Nuclear Magnetic Resonance and FTIR Structural Studies on Borosilicate Glasses Containing Iron Oxide

The structure of borosilicate glasses of composition 30Na ₂O-2Al ₂O ₃-25SiO ₂-xFe ₂O ₃ (43-x) B ₂O ₃ has been investigated in the composition range of 0.5 20 mol% Fe ₂O ₃. ²⁷Al, ¹¹B, ²⁹ Si MAS NMR and FTIR spectroscopies have been used to measure the fraction of different structural species in the glasses. It is evidenced from NMR data that both sodium and Fe₂O₃ (in low region up to 7 mol%) are the main glass modifier. Structural determination for borosilicate glasses with a high content of (Fe ₂O ₃) was carried out by FTIR spectroscopy, where both ¹¹B and ²⁹Si MAS NMR are impossible because of the high quantities of paramagnetic iron (III) species present. NMR analysis was performed on borosilicate glasses containing up to 7 mol% Fe ₂O ₃ and the N ₄ values obtained by FTIR spectroscopy agree within error with the ¹¹B NMR results of the same glass samples. Fe ₂O ₃ is a main glass modifier in the low-Fe ₂O ₃-content region (≤6 mol%). On other hand, it plays the role of glass former at higher content of Fe ₂O ₃. Increasing both N ₄ of boron tetrahedral units and chemical shift of silicon nuclei to reach maxima at 5 mol% Fe ₂O ₃ confirms the role of Fe ₂O ₃ as a glass modifier in the low composition region. On the other hand, fast decrease in N ₄ with further increasing Fe ₂O ₃ contents ≥6 mol%) is an evidence for iron oxide to inter the glass network as a network former.

• NMR-FTIR-Spectroscopy-Borosilicate-Glasses

1. Introduction

Ternary and quaternary borosilicate glasses (BSG) have an extended variety of industrial importance due to their different types of applications. They have been used as sealing components. For example, some of glass compositions have been simply used as sealants in television tubes or bulb lamps. On the other hand, increasing achievements of microelectronics and new demands necessitates new types of glasses. Recent studies have concentrated on borosilicate glasses, which have been developed as sealants, particularly, for molten carbonate fuel cells (MCFC) [1] [2] .

Because of the technological advantages, these glasses have long been the subject of structural studies, using different techniques, including ²⁹Si and ¹¹B MAS NMR and FTIR spectroscopy [3] - [7] to offer a quantitative analysis of different structural units forming the glass network. The structure of borosilicate glasses is based on Qⁿ [SiO₄] units, where n is the number of bridging oxygen atoms per tetrahedron, and [BO₃] and [BO₄] units. The [BO₃] units are triangle boron in both ring and non-ring configuration, while [BO₄] ones are boron atoms in tetrahedral coordination. The oxygen atoms can be bonded to one boron and three silicons (1B, 3Si) or four silicon atoms (0B, 4Si) and have a Na⁺ ion as a charge compensator.

The distribution of the borate and silicate structural units depends on two structural factors namely R and K. Generally, R is the ratios of Na₂O/B₂O₃ and K is the ratio of SiO₂/B₂O₃ [8] [9] . Dell et al. [8] proposed a structural model for which the variation of [BO₄] borate units (N₄) is represented as a function of the R parameter. For all K values, N₄ increases up to a maximum in between 0.5 and 0.75 for R = 1, and the value of this maximum increases with an increasing K values. Then, the proportion of N₄ slowly decreases for higher R values.

In present study, two individual composition regions of borosilicate glasses were investigated in order to test the possible quantitative use of NMR and FTIR spectroscopy. The first type contains a limited concentration from Fe₂O₃ as a paramagnetic material. In such a case NMR study can easily be applied to obtain values of Qⁿ and N₄ fraction in the borosilicate glass. On the other hand, the second type of borosilicate glasses contains further high level from Fe₂O₃ (>6 mol%). These concentrations are unsuitable for NMR measurements. This is because the variation in local fields due to the unpaired electron spins produces sufficient broadening of NMR lines [10] . As a result, different contributions cannot be resolved and some may even be unobservable.

FTIR spectroscopy is not affected by the presence of paramagnetic species and therefore it would be a very useful tool for quantitative measurement of structural fractions, such as Qⁿ in silicate and N₄ in borate networks where NMR is impracticable. In this study, FTIR spectroscopy can be applied for all glass compositions while MAS NMR technique is applied only on glasses of low Fe₂O₃ concentration (≤6 mol%).

2. Experimental

2.1. Sample Preparation

The modified borosilicate glasses were prepared by melting batches in alumina crucibles at a temperature ranging from 1250˚C to 1520˚C depending on composition. Then the melt was quenching over a stainless steel plate. The batches were prepared from reagent grade SiO₂, H₃BO₃, Na₂CO₃, Al₂O₃ and Fe₂O₃.

2.2. Nuclear magnetic Resonance

All samples were measured with JEOL GSX-500 high-resolution solid-state MAS NMR spectrometer in a magnetic field of 11.4 T. ²⁹Si MAS NMR spectra were recorded at a frequency of 99.3 MHz and a spinning rate of 6 kHz. A cylindrical zirconia sample holder was rotated at a speed depends on the type of the measured nuclei. The applied pulse length was 2.62 µs and a recycle delay of 30 s. typically 1000 - 2000 scans were acquired. ¹¹B MAS NMR spectra were recorded at a frequency of 160.4 MHz and spinning rate of 15 KHz. The glass samples were measured with a single pulse length of 0.5 - 1.0 ms and a pulse delay of 2.5 s, and an accumulation of 100 - 200 scans. ²⁷Al MAS NMR spectra were recorded at a frequency of 130.3 MHz and spinning rate of 6 KHz.

2.3. FTIR measurements

Fourier transform infrared absorption signals of the studied glasses were measured at room temperature (20˚C) in the wavelength range 4000 - 400 cm^-1 using a computerized recording FTIR spectrometer (Mattson 5000, USA). Fine powdered samples were mixed with KBr in the ratio 1:100 for quantitative analysis and the weighed mixtures were subjected to a load of 5 t/cm² in an evocable i.e. to produce clear homogenous discs. Then, the IR absorption spectra were immediately measured after preparing the discs to avoid moisture attack.

3. Results and Discussion

There is a wide range of compositions which can form glass in the silicate and borosilicate systems [3] [5] [9] [11] - [14] . The structure of these glasses depends on two factors, namely R and K (R = modifier/B₂O₃ and K = SiO₂/B₂O₃) [8] [9] . The structural factors of the base glass (30Na₂O-2Al₂O₃-25SiO₂-43B₂O₃) are R = 0.7 and K = 0.58. The experimental results are mostly obtained on modified borosilicate glasses having R > K in all cases, since R changes between 0.7 and 1 and K changes from 0.58 to 0.67. In this case, the modifier oxide is participated between both the borate and the silicate network.

3.1. Iron free Borosilicate Glass

It was known that Al₂O₃ plays the role of both glass former and modifier in the network of the glass [15] [16] . This depends upon the type of the glass and the content of Al₂O₃. In our case, the Al₂O₃ inters the glass as a network former, since the value of chemical shift of the Al spectrum figure 1 is 59.6 ppm which assures that the aluminum is found in tetrahedral coordination which means that Al₂O₃ is consumed as a glass former.

Figure 2 presents ²⁹Si NMR spectrum of iron free borosilicate glass. The chemical shift of base glass is appeared at −88.4 ppm which is related to Q² species. This leads that the modifier oxide is consumed to break two bridging bonds as an average per each tetrahedral SiO₄ unit. Figure 3 shows experimental ²⁹Si MAS NMR for glasses containing different concentrations from Fe₂O₃ (0, 5 mol%). Comparing the two spectra, it can be seen from this figure that there is remarkable shift in the peak position between the spectral lines of ²⁹Si NMR of iron free glass and one contains 5 mol%

Figure 1

27 Al NMR spectrum of iron free glass

Figure 2

29 Si NMR spectrum of iron free glass

Fe₂O₃. The substitution of 5 mol% B₂O₃ by 5 mol% Fe₂O₃ has a clear effect on the chemical sifts of silicon nuclei. of base glass is appeared at −88.4 ppm, but the chemical shift value of glass containing 5 mol% is increased to reach −84.4 ppm upon addition of 5 mol% Fe₂O₃. This means that most of all of Fe₂O₃ is consumed as suggested above as a modifier, which results in increasing the concentration of NBO in silicate network. Increasing NBO atoms in silicate network is reflected from increasing behavior of chemical shift with increasing Fe₂O₃ content figure 4. This confirms that silicate units are deshielded or depolymerized with increasing iron oxide content in the region up to 6 mol%.

Figure 3

Chemical shift of 29 Si NMR spectra of iron free (at bottom) and for glass contains 5 mol% iron oxide

Figure 5 shows experimental ¹¹B MAS NMR for glasses containing different concentrations from Fe₂O₃. The spectrum at the bottom was obtained from sodium borosilicate glass (free from Fe₂O₃) with R = 0.7 and K = 0.58. The second two spectra were obtained from a glass containing 3 and 5 mol% Fe₂O₃ with R around 0.9. It can be seen from this figure that there is no remarkable shift in the peak position of the spectral lines of ¹¹B NMR. Substitution of 5 mol% B₂O₃ by 5 mol% Fe₂O₃ has no effect on the chemical shifts of boron nuclei. Chemical shift of BO₄ is appeared at 0 ppm for all investigated samples. But the fraction of boron tetrahedral units (N₄) is only affected, since it changed from 0.52 to 0.65 upon addition of 5 mol% Fe₂O₃, see figure 6. This means that substitution of B₂O₃ by Fe₂O₃ results in increasing the concentration of bridging bonds in the borate network through transformation of BO₃ triangle units to BO₄ groups.

Fe₂O₃ is therefore played the role of glass modifier in the region between 0 and 5 mol%. This based on the difference between the values of N₄ for sample free from Fe₂O₃ and the other containing 5 mol%, since glasses containing Fe₂O₃ showed higher values of N₄ compared with that of iron free glass, see figure 6.

Higher concentrations of Fe₂O₃ (8 - 20 mol%) are unsuitable for both ¹¹B and ²⁹Si NMR measurements [10] . This is because the high level of paramagnetic oxides which can induce local fields due to the unpaired electron spins which in turns produces sufficient broadening of NMR lines. Therefore, structure of glasses in the higher composition region has to be investigated by FTIR absorbance spectroscopy.

Figure 4

Chemical shift of 29 Si NMR spectra as a function of Fe 2 O 3 contents

Figure 5

11 B NMR spectra of selected composition of modified borosilicate glasses containing Fe 2 O 3

Figure 6

Fraction of boron tetrahedral as a function of Fe 2 O 3 concentration

3.2. FTIR spectroscopic Analysis

Figure 7 shows FTIR spectra obtained for the studied sodium borosilicate glasses containing different concentrations from Fe₂O₃. It can be realized from this figure that there are well defined FTIR envelops appeared at about 800 and 1580 cm⁻¹ in glasses containing up to 6 mol% Fe₂O₃ concentrations. The FTIR envelop cited at about 800 cm^-1 is assigned to NBO atoms in Q² and Q¹ species. This envelop is totally disappeared and the new IR peak at about 1285 cm⁻¹ is clearly seen at higher Fe₂O₃ concentrations. The well-defined envelops at 800 cm⁻¹ and 1580 cm⁻¹ are assigned to vibration of Si-O (non-bridging bonds) in different silicate networks [11] [13] [17] as a result of addition of Fe₂O₃ (up to 6 mol%). Disappearing of these envelops and appearance of new one at 1285 cm^-1 leads that concentration of NBO in the silicate network is reduced upon more addition of Fe₂O₃ (≥6 mol%). The appearance of the envelop at about 1285 cm⁻¹ is assigned to stretching vibration of Fe-O in iron silicate and borate species, as a glass forming groups [18] - [21] . On other wards, Fe₂O₃ in such situation plays, as suggested above, the role of glass former. It can remove NBO from the silicate network and as a result it consumes part of the modifier to form FeO₄ groups. In such situation, Si-O-Fe bonds are constructed and it played the role of shielding silicate units.

Quantitavely, figure 8 presents change of the fitted areas representing NBO and concentration of FeO₄ as a glass former against Fe₂O₃ content. The relative area of the peak centered at about 800 cm⁻¹ is clearly seen to increase with increasing Fe₂O₃ content, see figure 8 (between 0 and 5 mol%). On the other hand, at higher Fe₂O₃ content the envelope at 800 cm⁻¹ is decreased and vanished and other envelop at 1280 cm⁻¹ is rather appeared. The increasing behavior of relative area (at 800 cm⁻¹) may indicate an increase of the number of non-bridging oxygen atoms upon addition of Fe₂O₃ in silicate network. Higher area means higher concentration of modifier in the silicate network upon addition of 6 mol% Fe₂O₃. In this case we suggest that most of Fe₂O₃ enter the borosilicate network as an effective modifier. The relative area of the envelope at 1280 cm⁻¹ (see figure 8) is remarked to increase upon addition of more Fe₂O₃ (>5 mol%). These features are suggested to be due to reducing the content of modifier and NBO in the network of the glasses due to changing the structure role of iron from modifier to former. Presence of the well resolved envelope at about 1280 cm⁻¹ confirms the presence of iron as a former units connected to BO₃ groups as basic former units. Presence of Fe₂O₃ as a glass former should in turns reduce the fraction of boron tetrahedral units and NBO in the silicate network.

Figure 7

FTIR absorbance spectra of borosilicate glasses containing different Fe 2 O 3 concentrations

Figure 8

Changes of relative area of both IR peaks centered at 800 cm − 1 and 1280 cm − 1 .

4. Conclusion

Borosilicate glasses containing different concentrations from Fe₂O₃ have been prepared and investigated with FTIR and NMR spectroscopy. At low concentration from Fe₂O₃ (≤6 mol%), iron oxide inters the glass network as a modifier. On the other hand, Fe₂O₃ is consumed as a glass former at higher concentration. The modifier part of iron oxide is consumed to form NBO in silicate and bridging bond in borate network. As a result Q² was found as the main units in silicate network. The fraction of boron tetrahedral N₄ was found to increase in glasses of up to 6 mol% iron oxide. At higher concentration N₄ decreases due to the action of Fe₂O₃ as a glass former.

References

1. Pascual, M.J., Duran, A. and Pascual, L. (2002) Viscosity and Thermal Properties of Two Series of Glasses in the System R2O-B2O3-SiO2, R = Li, K, Na. Physics and Chemistry of Glasses, 43, 25-31.

2. Pascual, M.J., Duran, A. and Pascual, L. (2002) Sintering Process of Glasses in the System Na2O-B 2O3-SiO2. Journal of Non-Crystalline Solids, 306, 58. <p>http://dx.doi.org/10.1016/S0022-3093(02)01049-9

3. Munoz, F., Montagne, L. and Delevoye, L. (2006) Phosphate Speciation in Sodium Borosilicate Glasses Studied by Nuclear Magnetic. Journal of Non-Crystalline Solids, 352, 2958- 2968.

4. Gautam, C. (2014) Synthesis, Structural and Optical Investigations of (Pb, Bi)TiO3 Borosilicate Glasses. Physics Research International, 2014, Article ID: 606709, 1-7. <p>http://dx.doi.org/10.1155/2014/606709

5. Affatigato, M., Feller, S.A., Howes, A.P. and Scales, C.R. (2008) Quantitative Measurement of Q 3 Species in Silicate and Borosilicate Glasses Using Raman Spectroscopy. Journal of Non-Crystalline Solids, 354, 1936-1942. <p>http://dx.doi.org/10.1016/j.jnoncrysol.2007.06.105

6. Dupree, R., Holland, D. and Mortuza, M.G. (1990) A MAS-NMR Investigation of Lithium Silicate Glasses and Glass Ceramics. Journal of Non-Crystalline Solids, 116, 148. <p>http://dx.doi.org/10.1016/0022-3093(90)90687-H

7. Stebbins, J.F. (1987) Identification of Multiple Structural Species in Silicate Glasses by 29Si NMR. Nature, 330, 465. <p>http://dx.doi.org/10.1038/330465a0

8. Dell, W.J., Bray, P.J. and Xiao, S.Z. (1983) 11 B NMR Studies and Structural Modeling of Na2O-B2O3-SiO2 Glasses of High Soda Content. Journal of Non-Crystalline Solids, 58, 1-16. <p>http://dx.doi.org/10.1016/0022-3093(83)90097-2

9. El-Damrawi, G., Müller-Warmuth, W. and Doweidar, H. (1992) Structure and Heat Treatment Effects of Sodium Borosilicate Glasses as Studied by 29Si and 11B NMR. Journal of Non-Crystalline Solids, 146, 137-144. <p>http://dx.doi.org/10.1016/S0022-3093(05)80485-5

10. Mackenzie, K.J.D. and Smith, M. (2002) Multinuclear. Solid-State NMR Inorg. Mater. Pergamon Materials Series 4, 218.

11. El Damrawi, G., Hassan, A.K., Kamal, H., Aboelez, M. and Labeeb, S. (2016) Structural Investigation on Na2O-CaO-V2O5-SiO2 Bioglass Ceramics. British Journal of Applied Science & Technology, 16, 1-9. <p> http://dx.doi.org/10.9734/BJAST/2016/26683

12. El-Damrawi, G., Doweidar, H. and Kamal, H. (2014) Microstructure and In Vitro Bioactivity of Metal Substituted Hydroxyapatite. Silcon, 1-7.

13. Du, L.S. and Stebbins, J.F. (2003) Nature of Silicon-Boron Mixing in Sodium Borosilicate Glasses: A High-Resolution 11B and 17O NMR Study. The Journal of Physical Chemistry B, 107, 10063-10076. <p>http://dx.doi.org/10.1021/jp034048l

14. El-Damrawi, G., Doweidar, H. and Kamal, H. (2013) Structure and Crystallization Behavior of Silicate-Based Bioactive Glasses. Australian Journal of Basic and Applied Science, 7, 573-582.

15. Doweidar, H., El-Damrawi, G. and Ramadan, R.M. (2008) Sites Distribution and Properties of Al2O3-PbO-B2O3 Glasses. Phys. Chem. Glasses: Eur. J. Glass Sci. Technol. B, 49, 271-277.

16. El-Damrawi, G., Müller-Warmuth, W., Doweidar, H. and Gohar, I.A. (1993) 11B, 29Si and 27Al Nuclear Magnetic Resonance Studies of Na2O-Al2O3-B2O3-SiO2 Glasses. Physics and Chemistry of Glasses, 34, 52.

17. Nesbitt, H.W., Bancroft, G.M., Davidson, R, Mclntyre, N.S. and Pratt, A.R. (1998) Sulfur and Iron Surface States on Fractured Pyrite Surfaces. American Mineralogist, 83, 1067-1076. <p>http://dx.doi.org/10.2138/am-1998-9-1015

18. Magdas, D.A., Cozar, O., Chis, V., Ardelean, I. and Vedeanuc, N. (2008) The Structural Dual Role of Fe2O3 in Some Lead-Phosphate Glasses. Vibrational Spectroscopy, 48, 251- 254. <p>http://dx.doi.org/10.1016/j.vibspec.2008.02.016

19. Moustafa, Y.M., Doweidar, H. and El-Damrawi, G. (1994) Utilisation of Infrared Spectroscopy to Determine the Fraction of the Four Coordinated Borons in Borate Glasses. Physics and Chemistry of Glasses, 35, 104-106.

20. Varshneya, A.K. (2008) Fundamentals of Inorganic Glasses. Society of Glass. Technology, (2006), J. Mater. Chem. 111, 455-462.

21. Qiu, D. and Guerry, P (2008) A High-Energy X-Ray Diffraction, 31P and 11B Solid-State NMR Study of the Structure of Aged Sodium Borophosphate Glasses. Materials Chemistry and Physics, 111, 3455-462. <p>http://dx.doi.org/10.1016/j.matchemphys.2008.04.045