Optical characterization of sol-gel glasses derived from Eu3+ complex-forming precursors

Abstract

Fabrication of new optical devices based upon the incorporation of rare earth ions via sol-gel methods depends on elimination of dopant ion clusters and residual hydroxyl groups from the final material. The optical absorption and/or luminescence properties of luminescent rare earth ions are influenced by the local bonding environment and the distribution of the rare-earth dopants in the matrix. Typically, dopants are incorporated into gel via dissolution of soluble species into the initial precursor sol. In this work, Eu3+ is used as optical probe, to assess changes in the local environment. Results of emission, excitation, fluorescence line narrowing and lifetimes studies of Eu3+-doped gels derived from Si(OCH3)4 and fluorinated/chelate Eu3+ precursors are presented. The precursors used in the sol-gel synthesis were: tris (6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedionate) Eu(III), Eu (III) trifluoromethanesulfonate, Eu(III) acetylacetonate hydrate, Eu (III) trifluoroacetate trihidrate, tris (2,2,6,6-tetramethyl-3,5- heptanedionate) Eu(III) and Eu(NO3)3.6H2O. The results were interpreted in terms of the evolution of the Eu3+ fluorescence in systems varying from solution to the gels densified to 800ºC. The lifetimes studies indicate that the fluorinated precursors are effective at reducing the water content in densified gels.

Eu3+; sol-gel; complex-forming precursors

ARTIGO

OPTICAL CHARACTERIZATION OF SOL-GEL GLASSES DERIVED FROM EU³⁺ COMPLEX-FORMING PRECURSORS

Centro de Desenvolvimento da Tecnologia Nuclear - CDTN - Rua Mário Werneck s/n - Cidade Universitária - Pampulha - Belo Horizonte - MG

Wander L. Vasconcelos

Depto. Eng. Metalúrgica - UFMG - Rua Espírito Santo, 35 - 2^o andar - 30160-030 - Belo Horizonte - MG

Kevin L. Bray

Department of Chemical Engineering - University of Wisconsin - Madison - WI 53706

E-mail: [e-mail]

Recebido em 2/12/96; aceito em 29/4/97

FABRICATION OF NEW OPTICAL DEVICES BASED UPON THE INCORPORATION OF RARE EARTH IONS VIA SOL-GEL METHOD DEPENDS ON ELIMINATION OF DOPANT ION CLUSTERS AND RESIDUAL HYDROXYL GROUPS FROM THE FINAL MATERIAL. The optical absorption and/or luminescence properties of luminescent rare earth ions are influenced by the local bonding environment and the distribution of the rare-earth dopants in the matrix. Typically, dopants are incorporated into gel via dissolution of soluble species into the initial precursor sol. In this work, Eu³⁺ is used as optical probe, to assess changes in the local environment. Results of emission, excitation, fluorescence line narrowing and lifetimes studies of Eu³⁺-doped gels derived from Si(OCH₃)₄ and fluorinated/chelate Eu³⁺ precursors are presented. The precursors used in the sol-gel synthesis were: tris (6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedionate) Eu(III), Eu (III) trifluoromethanesulfonate, Eu(III) acetylacetonate hydrate, Eu (III) trifluoroacetate trihidrate, tris (2,2,6,6-tetramethyl-3,5- heptanedionate) Eu(III) and Eu(NO₃)₃.6H₂O. The results were interpreted in terms of the evolution of the Eu³⁺ fluorescence in systems varying from solution to the gels densified to 800^oC. The lifetimes studies indicate that the fluorinated precursors are effective at reducing the water content in densified gels.

Keywords: Eu³⁺; sol-gel; complex-forming precursors.

INTRODUCTION

The sol-gel process is among the most promising of the developing techniques for synthesizing new optical materials. Some of the applications for these optical materials are optical fibers and gradient-index waveguides, antireflection coatings, solid state lasers, non-linear optics, luminescent solar concentrators and photochromic glasses^1-3. The potential advantages of the sol-gel method for preparing optical materials include obtaining new chemical compositions, better purity and more convenient processing conditions.

The rare earth ions have interesting emission properties for laser applications and have been studied as luminescent ions when incorporated inside sol-gel matrices. Sol-gel glasses doped with rare earth are an important kind of materials with applications including solid-state laser, fiber amplifiers and optical waveguides. The optical spectroscopy of rare earth ions can also be useful for characterization of glasses^4,5. Eu³⁺ has been most often used as an optical probe because of its particularly informative luminescence spectrum⁶.

The luminescence efficiency of rare earth in sol-gel host materials is compromised due to the tendency of rare earth ions to form clusters and by the presence of hydroxyl ions and remnant organic. Clustering results in concentration quenching of fluorescence due to nonradiative energy transfer between rare earth cations within the clusters. This effect is undesirable for application of sol-gel materials based on luminescence properties.

Hydroxyl quenching and dopant clustering are two complications that currently limit the luminescence efficiency of rare earth ions in sol-gel host materials^7,8. Hydroxyl quenching is caused by residual water, solvents, and silanol groups present in sol-gel glasses and leads to an enhancement of non-radiative decay pathways of rare earth ions. Dopant clustering is deleterious because it leads to concentration quenching of luminescence through cross-relaxation and energy transfer processes. In previous work, it was shown a method for detecting rare earth clustering in sol-gel glasses based on Eu³⁺ fluorescence line narrowing spectroscopy⁹. Fluorescence line narrowing (FLN) is a site-selective spectroscopy designed to combat inhomogeneous spectral broadening in glass. It was reported that the addition of metal cations co-dopants leads to some inhibition of the tendency for Eu³⁺ to aggregate in a gel-derived SiO₂ matrix⁹. Dopant ion encapsulation by complex-forming ligands has been achieved in order to inhibit clustering of the dopant ions^10,11. Some ligands, such as TTFAH (thenoyltrifluoroacetone), are highly asymmetric, reducing the local-field symmetry of the metal ion and making the radiative transitions somewhat more allowed¹⁰. Brecher et al. demonstrated that organometallic complexes of europium show intense fluorescence in a variety of solutions¹¹.

In order to minimize the hydroxyl content of sol-gel matrices, the use of organically modified alkoxide sol-gel precursors¹², dopant ion encapsulation and the addition of fluorine as dehydroxylating agent¹³, have been tried and demonstrated to reduce hydroxyl content in silica gel glasses.

In this work the spectroscopy of Eu³⁺ ions is used as probe of structural changes in sol-gel silica glasses to investigate the effect of the use of different Eu³⁺ precursors, some of them are fluorinated and/or chelate precursors, on clustering and quenching of Eu³⁺ ions. The fluorescence characterization consisted of measuring broadband fluorescence spectra, excitation spectra, fluorescence line narrowing and fluorescence lifetimes.

EXPERIMENTAL PROCEDURE

Sample Preparation. The gels were prepared by mixing TMOS and the solutions in which the europium precursors were previously dissolved. The Eu³⁺ precursors (from Aldrich) studied were: tris (6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedionate) Eu(III) (Eu(fod)₃), Eu (III) trifluoromethanesulfonate ((CF₃SO₃)₃Eu), Eu(III) acetylacetonate hydrate (Eu(acac)₃), Eu (III) trifluoroacetate trihidrate ((CF₃CO₂)₃ Eu.3H₂O), tris (2,2,6,6-tetramethyl-3,5- heptanedionate) Eu(III) (Eu(thd)₃) and Eu(NO₃)₃.6H₂O. DMF was used to dissolve the Eu(fod)₃ and methanol was the solvent for Eu(thd)₃ and Eu(acac)₃. The molar ratio solvent to TMOS was 4. The water to TMOS molar ratio was 16. HNO₃ was added to low the pH to 1.5. The samples contained 1 and 5 wt% Eu₂O₃. The europium concentration in a sample is expressed in terms of the weight percent of Eu₂O₃ that would be present in a fully densified SiO₂ glass in which water and organics are absent. Reaction and gelation occurred at room temperature. After gelation the gels were aged and dried at 60^oC and 90^oC for two days. Further heat treatment in air was carried out in a box furnace. The samples were ultimately heated to 800^oC.

All precursors were studied in solution (prepared with a precursor concentration of 0.1M). In all cases, the solvents and solutions were prepared immediately before the use. Spectra are presented for the sol or initial mixing stage (before gelation), the wet gel (corresponding to the material after aging at room temperature in the ambient atmosphere) and the gels heated to 90^oC and 800^oC.

The Eu³⁺ precursors studied that involved fluorinated ligands were: Eu(fod)₃, (CF₃SO₃)₃Eu, (CFS) and (CF₃CO₂)₃Eu.3H ₂O, (CFC) and the chelate ligands were: Eu(fod)₃, Eu(thd)₃ e Eu(acac)₃.

Spectroscopic Measurements. The broadband emission spectra were obtained, at room temperature, by exciting samples with the 466 nm and 514.5 nm lines of a Coherent Innova Series 70-5 argon ion laser. Luminescence was collected on a 1-meter monochromator and detected by a photomultiplier tube.

Fluorescence line narrowing measurements were carried out using a pulsed tunable dye laser pumped by a Q-switched Nd:YAG. The dye laser provided the tunable emission from 571-580 nm to selectively excite the ⁷F₀®⁵D₀ transition of Eu³⁺. Spectra were normalized to the ⁵D₀®⁷F₁ peak intensity and were collected at 77 K. Excitation spectra were measured by scanning the dye laser across the ⁷F₀®⁵D₀ transition while the strongest fluorescence band was detected. Luminescence decay measurements were recorded using a digital storage oscilloscope. The lifetime were obtained by calculating the area under the normalized decay curve.

RESULTS AND DISCUSSION

Fluorescence Spectra. Figures 1 and 2 presents broadband fluorescence spectra for the 5.0 wt% Eu₂O₃ doped samples prepared from the different Eu³⁺ precursors, in different stages of sol-gel reaction. The precursors used are indicated in the figures.

Important spectral changes related to the emission peaks relative intensity are observed. Lifetime shortening and the relative weakness of the ⁵D₀®⁷F₀ and ⁵D₀®⁷F₂ emission bands reflect a symmetric solvation shell of the Eu³⁺ bonding environment.

The spectra of samples prepared from Eu(fod)₃, show that the effect of this ligand is significant. The spectra show marked enhancement of the ⁵D₀®⁷F₂ fluorescence, indicative of low symmetry sites. Changes in intensity are more pronounced and the ⁵D₀®⁷F₀ emission, which is almost unobservable in the absence of chelating agent, becomes very significant upon chelation. The presence the chelate organic ligands in the inner-coordination shell reduces the nonradiative decay probability. This increased intensity comes from a reduction in the symmetry on the europium species that makes the transition 'more allowed'. The same considerations can be made for Eu(thd)₃, but in this case, the effect is not so pronounced.

Comparision of figures 1 and 2 reveals that the spectra of (CF₃SO₃)₃Eu, (CF₃CO₂)₃Eu and Eu(NO₃)₃.6H₂O show very similar intensity and behavior. Previous work¹⁴ has shown that the NO₃^- ion has the ability to form an inner sphere complex with Eu³⁺. In the nitrate solution, coordination will be shared by the NO₃^- ions and water molecules, resulting in increased asymmetry compared to coordination only by water. For (CF₃SO₃)₃Eu and (CF₃CO₂)₃Eu, the Eu³⁺ bonding environment most likely consists of coordinated water molecules and the organic ligands, in the first coordination sphere, as evidenced by the similarity between the wet gel and sol spectra and the aqueous solution.

The spectra of the wet (unheated) gel shows a (⁵D₀®⁷F ₂)/(⁵D₀®⁷ F₁) peak height ratio less than one, supporting arguments of a higher symmetry solvation shell.

Figure 2(b) shows the spectra after heating 2 days at 90^oC. Heating leads to increased solvent and water evaporation. The spectra, for all precursors, show a significantly enhanced ⁵D₀®⁷F₂ intensity, indicating the presence of low symmetry sites. Furthermore, the effect of line broadening with thermal treatment indicates a progressive increase in the distribution of site symmetries for Eu³⁺.

From figures 1 (a) and (b) and figure 2, it can be seen that the emission profile observed is very similar before and after gelation. This suggests that although the material is macroscopically rigid, the Eu³⁺ ions retains partial coordination by a liquid-like environment. This result can be supported by the lifetime values obtained.

Lifetime Measurements. Room-temperature luminescence lifetime of the ⁵D₀ state are presented in table I. Lifetime values are extremely sensitive to the constitution of the metal ion coordination sphere. Particularly, the lifetime of the ⁵D₀ level of the Eu³⁺ site is very sensitive to the hydroxyl groups concentration. These species lead to enhanced nonradiative decay of luminescent ion through hydroxyl quenching and so lifetime shortening is expected.

The lifetimes obtained for samples prepared from Eu (thd)₃ and Eu(fod)₃ were longer than the lifetimes obtained using the other precursors. The organic ligand shell shields the Eu³⁺ion from immediate local surroundings and the fluorinated ligand in Eu(fod)₃ impart hydrophobic character to that precursor. This can be seen from the emission spectra and the lifetime value in all samples prepared. For the other precursors, the lifetime is not affected until some heat treatment is performed. Although the sample is not dehydrated at 90^oC, water and solvent concentrations are decreased. The heat treatment lead to lenghtening of lifetime. The lifetime values for all precursors are longer if compared to lifetime of the sample prepared from europium nitrate.

Excitation Spectra. The ⁷F₀®⁵D₀ excitation spectroscopy can be used as diagnostic of the bonding environment of Eu^{3+ 15}. Figure 3 presents the ⁷F₀®⁵D₀ excitation spectra taken at room temperature for all Eu³⁺ precursors studied in the sol stage.

Figure 3

Figure 3 (e) shows two unresolved peaks for the Eu(thd)₃ sol. The higher energy component coincides with the single sharp line observed in the others precursors (579.1 nm). The existence of two distinct lines in the ⁵D₀ - ⁷F₀ region demonstrates that at least two complex species are present. In this case, probably, the coordination sphere is composed by water and alcohol (solvent) molecules.

Figure 3 (f) shows the excitation spectra for the sol prepared from the florinated and chelate precursor, Eu(fod)₃. The bandwidth is broader (~15 Å), that can be an evidence for the presence of different Eu³⁺ environment. DMF is a strong Lewis base used, to coordinate with the chelated Eu³⁺, which may compete with water molecules for the Eu³⁺ coordination. The bulky ligand groups would tend to keep unbonded molecules comparatively far from the central ion, and, in view of the random spatial arrangement of these molecule, a broader range of bonding environment for Eu³⁺.

Figure 3 (d) shows the excitation spectra for Eu(acac)₃ precursor. In this case, the presence of only methyl groups on the chelate ligands instead of bulkier groups may allow solvent and water molecules to approach more closely than in the other cases. This can result in less intense fluorescence and in shorter lifetime values than observed for the other chelate precursors.

Eu³⁺ doped silica gel partially densified. Room temperature luminescence broadband spectra of 1 wt% Eu₂O₃ doped samples prepared from the Eu³⁺ precursors and densified at 800^oC were measured (Figure 4) and compared to the spectrum of a sample prepared from Eu(NO₃).6H₂O. The spectrum of the sample prepared from (CF₃COO)₃Eu (is not presented) was similar to that of the non-fluorinated sample prepared from Eu(NO₃).6H₂O. The other two fluorinated samples had spectra that differed in two important ways from the spectrum of the non-fluorinated sample. First, the ⁵D₀®⁷F₀ transition was broader and second, the ⁵D₀®⁷F₂ transition intensity was much stronger. The first observation suggests that a wider range of Eu³⁺ bonding sites is present in the samples prepared from Eu(fod)₃ and (CF₃SO₃)₃Eu and the second observation implies that more highly distorted/less symmetrical Eu³⁺ bonding environment are present. Selectively excited fluorescence spectra measured at 77 K for the samples prepared from the fluorinated Eu³⁺ precursors revealed no line narrowing effect. Figure 5 shows broadband and selectively excited emission spectra for 1 wt % silica gel prepared from Eu(fod)₃. The FLN spectra of all precursors studied were very similar to those of the Eu(fod)₃ precursor and therefore are not presented. This result indicates that the use of fluorinated or chelate Eu³⁺ precursors do not inhibit Eu³⁺ clustering.

The fluorinated precursors had a pronounced effect on the Eu³⁺ lifetime. Table 2 summarizes 77 K lifetime results of samples containing 5 wt% Eu₂O₃ heated to 800^oC for two excitation (l_exc) and two detection (l_em) wavelengths. A clear increase in lifetime was observed when fluorinated precursors were used. Since the fluorescence line narrowing spectra indicate that appreciable clustering is present, the longer lifetimes indicate that the use of fluorinated Eu³⁺ precursors leads to lower hydroxyl content in the densified gels.

CONCLUSIONS

Fluorescence, excitation, fluorescence line narrowing, and lifetime studies were presented during different sol-gel stages of Eu³⁺-doped TMOS gels prepared using diferent Eu³⁺ precursors. The FLN studies indicate that significant Eu³⁺ clustering occurs in the gels. Much longer lifetimes, relative to samples prepared from Eu(NO₃).6H₂O, were observed in gels heated to 800 ⁰C that were prepared from the fluorinated Eu³⁺ precursors. This result suggests that the presence of fluorine in the pores of the gel significantly facilitates the removal of water during heat treatment and that the use of fluorinated precursors may lead to rare earth doped sol-gel materials with higher luminescence efficiency.

Chelate precursors provide a asymmetric environment and consequently enhanced Eu³⁺ emission and a general increase in a lifetime values for the initial stage of the sol-gel reaction.

ACKNOWLEDGMENT

The authors acknowledge financial support from National Science Foundation-Division of Earth Sciences, the UW Graduate School, the Brazilian National Council of Research, CNPq, Fapemig and FINEP/PADCT.

• 1. Mackenzie, J. D., Ed.; Sol-Gel Optics III, SPIE Proceedings Series. 1994, 2288
• 2. Reisfeld, R.; Jörgensen, C.; J. Structure and Bonding 1992, 77, 207.
• 3. Klein, L. C.; Annu. Rev. Mater. Sci. 1993, 23,437.
• 4. Ferrari, M.; Campostrini, R.; Carturan, G.; Montagna, M.; Phil. Mag. B 1992, 65, 251,
• 5. Matthews, L. R.; Wang, X.; Knobbe, E. T.; J. Non-Cryst. Solids. 1994, 178, 44
• 6. Levy, D.; Reisfeld, R.; Chem. Phys. Lett. 1984, 109, 593
• 7. Thomas, I. M.; Payne, S.A.; Wilke, G. D.; J. Non-Cryst. Solids 1992, 151, 183.
• 8. Fujiyama, T.; Yokoyama, T.; Hori, M.; Sasaki, M.; J. Non-Cryst. Solids. 1991, 135, 198.
• 9. Costa, V. C.; Lochhead, M. J.; Bray, K. L.; J. Chem. Mater. 1996, 8, 783.
• 10. Matthews, L. R.; Knobbe, E. T.; Chem. Mater 1993, 5, 1697.
• 11. Brecher, C.; Samelson, H.; Lempicki, A.; J. Chem. Phys 1965, 42, 1081.
• 12. Sanchez, C.; Lebeau, B.; Viana, B.; In Sol-Gel Optics III, SPIE Proceedings Series; Mackenzie, J. `D. Ed.; SPIE: Bellingham, 1991; 2288, p. 227.
• 13. Pope, E. J. A.; Mackenzie, J. D.; J. Am. Cer. Soc. 1993, 76, 1325.
• 14. Breen, P. J.; Horrocks, W. Dew. Jr.; Inorg. Chem 1983, 22, 536.
• 15. Albin, M; Horrocks, W. Dew. Jr.; Inorg. Chem 1985, 24, 895.

Publication Dates

• Publication in this collection
  04 Dec 2003
• Date of issue
  June 1998

History

• Accepted
  29 Apr 1997
• Received
  02 Dec 1996