Judd–Ofelt analysis of lanthanide doped silica–PEG hybrid sol–gels

Abstract

Lanthanide complexes of 1,10-phenanthroline (phen), 2,2′-bipyridine (bipyridyl, bpy) and dipicolinate (dpa), together with uncomplexed lanthanide ions (Ln = Nd, Sm, Dy, Ho, Er) were doped into silica–polyethylene glycol (SiO₂–PEG) inorganic–organic hybrid materials. The samples were prepared via a sol–gel process. The intensities of the f–f transitions in the absorption spectrum were analysed by application of the Judd–Ofelt theory. Sets of phenomenological Ω_λ intensity parameters have been extracted. The hypersensitive transitions are strongly influenced by the drying time of the sol–gel glass. The PEG molecules compete with the organic ligand for complex formation to the lanthanide ions complex. 1,10-Phenanthroline and bipyridyl complexes are not stable in the presence of PEG, but the dipicolinate complexes are.

---

1. Introduction

The sol–gel process allows preparation of silica-based materials at ambient temperature. Pure SiO₂ glass can be made by hydrolysis and polymerisation of silicon alkoxides. Lanthanide complexes of organic ligands which are soluble in the sol, can be easily doped without decomposition in this glass matrix. The resulting silica materials are transparent and have good mechanical properties. Bulk glass, coatings, fibres can be made,¹ but there are some drawbacks like the low solubility of the complexes at the low pH needed for the hydrolysis reaction. The cracking of the material and the water present in the pores also cause problems. One way to overcome these problems is to neutralise the solution after hydrolysis and to introduce organic parts in the material that make it more flexible. The properties of the inorganic–organic hybrid materials depend on the chemical nature of the different components.² Polyethylene glycol (PEG) can be introduced in the silica matrix by simple mixing of the PEG with the silica precursors.³ PEG-200 (average molecular weight = 200) is a viscous liquid that is soluble in water and is able to form complexes with metal cations. No strong bonds are formed between the PEG-chains and the silica backbone and these materials are therefore class I hybrids.² The silica–PEG sol–gels are transparent, are easily made, and have good mechanical and optical properties. The solubility of the ligands can be improved by using a buffered solution for the sol–gel synthesis. Other related hybrid materials called ureasils have PEG chains covalently bonded in the silica network through urea bridges.⁴ The urea bridges are able to bind the lanthanides and ureasils are therefore unsuited for this work.

The closed 5s- and 5p-shells efficiently shield the 4f-shell of lanthanides from the environment and this results in unique optical features. Lanthanide f–f-transitions show narrow lines in the near-infrared, visible and ultraviolet part of the spectrum, but the magnetic dipole (MD) and induced electric dipole (ED) transitions are weak.⁵ There is a lot of interest in the trivalent lanthanide ions because of their strong luminescence with high coloric purity and the antenna effect.⁶ When a lanthanide ion is surrounded by a ligand that contains a light-absorbing group, the excitation energy absorbed by the ligand can be transferred to the lanthanide ion (=antenna effect). The strong light absorption capacity by the ligand improves the luminescence yield. Complexes of Eu³⁺ and Tb³⁺ with 1,10 phenanthroline (phen) and 2,2′-bipyridine (bpy) have been studied in silica–PEG sol–gels.⁷ The absorption spectra of different lanthanide complexes doped in silica–PEG sol–gels are the subject of this paper. The lanthanide complexes [Ln(bpy)₂]Cl₃, [Ln(phen)₂]Cl₃, Na₃[Ln(dpa)₃] and LnCl₃ (bpy = bipyridyl, 2,2′-bipyridine, phen = 1,10-phenanthroline, dpa = dipicolinate, Ln = Nd, Sm, Dy, Ho, Er) are incorporated. The induced ED-transitions can be characterised by the three phenomenological intensity parameters Ω_λ(λ = 2, 4, 6). These Ω_λ-intensity parameters derived from the Judd–Ofelt theory represent the square of the charge displacement due to induced electric dipole transitions.^8,9 The parameters can be determined by a least-squares fit to the experimental dipole strengths extracted from the absorption spectra and can be used to rationalize the spectroscopic and structural properties of the complexes and the hybrid materials.¹⁰ The Ω₂-parameter, gives information on the intensities of the so-called hypersensitive transitions, i.e. transitions which are strongly influenced by small changes in the coordination sphere. Comparisons are made between the different ligands, the different lanthanides and the sol–gel materials. The influence of PEG and water will be discussed.

2. Experimental procedures

The silica–PEG sol–gel glasses were synthesized by hydrolysis and polymerisation of the monomeric precursor tetramethylorthosilicate (TMOS, purchased from Fluka) with water and PEG-200 (purchased from Fluka). First, a solution A was made by mixing 8 mL of TMOS with 2 mL of water containing hydrochloric acid or nitric acid to obtain pH = 2. This solution was stirred for 1 h. In solution B, PEG was mixed with water or a buffered water solution in a 80% (w/w) composition. An equivalent quantity of Na₂HPO₄ and NaH₂PO₄ was used to make the buffer (pH = 6.33).¹¹ The final solution C was prepared by mixing 8 mL of solution A with 40 mL of solution B. Stirring was continued for 20 min. A lanthanide complex can be added in this stage. 3 mL portions of solution C were poured in disposable PMMA (polymethylmethacrylate) UV-cuvettes. All volumes were exactly pipetted and the masses were exactly measured because the lanthanide concentrations have to be known for the Judd–Ofelt analysis. Parafilm® was used to seal the cuvettes for direct measurement and aluminium foil was used for the cuvettes in the oven (50 °C). The solutions gelled after a few hours. Ligands that do not dissolve or where the binding atom is protonated in strong acid environment can be used and the resulting silica–PEG sol–gels are transparent.

The preparation of lanthanide(III) phenanthroline and bipyridyl complexes [Ln(phen)₂]Cl₃·2H₂O and [Ln(bpy)₂]Cl₃·2H₂O (phen = 1,10-phenanthroline; bpy = 2,2′-bipyridine; Ln = Nd, Sm, Dy, Ho and Er) is described elsewhere.¹⁰ For the Na₃[Ln(dpa)₃] complexes, pyridine-2,6-dicarboxylic acid (dipicolinic acid) is dissolved in a warm 0.2 M NaOH solution. The pH must be around pH = 7 or 8. The resulting solution is further concentrated by evaporation of the water and the pH is measured again. Before cooling down, an equivalent quantity (Ln∶dpa 1∶3) of the lanthanide salt is added. Small crystals appear after a few minutes. The precipitate was filtered, then washed several times with cold water and dried in a vacuum oven. The dried complexes were characterised by IR spectrometry and CHN elemental analysis.

A different approach was used for the glasses containing LnCl₃ (Ln = Nd, Sm, Dy, Ho, Er). An exact quantity of Ln₂O₃ was dissolved in a concentrated solution of HCl. Repetitive evaporation of the solution and redissolving in water removed the excess of hydrogen chloride. Finally a neutral water solution with a known quantity of Ln³⁺ was obtained. This solution was added to the water fraction in solution B.

The buffered silica–PEG sol–gel method was used for the dipicolinate, bipyridyl and phenanthroline complexes. When LnCl₃ was added to a neutralised sol–gel lanthanide phosphates precipitated, therefore pH = 2 was used for these samples. Samples with different concentrations were prepared.

Another experiment was done with different equivalents (1–20) of phenanthroline or bipyridyl added to a fixed lanthanide concentration. Exact amounts of ligand were added to a solution B containing a fixed concentration of Ln³⁺.

Absorption spectra were recorded at room temperature using a Shimadzu UV-3100 spectrophotometer. Three absorption spectra of each sample were recorded: one before gelling, one after 24 h of drying at 50 °C and a third one after a month of drying at 50 °C. During this month of heat treatment the volume of the glasses shrunk further to between 85% and 95% of the starting volume. Also spectra of [Ln(phen)₂]Cl₃·2H₂O and [Ln(bipy)₂]Cl₃·2H₂O were recorded in solution B and in pure PEG.

3. Results

Judd–Ofelt calculations can be used to rationalize the transition intensities in disordered systems like silica–PEG sol–gels. The Ω_λ-parameters can be determined via a least-squares fit of calculated against experimental dipole strengths.⁵ Before calculation of these parameters, cluster forming was checked. This is shown in Fig. 1 for three transitions. At high lanthanide concentrations the lanthanide ions can form clusters and influence their neighbours, resulting in a change in dipole strength.¹² No such behaviour was observed in the concentration region. A concentration less than 0.05 M was chosen for the experiments to avoid solubility problems. This concentration in combination with the 1 cm cell path length was enough to obtain reproducible parameters.

Fig. 1 Dipole strength of three transitions: (a) ⁴F_5/2 ← ⁴I_9/2, (b) ⁴S_3/2 ← ⁴I_9/2 and (c) ⁴G_5/2 ← ⁴I_9/2 of Nd³⁺ at different concentration in PEG–silica sol–gels.

The relationships between experimental dipole strengths D_exp, the theoretical calculated dipole strengths D_th and the Judd–Ofelt parameters are given in eqn. (1)–(3).

D_exp = D_th (3)

Dipole strengths are expressed in Debye² (D²). The factor 10³⁶ is used for the conversion between D² units and esu² cm² (1 D = 10⁻¹⁸ esu cm). Correction for a dielectric medium (n is the refractive index of silica–PEG sol–gels) is incorporated in the factor (n²+2)²/9n. The origin of factor 108.9 is described in detail elsewhere.⁵ The degeneracy of the ground state equals 2J + 1 and the calculated |〈J‖U^(λ)‖J′〉|² are squared reduced matrix elements. The squared reduced matrix elements can be calculated and have been tabulated by Carnall et al.¹³ The dipole strengths of most transitions in lanthanide ions are little affected by the environment, but the hypersensitive transitions, that have high |〈J‖U⁽²⁾‖J′〉|² squared reduced matrix elements, are very sensitive to small changes in the ligand coordination sphere. Important luminescent transitions for optical application like the ⁵D₀ → ⁷F₂ of Eu³⁺ (ca. 615 nm) and ⁴I_11/2 → ⁴F_3/2 of Nd³⁺ (ca. 1064 nm) are hypersensitive and a lot of effort is being used to increase their intensities. Jørgenson and Judd noted that hypersensitive transitions obey the selection rules |ΔS| = 0, |ΔL| ⩽ 2 and |ΔJ| ⩽ 2 i.e. the same selection rules as for electric quadrupole transitions.¹⁴ The hypersensitivity is reflected by the value of the Ω₂-parameter, related to |〈J‖U⁽²⁾‖J′〉|². These Ω₂-parameters are therefore an important spectroscopic probe for the coordination changes. Hypersensitive absorption transitions in the glasses we have studied are the ⁴G_5/2 ← ⁴I_9/2 transition of Nd³⁺ (586 nm or 17 300 cm⁻¹), the ⁵G₆ ← ⁵I₈ transition of Ho³⁺ (451 nm or 22 100 cm⁻¹) and the ²H_11/2 ← ⁴I_15/2 and ⁴G_11/2 ← ⁴I_15/2 transitions of Er³⁺ (at 521 and 378 nm or 19 200 cm⁻¹ and 26 400 cm⁻¹ respectively). Sm³⁺ and Dy³⁺ also have hypersensitive transitions but the transition they are situated at is 1562 nm and 1298 nm where a broad absorption band of the silica–PEG sol–gel matrix is situated. It was impossible to calculate the Ω₂-parameter for these ions. The Ω₄- and Ω₆-parameters are related to transitions that are little affected by their surroundings.⁵

Phenanthroline and bipyridyl dissolve very well in these hybrid sol–gel materials. The influence of high concentrations of the ligands on the dipole strength, while Nd³⁺ concentration was kept constant, was investigated (Fig. 2). Extra amounts of 1,10-phenanthroline influence the dipole strength of the hypersensitive transition ⁴G_5/2 ← ⁴I_9/2, with a plateau when the ligand concentration is twelve times the lanthanide concentration. The intensity of the non-hypersensitive transitions does not change as a function of the phenanthroline content. Other lanthanides show similar behaviour. Little or no influence on the transition intensities is observed when the bipyridyl concentration is increased.

Fig. 2 Dipole strength of three transitions: ⁴F_5/2 ← ⁴I_9/2, ⁴S_3/2 ← ⁴I_9/2 and ⁴G_5/2 ← ⁴I_9/2 of Nd³⁺ at different metal to 1,10-phenanthroline ratio (Nd³⁺∶phen) in PEG–silica sol–gels.

There was no difference between the Judd–Ofelt parameters of the final solution C and the gelled samples that stayed in the oven for 24 h. The values of the Ω₂-parameters do not change when the starting solution was transformed into a gel. During a longer drying period, these parameters do change. In contrast, the Ω₄- and Ω₆-values vary little upon heating and no general trend can be observed. The Ω₄- and Ω₆-parameters both decrease when going from Nd³⁺ to Sm³⁺ and from Dy³⁺ to Er³⁺. When the Ω_λ-values for the different ligands are compared, we distinguish two groups. Between the parameters of [Ln(bpy)₂]Cl₃, [Ln(phen)₂]Cl₃, LnCl₃ little variation is found. The parameters for Na₃[Ln(dpa)₃] are quite different. In Fig. 3, the absorption spectrum of the hypersensitive transition of Nd³⁺ and two other transitions are compared for two ligands. There is crystal-field fine structure in the spectrum of Na₃[Nd(dpa)₃]. Spectra of the other complexes are similar and do not show fine structure.

Fig. 3 Comparison of the fine structure in the room temperature absorption spectrum of the hypersensitive transition ⁴G_5/2 ← ⁴I_9/2 of [Nd(bpy)₂]Cl₃ and Na₃[Eu(dpa)₃] in PEG–silica sol–gels.

For comparison the calculated Judd–Ofelt parameter of Ln³⁺ in pure PEG or in 80% PEG with 20% water are also summarised in Table 2 (see later). It is clear that the 20% water added affects the spectrum. No values are given for Dy³⁺ because the faults were too big. They are in the order of 1 for the Ω₄-parameter and 1.5 for the Ω₆-parameter.

4. Discussion

Previous experiments in lanthanide doped sol–gel materials reported cluster forming at high concentration levels.¹⁵ A deviation in the dipole strength by this cluster formation can lead to false parameters values. From the linear relationship between concentration and dipole strength of hypersensitive and non-hypersensitive peaks, it was deduced cluster forming was absent in this concentration interval (Fig. 1).

The Ω₂-parameter is associated with short-range coordination effects and its value increases with increasing coordination number, higher basic character of the ligand and higher covalence of the bonding.⁵ The intensity of the hypersensitive transition and the value of the Ω₂-parameter depends on the ligand type: oxygen donor > nitrogen–oxygen donor > sulfur donor > aqua ion. There is an inverse relationship between the length of the ligand–lanthanide bond and the Ω₂-parameter. The Ω₄- and Ω₆-parameters that are related to the other transitions depend on long-range effects. More rigid matrices have lower Ω₄- and Ω₆-values. Rigidity increases in the following order: coordination complexes < halide vapours < hydrated ions < viscous solutions < glasses < crystalline mixed oxides.⁵

No difference in parameter values is observed during the 24 h of drying. In this period the silica backbone was formed and the samples gelled. The long range-effects that influence Ω₄- and Ω₆-values are not changed. This is interesting because it indicates that the other components (water, PEG, ligand) are responsible for the transition intensities and not the silica. Results of the complexes in pure PEG and 80% PEG (+20% water) confirm this statement because silica is absent in this case. The Ω₄- and Ω₆-parameters of the 80% PEG are similar to those for solution C and the gels dried for a short period. The Ω₄- and Ω₆-parameters of the pure PEG are similar to those of the long dried gels.

The influence of the different ligands was studied. The spectra of [Ln(bpy)₂]Cl₃, [Ln(phen)₂]Cl₃ and LnCl₃ are very similar. On the other hand the spectrum and therefore the Ω_λ-parameters are totally different for Na₃[Ln(dpa)₃]. This is evident from Table 1. Luminescence of phenanthroline and bipyridyl complexes of Tb³⁺ and Eu³⁺ in silica–PEG sol–gels was previously studied.⁷ These organic ligands are used to improve luminescence yield via the antenna effect. This means that a part of the ligands has to be close enough to the lanthanide to transfer the excitation energy.¹⁶ Bipyridyl dissolves very well in PEG. But at high concentration (C > 0.08 M) the phenanthroline complex crystallises and there is no homogenous distribution of the lanthanide. In Table 1, little effect of complex formation is observed; the values are not much higher than those for LnCl₃ where the only interactions are ionic. This means that the ligands are close enough to the lanthanide to transfer energy at high concentration (C > 0.1 M bipy) but still do not bond strongly and are not able to influence the lanthanide ion much. Water and PEG dominate all the parameters. Only at higher concentrations does 1,10-phenanthroline influence the hypersensitive transition of Nd³⁺ (Fig. 2). Beyond a metal-to-ligand ratio (Ln³⁺∶ligand) 1∶12, further addition of the ligand gives little change in dipole strength and a plateau is reached. Therefore samples with ratio 1∶12 were prepared. For bipyridyl, which is a more flexible molecule than 1,10-phenanthroline and which is better soluble in PEG, little or no change was observed. It appears that the cation complexing capacity of PEG is too high in comparison to that of bipyridyl so that the bipyridyl ligand cannot efficiently compete with the PEG molecules for the lanthanide ion.

Table 1 Judd–Ofelt parameters of the silica sol–gel glasses, before (1) and after a heat treatment of one month at 50 °C (2)^a

Ion	Matrix	Ligand^b		Ω 2	Ω 4	Ω 6
a All parameters are expressed in 10^−20 cm^2. b Phen 1∶12 means that the lanthanide to 1,10-phenanthroline ratio was 1 to 12.
Nd^3+	Si–PEG	dpa	(1)	4.2 ± 0.9	6.9 ± 1.5	11.1 ± 1.3
			(2)	3.9 ± 0.9	6.8 ± 1.5	11.5 ± 1.1
		phen 1∶12	(1)	4.1 ± 0.7	5.9 ± 1.0	7.8 ± 0.5
			(2)	4.4 ± 1.3	6.1 ± 1.7	8.1 ± 0.9
		phen 1∶2	(1)	2.2 ± 0.8	6.6 ± 1.1	8.3 ± 0.5
			(2)	2.8 ± 0.6	6.2 ± 1.0	7.9 ± 0.6
		bipy 1∶12	(1)	1.6 ± 0.6	5.8 ± 0.9	7.7 ± 0.4
			(2)	2.4 ± 1.0	6.9 ± 1.2	7.8 ± 1.0
		bipy 1∶2	(1)	1.2 ± 0.7	5.9 ± 1.0	7.8 ± 0.5
			(2)	2.7 ± 0.5	5.1 ± 0.7	6.5 ± 0.4
		chloride	(1)	0.7 ± 0.6	6.5 ± 0.7	7.7 ± 0.4
			(2)	2.3 ± 0.6	6.5 ± 0.8	7.3 ± 0.5
	PEG	chloride	(1)	0.9 ± 0.9	6.5 ± 1.2	7.8 ± 0.7
			(2)	2.7 ± 0.7	5.8 ± 0.9	6.3 ± 0.5
Ho^3+	Si–PEG	dpa	(1)	6.4 ± 0.9	4.6 ± 1.4	5.0 ± 0.8
			(2)	6.6 ± 0.6	4.8 ± 0.9	5.4 ± 0.6
		phen 1∶12	(1)	5.1 ± 0.7	5.5 ± 1.2	2.4 ± 0.5
			(2)	5.7 ± 0.8	4.9 ± 1.5	3.2 ± 0.6
		phen 1∶2	(1)	3.4 ± 0.4	3.2 ± 0.6	3.2 ± 0.3
			(2)	3.9 ± 0.6	3.5 ± 0.9	3.8 ± 0.5
		bipy 1∶12	(1)	2.8 ± 0.5	3.5 ± 0.8	3.7 ± 0.5
			(2)	3.7 ± 1.1	4.6 ± 1.5	4.1 ± 0.9
		bipy 1∶2	(1)	1.6 ± 0.3	4.2 ± 0.5	2.7 ± 0.3
			(2)	3.3 ± 0.4	4.0 ± 0.7	2.3 ± 0.4
		chloride	(1)	1.4 ± 0.4	3.3 ± 0.4	3.4 ± 0.3
			(2)	4.1 ± 0.5	3.3 ± 0.8	3.6 ± 0.5
	PEG	chloride	(1)	1.5 ± 0.5	3.3 ± 0.7	3.6 ± 0.5
			(2)	3.9 ± 0.6	3.3 ± 0.9	3.4 ± 0.5
Er^3+	Si–PEG	dpa	(1)	6.5 ± 0.9	1.9 ± 1.2	3.2 ± 1.1
			(2)	6.7 ± 1.2	1.7 ± 1.0	4.0 ± 1.0
		phen 1∶12	(1)	6.2 ± 0.5	1.7 ± 0.6	2.2 ± 0.4
			(2)	7.0 ± 1.2	1.8 ± 1.2	2.6 ± 1.0
		phen 1∶2	(1)	3.3 ± 0.1	2.5 ± 0.2	1.4 ± 0.1
			(2)	5.2 ± 0.4	2.5 ± 0.5	1.4 ± 0.4
		bipy 1∶12	(1)	4.2 ± 0.3	2.5 ± 0.3	2.1 ± 0.2
			(2)	4.7 ± 0.6	2.9 ± 0.7	2.5 ± 0.4
		bipy 1∶2	(1)	3.2 ± 0.2	2.3 ± 0.3	1.6 ± 0.2
			(2)	5.1 ± 0.2	2.1 ± 0.3	1.5 ± 0.2
		chloride	(1)	2.3 ± 0.2	1.9 ± 0.3	1.5 ± 0.2
			(2)	5.5 ± 0.4	2.8 ± 0.6	2.1 ± 0.4
	PEG	chloride	(1)	2.9 ± 0.4	1.5 ± 0.6	1.6 ± 0.4
			(2)	4.8 ± 0.4	1.5 ± 0.5	1.3 ± 0.3

---

Table 2 Judd–Ofelt parameters of the chloride complexes of different lanthanides in 80% PEG + 20% water (A) and 100% PEG^a

		Ω 2	Ω 4	Ω 6
a All parameters are expressed in 10^−20 cm^2. b These parameters could not be determined (see text).
Nd^3+	A	0.9 ± 0.8	6.5 ± 1.2	7.8 ± 0.7
	B	2.7 ± 0.7	5.8 ± 0.9	6.3 ± 0.5
Sm^3+	A	^b	4.6 ± 0.4	1.9 ± 0.2
	B	^b	5.9 ± 0.5	2.1 ± 0.3
Ho^3+	A	1.5 ± 0.5	3.3 ± 0.7	3.6 ± 0.5
	B	3.9 ± 0.6	3.3 ± 0.9	3.4 ± 0.5
Er^3+	A	2.9 ± 0.4	1.5 ± 0.6	1.6 ± 0.4
	B	4.8 ± 0.4	1.5 ± 0.5	1.3 ± 0.3

---

The parameters for the bipy, phen and chloride complexes in pure silica sol–gels have been published.¹⁷ Because PMMA cuvettes were used in this study, the samples could not be heated to 100 °C as in ref. 17. It is therefore difficult to compare the final results. As the long-range parameters of the PEG–silica sol–gels are more determined by PEG than by the silica one can expect that the final parameters will be different. But in both cases the Ω₂-parameter value increases when water is removed.

As can be deduced from the parameter values and from Fig. 3, the dipicolinate complex molecule gives a totally different spectrum. The dipicolinate ligand crystallises with lanthanides in a complex of D₃-symmetry.¹⁸ According to model calculations of Na₃[Ln(dpa)₃] (Ln = Nd, Ho or Er), or complexes with similar tridentate ligands like oxydiacetate in aqueous solution, the most important contribution to the intensities of hypersensitive transitions in these ions can be correlated with the ligand polarizability.¹⁹ Absorption and IR-luminescence of Na₃[Nd(dpa)₃] has been studied in pure silica sol–gels.¹¹ No antenna effect was observed because the gap between energy levels of the ligand and the lanthanide ion is too large. The luminescence of Na₃[Eu(dpa)₃] has also been studied in the past.²⁰ In the present study other lanthanide complexes are doped in a silica–PEG sol–gel and Judd–Ofelt analysis has been done. The spectra of the Na₃[Ln(dpa)₃] complexes (Ln = Nd, Sm, Dy, Ho, Er) in silica–PEG sol–gels still show some crystal-field fine structure and the Ω_λ-parameters differ from the other complexes. This clearly shows that dipicolinate is a better ligand than phenanthroline or bipyridine. The Ω₂-parameters are higher than in the other systems. The tridentate dipicolinate ligand is also strong enough to shield the lanthanide from water. The Ω₄- and Ω₆-parameters are higher than those of the other ligands. The parameters are more determined by the low rigidity of the surrounding complex than by the rigidity of the silica–PEG matrix. This can be explained by efficient shielding over a long range.

Some general statements can be made for both the Ω₄- and Ω₆-parameters. The values for the lighter lanthanides are larger than those for the heavy lanthanides, with a minimum between Sm³⁺ and Dy³⁺. The radial integrals in the expressions of the Judd–Ofelt parameters decrease from over the lanthanide series (no reliable Ω_λ-parameter can be determined for Eu³⁺, Gd³⁺ and Tb³⁺).

5. Conclusions

Monolithic silica–PEG hybrid sol–gels were made at a neutral pH. The glasses were doped with [Ln(bpy)₂]Cl₃, [Ln(phen)₂]Cl₃, Na₃[Ln(dpa)₃] and LnCl₃ (bpy = bipyridyl, 2,2′-bipyridine, phen = 1,10-phenanthroline, dpa = dipicolinate, Ln = Nd, Sm, Dy, Ho, Er) complexes. The Judd–Ofelt theory has been used to compare the intensities of lanthanide complexes in sol–gel materials. The PEG adds flexibility to the sol–gel glass but is also able to form complexes with lanthanide ions. The Ω₂-parameters values are small in PEG. When synthesising hybrid materials, it is important to use ligands that form stable complexes with lanthanide ions. It is also important that the organic phase of the hybrid materials have little complex formation capacity. Ligands like 1,10-phenanthroline and 2,2′-bipyridyl are not strong enough to remove the water and PEG molecules from the first coordination sphere of the lanthanide ions. On the other hand, dipicolinate ligands bind strongly to lanthanide ions and shield the lanthanides from PEG and water molecules.

Acknowledgements

K.D. (research assistant) and K.B. (postdoctoral fellow) thank the F.W.O.-Flanders (Belgium) for financial support. Support by the K.U. Leuven (GOA 98/03) is acknowledged.

References

1. C. J. Brinker and G. W. Scherrer, Sol-Gel Science, The Physics and Chemistry of Sol-Gel Processing, Academic Press, San Diego, 1990.
2. C. Sanchez and F. Ribot, New J. Chem., 1994, 18, 1007.
3. V. Bekiari, G. Pistolis and P. Lianos, J. Non-Cryst. Solids, 1998, 226, 200.
4. V. de Zea Bermudes, L. Carlos, M. Duarte, M. Silva, C. Silva, M. Smith, M. Assunçao and L. Alcacer, J. Alloys Compd., 1998, 275–277, 21.
5. C. Görller-Walrand and K. Binnemans, Spectral Intensities of f-f Transitions on the Physics and Chemistry of Rare Earths, ed. K. A. Gschneidner, Jr and L. Eyring, North-Holland, Amsterdam, 1998, vol. 25, ch. 167, 101.
6. M. Xiao and P. Selvin, J. Am. Chem. Soc., 2001, 123, 7067.
7. V. Bekiari, G. Pisolis and P. Lianos, Chem. Mater., 1999, 11, 3189.
8. B. R. Judd, Phys. Rev., 1962, 127, 750.
9. G. S. Ofelt, J. Chem. Phys., 1962, 37, 511.
10. K. Driesen, K. Binnemans and C. Görller-Walrand, Mater. Sci. Eng. C, 2001, 18, 255.
11. D. Lai, B. Dunn and J. Zink, Inorg. Chem., 1996, 35, 2152.
12. Rare Earth Doped Fiber Lasers and Amplifiers on Optical Engineering, ed. M. Digonnet, vol. 37, Marcel Dekker Inc., New York, 1993.
13. W. T. Carnall, P. R. Fields and K. Rajnak, J. Chem. Phys., 1968, 49, 4443.
14. C. K. Jørgenson and B. R. Judd, Mol. Phys., 1964, 8, 281.
15. W. Strek, J. Legendziewicz, E. Lukowaik, K. Maruszewski, J. Sokolnicki, A. Boiko and M. Borzechowska, Spectrochim. Acta A, 1998, 54, 2215.
16. X. Chuai, H. Zhang, F. Li, S. Wang and G. Zhou, Mater. Lett., 2000, 46, 244.
17. K. Driesen, P. Lenaerts, K. Binnemans and C. Görller-Walrand, Phys. Chem. Chem. Phys., 2002, 4, 552.
18. L. Van Meervelt, K. Binnemans, K. Van Herck and C. Görller-Wallrand, Bull. Soc. Chim. Belg., 1996, 106, 25.
19. M. T. Devlin, E. M. Stephens and F. S. Richardson, Inorg. Chem., 1988, 27, 1517.
20. E. Huskowska, P. Gawryszewska, J. Legendziewicz, C. Maupin and J. Riehl, J. Alloys Compd., 2000, 303, 325.

---