Structure and Viscosity of CaO–Al₂O₃–B₂O₃ Based Mould Fluxes with Varying CaO/Al₂O₃ Mass Ratios

Abstract

The effect of CaO/Al₂O₃ ratio on the structure of CaO–Al₂O₃–B₂O₃ based glassy mould fluxes was investigated by employing ²⁷Al and ¹¹B Triple Quantum Magic-angle spinning nuclear magnetic resonance (3QMAS-NMR) and Raman spectroscopy. ²⁷Al and ¹¹B 3QMAS-NMR spectrums showed that Al³⁺ mainly forms [AlO₄] as a network former and B³⁺ mainly forms [BO₃] groups in CaO–Al₂O₃–B₂O₃ based glasses. Raman spectrum showed existences of different [AlO₄] structure units and BO₃ pyro-borate units. In addition, deconvolution results on Raman spectrums indicate that the degree of polymerization of aluminate network in CaO–Al₂O₃–B₂O₃ based glasses decreases with the increase of CaO/Al₂O₃ ratio. The effect of CaO/Al₂O₃ ratio on viscosity of CaO–Al₂O₃–B₂O₃ based glassy mould fluxes was investigated by employing the rotating-cylinder method. The viscosity decreases with increasing CaO/Al₂O₃ ratio in CaO–Al₂O₃–B₂O₃ based mould flux. Correlation between viscosity and structural information of investigated mould fluxes was explored.

1. Introduction

Recently, high Al steel (e.g. high Al-TRIP and TWIP steel) has received increasing attention as it has the perfect combination of strength to weight ratio and excellent ductility.¹⁾ Significant amounts of aluminum are added into high Al-TRIP and TWIP steel to improve properties of steel.²⁾ However, high levels of Al in the steel may cause great process control problems during the continuous casting, it is due to the chemical reactions between SiO₂ in mould flux and [Al] in steel. This reaction can be illustrated by following equation:^3,4)   

$$4[\mathrm{A}\mathrm{l}]+3\mathrm{S}\mathrm{i}{\mathrm{O}}_2=2\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3+3[\mathrm{S}\mathrm{i}]$$(1)

The above reaction between the mould flux and steel will lead to a gradual increase of alumina content and decrease of silica content in the mould fluxes. The dynamically changing of chemistry for the mould flux leads to varied physicochemical properties during the casting. As a result, there will be several casting problems presented, such as increased crack frequency, non-uniform heat transfer across the mould flux, reduced lubrication, etc., and poor casting performance could be expected.⁴⁾

CaO–Al₂O₃ based mould flux is proposed to solve the problem on chemical composition change of mould fluxes during high Al steel casting.^4,5,6) Some researchers have suggested some additives such as BaO, MnO and TiO₂ to adjust the properties of mould fluxes to meet the demand for casting of high Al steel.^5,6,7) However, there are many problems need to be solved for application of CaO–Al₂O₃ based mould flux, e.g. strong crystallization and poor lubrication between steel shell and mould.^4,8) The study on physiochemical properties of the CaO–Al₂O₃ based system is still limited to meet the demand for development of optimal mould fluxes for casting high Al steel.

In our previous work, properties of a new type of mould flux based on CaO–Al₂O₃–B₂O₃ system were investigated.⁹⁾ The viscosity at 1300°C, melting characteristics, and crystallization temperature was measured to compare with traditional mould fluxes. However, the viscosity at different temperature and structure of mould fluxes were not reported. In order to evaluate the influence of CaO/Al₂O₃ ratio on the viscosity and structure of CaO–Al₂O₃–B₂O₃ based mould flux, measurements on viscosity and structure of mould fluxes were carried out. The viscosity values of CaO–Al₂O₃–B₂O₃ based mould fluxes were measured by using the rotating spindle method. The water-quenched glassy mould fluxes subjected to Raman spectroscopy and triple quantum Magic Angle Spinning Nuclear Magnetic Resonance (3QMAS-NMR) in order to analyze the structural of glassy mould fluxes. ²⁷Al and ¹¹B 3QMAS-NMR were applied to study the local environment of Al and B in the CaO–Al₂O₃–B₂O₃ based glasses. All approaches provide a better knowledge of the structure of the CaO–Al₂O₃–B₂O₃ based glasses that should improve the understanding of the physicochemical properties of the CaO–Al₂O₃–B₂O₃ based melts. A correlation between viscosity and structure of mould fluxes would be established.

2. Experimental

2.1. Sample Preparation

Analytical grade CaCO₃, Na₂CO₃, Li₂CO₃, Al₂O₃, SiO₂ and H₃BO₃ (all purity > 99.5%) were taken as raw materials, with the CaCO₃ being substituted for CaO. CaCO₃ were calcined at 1323 K to obtain CaO in a muffle furnace for 10 hours. Na₂CO₃, Li₂CO₃, SiO₂ and Al₂O₃ powders were also calcined at 873 K to remove moisture. After carefully weighed, the powder mixtures were ground in a mortar, about 140 g fluxes for each sample was prepared for the viscosity measurement and about 7 g fluxes for each sample were subject to the Raman and NMR measurement.

2.2. Structure Analysis

All samples were heated in a Pt crucible at 1723 K for 3 h to ensure homogeneous melting. High temperature melts were quenched into water to form the glasses. After quenching, the samples were subject to Electron Probe Micro-Analyzer (EPMA, JEOL JXA-8230, Japan) to determine the composition. Nominal Compositions of samples and analyzed compositions of samples after melting by EPMA are listed in Table 1. The water-quenched samples were subjected to XRD measurements as shown in Fig. 1. It was confirmed that all samples are in glassy state. Glassy samples subjected to non-polarized Raman spectroscopy and Nuclear Magnetic Resonance Spectrum (NMR) measurements to determine the structure of fluxes.

Table 1. Chemical composition of CaO–Al₂O₃–B₂O₃ based mould fluxes.

Sample No.	CaO/Al2O3		Composition (mass%/mol%)
			CaO	Al2O3	B2O3	Na2O	Li2O	Tb (°C)
1	0.9	Nominal	36/41.6	40/25.4	8/7.4	8/8.4	8/17.3	1240
		Analyzed	36.0/41.5	38.1/24.1	9.2/8.5	8.3/8.6	8/17.2
2	1	Nominal	38/43.5	38/24.1	8/7.3	8/8.3	8/17.1	1211
		Analyzed	36.9/43.4	35.4/22.8	8.8/8.3	7.5/8.0	8/17.6
3	1.11	Nominal	40/45.3	36/22.4	8/7.2	8/8.2	8/16.9	1155
		Analyzed	40.1/46.5	33.2/21.1	7.7/7.1	7.6/8.0	8/17.3

Fig. 1.

XRD patterns of glassy mould fluxes.

Raman spectra were recorded at room temperature in the frequency range of 20–1650 cm⁻¹ using a laser confocal micro-Raman spectrometer (Horiba, LabRAM HR Evolution). The experiments were performed in room temperature using He–Cd laser with an excitation wavelength of 532 nm. All spectra were recorded between 20 and 1650 cm⁻¹. Finally, the spectra were deconvolved using Peakfit software. During the curve-fitting procedure, widths and intensities are independent and unconstrained variables. Raman spectra were fitted by assuming Gaussian line shapes for peaks of different structural units.

The Triple Quantum Magic Angle Spinning Nuclear Magnetic Resonance (3QMAS-NMR) spectrums have been acquired using the Bruker BioSpin GmbH (400 MHz). ¹¹B 3QMAS and ²⁷Al MQMAS were used to obtained the structure information of B and Al. ¹¹B and ²⁷Al isotropic chemical shifts δiso are externally referenced relative to BF₃ · Et₂O and a 1 M aqueous Al(NO₃)₃ solution, respectively. ¹¹B and ²⁷Al 3QMAS experiments were conducted at a spinning frequency of 15 kHz. A 3-pulse z-filter sequence¹⁰⁾ was employed for the ¹¹B 3QMAS-NMR and ²⁷Al 3QMAS-NMR measurements. The duration of the first, second and third pulse was set to 3,1 and 11.6 μs, respectively.

2.3. Viscosity Measurements

Viscosity measurements were carried out using a rotating cylinder method with a rotary viscometer (model: RTW-10). The details about the apparatus, Mo crucible and spindle have been described in our previous works.¹¹⁾ Before measurements, the viscometer was calibrated at room temperature by using standard castor oil with known viscosity. After carefully weighed, the powder mixture was ground in a mortar, about 140 g slag for each sample was subject to the viscosity measurement which was carried out in a molybdenum crucible and a graphite crucible was used to protect the molybdenum crucible on the outside of the molybdenum crucible. Firstly, the samples were held in the furnace at 1673 K for 30 min to be pre-melted, and then quenched in air. During the procedure of viscosity measurement, all samples were held at 1673 K for 30 min to ensure that slags were completely melted and homogeneous. Subsequently, the molybdenum spindle with a size was immersed in the molten slag and rotated at 200 r/min. After stability of the measurement, the furnace was cooled at a rate of 5 K/min and the measurement of viscosity started. Argon was used by 50 ml/min as the protective gas during the whole experimental process.

3. Results and Discussions

3.1. Effect of the CaO/Al₂O₃ Mass Ratio on Structure of Glassy Mould Fluxes

Figures 2, 3 showed ²⁷Al and ¹¹B 3QMAS spectra for glassy mould fluxes samples with CaO/Al₂O₃ = 0.9, 1 and 1.11. In ²⁷Al and ¹¹B NMR studies on glasses, resolution among various [AlO₄], [AlO₅], [AlO₆], [BO₃] and [BO₄] species in conventional ²⁷Al and ¹¹B MAS spectra of such glasses is poor due to the quadrupolar broadening.¹²⁾ 3QMAS-NMR can provide much better resolutions for structural analyses. In 3QMAS-NMR spectroscopy, the evolution of multiple quantum coherence is correlated with the evolution of a single quantum coherence under the conditions of fast MAS, resulting in a separation of the isotropic chemical shifts and the quadrupolar broadening in a 2D experiment.^10,13,14) In the present work, structures of several aluminoborate glasses were explored by 3QMAS-NMR technique. For the CaO–Al₂O₃–B₂O₃ based glasses with differing CaO/Al₂O₃ ratios, 3QMAS-NMR experiments clearly revealed the bonding information of Al and B in different coordination.

Fig. 2.

²⁷Al 3QMAS-NMR spectra for CaO/Al₂O₃ = 0.9, 1, 1.11. The peak assignable to [AlO₄] sites is labeled. The low-intensity feature to the top of the main peak is a spinning sideband (SSB). (Online version in color.)

Fig. 3.

1D projections of ²⁷Al 3QMAS-NMR spectra for CaO/Al₂O₃ = 0.9, 1, 1.11. (Online version in color.)

As shown in Fig. 2, the resolution of a sole and obvious signal around 65 ppm region was observed in ²⁷Al 3QMAS spectra for CaO/Al₂O₃ = 0.9, 1 and 1.11. The 1D spectrum on top is the projection onto the isotropic dimension. The previous references revealed the presence of four-, five-, and six-fold coordinated aluminum sites with resonance shifts near 60–70, 30, and 0 ppm, respectively.¹²⁾ Based on the earlier assignments,^14,15) the signal around 65 ppm belongs to the tetrahedral coordinated species [AlO₄]. There is no detectable intensity at [AlO₅] and [AlO₆] groups at the corresponding positions in spectrums of the investigated glasses. From 1D projection of ²⁷Al 3QMAS spectra for CaO/Al₂O₃ = 0.8, 1, 1.11 in Fig. 3, it can be seen that the spectrum has a strong and broad peak centered at about 70 ppm, which is in close proximity to the chemical shift of [AlO₄]. Therefore, most of Al is exclusively connected by four oxygen of the structure in present glassy slags. It was reported that [AlO₄] is the main bond of Al in the investigated aluminoborate glasses.^13,16,17) The molar ratio of CaO/Al₂O₃ calculated from chemical compositions of mould fluxes is increased from 1.64 to 2.02 in present glassy slags. The molar ratio of CaO/Al₂O₃ is much larger than 1 which guarantees that Al³⁺ can be compensated completely to form [AlO₄] groups. Therefore, it can be concluded that Al³⁺ is mainly to form [AlO₄] as a network former. McMillan et al., investigated the structure of CaO–Al₂O₃ glasses by ²⁷Al MAS-NMR, showed that Al formed fourfold (tetrahedral) coordination to oxygen in the glass (CaO: Al₂O₃ ≥ 1).¹⁸⁾ Neuville et al.,¹⁹⁾ investigated the Al coordination and speciation in calcium aluminosilicate glasses by ²⁷Al MQ-MAS NMR, found that only AlO₄ tetrahedra are detected in calcium aluminate with low silica and high CaO content. Licheron et al.,²⁰⁾ reported that only a characteristic of the tetrahedrally-coordinated aluminum was identified in CaO–Al₂O₃ glasses with CaO/Al₂O₃ > 1. These results are consistent with our results.

¹¹B 3QMAS-NMR spectrums for the samples with increasing CaO/Al₂O₃ ratio were shown in Fig. 4. [BO₃]-ring groups and [BO₃]-nonring groups at 12 ppm and 6 ppm can be discriminated by ¹¹B 3QMAS-NMR spectrums.²¹⁾ No site of [BO₄] group around 0 ppm can be found in ¹¹B 3QMAS-NMR spectrums. This indicated that almost all B³⁺ is to form [BO₃] group.

Fig. 4.

¹¹B 3QMAS-NMR spectra for CaO/Al₂O₃ = 0.9, 1, 1.11. The peak assignable to [BO₃] sites is labeled. (Online version in color.)

According to previous studies,²²⁾ within a boron-containing glass, B³⁺ may be coordinated by four oxygen atoms ([BO₄] groups), or coordinated by three oxygen atoms ([BO₃] groups).Three coordinated boron may also have NBOs (asymmetric BO₃ groups). Unlike in silicates, when modifiers oxides are initially added to B₂O₃, most, if not all, [BO₃] is converted to [BO₄] without the formation of NBO. Since an increase in modifier species could introduce NBO into the system, this change has been hypothesized by the structural reaction:²²⁾   

$$[B{O}_3]+NBO\iff [B{O}_4]$$(2)

At low contents of modifier oxides in boron-rich glasses, this reaction is nearly complete.^23,24,25) If more modifiers are added, four coordinated boron begin to convert to asymmetric BO₃ groups. However, both the beginning point and the rate of this process depend strongly on the composition.²²⁾ In boroaluminate glass, at high modifier contents, [BO₄] transforms into [BO₃] in part to avoid formation of ^[4]B–O–^[4]B and ^[4]B–O–^[4]Al species, which is relatively energetically unfavorable since the bridging oxygen atoms have higher net negative charges.^12,15,16) Higher modifier cation field strength favors the formation of NBO and convert [BO₄] into [BO₃].^15,26,27,28) In present glassy slags, the molar ratio of CaO/B₂O₃ increases from 5.62 to 6.29. It can be seen that CaO is much higher amount than B₂O₃. Due to this phenomenon, [BO₃] is conclusively formed in present slags. Alkali metals and alkaline earths play dual structural roles in alkali and alkaline earth boroaluminate glass. One of these roles is to serve as network-modifiers to decrease the network degree of polymerization. In this sense, these cations are linked to terminal oxygens on tetrahedra. The other role is to charge compensate cations such as tetrahedral B³⁺ and Al³⁺. Hayek et al.,²⁹⁾ investigated the structure of lime alumino-borate glasses by ¹¹B NMR and ²⁷Al NMR spectroscopies and found that B³⁺ and Al³⁺ are mainly formed [BO₃] and [AlO₄] when the concentration of B₂O₃ is less than 8.8 mass% and the CaO/Al₂O₃ is approximately equal to 1.13. The composition of lime alumino-borate glass is very close to sample 3 in our work. This further proves the correctness of the results in our work.

In order to analyze the change of degree of polymerization of the aluminate or boroaluminate network, Raman spectrums were collected for the present glassy mould fluxes. Original Raman spectrums of all samples were shown in Fig. 5. It could be seen that there are three main intensity peak envelopes of the center shift at about 540 cm⁻¹, 770 cm⁻¹, 920 cm⁻¹, respectively. According to the Licheron et al.,²⁰⁾ the Raman spectrum of C50Al50 glass consists of strong bands at 520 cm⁻¹ and 790 cm⁻¹, which were identified by the transverse motions of bridged oxygens within the Al–O–Al linkages and the symmetric stretching vibration of Al–O. This is consistent to our study. These main peaks could show the information of Al–O–Al oxygen bridged bond, the network structure formed by [AlO₄] and [BO₃] structure. Three regions in the Raman spectra were detected: the low-frequency region (300–700 cm⁻¹), the mid-frequency region (700–900 cm⁻¹) and the high-frequency region (900–1000 cm⁻¹). The band near 550 cm⁻¹ in Al₂O₃-rich glasses is due to the presence of Al–O–Al bridges in the low-frequency region.^6,19) The bands at 770 cm⁻¹ were ascribed to the Al–O stretching in differing Qⁿ species (n = 0.1.2.3.4, it is the number of bridging oxygen in [AlO₄]) in the mid-frequency region (700–900 cm⁻¹).^19,30) The Raman band at around 770 cm⁻¹ shifts to lower wavenumbers with the increase of CaO/Al₂O₃ ratio. The bands at 920 cm⁻¹ were attributed to the symmetric stretching vibrations of terminal oxygen atoms in orthoborate units [BO₃].^19,31,32) That is to say, B³⁺ is mainly to form [BO₃] groups in present slags. It is in consistence with ¹¹B 3QMAS spectra results.

Fig. 5.

Original Raman spectra for samples for CaO/Al₂O₃ = 0.9, 1, 1.11 at room temperature.

The de-convoluted results of Raman spectra are showed in Figs. 6(a)–6(c) through the Gaussian-Deconvolution method similar to method by Mysen et al.³³⁾ with the minimum correlation coefficient r² ≥ 0.998. All spectrums were successfully fitted to obtain quantitative information of structure. The low-frequency region (300–700 cm⁻¹) has been fitted to two peaks in all glasses. The band near 550 cm⁻¹ was due to the presence of Al–O–Al bridges. The intensity of Raman band at around 550 cm⁻¹ drastically decreased with increasing CaO/Al₂O₃ ratio. The decrease of the relative intensity of Raman band at 550 cm⁻¹ indicated that the number of bridging oxygen decreased by increasing the CaO/Al₂O₃ ratio. That is to say, the aluminate structure is partly modified by the increase in the CaO/Al₂O₃ ratio. Next to the peak, the bands at 630 cm⁻¹ presented a small peak, it was ascribed to B_III–O–Al_IV bending vibration.^34,35) The relative intensity of Raman band at 630 cm⁻¹ had no significant changes in all glasses. Cormier et al.,^36,37) have reported that there have various Qⁿ(Al) (n = 0, 1, 2, 3, 4) species in 61 mol%CaO and 39 mol% Al₂O₃ glass. The CaO/Al₂O₃ ratio for their glass is close to that of the sample 3. Therefore, the Raman bands at 770 cm⁻¹ in the mid-frequency region (700–900 cm⁻¹) has been assumed to be fitted to five kinds of Qⁿ structure units in present work. The Raman bands at 720 cm⁻¹, 740 cm⁻¹, 760 cm⁻¹, 780 cm⁻¹ and 830 cm⁻¹ correspond to the Al–O stretching vibration in [AlO₄] tetrahedral units with the NBO/Al = 4 (Q⁰_Al), NBO/Al = 3 (Q¹_Al), NBO/Al = 2 (Q²_Al), NBO/Al = 1 (Q³_Al) and NBO/Al = 0 (Q⁴_Al), based on the assignments of Raman bands in spectral of aluminate glass.^30,38,39) The deconvolution results were shown in Table 2. As seen in Table 2, Q²_Al replaces Q³_Al to become the strongest band with the increase of CaO/Al₂O₃ ratio. The intensity of Q³_Al and Q⁴_Al were decreased, while the intensity of Q⁰_Al, Q¹_Al and Q²_Al were increased with increasing CaO/Al₂O₃ ratio.

Fig. 6.

The deconvoluted results of Raman spectra for CaO/Al₂O₃ = 0.9, 1, 1.11. (Online version in color.)

Table 2. Deconvolved results of Raman spectra for CaO–Al₂O₃–B₂O₃ based glasses.

Sample No.	CaO/Al2O3	Percentage of each unit (%)
		Q0Al	Q1Al	Q2Al	Q3Al	Q4Al
1	0.9	6.83	13.22	19.60	39.92	20.43
2	1	7.47	14.36	30.77	32.66	14.75
3	1.11	7.75	28.15	44.26	15.52	4.32

According to Frantz and Mysen,⁴⁰⁾ the mole fractions of different structure units are related to the band areas in Raman spectra. In this work, the average number of bridging oxygen of each sample is used to explain the change of the aluminates structure network, which could be estimated by summation of the area ratio of all structural units (Q⁰, Q¹, Q², Q³ and Q⁴) multiplied by the number of its bridging oxygen. The effects of CaO/Al₂O₃ ratio on the estimated average number bridging oxygen were shown in Fig. 7. It can be seen that the average number of bridging oxygen just decreased from 2.54 to 1.81 as the CaO/Al₂O₃ ratio in mass increases from 0.9 to 1.11, indicating degree of polymerization of aluminate network is decreased. Meanwhile, the NBO/T (the ratio of Non-bridging Oxygen/Oxygen in Tetragonal bonding) values have been estimated using a mathematical model reported by Mills et al.,⁴¹⁾ It was obtained that the values of NBO/T for slags with CaO/Al₂O₃ ratio of 0.9, 1.0 and 1.1 are approximately equal to 1.05, 1.19 and 1.37 respectively. The values of BO/T for slags with CaO/Al₂O₃ ratio of 0.9, 1.0 and 1.1 can be deduced to be 2.95, 2.81 and 2.63 respectively, which are shown in Fig. 7. It can be seen that the values between the Raman calculation and the mathematical model are close to each other, indicating that the estimation method for the average number of bridging oxygen in present work is reliable.

Fig. 7.

The change of the average number of bridging oxgyen for CaO/Al₂O₃ = 0.9, 1, 1.11. (Online version in color.)

3.2. Effects of the CaO/Al₂O₃ Mass Ratio on Viscosity

As can be seen from the viscosity curves of the mould fluxes with the w(CaO/Al₂O₃) ratio decreased from 0.8 to 1.11 in Fig. 8, the viscosity of CaO–Al₂O₃–B₂O₃ based mould fluxes decreased with the CaO/Al₂O₃ ratio increasing from 0.9 to 1.11 at the same temperature. Additionally, it can be found that with the increase of temperature, the measured viscosity decreased because of a weaker interaction among flow units caused by temperature increase. Below a critical temperature, the viscosity of mould fluxes rapidly increases with decreasing temperature. This critical temperature could be defined as a breaking temperature for mould fluxes. It can be seen from the Fig. 8 that, the breaking temperature for mould fluxes decreases with increasing CaO/Al₂O₃ ratio. The breaking temperatures from viscosity curves are mainly dependent on the thermodynamic liquidus temperature and crystallization behavior of slag. The thermodynamic liquidus temperature for slag cannot be calculated from thermodynamic package due to the limiting thermodynamic data for this new slag system. But according to our previous study on crystallization of the same slag,⁴²⁾ the slag with the lowest CaO/Al₂O₃ ratio (CaO/Al₂O₃ = 0.9) has the fastest crystallization, which could also contribute to the highest breaking temperature.

Fig. 8.

Viscosity of mould fluxes with varying CaO/Al₂O₃ mass ratio as functions of temperature.

Correlation between structural information and physiochemical properties of slags could be explored. Many researchers have reported that viscosity of slags has closely relationship with structure.^6,34,43,44) In this work, with the increase of CaO/Al₂O₃ ratio, more excess free oxygen ions (O²⁻) supplied from the dissociation of CaO would lead to a gradual decrease of the bridging oxygen structure (BO) in aluminate structure and an increase of the non-bridging oxygen (NBO) in aluminate structure. This would promote the de-polymerization of aluminate structure, leading to smaller and simpler aluminate structure units. On the other hand, with increasing the CaO/Al₂O₃ ratio, more excess free oxygen ions (O²⁻) supplied from the dissociation of CaO were introduced into borate structural units to create more NBO for borates. This also leads to the formation of many small and simple structure units of borate which is beneficial to viscous flow of slags. The modified network structures of boroaluminates have comparably smaller unit network structures resulting in lower resistance for shearing the liquid layers to decrease the viscosity. The 3QMAS-NMR and Raman spectral study in section 3.1 confirmed that the degree of polymerization of aluminates is decreased with the increase of CaO/Al₂O₃ ratio. Due to above reasons, with the increase of CaO/Al₂O₃ ratio, the viscosity of mould fluxes are decreased.

4. Conclusions

The effect of CaO/Al₂O₃ ratio on viscosity and structure of CaO–Al₂O₃–B₂O₃ based mould fluxes were investigated by employing the rotating-cylinder viscosity measurement in conjunction with ²⁷Al 3QMAS-NMR, ¹¹B 3QMAS-NMR and Raman spectroscopy measurements. Furthermore, the relationship between viscosity and structure has been explored. The following conclusions have been obtained.

(1) ²⁷Al 3QNMR and ¹¹B 3QNMR investigation results have shown that Aluminum mainly forms [AlO₄] as a network former and Boron mainly forms [BO₃] groups in CaO–Al₂O₃–B₂O₃ based glasses.

(2) Raman spectral investigation confirmed that boron mainly forms pyroborate (BO₃) structural units in CaO–Al₂O₃–B₂O₃ based glasses. In addition, deconvolution results on Raman spectral showed the degree of polymerization of aluminate network in CaO–Al₂O₃–B₂O₃ based glasses decreased with the increase of CaO/Al₂O₃ ratio.

(3) The viscosity of CaO–Al₂O₃–B₂O₃ based mould fluxes decreased as increasing CaO/Al₂O₃ ratio in CaO–Al₂O₃–B₂O₃ based mould fluxes. Correlation between viscosity and structural of mould fluxes could be obtained.

Acknowledgements

Financial support from the Natural Science Foundation of China (NSFC contract No. 51774026 and 51704050) and the Academy of Finland for Genome of steel grant (No. 311934) is gratefully acknowledged.

References

• 1)   A.  Grajcar,  R.  Kuziak and  W.  Zalecki: Arch. Civ. Mech. Eng., 12 (2012), 334.
• 2)   O.  Grässel,  L.  Krüger,  G.  Frommeyer and  L. W.  Meyer: Int. J. Plast., 16 (2000), 1391.
• 3)   M. S.  Kim,  S. W.  Lee,  J. W.  Cho,  M. S.  Park,  H. G.  Lee and  Y. B.  Kang: Metall. Mater. Trans. B, 44 (2013), 299.
• 4)   J.  Cho,  K.  Blazek,  M.  Frazee,  H.  Yin,  J. H.  Park and  S.  Moon: ISIJ Int., 53 (2013), 62.
• 5)   D.  Xiao,  W.  Wang and  B.  Lu: Metall. Mater. Trans. B, 46 (2015), 873.
• 6)   J. L.  Li,  Q. F.  Shu and  K. C.  Chou: Can. Metall. Q., 54 (2015), 85.
• 7)   H.  Zhao,  W.  Wang,  L.  Zhou,  B.  Lu and  Y.  Kang: Metall. Mater. Trans. B, 45 (2014), 1510.
• 8)   J.  Li,  Q.  Shu,  X.  Hou and  K.  Chou: ISIJ Int., 55 (2015), 830.
• 9)   J.  Li,  B.  Kong,  B.  Galdino,  J.  Xu,  K.  Chou,  Q.  Liu and  Q.  Shu: Steel Res. Int., 88 (2017), 1600485.
• 10)   V.  Lacassagne,  P.  Florian,  V.  Montouillout,  C.  Gervais,  F.  Babonneau and  D.  Massiot: Magn. Reson. Chem., 36 (1998), 956.
• 11)   J. L.  Li,  Q. F.  Shu and  K.  Chou: Ironmaking Steelmaking, 42 (2015), 154.
• 12)   B. C.  Bunker,  R. J.  Kirkpatrick,  R. K.  Brow,  G. L.  Turner and  C.  Nelson: J. Am. Ceram. Soc., 74 (1991), 1430.
• 13)   K. U.  Gore,  A.  Abraham,  S. G.  Hegde,  R.  Kumar,  J.  Amoureux and  S.  Ganapathy: J. Phys. Chem. B, 106 (2002), 6115.
• 14)   J. H.  Baltisberger,  Z.  Xu,  J. F.  Stebbins,  S. H.  Wang and  A.  Pines: J. Am. Ceram. Soc., 118 (1996), 7209.
• 15)   L.  Du and  J. F.  Stebbins: J. Non-Cryst. Solids, 351 (2005), 3508.
• 16)   L.  Züchner,  J. C. C.  Chan,  W.  Müller-Warmuth and  H.  Eckert: J. Phys. Chem. B, 102 (1998), 4495.
• 17)   H.  Deters,  A. S.  de Camargo,  C. N.  Santos,  C. R.  Ferrari,  A. C.  Hernandes,  A.  Ibanez,  M. T.  Rinke and  H.  Eckert: J. Phys. Chem. C, 113 (2009), 16216.
• 18)   P. F.  McMillan,  W. T.  Petuskey,  B.  Coté,  D.  Massiot,  C.  Landron and  J.  Coutures: J. Non-Cryst. Solids, 195 (1996), 261.
• 19)   D. R.  Neuville,  L.  Cormier and  D.  Massiot: Chem. Geol., 229 (2006), 173.
• 20)   M.  Licheron,  V.  Montouillout,  F.  Millot and  D. R.  Neuville: J. Non-Cryst. Solids, 357 (2011), 2796.
• 21)   L.  Du and  J. F.  Stebbins: J. Phys. Chem. B, 107 (2003), 10063.
• 22)   J.  Wu: Ph.D. thesis, Stanford University, (2011), 3.
• 23)   R. J.  Araujo: J. Non-Cryst. Solids, 58 (1983), 201.
• 24)   B. C.  Bunker,  D. R.  Tallant,  R. J.  Kirkpatrick and  G. L.  Turner: Phys. Chem. Glasses, 31 (1990), 30.
• 25)   W. J.  Dell,  P. J.  Bray and  S. Z.  Xiao: J. Non-Cryst. Solids, 58 (1983), 1.
• 26)   T. J.  Kiczenski,  L.  Du and  J.  Stebbins: J. Non-Cryst. Solids, 351 (2005), 3571.
• 27)   S.  Sen,  Z.  Xu and  J. F.  Stebbins: J. Non-Cryst. Solids, 226 (1998), 29.
• 28)   J. F.  Stebbins and  S. E.  Ellsworth: J. Am. Ceram. Soc., 79 (1996), 2247.
• 29)   R.  El Hayek,  F.  Ferey,  P.  Florian,  A.  Pisch and  D. R.  Neuville: Chem. Geol., 461 (2017), 75.
• 30)   T. S.  Kim and  J. H.  Park: ISIJ Int., 54 (2014), 2031.
• 31)   L. M.  Osipova,  A. A.  Osipov and  V. N.  Bykov: Glass Phys. Chem., 33 (2007), 486.
• 32)   G. D.  Chryssikos,  E. I.  Kamitsos and  A. P.  Patsis: J. Non-Cryst. Solids, 202 (1996), 222.
• 33)   B. O.  Mysen,  L. W.  Finger,  D.  Virgo and  F. A.  Seifert: Am. Mineral., 67 (1982), 686.
• 34)   G. H.  Kim and  I.  Sohn: Metall. Mater. Trans. B, 45 (2014), 86.
• 35)   I.  Polyakova,  V.  Klyuev,  B.  Pevzner,  V.  Goncharuk,  O.  Yanush,  T.  Markova,  V.  Maksimov and  V.  Kabanov: Phys. Chem. Glasses B, 51 (2010), 52.
• 36)   L.  Cormier,  D.  Ghaleb,  D. R.  Neuville,  J.  Delaye and  G.  Calas: J. Non-Cryst. Solids, 332 (2003), 255.
• 37)   L.  Cormier,  D. R.  Neuville and  G.  Calas: J. Am. Ceram. Soc., 88 (2005), 2292.
• 38)   P.  McMillan and  B.  Piriou: J. Non-Cryst. Solids, 55 (1983), 221.
• 39)   L.  Hwa,  S.  Hwang and  L.  Liu: J. Non-Cryst. Solids, 238 (1998), 193.
• 40)   B. O.  Mysen and  J. D.  Frantz: Am. Mineral., 78 (1993), 699.
• 41)   K. C.  Mills,  L.  Yuan,  Z.  Li and  G.  Zhang: High Temp. High Press., 42 (2013), 3.
• 42)   Q.  Shu,  J.  Klug and  Q.  Li: ISIJ Int., 59 (2019), 1057.
• 43)   C.  Feng,  J.  Tang,  L.  Gao,  Z.  Liu and  M.  Chu: ISIJ Int., 59 (2019), 31.
• 44)   Z. J.  Wang,  Q. F.  Shu,  S.  Sridhar,  M.  Zhang,  M.  Guo and  Z. T.  Zhang: Metall. Mater. Trans. B, 46 (2015), 758.