A 125Te and 23Na NMR investigation of the structure and crystallisation of sodium tellurite glasses

https://doi.org/10.1016/j.ssnmr.2004.06.010Get rights and content

Abstract

125Te static nuclear magnetic resonance (NMR) and 23Na and 125Te magic angle spinning (MAS) NMR have been used, in conjunction with X-ray diffraction, to examine the structure and crystallisation behaviour of glasses of composition xNa2O·(1−x)TeO2 (0.075x0.4). The MAS NMR 23Na spectra from the glasses are broad and featureless but shift by ∼+5 ppm with increased x, i.e. as the system becomes more ionic. The static 125Te NMR spectra show an increase in axial symmetry with increasing x, indicating a shift from predominantly [TeO4] to [TeO3] structural units. The 23Na and 125Te spectra from the crystallised samples have been fitted to obtain information on the sites in the metastable crystal phases, which are the first to form on heating and which are therefore more closely related to the glass structure than thermodynamically stable crystal phases. New sodium tellurite phases are reported, including a sodium stabilised, face centred cubic phase related to δ-TeO2; a metastable form of Na2Te4O9 containing 3 sodium and 4 tellurium sites; and a metastable form of Na2Te2O5 containing 2 sodium sites. There is evidence of oxidation of TeIV to TeVI occurring in glasses with high values of x and, at x=0.40 and 0.50 (outside the glass forming range), some sodium metatellurate (Na2TeO4) is formed at the same time as sodium metatellurite (Na2TeO3). The 125Te shift is very sensitive to environment within the sodium tellurite system, covering more than 320 ppm, with anisotropies varying from 640 to 1540 ppm. The lack of features in the 125Te spectra of the glass phases, combined with the large shift range and high but variable anisotropy, means than it is not possible to obtain a unique fit to any presumed species present. Furthermore, the chemical shift anisotropy parameters for three of the four Te sites in the Na2Te4O9 phase are found to lie outside the range used for previous simulations of glass spectra.

Introduction

Glasses based on tellurium oxide have received interest because of their potential for use in optical devices. They combine high non-linear optical transmittance (especially in the infrared region); excellent chemical durability; and high refractive indices and dielectric constants. There have been several structural studies, largely of binary alkali tellurite glasses, using a variety of structural techniques. However, the variability of the tellurium–oxygen polyhedron makes it difficult to characterise the change in the distribution of structural environments with composition.

Neov et al. [1] concluded from neutron diffraction data that, for high TeO2 content glasses (<4 mol% modifier) the tellurium structural unit is a four-coordinated trigonal-bipyramid as found in the paratellurite (α-TeO2) polymorph of crystalline TeO2. They also concluded that, as the TeO2 content was reduced, the sites modified to a three-coordinated structure by the extension of one of the Te–O bonds. Other investigations into alkali-tellurite glass structures [2], [3], [4], [5], [6], [7], [8] have adopted a similar approach; assuming the tellurium polyhedra present within a glass correspond to those found in the known stoichometric crystalline phases. Five tellurium polyhedra have been identified from crystalline phases, and all have been suggested to be present to some extent within tellurium glass systems [2], [3], [4] (Fig. 1). McLaughlin et al. [3] introduced the notation Qmn, where n represents the number of bridging oxygens and m is the co-ordination number and they suggested [4] that the abundance of each tellurite polyhedron is dependent on the charge on the species, resulting in a decrease in the numbers of non-charged tellurite polyhedra (Q44, Q32) with increasing modifier, and an increase in the numbers of charged polyhedra (Q43, Q31, Q30). Sakida et al. [2], [5] conducted 125Te NMR studies of tellurite crystals and glasses. They produced a plot of asymmetry parameter η versus chemical shift anisotropy Δδ for a range of tellurite crystals which enabled them to map areas which contained specific structural units. This was then used to identify the units in alkali tellurite glasses as TeO4 trigonal bipyramids (Q4) and TeO3 trigonal pyramids (Q3). They fitted their static 125Te NMR glass spectra to contributions from these two species.

Zwanziger and co-workers have published several papers on the crystallography and 23Na nuclear magnetic resonance (NMR) of sodium tellurites and the relationship of the crystal structures to possible structural units in the corresponding xNa2O·(1−x)TeO2 glasses [3], [4], [9], [10], [11], [12], [13]. In [10] they noted that the x=0.20 composition produces the most stable glass in the system despite also producing a stoichiometric crystal phase Na2Te4O9. They reported the crystal structure of this phase and ascribed the stability of the tellurite glasses in general to the wide range of tellurium polyhedra present and the significant reconstruction necessary in order to produce a transformation to the stable crystal phases. In [11] they showed that the coordination number of sodium decreases from about 6 for x<0.20 to about 5 at x=0.35, with the sodium environment at x=0.20 being very different for the crystal and glass. In [12] they used spin-echo 23Na NMR to obtain information on the spatial distribution of sodium in the glass system. Whilst the sodium ions appear to be randomly distributed at low concentrations, there is significant intermediate range order at higher concentrations, being most pronounced at x=0.20. The other reported crystal phases in the sodium tellurite system are Na4Te4O10 [13] and Na2TeO3. The end-member TeO2 has been reported to have several polymorphs, the stable paratellurite (α-TeO2), β-TeO2 and the more recently reported γ- and δ-forms, the latter being formed from TeO2 containing significant levels of impurity [14], [15], [16], [17].

Glasses undergoing devitrification are likely to transform first to the crystal phase whose structure is most closely related to that of the glass. This metastable crystal structure is often a high temperature polymorph of the phase which is thermodynamically stable under ambient conditions, and to which it transforms on further heating, or may even be a polymorph which has no stable existence range at atmospheric pressure. The simplest examples of this are vitreous silica, which forms cristobalite on devitrification rather than quartz or tridymite [18] and vitreous antimony trioxide which forms valentinite at about 340 °C instead of senarmontite which is the stable phase up to 570 °C [19]. Further insight into the environments of Na and Te in sodium tellurite glasses may therefore be obtained by studying the crystal phases which are formed during the first devitrification event in these glasses. In this paper we report the results of 23Na and 125Te NMR studies of sodium tellurite glasses and their devitrification products.

Section snippets

Sample preparation

Samples with nominal composition xNa2O·(1−x)TeO2, where x=0.05, 0.075, 0.1, 0.15, 0.2, 0.25, 0.3, 0.33 0.35, 0.4, and 0.5 were prepared from Na2CO3, and TeO2, using a total batch weight of 50 g. The batch components were thoroughly mixed by tumbling for at least 24 h and then melted, in a 90Pt/10Rh crucible, by heating at 5 °C a minute to 800 °C and holding at temperature for 18 min. The melt was quenched between steel plates and the resulting material stored in a vacuum desiccator to prevent any

Samples

The sample with x=0.5 was fully crystalline whilst samples with x=0.05, 0.075 and 0.4 partially crystallised on cooling. The remaining compositions gave glasses with no visible crystallisation or crystal phase detectable by XRD. Fig. 2 shows the XRD patterns from all of the samples as formed. The diffraction patterns of the crystalline phases formed on cooling the x=0.05 and 0.075 samples matched the pattern for paratellurite (α-TeO2). Similarly, the phase formed in the x=0.40 and 0.50 samples

x=0.00

The fit to the 125Te spectrum from α-TeO2 shown in Fig. 8a yielded CSA parameters: δ=754ppm, Ω=1331ppm, κ=0.28 (Table 3) which can be compared with those obtained by Sakida et al. [5] (δ=776ppm; Ω=1509ppm; κ=0.22). The significant difference in Ω is likely to come from the better delineated spectrum here.

x=0.10

Fig. 4 shows the XRD patterns obtained by heating the x=0.10 sample at 325, 367 and 423 °C. They are compared with the pattern from α-TeO2. The sample crystallised at 325 °C is almost fully

Conclusions

It is readily apparent that phase development in the Na2O–TeO2 system is complex and that several metastable crystalline phases can be obtained by controlled devitrification of glasses of different composition. We have produced a number of these new metastable phases by heating at the first exotherm in the DTA for various compositions of the sodium tellurite system. These include a Na stabilised δ-TeO2 and Na2Te4O9 which has 3 sodium and 4 tellurium sites where detailed information has been

Acknowledgements

The authors thank the referees for stimulating us into deriving the full CSA parameters for 125Te.

References (24)

  • S. Neov et al.

    J. Non-Cryst. Solids

    (1995)
  • S. Sakida et al.

    J. Non-Cryst. Solids

    (1999)
  • J.C. McLaughlin et al.

    J. Non-Cryst. Solids

    (2000)
  • S. Sakida et al.

    J. Non-Cryst. Solids

    (1999)
  • T. Sekiya et al.

    J. Non-Cryst. Solids

    (1992)
  • J.W. Zwanziger et al.

    J. Non-Cryst. Solids

    (1995)
  • K. Shioya et al.

    J. Non-Cryst. Solids

    (1995)
  • A.P. Mirgorodsky et al.

    J. Phys. Chem. Solids

    (2000)
  • O. Noguera et al.

    J. Non-Cryst. Solids

    (2003)
  • R.T. Hart et al.

    J. Non-Cryst. Solids

    (2004)
  • M. Trömel et al.

    J. Less-Common Met.

    (1985)
  • J.C. McLaughlin et al.

    J. Phys. Chem.

    (2001)
  • Cited by (24)

    • Structure of BaO-TeO<inf>2</inf> glasses: A two-dimensional <sup>125</sup>Te NMR spectroscopic study

      2018, Journal of Non-Crystalline Solids
      Citation Excerpt :

      In this notation system, n is the coordination number of the central Te atom, i.e., the number of nearest neighbor oxygen atoms and m represents the number of bridging oxygen and hence, m ≤ n. Consequently, each of these Q species carries (n - m) non-bridging oxygen (NBO) atoms. A total of five different types of Te species are found in alkali tellurite crystals (Fig. 1), two of which are fourfold coordinated (trigonal bipyramids): Q44 and Q43, and three are threefold coordinated (trigonal pyramids): Q32, Q31 and Q30 [3,7,9–11]. It may be noted here that the “Q” in the “Q species” notation stems from “quaternary”, meaning four bonds.

    View all citing articles on Scopus
    View full text