from Tb3þ only slightly; no significant concentration quenching occurred in the case of Tb3þ [62]. Figure 18.18 shows that, when the temperature rose, the emission from Tb3þ came to take precedence over Sm3þ for the system containing Tb3þ and Sm3þ. For cooling, the mechanism is described as follows. Energy used for the emission from Sm3þwas supplied by nonradiative relaxation from Tb3þ. This nonradiative relaxation is called multiphonon relaxation, which relates to the lattice vibration and dipole-dipole interaction [63,64]. This multiphonon relaxation occurred only slightly around 77K. Therefore, the energy cannot 5.0 4.0 3.0 2.0 1.0 0 0 5 10 15 20 25 30 35 40 45 x / mol% In te ns ity ra tio I 5 40 /6 00 Figure 18.17 Relation between the intensity ratio (I540/600) and contents of TbF3 in xTbF3 – 20BaF2 – 10AlF3 – (70–x)GeO2 /mol% þ 0.05 / wt% SmF3 glasses. Reproduced from [2] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 565 be supplied from Tb3þ to Sm3þ when the sample temperature was lower than 77K. Consequently, the emission from Tb3þ was predominant and the emission colour was green around 77K. It is expected that these phenomena are useful to probe the energy transfer mechanism in the glass matrix. 18.3.4 Magnetic Property of TbF3 Containing Oxide Fluoride Glasses In fact, Tb3þ is a rare-earth ion having a large effective magnetic moment, 9.72 mB, which is third among the rare earth ions. Considering the transparency of Tb3þ in the visible light region, the Tb3þ-containing compounds are very interesting to prepare material for optical and magnetic applications [22,54, 65, 66]. Using a TbF3-BaF2-AlF3-GeO2 system, glasses containing a large amount of Tb3þ (50mol% as TbF3) were prepared as described above. Magnetic measurements were conducted at 77–273K using a vibrating sample magnet- ometer (VSM) (VSM-3; Toei Industry Co., Ltd.) with a maximum magnetic field of �10000 – 10000 Oe. A Physical Properties Measurement System (PPMS) susceptometer (Quantum Design Co.) is also used to investigate the relation between magnetization and temperature at 2–300K. Figure 18.19 portrays the relation between TbF3 contents in xTbF3-20BaF2-10AlF3-(70-x)GeO2 /mol% glass and magnetic susceptibility at room temperature. Generally, the relation between magnetization and the magnetic field is shown as M ! ¼�H! (18:4) where w represents the magnetic susceptibility (emumol�1). Themagnetic susceptibility is expressed as Equation (18.5) from the Curie–Weiss laws �¼ c T � � (18:5) 0 100 200 300 400 500 600 700 Temperature / K 1.2 1.0 0.8 0.6 0.4 0.2 0 In te ns ity ra tio I 6 00 /5 40 Figure 18.18 Temperature dependence of the intensity ratio (I600/540) of 20TbF3 – 20BaF2 – 10AlF3 – 50GeO2 þ 0.05wt%SmF3 glass. Reproduced from [2] by permission of Elsevier 566 Functionalized Inorganic Fluorides where � is the Weiss temperature and C is the Curie constant; C is written as C ¼ ng 2m2gJ J þ 1ð Þ 3kb ¼ nMaf f 2 3kB (18:6) where n, g, mB, kB, J, andMeff respectively represent the number of magnetic ions per mol, the Lande´ g-factor, Bohr magneton, total angular momentum, and the effective magnetic moment. The solid line in Figure 18.19 was calculated from Equations (18.4)�(18.6). The magnetic susceptibilities are proportional to the contents of TbF3 in the glasses obtained in this study. Their values fit the calculated values. Figure 18.20 portrays the magnetic suscept- ibility and the reciprocal magnetic susceptibility (w�1) of 30TbF3-20BaF2-10AlF3-40GeO2 glass of 2–300K. The relation between w�1 and temperature is linear, as presented in Figure 18.20. TheWeiss temperature – which is an intersection point of the w�1 axis and temperature axis – is almost zero, meaning that the spin–spin interaction was negligible at 100�275K in the glasses obtained here. From Equations (18.2) and (18.3), the Curie constant (C) and the effective magnetic moment (Meff) were calculated. These values are presented in Table 18.8. The Curie constant increased concomitantly with increasing glass TbF3 contents. In addition, the effective magnetic susceptibility is consistent with the theoretical value of 9.78mB. It is therefore readily apparent that the magnetic moment is derived only from a trivalent terbium ion. The atomic content of Tb3þ in 40TbF3–20BaF2–10AlF3–30GeO2 was calculated as 11 %, which is comparable to that reported for oxide glasses as the maximum one, 12 % [67]. In addition, the saturation behaviour might result from the Tb3þ(#)–O2�–Tb3þ(") superex- change interaction, which prevented orientation of Tb3þ magnetic moments to the applied magnetic fields in the case of the oxide glasses [68]. Using fluoride instead of oxide is one way to restrain this saturation behaviour. The atomic content of Tb3þ in 50TbF3–20BaF2– 10AlF3–20GeO2 glass obtained in this study was calculated as 14 %. Therefore, it might have high potential for use as a material for Faraday devices [69]. 0.02 0.015 0.01 0.005 0 10 20 30 40 50 M ag ne tic s us ce pt ib ilit y χ / e m u m ol – 1 x / mol% Figure 18.19 Relationship between TbF3 contents in the glass and magnetic susceptibility at room temperature. x is in xTbF3-20BaF2-10AlF3-(70-x)GeO2 (mol%). The solid line represents the theoretical relationship. Reproduced from [2] by permission of Elsevier Rare-earth-Containing Oxide Fluoride Glasses 567 18.4 Conclusion Rare-earth-containing oxide fluoride glasses LnF3 (Ln: Y through Lu)-BaF2-AlF3-GeO2 (or SiO2) were produced in which the nominal content of LnF3 reached 60mol% max- imum. Their basic properties, such as density, refractive index, and glass transition temperature were investigated and summarized in detail. A CO atmosphere is effective to prepare glasses containing a trivalent ion, whose valency might change during the preparation process, such as Ce3þ/Ceþ4. In particular, to discuss the local structure surrounding the rare-earth ion in the glass, a Judd–Ofelt analysis (discussion with O parameters) of the HoF3-BaF2-AlF3-GeO2 glasses was conducted. The fluorescent beha- viour and the magnetic properties of LnF3-BaF2-AlF3-GeO2 glasses (Ln ¼ Tb and/or Sm) were also studied to characterize the glasses. Their magnetic and optical properties are attractive for some applications. This glass system has much compositional variety. It might be interesting for applications and for fundamental studies of the lanthanides’ optical properties. 0 50 100 150 200 250 χ / e m u m o l–1 Temperature / K 1.0 0.8 0.6 0.4 0.2 100 80 60 40 20 χ –1 / m ol em u –1 Figure 18.20 Temperature dependence of magnetic susceptibility of 30TbF3-20BaF2-10AlF3- 40GeO2 glass. Reproduced from [2] by permission of Elsevier Table 18.8 Curie constant and effective magnetic susceptibility of xTbF3-20BaF2-10AlF3- (70�x)GeO2 glasses (x ¼ 10-40 / mol%). Reproduced from [1] by permission of Elsevier Concentration of TbF3/ mol% Curie constant / K Effective magnetic 10.0 1.00 8.83 20.0 2.47 9.77 27.5 3.12 9.37 30.0 4.49 10.7 40.0 5.13 9.96 568 Functionalized Inorganic Fluorides References [1] H.Takahashi, S. Yonezawa, M. Kawai, M. Takashima, J. Fluorine Chem. 129 (2008), 1114– 1118. [2] S. Yonezawa, S. Nishibu, M. Leblanc, M. Takashima, J. Fluorine Chem. 128 (2007) 438–447. [3] S. Nishibu, T. Nishio, S. Yonezawa, J. H. Kim, M. Takashima, H. Kikuchi, H. Yamamoto, J. Fluorine Chem. 127 (2006) 821–823. [4] S. Nishibu, S. Yonezawa, M. Takashima, J. Fluorine Chem. 126 (2005) 1048–1053. [5] M. Takashima, S. Yonezawa, Jeo-ho Kim, J. Alloys Compd. 408–412 (2006) 468–473. [6] T. Tamaoka, S. Tanabe, S. Ohara, H. Hayashi, N. Sugimoto, J. Alloys Compd. 408–412 (2006) 848–851. [7] L. A. Bueno, A. S .L. Gomes, Y. Messaddeq, C. V. Santilli, J. D expert-Ghys, S. J. L. Ribeiro, J. Non-Cryst. Solids 351 (2005) 1743–1746. [8] S. Tanabe, T. Kouda, T. Hanada, J. Non-Cryst. Solids 274 (2000) 55–61. [9] J. A. Medeiros-Neto, J. A. M. Neto, E. R. Taylor, B. N. Samson, J. Wang, D. W. Hewak, R. I. Laming, D. N. Payne, E. Tarbox, P. D. Maton, G. M. Roba, B. E. Kinsman, R. Hanney, J. Non-Cryst. Solids 184 (1995) 292–296. [10] D. Raasch, H. Wierenga, J. Magn. Magn. Mater. 168 (1997) 336–346. [11] S. Yonezawa, J. H. Kim, M. Takashim, Solid State Sciences 4 (2002) 1481–1485. [12] M. R. Reddy, S. B. Raju, N. Veeraiah, J. Phys. Chem. Solids 61 (2000) 1567–1571. [13] New Glass Forum INTERGLAD ver. 5 [http://www.interglad.ip/interglad6/trial.html] (2004). [14] R. B. Sosman, The Properties of Silica, Reinhold, New York (1927). [15] T. Kano, H. Yanagida, Rare Earth – Properties and Applications, Gihodo Shuppan, Tokyo, (1990) p. 159. [16] K. Soga, J. Kaga, H. Inoue, A. Makishima, J. Non-Cryst. Solids 315 (2003) 1–6. [17] Y. Jestin, A. Le Sauze, B. Boulard, Y. Gao, P. Baniel, J. Non-Cryst. Solids 320 (2003) 231–237. [18] A. Boutarfaia, M. Poulain, Solid State Ionics 144 (2001) 117–121. [19] C. C. Chen, Y. J. Wu, L. G. Hwa, Mater. Chem. Phys. 65 (2000) 306–309. [20] Y. Nishida, T. Kanamori, T. Sakamoto, Y. Ohishi, S. Sudo, J. Non-Cryst. Solids 221 (1997) 238–244. [21] F. Gan, J. Non-Cryst. Solids 184 (1995) 9–20. [22] K. K. Mahato, S. B. Rai, Spectrochimica Acta Part A 56 (2000) 2333–2340. [23] S. B. Rai, Spectrochimica Acta Part A 58 (2002) 1559–1566. [24] L. F. Santos, R. M. Almeida, V. K. Tikhomirov, A. Jha, J. Non-Cryst. Solids 284 (2001) 43–48. [25] R. El-Mallawany, A. H. Khafagy, M. A. Ewaida, I. Z. Hager, M. A. Poulain, M. J. Poulain, J. Non-Cryst. Solids 184 (1995) 141–146. [26] R. R. Reddy, Y. N. Ahammed, P. A. Azeem, K. R. Gopal, T. V. R. Rao, S. Buddhudu, N. S. Hussain, J. Quantitative Spectroscopy and Radiative Transfer 77 (2003) 149–163. [27] L. R. Moorthy, T. S. Rao, K. Janardhanam, A. Radhapathy J. Alloys Compd. 298 (2000) 59–67. [28] Y. Zhanci, H. Shihua, L. Shaozhe, C. Baojiu, J. Non-Cryst. Solids 343 (2004) 154–158. [29] M. Takashima, S. Yonezawa Y. Ukuma, J. Fluorine Chem. 87 (1998) 229–234. [30] B. R Judd, Phys. Rev. 127 (1962) 750–761. [31] E. Rukmini, C. K. Jayasankar, Opt. Mater. 4 (1995) 529–546. [32] S. B. Rai, A. K. Singh, S. K. Singh, Spectrochimica Acta Part A 59 (2003) 3221–3226. [33] K. Driesen, C. Gorller-Walrand, K. Binnemans, Mater. Sci. Eng. C 18 (2001) 255–258. [34] M. Malinowski, A. Wnuk, Z. Frukacz, G. Chadeyaron, R. Mahiou, S. Guy, M. F. Joubert, J. Alloys Comp. 323–324 (2001) 731–735. [35] P. J. Deren, J.-C. Krupa, J. Alloys Compd. 380 (2004) 362–367. [36] A. Lorenzo, L. E. Bausa, J. A. Sanz-Garcia, J. Garcia-Sole, J. Phys. Condens. Matter 8 (1996) 5781–5791. [37] G. S. Ofelt, J. Chem. Phys. 37 (1962) 511–520. [38] M. A. Khashan, A. Y. Nassif, Opt. Commun. 198 (2001) 247–256. [39] R. T. Genova, I. R. Martin et al., J. Alloys Compd. 380 (2004) 167–172. [40] X. Qiao, X. Fan, M. Wang, X. Zhang, Opt. Matter. 27 (2004) 597–603. Rare-earth-Containing Oxide Fluoride Glasses 569 [41] W. T. Carnall, P. R. Fields, K. Rajnak, J. Chem. Phys. 49 (1968) 4424–4442. [42] C. Gorller-Walrand, K. Binnemans,Handbook on the Physics and Chemistry of Rare Earths, 25 Elsevier, Science B. V., North-Holland, Amsterdam (1998) p. 101; and references therein. [43] H. Kobayashi, Hakko no Butsuri, Asakura Publishing Co., Ltd., Tokyo, 2002, pp. 58–59. [44] H. Matsui, C. N. Xu, Y. Liu, T. Watanabe, J. Ceram. Soc. Japan 108 (2000) 1003–1006. [45] J. Sytsma, D. Piehler, N. M. Edelstein, L. A. Boatner, M. M. Abraham, Phys. Rev. B 47 (1993) 14786–14794. [46] Y. Kojima, K. Machi, T. Yasue, Y. Arai, J. Ceram. Soc. Japan 108 (2000) 836–841. [47] L. Huang, X. Wang, H. Lin, X. Liu, J. Alloys Compd. 316 (2001) 256–259. [48] M. Yamaga, N. Kodama, J. Alloys Compd. 316 (2001) 256–259. [49] Y. Hayashi, M. Kudo, J. Surf. Sci. Soc. Jpn. 221 (2001) 64–71. [50] Y. Uwamino, T. Ishizuka, A. Tsuge, H. Yamatera, Japan Analyst 34 (1985) 166–170. [51] D. R. Mullins, S. H. Overbury, D. R. Huntley, Surf. Sci. 409 (1998) 307–319. [52] Y. Yang, S. Zhang, Q. Su, Mater. Res. Bull. 40 (2005) 1010–1017. [53] H. Lin, E. Y. Pun, X. Wang, X. Liu, J. Alloys Compd. 390 (2005) 197–201. [54] X. Zou, H. Toratani, J. Non-Cryst. Solids 195 (1996) 113–124. [55] M. Takashima, S. Yonezawa, T. Tokuno, H. Umehara, T. Kato, J. Fluorine Chem. 112 (2001) 241–246. [56] S. Nishibu, S. Yonezawa, M. Takashima, J. Non-Cryst. Solids 351 (2005) 1239–1245. [57] Y. Zhou, J. Lin, S. Wang, J. Solid State Chem. 171 (2003) 391–395. [58] E. Malchukova, B. Boizot, D. Ghaleb, G. Petite, Nuclear Instruments and Methods in Physics Research Section A, 537 (2005) 411–414. [59] R. M. Clegg, Curr. Opin. Biotechnol. 6 (1995) 103–110. [60] G. A. Kunar, N. V. Unnikrishnan, J. Photochem. Photobiol. A 144 (2001) 107–117. [61] M. Nikl, N. Solovieva, M. Dusek, A. Yoshikawa, Y. Kagamitani, T. Fukuda, J. Ceram. Proc. Res. 4 (2003) 112–114. [62] L. A. Riseberg, H. W. Moos, Phys. Rev. 174 (1968) 429–438. [63] K. K. Mahato, D. K. Rai, S. B. Rai, Solid State Commun. 108 (1998) 671–676. [64] D. J. Robbins, B. Cockayne, B. Lent, J. L. Glasper, Solid State Commun. 20 (1976) 673–676. [65] D. Kaczorowski, K. Gofryk, L. Romaka, Ya. Mudryk, M. Konyk, P. Rogl, Intermetallics 13 (2005) 484–489. [66] Y. Isikawa, D. Kato, A. Mitsuda, T. Mizushima, T. Kuwai, J. Magn. Magn. Mater. 272 (2004) 635–636. [67] T. Hayakawa, M. Nogami, N. Nishi, N. Sawanobori, Chem. Mater. 14 (2002) 3223–3225. [68] D. Imaizumi, T. Hayakawa, M. Nogami, J. Lightwave Technol. 20 (2002) 740–744. [69] S. Nishibu, T. Nishio, S. Yonezawa, J. H. Kim, M. Takashima, H. Kikuchi, H. Yamamoto, J. Fluorine Chem. 127 (2006) 821–823. 570 Functionalized Inorganic Fluorides Preparation and Properties of Rare-earth-Containing Oxide Fluoride Glasses Introduction Preparation and Basic Characteristics of Oxide Fluoride Glasses Containing LnF3 Preparation of Oxide Fluoride Glasses Containing LnF3 Density and Refractive Index Glass Transition Temperature Optical and Magnetic Properties of LnF3-BaF2-AlF3-GeO2 (SiO2) Glasses Optical Properties of HoF3-BaF2-AlF3-GeO2 Glasses Optical Properties of CeF3-BaF2-AlF3-SiO2 Glasses Optical Properties of the Glasses Co-doped with TbF3 and SmF3 Magnetic Property of TbF3 Containing Oxide Fluoride Glasses Conclusion