Spectral Resolution and Raman Q3 and Q2 cross sections in ~40 mol% Na2O glasses

doi:10.1016/j.chemgeo.2020.120040

Chemical Geology

Volume 562, 20 February 2021, 120040

https://doi.org/10.1016/j.chemgeo.2020.120040 Get rights and content

Highlights

•
Raman cross sections for Q² and Q³ Raman bands of alkali-silicate glasses are determined.
•
The cross sections for the two bands are similar.
•
Quantitative estimates of Q species concentrations can now be performed.

Abstract

There are four ²⁹Si NMR experimental studies published for glasses containing ~40 mol% Na₂O and the reported Q species abundances are remarkably consistent. These results have been used to determine accurate Raman Q³ and Q² cross sections for a 40.1 mol% Na₂O glass. The Q³ and Q² cross sections are respectively 1.10 and 0.95 (±0.03). The Q² band of the Raman spectrum is dominantly Lorentzian in shape whereas the Q³ band is asymmetric and based on other studies, the asymmetry increases with alkali content, due primarily to alkali-BO interactions. As explanation, increase in alkali content enhances preferentially electron density over Si atoms of tetrahedra, thereby weakening Si-O coulombic interactions (i.e., force constants), which shifts the symmetric stretch of the Q³ species to lower frequencies, producing asymmetric Q³ line shapes. With Q² and Q³ cross sections established, the high resolution of Raman spectroscopy now can be used to provide highly accurate estimates of Q species in silicate glasses and melts.

Introduction

Insights into the short-range structure of silicate glasses and melts have been obtained primarily from, ²⁹Si and ¹⁷O Nuclear Magnetic Resonance (NMR), Raman, and more recently, O 1s X-ray Photoelectron Spectroscopy (XPS) (e.g., Hehlen et al., 2017; Nesbitt et al., 2017a, Nesbitt et al., 2017b, Nesbitt et al., 2017c, Nesbitt et al., 2017d, Nesbitt et al., 2015a, Nesbitt et al., 2015b; Sawyer et al., 2015, Sawyer et al., 2012; Stebbins and Xue, 2014; Nesbitt and Bancroft, 2014; Nesbitt et al., 2011; Bancroft et al., 2009; Lee and Stebbins, 2009; Dalby et al., 2007; Neuville et al., 2014; Neuville, 2006; Xue et al., 1994; Mysen and Frantz, 1993; Maekawa et al., 1991; Stebbins, 1988, Stebbins, 1987; McMillan, 1984; Matson et al., 1983; Furukawa et al., 1981; Bruckner et al., 1980; Virgo et al., 1980; Brawer and White, 1975). These studies have sometimes allowed quantification of the five tetrahedrally coordinated Si species (Q species) in alkali silicate melts and glasses where each Q species hosts a different number of ‘Bridging Oxygen’ atoms (BOs) and ‘Non-Bridging Oxygen’ (NBO). The Si tetrahedra are referred to as Qⁿ species, where the ‘n’ represents the number of BO's associated with the individual tetrahedra.

The number and types of Qⁿ species strongly affect the chemical and physical properties of melts and glasses (e.g., Le Losq and Neuville, 2017) and the mechanisms of melting (e.g., Nesbitt et al., 2017d, Nesbitt et al., 2015a; Mysen and Richet, 2019; Richet et al., 1998, Richet et al., 1996, Richet et al., 1994), as well as reaction mechanisms (Nesbitt et al., 2020, Nesbitt et al., 2015a; Magnien et al., 2008). There is, therefore, good reason to obtain accurate Qⁿ abundances in glasses and melts. In Raman spectra the abundance of such species is often estimated by comparing relative peak intensities or areas of the vibrational bands associated with the Qⁿ species (i.e., Q⁴, Q³, Q², Q¹ and Q⁰). Furthermore, these intensities are often obtained from curve fits of the spectral envelope (cf., O'Shaughnessy et al., 2020). However, the area of the fitted component is not necessarily directly proportional to the actual fraction of a given structural species. This is because scattering cross sections for different Q species are expected to be affected to some degree by the short-range order (SRO) and the structure itself.

Previously, estimation of Raman cross-sections and quantitative Qⁿ concentrations have been determined by calibration using the Q species numbers obtained by other methods (e.g., NMR spectroscopy) cf., Mysen and Richet, 2019. A problem arises, however, if the spectral envelope in the Raman spectra (~900–1300 cm⁻¹) from which quantitative Qⁿ numbers are being determined, contains peaks that are not due to Qⁿ vibrations. This is the case for previous studies. Fits to the high frequency envelope always require extra peaks to be added to the fit (there are always more peaks than can be assigned to Qⁿ vibrations), see below (cf., Henderson and Stebbins, 2021).

The objectives of the study are to establish the abundances of Q species in ~40 mol% Na₂O glasses and to determine accurate Raman cross sections for the Q³ and Q² species in these same glasses. Quantitative estimates of the Q species abundances derived from Raman spectra are contentious due to the difficulty of interpreting the spectra, as noted above (e.g., Nesbitt et al., 2017c; Sawyer et al., 2015; Malfait, 2015; Nesbitt et al., 2015b; Nesbitt et al., 2011). There is, however, good reason to quantify Raman spectra because it has superior resolution relative to other spectroscopic techniques (see below). If accurate Q³ and Q² cross sections can be established for Raman spectra, it will be possible to obtain highly accurate estimates of Q species in glasses and more importantly, in high temperature silicate melts where most other spectroscopic techniques have very low resolution or cannot be used.

We undertake a detailed analysis of Q species abundances in glasses containing ~40 mol% Na₂O, which necessarily includes evaluation of the properties of the Q³ line shape. As background, McMillan et al. (1992) argued that the Q³ line shapes was asymmetric and dependent on the local environment. The implication is that the asymmetry changes as the local environment changes. Nesbitt et al. (2017a), Bancroft et al. (2018), Nesbitt et al. (2019) and O'Shaughnessy et al. (2020) investigated line shapes using a variety of techniques and concluded that indeed the Q³ Raman signal was asymmetric and that its shape was a function of melt composition. With this aspect in mind, the Q² and Q³ spectral intensities (peak areas) are evaluated for a highly resolved Raman spectrum containing 40.1 mol% Na₂O glass. The implications of the results are then addressed.

Section snippets

Raman experimental considerations

The 40.1 mol% Na₂O Raman spectrum used in this study was originally published by Neuville (2006) and he provides details of experimental methods. These are repeated here for the reader's convenience. A T64000 Jobin-Yvon confocal micro Raman spectrometer equipped with a CCD detector was used to collect data. Sample excitation was by a 514.532 nm line of a Coherent 70-C5 Ar+ laser operating at 2.8 W at the sample surface, resulting in a signal-to-noise ratio of 80/1. Integration time was 60 s.

Measurement of resolution

The comparative resolution of Raman, ²⁹Si MAS NMR, ¹⁷O NMR and O 1s XPS spectroscopic techniques has not been evaluated previously but is required if uncertainties associated with reported Si and O species abundances are to be established across techniques. Resolution usually relates to FWHM or ‘linewidth’ of individual peaks relative to peak separation. This parameter is inevitably dependent on the peak shape (e.g., %Lorentzian; Hesse et al., 2007) used to fit spectral lines. Where two peaks

Q species abundances from ²⁹Si NMR spectra

Four ²⁹Si NMR studies of ~40 mol% Na₂O glass (Fig. 1b) have been published (Nesbitt et al., 2011; Zhang et al., 1996; Maekawa et al., 1991; Stebbins, 1987) and their reported Q⁴, Q³, Q² and Q¹ abundances are provided in Table 1. Q⁰ species were not detected. The best resolved and best quality spectrum is that of Stebbins (1987) for the 41.1 mol% Na₂O glass. He measured the Q⁴ species abundance using the more accurate ²⁹Si Static NMR technique and reported a K_D value of 0.011 (±0.005) for the

Establishing line shapes and widths

Ambiguities concerning interpretation of Raman spectra (noted above) have limited the use of the technique to obtain accurate Q species abundances (Bancroft et al., 2018; Neuville et al., 2014; Neuville, 2006; Mysen and Frantz, 1992, Mysen and Frantz, 1993; McMillan et al., 1992; Furukawa et al., 1981; Brawer and White, 1975). The Raman spectrum of the 40.1 mol% Na₂O glass spectrum shown in Fig. 2a is composed of two well resolved bands centered at ~950 cm⁻¹ and ~1100 cm⁻¹. To establish line

Raman cross sections of Q² and Q³ species

The spectral intensity of a Qⁿ species (I_Qⁿ) is the product of its cross-section (σ_Qⁿ) and mole fraction (X_Qⁿ): ${I_{Q}}^{n} = {σ_{Q}}^{n} {X_{Q}}^{n}$

The Q³ and Q² Raman spectral intensities are respectively 65.2 and 32.2 mol% for the 40.1 mol% Na₂O glass (Table 1, Raman Intensities) whereas the best estimates of Q³ and Q² abundances evaluated from the ²⁹Si NMR results are respectively 68.8 and 28.8 mol% (Table 1). The Q³ cross section consequently is 0.95 (i.e., 65.2/68.8) and the Q² cross section is 1.12 (i.e.,

Conclusions

The Q species abundances of four independent ²⁹Si NMR studies of ~40 mol% Na₂O glasses are in accord, as illustrated in Table 1. These results have been used to evaluate, for the first time, accurate Raman Q² and Q³ cross sections in a glass containing 40.1 mol% Na₂O. The asymmetric nature of the Q³ band is clearly apparent and must be addressed where Raman spectra are to be quantified. The results of Sen and Youngman (2003) and Olivier et al. (2001) demonstrate that Q species peaks (Q³ and Q⁴)

Declaration of Competing Interest

None.

Acknowledgements

The authors thank their respective Universities and departments for logistical support during the conduct of this research. GSH acknowledges funding through a NSERC discovery grant.

References (54)

G.M. Bancroft et al.
Lorentzian dominated lineshapes and linewidths for Raman symmetric stretch peaks (800–1200 cm−1) in Qⁿ (n=1–3) species of alkali silicate glasses/melts
J. Non-Cryst. Solids
(2018)
K.N. Dalby et al.
Resolution of bridging oxygen signals from O 1s spectra of silicate glasses using XPS: Implications for O and Si speciation
Geochim. Cosmochim. Acta
(2007)
J. Du et al.
The medium range structure of sodium silicate glasses: A molecular dynamics simulation
J. Non-Cryst. Solids
(2004)
M.J. Duer et al.
An investigation of the structural units in sodium disilicate glass: A 2-D ²⁹Si NMR study
J. Non-Cryst. Solids
(1995)
B. Hehlen et al.
Bimodal distribution of Si–O–Si angles in sodo-silicate glasses
J. Non-Cryst. Solids
(2017)
E.I. Kamitsos et al.
Vibrational spectra of single and mixed alkali pentasilicate glasses
J. Non-Cryst. Solids
(1984)
O.N. Koroleva et al.
Structure of Na₂O–SiO₂ melt as a function of composition: In situ Raman spectroscopic study
J. Non-Cryst. Solids
(2013)
C. Le Losq et al.
Molecular structure, configurational entropy and viscosity of silicate melts: Link through the Adam and Gibbs theory of viscous flow
J. Non-Cryst. Solids
(2017)
S.K. Lee et al.
Effects of the degree of speciation on the structure of sodium silicate and aluminosilicate glasses and melts: An ¹⁷O NMR study
Geochim Cosmochim. Acta
(2009)
H. Maekawa et al.
The structural groups of alkali silicate glasses determined from ²⁹Si MAS-NMR
J. Non-Cryst. Solids
(1991)

P. Zhang et al.

Silicon site distributions in an alkali silicate glass derived by two-dimensional ²⁹Si nuclear magnetic resonance

J. Non-Cryst. Solids

(1996)

G.M. Bancroft et al.

Toward a comprehensive understanding of solid state core-level XPS linewidths: Experimental and theoretical studies on the Si 2p and O 1s linewidths in silicates

Phys. Rev. B

(2009)

Cited by (13)

The influence of Al<inf>2</inf>O<inf>3</inf>, CaO, MgO and TiO<inf>2</inf> content on the early-age reactivity of GGBS in blended cements, alkali-activated materials and supersulfated cements
2024, Cement and Concrete Research
GGBS composition and the choice of the activation systems have a large impact on the reactivity of GGBS-based binders. Here, the reactivity of 16 artificially-modified GGBSs was investigated in blended cements, alkali-activated binders and supersulfated cements, using isothermal calorimetry for hydration times between 24 h and 120 h. Lower glass network polymerization by addition of CaO or MgO increased the reactivity at all ages and in all activation systems. Increased Al₂O₃ content mainly resulted in higher early reactivity. This effect was more pronounced in blended and supersulfated cements. TiO₂ addition decreased GGBS reactivity in all activation systems but the negative effect was reduced at high Al₂O₃ contents, especially at later ages in supersulfated cements. In alkali-activation, the hydration was delayed by several hours for some compositions. In summary, the results suggest that it is possible to increase the reactivity of GGBS by choosing an activation system optimized for a given GGBS composition.
A critical evaluation of barium silicate glass network polymerization
2022, Journal of Non-Crystalline Solids
The state of polymerization in barium silicate glasses (xBaO·(1-x)SiO₂, 0.28 ≤ x ≤ 0.45) has been studied by both experiment (Raman, static and MAS ²⁹Si NMR spectroscopy) and molecular dynamics simulations (Vashishta-Rahman and Du-Cormack potentials). No a priori assumptions were made about the number of Qⁿ species, specific mean polymerization nor presence or absence of ‘free’ oxygen species. This allowed us to test the relationships between these structural aspects.
The Qⁿ site interconversion equilibrium constants K_Q2 and K_Q3 were found to be 0.07 and 0.05±0.02 from static ²⁹Si NMR, and 0.14±0.07 and 0.07±0.01 from ²⁹Si MAS NMR, respectively. The Qⁿ species distribution was used to determine the Raman scattering cross-sections for Q¹, Q² and Q³ as 1.01±0.15, ∼1.1, and 1.14±0.04, respectively, although some compositional dependence was found. The Q⁴ cross-section displays a unique behavior approaching 0 at 38% BaO. This more accurate structural framework paves the way for future studies to establish meaningful structure−property relationships for these glasses.
Discoveries about the structure of alkaline earth-bearing borosilicate glasses doped with TiO<inf>2</inf> revealed by Raman spectroscopy
2022, Journal of Non-Crystalline Solids
The influence of TiO₂ on the glass structure in the CaO-SrO-B₂O₃-SiO₂ (CSrBS) system was examined by Raman spectroscopy. The Raman spectra of these glasses presented bending vibrational modes and stretching vibration of Qⁿ species (where n is the number of bridging oxygen atoms shared between tetrahedra), leading to the identification of Q¹, Q², Q²–A, and Q³ species. The addition of Ti produced two new Raman bands at 756 and 825 cm⁻¹ which were attributed, respectively, to the stretching vibrations of Ti-O_NBO and Ti-O_BO. Our results suggest that titanium is coordinated mainly in Ti^[⁴^] species, with negligible amounts of Ti^[⁵^] units. The intensity of Si-O-Si bending bands shows a decrease in intensity, which is attributed to the restriction of bending motions, possibly due to titanium bonding to the network structure through both vertices and edges.
Speciation and polymerization in a barium silicate glass: Evidence from <sup>29</sup>Si NMR and Raman spectroscopies
2021, Chemical Geology
Citation Excerpt :
In total, there are 14 adjustable parameters in this fit. In fitting the Na40 Raman spectrum, Nesbitt et al. (2021) added a third Q3 peak to their fit, however, in the Ba40 case, this addition mainly influences the width of the Q2 band and changes the sum of the Q3 peak areas by a total of only ~2%. Given the uncertainties discussed below, this does not represent a statistically significant improvement in our fit.
Raman and NMR spectroscopies have long been employed in pursuing quantitative relationships between polymerization of silicate glasses and melts. However, varying assumptions and ambiguities have often meant structural models vary significantly depending on the probe used. This study investigates a 40BaO·60SiO₂ glass to determine its polymerization using static and MAS ²⁹Si NMR, and Raman spectroscopies to construct a coherent structural interpretation. The challenges and advantages of each technique are discussed as the structural model of 40BaO-60SiO₂ glass required partial evidence derived from each of these spectroscopies. The Raman scattering cross sections for Q² and Q³ were found to be 0.85 and 1.22, respectively. The Qⁿ distribution was shown to be 3, 35, 57, 6% (±4%) for Q¹ to Q⁴, respectively. This distribution yields a polymerization consistent with the stoichiometry, Ba_0.66SiO_2.66, having a O:Si ratio of 2.66. The agreement between all Qⁿ distributions, the composition, the polymerization, and equilibrium constants strongly indicate that this structural interpretation is robust.
Structure and properties of alkali aluminosilicate glasses and melts: Insights from deep learning
2021, Geochimica et Cosmochimica Acta
Aluminosilicate glasses and melts are of paramount importance for geo- and materials sciences. They include most magmas, and are used to produce a wide variety of everyday materials, from windows to smartphone displays. Despite this importance, no general model exists with which to predict the atomic structure, thermodynamic and viscous properties of aluminosilicate melts. To address this, we introduce a deep learning framework, ‘i-Melt’, which combines a deep artificial neural network with thermodynamic equations. It is trained to predict 18 different latent and observed properties of melts and glasses in the K₂O-Na₂O-Al₂O₃-SiO₂ system, including configurational entropy, viscosity, optical refractive index, density, and Raman signals. Viscosity can be predicted in the 10⁰–10¹⁵ log₁₀ Pa·s range using five different theoretical frameworks (Adam-Gibbs, Free Volume, MYEGA, VFT, Avramov-Milchev), with a precision equal to, or better than, 0.4 log₁₀ Pa·s on unseen data. Density and optical refractive index (through the Sellmeier equation) can be predicted with errors equal or lower than 0.02 and 0.006, respectively. Raman spectra for K₂O-Na₂O-Al₂O₃-SiO₂ glasses are also predicted, with a relatively high mean error of ∼25% due to the limited data set available for training. Latent variables can also be predicted with good precisions. For example, the glass transition temperature, T_g, can be predicted to within 19 K, while the melt configurational entropy at the glass transition, S^conf(T_g), can be predicted to within 0.8 J mol⁻¹ K⁻¹.
Applied to rhyolite compositions, i-Melt shows that the rheological threshold separating explosive and effusive eruptions correlates with an increase in the fraction of non-bridging oxygens in rhyolite melts as their alkali/Al ratio becomes larger than 1. Exploring further the effect of the K/(K + Na) ratio on the properties of alkali aluminosilicate melts with compositions varying along a simplified alkali magmatic series trend, we observe that K-rich melts have systematically different structures and higher viscosities compared to Na-rich melts. Combined with the effects of the K/(K + Na) ratio on other parameters, such as the solubility, solution mechanisms and speciation of volatile elements, this could ultimately influence the eruptive dynamics of volcanic systems emitting Na-rich or K-rich alkali magmas.
Effect of the Na/Mg Mixing on the Structure and Properties of Aluminosilicate Melts
2024, SSRN

View all citing articles on Scopus

View full text

Spectral Resolution and Raman Q3 and Q2 cross sections in ~40 mol% Na2O glasses

Highlights

Abstract

Introduction

Section snippets

Raman experimental considerations

Measurement of resolution

Q species abundances from 29Si NMR spectra

Establishing line shapes and widths

Raman cross sections of Q2 and Q3 species

Conclusions

Declaration of Competing Interest

Acknowledgements

J. Non-Cryst. Solids

Geochim. Cosmochim. Acta

J. Non-Cryst. Solids

J. Non-Cryst. Solids

J. Non-Cryst. Solids

J. Non-Cryst. Solids

J. Non-Cryst. Solids

J. Non-Cryst. Solids

Geochim Cosmochim. Acta

J. Non-Cryst. Solids

Geochim. Cosmochim. Acta

J. Non-Cryst. Solids

Chem. Geol.

Chem. Geol.

J. Non-Cryst. Solids

J. Non-Cryst. Solids

Chem. Geol.

Chem. Geol.

Chem. Geol.

J. Non-Cryst. Solids

Earth Planet. Sci. Lett.

J. Non-Cryst. Solids

J. Non-Cryst. Solids

J. Non-Cryst. Solids

Chem. Geol.

J. Non-Cryst. Solids

Toward a comprehensive understanding of solid state core-level XPS linewidths: Experimental and theoretical studies on the Si 2p and O 1s linewidths in silicates

Phys. Rev. B

Spectral Resolution and Raman Q³ and Q² cross sections in ~40 mol% Na₂O glasses

Q species abundances from ²⁹Si NMR spectra

Raman cross sections of Q² and Q³ species