Structure and Bonding in TeO2 Melt and Glass (Talk Slides) - - PDF document

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/343449983

Structure and Bonding in TeO2 Melt and Glass (Talk Slides)

Presentation · August 2020

CITATIONS READS

37

1 author: Some of the authors of this publication are also working on these related projects: Nuclear Materials View project Germanate glasses and melts View project Oliver L. G. Alderman Science and Technology Facilities Council

91 PUBLICATIONS 329 CITATIONS

SEE PROFILE

All content following this page was uploaded by Oliver L. G. Alderman on 05 August 2020.

The user has requested enhancement of the downloaded file.

Structure and Bonding in TeO2 Melt and Glass

Oliver L.G. Alderman

ISIS Neutron and Muon Source

• liver.alderman@stfc.ac.uk

1

ACerS Virtual Glass Summit August 3rd - 5th 2020

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

Main collaborators

Chris J. Benmore

Advanced Photon Source 2

Rick Weber Anthony Tamalonis Vrishank Walia

Materials Development Inc.

Steven Feller Martha Jesuit Makayla Boyd Michael Packard

Coe College

• E. I. Kamitsos
• E. Simandiras
• D. G. Liakos

National Hellenic Research Foundation

Funding

U.S. DOE SBIR DE-SC0015241 SBIR DE-SC0018601 NSF-DMR 1746230 NSRF 2014-2020 MIS 5002409

Special Thanks: Emma Barney

University of Nottingham

3

Single-oxide glass formers

Angell Plot VFT fits

• Oxides of only 8 elements from glasses B2O3, SiO2,

GeO2, P2O5, V2O5, As2O5, As2O3, Sb2O3, TeO2

• TeO2 vastly more fragile cf. canonical oxide glass

formers SiO2, GeO2, B2O3

• Correlates with relatively poor glass forming ability
• So what makes TeO2 so different?

What is the structure of the liquid and its glass?

0.0 0.2 0.4 0.6 0.8 1.0

• 4
• 2

2 4 6 8 10 12

log10 viscosity (Pa s) Tg/T (K/K) SiO2 m = 25 B2O3 m = 37 TeO2 m = 145 Veber & Mangin, Mater. Res. Bull. 2008, 43, 3066 Macedo & Napolitano, J. Chem. Phys. 1968, 49, 1887 Doremus, J. Appl. Phys. 2002, 92, 7619

4

Single-oxide glass formers

• Zachariasen predicted most single-oxide glass formers, based on crystal

structures known in 1932!

• Corner-sharing MO3 triangles or MO4 tetrahedra, oxygen 2-coordinated (OM2)
• M–O–M bond angles deform easily (with little energy cost) → small

thermodynamic driving force for crystallization

N = Group Number Zachariasen Predicted Confirmed Glass Cation Valence N N – 2

• W. H. Zachariasen J. Am. Chem. Soc., 1932. 54: 3841
• The ones Zachariasen got wrong were only due to lack
• f knowledge of crystal structures

5

Lone-pair-oxides & glass formation

N = Group Number Zachariasen Predicted Confirmed Glass Cation Valence N N – 2 = Cationic electron lone-pair in oxide

• TeO2 is an N – 2 glass former Te4+ is a lone-pair cation
• Electron lone-pairs tend to be stereochemically active which leads to low

coordination numbers to oxygen, & increased glass forming tendency

• Excellent non-linear optical properties high index, infrared transmission etc.

6

TeO2 crystal polymorphs

• Stable phase
• 2 bond lengths
• 1 asymmetric
• xygen bridge
• ‘Paratellurite’

mineral

α P412121 β Pbca γ P212121 δ Fm-3m

• Layered structure
• 4 bond lengths
• 2 asymmetric
• xygen bridges
• ‘Tellurite’ mineral
• 4 bond lengths
• 2 asymmetric
• xygen bridges
• Less stable than α

& β – obtained by crystallization of glass/doped-glass

• Oxygen sublattice

disorder

• Many bond lengths
• Many asymmetric
• xygen bridges
• crystallization of

doped-glass

7

TeO2 crystal polymorphs

α P412121 β Pbca γ P212121 δ Fm-3m

γ-TeO2 Showing connectivity of the [TeO4] polyhedra through asymmetric Te-O-Te bridges

8

Trigonal pyramidal site in Na2TeO3 crystal

Na2TeO3 Showing isolated [TeO3]2- polyanion Terminal bonds If present in pure TeO2, such a site could involve [(Te=O)O2/2] neutral groups with terminal

• xygen

Te O O O

:

Te O O O

:

9

TeO2 glass & liquid

• What about the disordered liquid & glassy structures?
• Te–O coordination number varies between reports 3.6 < nTeO ≲ 4.0 in the glass
• Where nTeO < 4, there must exist low coordination sites, e.g. 3-fold TeO3

not seen in the crystal polymorphs

• Plan to measure Te–O coordination in glass & liquid by HEXRD High-Energy X-ray Diffraction
• Expected dominant species as follows:
• Raman evidence suggests increase in 3-fold TeO3 with liquid temperature

Te O O O

:

Te O O O

:

O Te O O

:

Solids Liquid Gas

10

Aerodynamic levitation – containerless melting

6.0 mm 2.5 mm

Recalescence Glass formation

• Float sample on gas stream & heat with CO2 laser 10.6μm
• Access to high T refractory melts Avoid contamination, alloying

& background scattering of x-ray/neutron probe beams

• Extend glass forming regions & supercooling heterogeneous

nucleation at melt-container interface eliminated

Alderman, O.L.G., et al., J. Am. Ceram. Soc., 2018. 101: 3357

1 2 3 4 5 6 7 8 0.0 0.7 1.4 2.1 2.8 3.5

Polycrystalline Glass

Ba

11B4O7 r / Å Neutron T(r) / barns per Å

2

11

Diffraction from amorphous substances

Sine Fourier transform ℱ{Q·S(Q) – 1} = D(r) r = interatomic separation

• Short & medium range order very

similar in this case

• Resolution Δr = 3.791/Qmax ∼ 1/Qmax

5 10 15 20

• 0.5

0.0 0.5 1.0 1.5

Ba

11B4O7

Polycrystalline Glass

Q / Å

• 1

i(Q) / Barns per atom per steradian

Pulsed neutron diffraction measurement GEM, ISIS

B–O O–O |Q| = (4π/λ) sinϑ scalar momentum transfer/ħ (glass is isotropic) 2ϑ = scattering angle λ = wavelength

• Both crystals and glass show diffuse

scattering

• Only crystals show Bragg peaks

Hannon, A.C., Nucl. Instrum. Meth. A, 2005. 551: 88

12

High-energy x-ray diffraction (HEXRD)

• 90 to 100 keV synchrotron x-rays + rapid large-area

detector APS beamline 6-ID-D, λ ∼ 0.14 to 0.12 Å

• High resolution pair distribution functions X-ray weighted

distributions of the interatomic separations

• Each measurement made in seconds or minutes

Advanced Photon Source (APS) Argonne National Laboratory

2ϑ

|Q| = (4π/λ) sinϑ Momentum transfer = ħQ Energy = hc/λ Alderman, O.L.G., et al., J. Am. Ceram. Soc., 2018. 101: 3357

100 101 102 103 10-2 100 102 104

1.47

Mass attenuation coefficient / cm2g-1 Photon energy / keV

Coherent Incoherent Photoelectric Total

TeO2

Cu K 1 100 keV 215 5 10 15 20 25

0.0 0.2 0.4 0.6 0.8

Wij(Q) Q / Å-1

Te-Te Te-O O-O

Pair weightings

13

Findings

• Observe only subtle differences between

glass & melt in terms of Te–O units

• Shift of Te–Te peak to longer distance

qualitatively consistent with thermal expansion

• Shaded region represents uncertainty in

liquid density

2 4 6 8 10 2 4 6 8

O-O Te-O T(r) (Å-2) r (Å)

TeO2

Glass Melt Te-Te

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

14

Te–O bond length distribution

• More short, strong Te–O bonds in the melt fewer long, weak bonds
• But very similar mean nTeO ≈ 4 poorly defined due to lack of plateau

in nTeO(r) as seen for all ordered crystals (& e.g. SiO2)

• Qualitatively consistent with Raman spectroscopic observations
• Good agreement with ab-initio cluster calculations 100 atom, high

level of accuracy (blue open points )

• The long weak bonds are more important in satisfying the bonding

requirements in the solid

• Difficult to distinguish Te=O from Te-O·····Te by any means
• Bond valence sum asymptote VTe(r→∞) > 4 due to lone pair
• O–O contribution not important (neutron-x-ray difference analysis)

4 8 12 2 4 6 2.0 2.5 3.0 2 4 Glass Melt g-TeO2

cOrTTeO(r) (Å-1)

Ab initio clusters

nTeO(r) VTe(r) or vTe(r) (e) r (Å)

Radial Distribution Function (RDF) Running Te–O coordination number nTeO(r) Running Te–O bond-valence sum VTeO(r)

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

15

Conclusions

• TeO2 glass & liquid incorporate short-range disorder as well as long-range disorder
• TeO2 is therefore distinct from the canonical glass forming oxides SiO2, B2O3…

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427 Questions? Email me: oliver.alderman@stfc.ac.uk

16

α-TeO2 recovered and starting materials

Figure S1: X-ray diffraction patterns for polycrystalline TeO2 beads recovered after melting and liquid diffraction measurement in O2 gas, and prior to the same in Ar gas. Both patterns indicate the presence of α-TeO2, although with considerable preferred orientation, especially in the case of the former. In each case 2-dimensional diffraction patterns were averaged

• ver the full azimuthal range to reduce the effects of preferred orientation,

no background subtraction has been made and arbitrary scaling factors have been applied. The calculated powder diffraction patterns for elemental Te, α-TeO2 and γ-TeO2 are shown for comparison.

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 2000 4000 6000 8000 10000 12000

I(d) (a.u.) d-spacing (Å)

Recovered after melting in O2 Prior to melting in Ar Elemental Te g-TeO2 -TeO2

17

TeO2 melt and glass densities

Figure S2: Temperature dependent density of α-TeO2 derived from x-ray crystallography, along with least-squares linear fit extrapolated up to the melting point of Tm = 1005 K. The ambient glass density is marked (dashed line) and taken to be 5.65 gcm-3, from linear extrapolation of sodium tellurite glass densities. This value for the glass density is in reasonable agreement with extrapolations based upon measured densities in other binary tellurite glass series, and with direct measurements on TeO2 glass of 5.62 gcm-3, 5.61 gcm-3, and 5.57 gcm-3 (the authors also reported on a shell model molecular dynamics (MD) model with glass density of 5.53 gcm-3). A value of 5.57(3) gcm-3 has also been obtained at Coe College by means of helium pycnometry. A higher glass density of 5.84 gcm-3 has been reported and used to generate first principles MD models of TeO2 melt and glass, but this value appears to be an outlier, and is similar to the densities of the metastable crystalline polymorphs reported in the same work (plotted as triangles above). The blue curve indicates the glass and melt density assuming the same expansion rate as for α-TeO2, while the dashed curve illustrates a more likely scenario where the liquid has a larger expansion rate than the glass due to the additional configurational degrees of

• freedom. The three stars indicate the densities used for the analysis in the

main paper.

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

400 600 800 1000 1200 0.056 0.060 0.064 0.068 4.95 5.30 5.65 6.01

-TeO2 -TeO2 - g-TeO2 80% of -TeO2 at Tm

Tg = 576K

Mass density (g cm

• 3)

Krishna Rao et al. -TeO2 Glass + melt estimate Melt (Melt > Glass) Melt densities for analysis

Atom number density (Å

• 3)

T (K)

Glass at 298K Tm = 1005K

90% of ambient glass

18

Comparison to unpublished Barney et al. X-ray data

Figure S3: X-ray interference functions, Q(S(Q) – 1), for glassy TeO2. The structure factors, S(Q) – 1, are shown inset for Q ≤ 5 Å-1. The magenta curve is taken from Barney et al., as measured at the Diamond synchrotron.

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

1 2 3 4 5

• 2
• 1

1

S(Q) -1 Q (Å

• 1)

5 10 15 20 25

• 2
• 1

1 2 3

This work Barney et al. Q(S(Q) -1) (Å

• 1)

Q (Å

• 1)

TeO2 Glass

2 4 6 8 10 2 4 6 8

Qmax = 21.39 Å

• 1

Qmax = 25.57 Å

• 1

Barney et al. (Qmax = 21.39 Å

• 1)

O-O Te-O T(r) (Å

• 2)

r (Å)

TeO2 Glass

Te-Te Figure S6: X-ray total correlation functions obtained by sine Fourier transform of the interference functions of Fig. S3, with Qmax indicated in the legend, and a step modification function which yields the highest possible resolution.

19

D(r) functions, intermediate range order

Figure S4: X-ray differential distribution functions obtained by sine Fourier transform of the interference functions of Fig. 1, using a Qmax = 21.39 Å-1 and step modification function which yields the highest possible resolution.

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

Figure S5: X-ray differential distribution functions obtained by sine Fourier transform of the interference functions of Fig. S3, with Qmax indicated in the legend, and a step modification function which yields the highest possible resolution.

5 10 15 20

• 1.5
• 1.0
• 0.5

0.0 0.5 1.0 1.5 2.0

Qmax = 21.39 Å

• 1

TeO2

D(r) (Å

• 2)

r (Å)

Glass Melt

5 10 15 20

• 1.5
• 1.0
• 0.5

0.0 0.5 1.0 1.5 2.0

Qmax = 21.39 Å

• 1

Qmax = 25.57 Å

• 1

Barney et al. (Qmax = 21.39 Å

• 1)

TeO2 Glass

D(r) (Å

• 2)

r (Å)

20

Te–O bond length distribution

• Removing the O–O pair term by x-ray-neutron difference does not

affect conclusions red curve

• Figure S7: Upper panel: X-ray radial distribution functions (RDFs) with the Q-dependent

Te–O pair weighting divided out, prior to sine Fourier transform of the interference functions of Fig. S3, with Qmax indicated in the legend, and a step modification function which yields the highest possible resolution. The O–O pair term has also been eliminated from the red curve using the neutron diffraction data of Barney et al. and the O–O contribution is shown separated (chained curve). Middle panel: Running Te–O coordination numbers (nTeO(r)) obtained by integration of the RDFs with the lower limit

• f the integral set to the zero-crossing at circa 1.72(1) Å. Results of our ab initio

amorphous cluster calculations are also shown (blue circles). Lower panel: Running bond valence sums for Te obtained similarly to the nTeO(r), but after weighting by vTe(r) = exp(RTeO – r)/b (dashed line) prior to integration.

Radial Distribution Function (RDF) Running Te–O coordination number nTeO(r) Running Te–O bond-valence sum VTeO(r)

Alderman, O.L.G., et al., J. Phys. Chem. Lett., 2020. 11: 427

4 8 12 2 4 6 2.0 2.5 3.0 2 4

TeO2 Glass (Qmax (Å

• 1))

21.39 25.57 25.57, O-O eliminated 25.57 O-O contribution Barney et al. (21.39)

cOrTTeO(r) (Å

• 1)

Ab initio clusters

nTeO(r) VTe(r) or vTe(r) (e) r (Å)

21

Resolving power

P2O5 glass 140 keV 22 keV

Hoppe, et al. Solid State Commun. 2000 115:559

Sine Fourier transform ℱ{Q·S(Q) – 1} = T(r) Interatomic separation = r

• Resolution Δr = 3.791/Qmax
• Terminal & bridging P–O bonds 1.43 Å & 1.58 Å, ΔrPO = 0.15 Å
• Ambient thermal + static broadening typically ∼0.05 Å

Δr (Å) 0.20 0.13 0.08