Effect of complex nucleation agents on preparation and crystallization of CaO-MgO-Al2O3-Si

Journal of Non-Crystalline Solids 450 (2016) 6–11

Contents lists available at ScienceDirect

Journal of Non-Crystalline Solids journal homepage: www.elsevier.com/locate/jnoncrysol

Effect of complex nucleation agents on preparation and crystallization of CaO-MgO-Al2O3-SiO2 glass-ceramics for float process Weihong Zheng a,b, Jingjing Cui a,⁎, Li Sheng a, Hua Chao a, Zhigang Peng b, Chunhua Shen c a b c

State Key Laboratory of Silicate Materials for Architectures, Wuhan University of Technology, 122 Luoshi Road, 430070 Wuhan, Hubei, PR China Glass and Technology Research Institute of Shahe, 054100 Shahe, Hebei, PR China The Center for Materials Research and Analysis, Wuhan University of Technology, 430070 Wuhan, Hubei, PR China

a r t i c l e

i n f o

Article history: Received 28 March 2016 Received in revised form 19 July 2016 Accepted 22 July 2016 Available online 28 July 2016 Keywords: CaO-MgO-Al2O3-SiO2 Glass-ceramics Preparation Crystallization

a b s t r a c t CaO-MgO-Al2O3-SiO2 flat glass-ceramic produced by float process with excellent properties can be widely used in architectural decoration, electronics, environmental protection and other fields. The effects of complex nucleation agents (TiO2, CaF2 and P2O5) on melting and crystallization behaviors of glasses were investigated by DSC, XRD, SEM, FESEM, Raman spectroscopy and viscosity test. The results showed that the practical melting temperature and forming temperature of T specimen doped with 5 wt% TiO2 were about 1489 °C and 1286 °C, and those of TF specimen doped with 5 wt% TiO2 and 2 wt% CaF2 were decreased to about 1463 °C and 1260 °C, respectively. However, the practical melting temperature, forming temperature and Tp of TP and TFP specimens doped with 5 wt% TiO2 + 2 wt% P2O5 and 5 wt% TiO2 + 2 wt% CaF2 + 2 wt% P2O5 were shifted to higher temperature, which resulted in the difficulty of float process and crystallization. Diopside as the major crystalline phase was precipitated in four specimens with different nucleation agents. Furthermore, the crystallinity of TF specimen was increased with the addition of 2 wt% CaF2. Nevertheless, the addition of P2O5 resulted in the decrease of diopside content and the precipitation of calcium phosphate. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Glass-ceramic materials are polycrystalline materials produced through controlled crystallization of parent glass. Glass-ceramics obtained from glasses in the quaternary system CaO-MgO-Al2O3-SiO2 (CMAS) have remarkable mechanical properties together with excellent chemical resistance and have a wide variety of applications. In recent years, numerous applications require flat CMAS glass-ceramics, such as building materials and decorative materials [1–3]. CMAS flat glass-ceramics are produced by rolling process or sintering process [4–7]. Compared with rolling and sintering process, float process with high output, excellent quality and low energy consumption is the optimum technology for flat glass-ceramics [8]. Therefore, float process for CMAS glass-ceramics production is of great scientific and practical importance. In recent years, glass-ceramics produced by float process have caught the attention of researchers [9–11]. Compared with commercial soda-lime-silica glass (NCS glass), there are some technical difficulties needed to be resolved in float process of CMAS glass-ceramics [12–14]. Firstly, it will be difficult to realize float forming of CMAS glass-ceramics, which has the high practical melting temperature and viscosity. Secondly, it is hard to control the crystallization of CMAS ⁎ Corresponding author. E-mail address: [email protected] (J. Cui).

http://dx.doi.org/10.1016/j.jnoncrysol.2016.07.026 0022-3093/© 2016 Elsevier B.V. All rights reserved.

glass-ceramics. Therefore, it is necessary to add proper flux and nucleation agent into glass to decrease practical melting temperature and viscosity and to control crystallization. According to some literatures [15–19], the presence of TiO2 is highly effective as catalyst for initiating crystallization centers that induce volume crystallization. CaF2 as an effective flux is used to decrease viscosity and control the formation of CMAS glass-ceramics. Furthermore, CaF2 is a nucleation agent, which is used to suppress spontaneous crystallization and control crystallization. However, P2O5 added as flux induces phase-separation in CMAS glasses. In addition, the incorporation of P2O5 suppresses the surface crystallization and promoted bulk crystallization. The co-doping of CaF2 and P2O5 promotes the crystallization of CMAS glasses, and affects crystallization temperature and thermal expansion coefficient. However, these researches have not investigated the effects of complex nucleation agents (TiO2, CaF2 and P2O5) on float forming process of glasses. Therefore, we choose TiO2, CaF2 and P2O5 as nucleation agents to decrease practical melting temperature and viscosity and to improve crystallization. In this work, based on constant nucleation agent (TiO2), others nucleation agents such as CaF2 and P2O5 are added. The effect of dopants on viscosity and crystallization of glass-ceramics is investigated by differential scanning calorimeter (DSC), Raman spectroscopy, X-ray diffraction (XRD), scanning electron microscope (SEM) and field emission scanning electron microscope (FESEM). The aim of present work is to determine the proper dopants and to prepare CMAS glass-

W. Zheng et al. / Journal of Non-Crystalline Solids 450 (2016) 6–11

ceramics with low practical melting temperature and proper crystalline phase, which is suitable for float process.

7

3. Results 3.1. The viscosity of glasses

2. Experimental 2.1. Materials preparation The compositions of the parent glass were listed in Table 1. Reagent grades SiO2, MgO, CaCO3, Na2CO3, K2CO3, Al2O3, ZnO, TiO2, CaF2 and P2O5 were used as raw materials for glass preparation. Glass batches (200 g) were mixed thoroughly, and then melted in a platinum crucible at 1470–1520 °C for 3 h in an electric furnace. The melts was then poured into preheated steel molds and annealed at 600 °C for 30 min to relax internal stress. About 10 g of glass was quenched in water, dried in a dry box and milled in a milling machine for the DSC measurement. The shaped glass was cut to a size of (30 × 10 × 5 mm) for heat treatment and test. Heat-treatment schedule of glass was determined by the results of the DSC curves. Glass samples were heat-treated under various conditions for the studies of phase transition and crystallization. 2.2. Sample characterization The samples were placed in defined amount (about 200 g) into the THETA Rheotronic II to measure viscosity of glass above the softening point. The viscosity (η; in Pa·s) of glasses can be fitted to the Fulcher (VFT; Vogel-Fulcher-Tamman) equation: logη ¼ A þ B=ðT−T0 Þ

Glass viscosity is one of the key properties for melting, fining, processing optimization, glass forming and annealing process. Float process requires a very precise control of the viscosity throughout the glass melting and forming process in order to achieve high throughput and high yield of acceptable products. For float process, high forming temperature will lead to the acceleration of the tin diffusion rate, which is harmful to flat glass. The viscosity of the glass specimens is plotted as a function of temperature in Fig. 1. It is the natural behavior of glass for the decrease of viscosity as temperature increases; however, the decreasing rate can be changed by the addition of CaF2 and P2O5. Some characteristic temperature points [20,21] of CMAS glass and commercial soda-lime-silica glass (NCS glass) are shown in Table 2, in which characteristic temperature points of float NCS glass are measured by THETA Rheotronic II. For T glass, the practical melting temperature and forming temperature are about 1489 °C and 1286 °C. With the addition of 2 wt% CaF2, the practical melting temperature and forming temperature of TF specimen are decreased to about 1463 °C and 1260 °C, which are close to float NCS glass. Adding 2 wt% P2O5, characteristic temperatures of TP specimen are increased. In the case of simultaneous presence of CaF2 and P2O5, the viscosities and characteristic temperatures of TFP specimen are close to T specimen during the temperature region of 1100–1500 °C. It indicates that the addition of CaF2 results in the decrease of viscosity. On the contrary, the presence of P2O5 leads to the increase of viscosity. 3.2. The structure of glasses

wherein, T is the temperature (°C) and A, B and T0 are the adjustable constants. Raman spectra of T, TF and TFP glasses were excited with argon ion laser at wavelength of 514.5 nm and recorded at room temperature in back (180°) scattering configuration under a microscope by using a Raman Spectrometer (Type: Renishaw RM-1000). The resolution of the Raman spectra was 1 cm−1. Crystallization temperature of the glasses was studied by differential scanning calorimeter (DSC). DSC of quenched glass was carried out in a Netzsch DSC 404F3. After crushing quenched glass to about 100– 200 μm, non-isothermal experiments were performed by heating 30 mg samples in a Pt crucible with Al2O3 as the reference material in the temperature range between 20 and 1200 °C with a heating rate of 5–20 °C/min. X-ray diffraction (XRD) analysis (0.02°/2θ step, 2.5 s/step) was performed using a Rigaku Ultima IV X-ray Diffractometer (Tokyo, Japan) with Cu Kα radiation to evaluate the development of crystalline phase in heat-treated glass specimens. The samples for XRD analysis were the heat-treated samples without etching. The fine microstructure of the glasses was observed by a scanning electron microscope (SEM) with a JSM-5610LV and the field emission scanning electron microscope (FESEM) (ULTRA PLUS-43-13, Carl Zeiss AG). The heat-treated samples were etched with 5% HF solution at room temperature for 1 min before SEM and FESEM analysis. The samples were washed ultrasonically with ethanol for 5 min after etching. Etched glass samples were coated with a thin layer of carbon.

Structural characterization of the glassy samples was studied by Raman spectroscopy. According to the results of viscosity tests, the viscosities and characteristic temperatures of TFP specimen are close to T specimen during the temperature region of 1100–1500 °C, so only Raman spectra of T, TF, TP glasses are analyzed. Raman spectra of glasses at room temperature provide the most important configurational information on the structures. Fig. 2 shows deconvoluted Raman spectra at 800–1250 cm− 1 of CMAS glasses. The high frequency region of the Raman spectrum of glass is deconvoluted with Gaussian fitting method [22]. The band at around 800–1250 cm−1 range is assigned to Si\\O stretch vibration of Qn (n = 1, 2, 3, 4) tetrahedral units [23]. The band at 870 cm−1 is attributed to the Si\\O\\ stretching in tetrahedron with three non-bridging oxygen per silicon (Q1; Si2O67 − dimer). The band near 950 cm−1 is assigned to Si\\O\\ stretching in structural units with two non-bridging oxygen per silicon (Q2; SiO2− chain). The third 3

Table 1 Oxide composition (wt%) of CMAS glass-ceramics. Sample

CaO

MgO

Al2O3

SiO2

K2O

Na2O

ZnO2

TiO2

CaF2

P2O5

T TF TP TFP

9.0 9.0 9.0 9.0

6.0 6.0 6.0 6.0

21.0 21.0 21.0 21.0

49.0 49.0 49.0 49.0

4.0 4.0 4.0 4.0

4.0 4.0 4.0 4.0

2.0 2.0 2.0 2.0

5.0 5.0 5.0 5.0

0 2.0 0 2.0

0 0 2.0 2.0

Fig. 1. Viscosity-temperature curves of CMAS glasses.

8

W. Zheng et al. / Journal of Non-Crystalline Solids 450 (2016) 6–11

Table 2 Parameters for VFT equation and characteristic temperatures of forming process in tin bath for glass samples.

VFT equation parameter A B T0 Characteristic temperature (°C) (±5 °C) Practical melting (η = 10 Pa·s) Forming (η = 102 Pa·s) Polishing (η = 102.7 − 103.2 Pa·s)

T

TF

TP

TFP

Float NCS glass

−3.39383 4794.09931 398.0375

−3.39489 4806.98797 369.7705

−3.49798 5020.77046 398.0062

−3.15637 4538.65851 400.9342

– – –

1489 1286 1185–1125

1463 1260 1158–1098

1512 1309 1206–1145

1492 1281 1176–1115

1450 1200 1077–1007

Fig. 2. Deconvoluted Raman spectra at 800–1250 cm−1 of CMAS glasses, (a) T; (b) TF; (c) TP.

band at 1050 cm− 1 is due to the Si\\O\\ stretching in units with one non-bridging oxygen per silicon (Q3; Si2O2− sheet). The band around 5 1150 cm−1 results from the presence of fully polymerized units (Q4; SiO2 three-dimensional network). Frequencies (V), areas (A), and area % (A %) of Raman bands obtained from the deconvolution fits are shown in Table 3. Variation of the content of Qn (n = 1, 2, 3, 4) for TF and TP specimens is very different from the T specimen without CaF2 and P2O5. From Table 3, the content of Q1, Q2, Q3 and Q4 of T glass is 9.54%, 21.97%, 49.29%, and 19.20%, respectively, which shows that the network is mainly formed by SiO23 − chain, Si2O25 − sheet and SiO2 threedimensional network. Furthermore, in TF glass, the content of Q2 units is increased to 24.59%; however, the number of Q4 units is decreased to 15.26%. The content of Q3 in TP specimen doped with 2 wt% P2O5 is increased to 53.76%, and the number of Q4 units is decreased to 17.81%. 3.3. Influence on crystallization Fig. 3 shows the DSC curves of the CMAS glasses. All curves present a small endothermic peak and a more intense exothermic peak. The first one is related to the glass transition and Tg is the glass transition temperature. The second one is due to the crystallization of the sample and the corresponding peak occurs at temperature Tp. Table 4 presents

these temperatures for the glasses as well as the crystallization onset temperatures, Tx. With addition of CaF2 (TF glass), one observes a decrease in the Tg and Tp values. The addition of P2O5 to the T glass (TP glass) results in a broadening of the whole crystallization peak and an

Table 3 Frequencies (V), area (A), and area% (A %) of Raman bands obtained from the deconvolution fits.

V1 V2 V3 V4 A1 A2 A3 A4 A1% A2% A3% A4%

T

TF

TP

865 953 1056 1150 8473 19,517 43,786 17,057 9.54 21.97 49.29 19.20

875 959 1059 1155 11,343 27,065 54,850 16,788 10.31 24.59 49.84 15.26

866 951 1057 1156 8494 20,886 50,935 16,228 8.80 21.63 53.76 17.81

Here, V1, A1 and A1% correspond of the frequency, the area and the ratio of the area for the Q1 species. This ratio of the area represents the content of the Qn species.

W. Zheng et al. / Journal of Non-Crystalline Solids 450 (2016) 6–11

Fig. 5. XRD patterns of samples treated at 700 °C for 30 min.

Fig. 3. DSC curves of CMAS initial glasses at 10 °C/min.

Table 4 Tg, Tx and Tp temperatures (°C) of all specimens at 10 °C/min. (±2 °C).

T TF TP TFP

9

Tg

Tx

Tp

698 668 701 671

861 854 872 850

880 875 898 878

increase of its Tp. In TFP glass, the simultaneous presence of CaF2 and P2O5 results in a Tp which is lower than that for the TP glass. Fig. 4 shows FESEM images of CMAS glasses treated at 700 °C for 30 min. According to some literatures [24–26], the temperature at which phase separation starts to occur is thought to be at about Tg temperature, and from DSC curves, it can be seen that Tg of all glass samples is approximately 700 °C. Therefore, so all glass samples are treated at 700 °C for 30 min to study phase separation. From FESEM images, it can be observed that in T glass, there are some 150 nm scaled droplets in the amorphous matrix. Adding 2 wt% P2O5 to T glass, the droplets

size is decreased to 100 nm. Adding 2 wt% CaF2 and 2 wt% P2O5 to T glass, the droplets with 200 nm in diameter are loosely dispersed in the matrix. However, adding 2 wt% CaF2 to T glass, it is very interesting to note that the nucleated particles with the size of 200 nm develop from the continuous matrix instead of the droplet phase. Fig. 5 shows the XRD patterns of glass samples treated at 700 °C for 30 min. The XRD pattern of TF sample shows there are tiny diffraction peaks corresponding to diopside crystal phase appearing near 30° and 35°. However, T, TP and TFP samples only have a broad scattering peak, which suggests there is no crystal precipitated in glass matrix. Fig. 6 shows the XRD patterns of glass samples treated at 730 °C for 30 min and at 930 °C for 30 min. It can be seen that T, TF, TP and TFP samples have the same major crystalline phase (diopside, Ca(Mg,Al)(Si,Al)2O6, JCPDS-PDF 41-1370). It demonstrates that the major crystalline phase is not sensitive to the addition of CaF2 and P2O5. It can be also observed that with 2 wt% CaF2, diopside as the major crystal phase is precipitated, and fluorpargasite phase ((Na,Ca)Mg3(Mg,Al)2 (Ca,Mg)2(Si,Al)8O22F2, JCPDS-PDF 86-0896) and pargasite phase (NaCa2Mg4Al(Si6Al2)O22(OH)2, JCPDS-PDF 41-1430) are

Fig. 4. FESEM images of CMAS glass ceramics treated at 700 °C for 30 min, (a) T; (b) TF; (c) TP; (d) TFP.

10

W. Zheng et al. / Journal of Non-Crystalline Solids 450 (2016) 6–11

4. Discussion

Fig. 6. XRD patterns of samples nucleated at 730 °C for 30 min followed by crystallization at 930 °C for 30 min.

also observed. With 2 wt% P2O5, the content of diopside phase decreases markedly and calcium phosphate (Ca3 (PO4)2, JCPDS-PDF 20-0359) and anorthite (CaAl2Si2O8, JCPDS-PDF 41-1486) are precipitated. With 2 wt% CaF2 and 2 wt% P2O5, compared with TF sample, the content of pargasite and fluorpargasite is increased and diopside is decreased. Meanwhile, calcium phosphate as minor crystal phase is precipitated. Fig. 7 shows the SEM results of CMAS glass-ceramics nucleated at 730 °C for 30 min followed by crystallization at 930 °C for 30 min. It can be seen that the shape and size of crystals can be affected by CaF2 and P2O5. According to the image of T sample, the granular crystals irregularly arrange with low crystallinity, and the crystal size is about 1 μm. With the addition of 2 wt% CaF2 (TF), a large number of grains are precipitated in TF specimen with high crystallinity, and the mean crystal size is decreased to 150 nm. Flake-like crystals with the size of 0.8 μm are precipitated in TP specimens doped with 2 wt% P2O5. With the addition of 2 wt% CaF2 and 2 wt% P2O5, disk-like crystals with the size of 0.4 μm are intertwined together in TFP specimen. Furthermore, small needle-shaped crystals are distributed between disk-like crystals.

The importance of viscosity as a property of molten glasses has long been recognized. For practical purposes, a number of important stages in the variation of viscosity with temperature have been named; these names serve to label the temperatures at which particular glassmaking operations can be carried out. The characteristic temperatures of CMAS glass and float NCS glass are shown in Table 2. It can be seen that the addition of CaF2 results in the decrease of viscosity and the presence of P2O5 leads to an increase of viscosity. It is well-known that the viscosity of the glass at a certain temperature depends on the degree of polymerization of glass network, i.e. they are positively correlated. According to the results of Raman spectroscopy (seen in Fig. 2 and Table 3), it can be seen that with addition of CaF2, the number of Q4 units is decreased, which suggests SiO2 three-dimensional network is broken. It is known that Si4+ has high bond energy and covalency, which will strengthen the polymerization of glass network. However, F− acts as network breakers, weakening glass structure. The bridging oxygen can be replaced by F− due to 2− their similarity in radius (R− F = 0.133 nm; RO = 0.140 nm) [27]. Consequently, the strong `Si\\O\\Si` linkage is replaced by a pair of the weak `Si\\F, which leads to the weakening of the polymerization of glass network and the decreasing of viscosity. With the addition of P2O5, the number of Q3 units is increased, which indicates the network structure is strengthened. It can be related to the high bond energy of P5 +. P5 + has higher bond energy and covalency than Si4 +, which makes it easier to tighten glass network, as a result, increasing the viscosity. The broken [SiO4] caused by F− can not only decrease viscosity but also make Tg and Tp shifts to a lower temperature region. When the network structure is broken, the mobility and diffusion of the different ions during the crystallization process are markedly increased, leading to higher crystallizability [14]. In addition, the decrease of Tg and Tp is also related to phase separation. Fluorides are known to be immiscible in silicate melts, two-phase separation comprising numerous droplets of one phase dispersed in another. This type of phase separation decreases the energy barriers necessary for crystallization. Meanwhile, from FESEM images, it can be seen that the addition of CaF2 also decreases nucleation temperature.

Fig. 7. SEM images of CMAS glass-ceramics nucleated at 730 °C for 30 min followed by crystallization at 930 °C for 30 min, (a) T; (b) TF; (c) TP; (d) TFP.

W. Zheng et al. / Journal of Non-Crystalline Solids 450 (2016) 6–11

The FESEM images of T, TP, TFP glasses treated at 700 °C for 30 min show phase separation into droplet-like zones and SiO2-rich glass matrix, and these droplets are the starting point of nucleation. According to some literatures [28–30], the droplets in T glass are rich in [TiO4], and in TP and TFP glasses, the droplets are rich in [TiO4] and [PO4]. However, in TF glass, the nucleated particles develop from the continuous matrix instead of the droplet phase, and XRD pattern of TF shows that a small amount of diopside crystals are precipitated, which illustrates that CaF2 can decrease nucleation temperature and promotes crystallization. Nevertheless, the introduction of P2O5 increases Tp and delays crystallization according to Tulyaganov and Marques researches [31,32]. The addition of CaF2 and P2O5 can also affect the nucleation and crystallization of specimens. XRD patterns and SEM images of specimens treated at 730 °C for 30 min and at 930 °C for 30 min shows that the crystallinity is increased and crystal size is decreased with the addition of CaF2. According to some literatures [33,34], at the nucleation stage, there are some small CaF2 nuclei precipitated in the glass matrix as the crystal nucleation centers, consequently, decreasing the energy barriers necessary for nucleation, which promotes the forming of numerous nuclei. The interaction between nuclei restrains the growth of crystals, resulting in a large number of tiny crystals. With the addition of P2O5, the content of diopside phase is decreased. Furthermore, calcium phosphate and anorthite as minor phases are precipitated. The formation of calcium phosphate is due to the crystallization of the separate orthophosphate groups that are charge balanced by Ca2+. As a result, the content of Ca2+ in specimens is decreased, leading to the decrease of diopside phase. 5. Conclusions In the present work, CaO-MgO-Al2O3-SiO2 (CMAS) glass-ceramics doped with TiO2, CaF2 and P2O5 as composite nucleation agents were prepared by melting process. Based on the results reported in the investigation, the conclusions can be drawn as follows: (1) The addition of CaF2 results in the decrease of viscosity. On the contrary, the presence of P2O5 and CaF2 + P2O5 leads to the increase of viscosity. Furthermore, Tg and Tp of TF specimen doped with 2 wt% CaF2 are decreased. However, the addition of P2O5 and CaF2 + P2O5 leads to a shift of Tp to higher temperature region and a widening of crystallization peak. (2) The practical melting temperature and forming temperature of TF specimen doped with 2 wt% CaF2 are 1463 ± 5 °C and 1260 ± 5 °C, respectively, which is suitable for float process. (3) Diopside as major crystalline phase is precipitated in T, TF, TP and TFP specimens, which is not sensitive to the dopants of CaF2, P2O5 and CaF2 + P2O5. Furthermore, the addition of CaF2 leads to the increase of crystallinity and the decrease of crystal sizes. However, the content of diopside phase is decreased and calcium phosphate as minor phase is precipitated with the addition of P2O5.

Acknowledgement This work is supported by the National Natural Science Foundation of China (No. 50802065, 51202172) and the Fundamental Research Funds for State Key Laboratory of building materials silicate (Wuhan University of Technology) (No·SYSJJ2015). References [1] W.H. Zheng, H. Cao, J.B. Zhong, S.Y. Qian, Z.G. Peng, C.H. Shen, CaO-MgO-Al2O3-SiO2 glass-ceramics from lithium porcelain clay tailings for new building materials, J. Non-Cryst. Solids 409 (2015) 27–33. [2] I.V. Vladimir, N.S. Svetlana, Granulated foam glass-ceramic material from zeolitic rocks, Constr. Build. Mater. 22 (2008) 999–1003.

11

[3] B.E. Yekta, P. Alizadeh, L. Rezazadeh, Floor tile glass-ceramic glaze for improvement of glaze surface properties, J. Eur. Ceram. Soc. 26 (2006) 3809–3812. [4] E.Y. Guseva, M.N. Gulyukin, Deformation resistance of glass ceramic material, Glas. Ceram. 58 (2001) 17–20. [5] H.Y. Liu, H.X. Lu, D.L. Chen, H.L. Wang, H.L. Xu, R. Zhang, Preparation and properties of glass-ceramics derived from blast-furnace slag by a ceramic-sintering process, Ceram. Int. 35 (2009) 3181–3184. [6] W.Y. Zhang, H. Gao, Y. Xu, Sintering and reactive crystal growth of diopside-albite glass-ceramics from waste glass, J. Eur. Ceram. Soc. 31 (2011) 1669–1675. [7] V.M.F. Marques, D.U. Tulyaganova, S. Agathopoulos, V.K. Gataullin, G.P. Kothiyal, J.M.F. Ferreira, Low temperature synthesis of anorthite based glass-ceramics via sintering and crystallization of glass-powder compacts, J. Eur. Ceram. Soc. 26 (2006) 2503–2510. [8] L.B.A. Pilkington, Review lecture. The float glass process, Proc. R. Soc. Lond. A Math. Phys. Sci. 314 (1969) 1–25. [9] J.P. Wan, J.S. Cheng, P. Lu, Study on float technics of borosilicate glass and silicate glass, Bull. Chin. Ceram. Soc. 26 (2007) 1197–1200. [10] S. Takeda, R. Akiyama, H. Hosono, Precipitation of nanometer-sized SnO2 crystals and Sn depth profile in heat-treated float glass, J. Non-Cryst. Solids 311 (2002) 273–280. [11] S. Cava, T. Sequinel, S.M. Tebcherani, S.R. Lazaro, S.A. Pianaro, J.A. Varela, Effect of temperature on glass-ceramic films prepared by impregnation of commercial float glass surfaces with oxide powders under pressure, Thin Solid Films 518 (2010) 5889–5891. [12] C.S.J. Shaw, Effects of melt viscosity and silica activity on the rate and mechanism of quartz dissolution in melts of the CMAS and CAS systems, Contrib. Mineral. Petrol. 151 (2006) 665–680. [13] M. Calvo-Dahlborg, J.M. Ruppert, E.D. Tabachnikova, V.Z. Bengus, U. Dahlborg, F. Haussler, Influence of the heat treatment of the melt on the structure and mechanical behaviour of metallic glass ribbons, J. Phys. IV 11 (2001) 41–49. [14] S. Jang, S. Kang, Influence of MgO/CaO ratio on the properties of MgO-CaO-Al2O3SiO2 glass-ceramics for LED packages, Ceram. Int. 38 (Supplement 1S) (2012) 543–546. [15] G.A. Khater, Influence of Cr2O3, LiF, CaF2 and TiO2 nucleants on the crystallization behavior and microstructure of glass-ceramics based on blast-furnace slag, Ceram. Int. 37 (2011) 2193–2199. [16] D.S. Brauer, M.N. Anjum, M. Mneimne, R.M. Wilson, H. Doweidar, R.G. Hill, Fluoridecontaining bioactive glass-ceramics, J. Non-Cryst. Solids 358 (2012) 1438–1442. [17] Q.B. Tian, X.H. Wang, L. Xu, X.J. Lin, H. Gao, Glass constitution and crystallization characteristics of CaO-MgO-Al2O3-SiO2 glass-ceramic with P2O5 and F additions, Rare Metal Mater. Eng. 36 (2007) 316–318. [18] S.N. Salama, E.A. Saad, H. Darwish, H.A. Abo-Mosallam, Formation of glass-ceramic materials based on pyroxene solid solution-fluorapatite phases and their thermal expansion properties, Ceram. Int. 31 (2005) 559–566. [19] X.Z. Guo, X.B. Cai, J. Song, G.Y. Yang, H. Yang, Crystallization and microstructure of CaO-MgO-Al2O3-SiO2 glass-ceramics containing complex nucleation agents, J. NonCryst. Solids 405 (2014) 63–67. [20] E. Le Bourhis, Glass: Mechanics and Technology [M], France, 2007 53–106. [21] L.D. Pye, I. Joseph, A. Montenero, Properties of Glass-forming Melts [M], USA, 2005 75–138. [22] D.R. Neuville, Viscosity, structure and mixing in (Ca, Na) silicate melts, Chem. Geol. 229 (2006) 28–41. [23] D.R. Neuville, Advances in Raman spectroscopy applied to earth and material sciences, Rev. Mineral. Geochem. 78 (2014) 509–541. [24] R.H. Doremus, Glass Science [M], USA, 1994 48–70. [25] P.F. James, Review: liquid-phase separation in glass-forming systems, J. Mater. Sci. 10 (1975) 1802–1825. [26] D.U. Tulyaganov, S. Agathopoulos, J.M. Ventura, M.A. Karakassides, O. Fabrichnaya, J.M.F. Ferreira, Synthesis of glass-ceramics in the CaO-MgO-SiO2 system with B2O3, P2O5, Na2O and CaF2 additives, J. Eur. Ceram. Soc. 26 (2006) 1463–1471. [27] D.R. Lide, Handbook of Chemistry and Physics [M], USA, 2001–2002 12–16. [28] W. Höland, V. Rheinberger, M. Frank, Mechanisms of nucleation and controlled crystallization of needle-like apatite in glass-ceramics of the SiO2-Al2O3-K2O-CaO-P2O5 system, J. Non-Cryst. Solids 253 (1999) 170–177. [29] M. Mirsaneh, I.M. Reaney, P.V. Hatton, S. Bhakta, P.F. James, Effect of P2O5 on the early stage crystallization of K-fluorrichterite glass–ceramics, J. Non-Cryst. Solids 354 (2008) 3362–3368. [30] H.R. Fernandes, D.U. Tulyaganov, J.M.F. Ferreira, The role of P2O5, TiO2 and ZrO2 as nucleating agents on microstructure and crystallization behaviour of lithium disilicate-based glass, J. Mater. Sci. 48 (2013) 765–773. [31] D.U. Tulyaganov, S. Agathopoulos, H.R. Fernandes, J.M.F. Ferreira, Processing of glassceramics in the SiO2-Al2O3-B2O3-MgO-CaO-Na2O-(P2O5)-F system via sintering and crystallization of glass powder compacts, Ceram. Int. 32 (2006) 195–200. [32] V.M.F. Marques, D.U. Tulyaganov, G.P. Kothiyal, J.M.F. Ferreira, The effect of TiO2 and P2O5 on densification behavior and properties of Anortite-Diopside glass-ceramic substrates, J. Electroceram. 25 (2010) 38–44. [33] M. Poulain, Overview of crystallization in fluoride glasses, J. Non-Cryst. Solids 140 (1992) 1–9. [34] T. Hamasaki, K. Eguchi, Y. Koyanagi, A. Matsumoto, T. Utsunomiya, K. Koba, Preparation and characterization of machinable mica glass-ceramics by sol-gel process, J. Am. Ceram. Soc. 71 (1988) 1120–1124.