WorksPhysics and Astronomy
Oxyfluoride glasses obtained through incorporation of CaF into photovoltaic cover glass melts
Abstract
The glass industry has limited options to mitigate its environmental footprint, and the demand for cover glass to produce photovoltaic panels is increasing. Currently, the majority of this special type of glass is not being recycled, and in this work, we propose to reuse it as raw material to obtain oxyfluoride glasses. The incorporation of CaF and the increasing NaCO content resulted in a melting temperature of about 1200C, significantly lower than in soda-lime glasses, which adds up to the environmental benefits of reusing end-of-life cover glass. The obtained samples show high transparency and thermal stability, allowing the cover glass to make up to 80% of its weight. XRF analysis was employed to determine the elemental composition of the samples, while XRD and Raman indicated that by adding CaF, the glass network was depolymerized. In situ XRD as a function of temperature showed the formation of a few crystalline phases in these oxyfluoride samples, evidencing that it can be explored as a matrix to obtain different glass-ceramics. The combination of the glass properties indicates that this method and the resulting material can contribute to reducing the environmental impact of the glass industry, by creating new glass or glass-ceramic materials that can be obtained at a reduced temperature compared to the soda-lime glass, while cover glass being the primary raw material could reduce the need to extract minerals from nature.
Graphical Abstract
Oxyfluoride glasses obtained through incorporation of into photovoltaic cover glass melts
Rafaela Valcarenghi; Brenno Silva Greatti; Robson Ferrari Muniz; Vitor Santaella Zanuto; Anna Paulla Simon; Ricardo Schneider; Raquel Dosciatti Bini; Márcio Antônio Fiori; Maxence Vigier; Emmanuel Veron; Mathieu Allix; Marcelo Sandrini; Marcos Paulo Belançon
Highlights
Glasses containing up to 80% cover glass content were obtained
Analyses have shown that introducing depolymerizes the material
These oxyfluoride glasses can be crystallized in several different crystalline phases
The samples exhibit transparency and thermal stability suitable for optical applications
reduce the need to extract minerals from nature.
keywords: glass waste management; glass recycling; waste reduction; glass-ceramics; silicate
Introduction
The world consumes over 130 million tons of glass annually, or about 16 kg per capita. Container and flat glass account for about 80% of global glass production and are estimated to release more than 60 million tons of carbon dioxide into the atmosphere, primarily due to intensive heating of raw materials [1, 2]. Soda-lime is the most common glass, and the industry has limited options to mitigate its environmental footprint. To produce 1 kg of soda-lime glass, the carbonates present in the raw materials will emit around 0.15 kg of CO, while the huge amount of energy required to heat and melt (C) the glass results in another 0.45 kg, primarily due to the combustion of fossil fuels [1, 3]. Although clean electricity and hydrogen can be employed, these alternatives have yet to be demonstrated on an industrial scale [2, 4].
Though glass can be recycled indefinitely, doing so is often not feasible or attractive due to technical and economic reasons [5, 6]. As discussed in our previous work [7], highly transparent soda-lime is used as cover glasses (CG) in solar energy applications, and these high-quality materials are facing the risk of being dumped in landfills once their demand continues to rise, while the options to reuse them are pretty limited. Additionally, glass production requires vast amounts of raw materials and mining is one of the most energy-intensive industries [8], resulting in a severe environmental impact. These factors highlight the importance of recycling and reusing CG, turning end-of-life products into raw materials to keep them on the market, in a context of circular economy [9, 10, 11].
Glasses other than soda-lime have been proposed as alternatives for solar energy applications, which could include new features such as spectral conversion to enhance silicon solar panel technology [7]. However, among dozens of proposed glass systems, silicates are the only ones based on abundant, affordable, and non-toxic minerals. The resulting glasses often exhibit high transparency and adequate chemical and mechanical resistance for solar energy applications. Reviewing the literature recently [7, 12], we concluded that modified soda-lime silicates are the most promising system that could meet the requirements needed for mass-scale industrial production, while providing new or enhanced properties that could reduce industry energy consumption, expand solar power production, and contribute significantly to sustainability goals.
An interesting silicate system, which we have already investigated, is achieved when is added into silica melts [13, 14]. While modifiers such as and are well-known for reducing the melting temperature of silicates, the F provided by often replaces O, affecting the silica network [15]. Several studies have shown that incorporation reduces the melting temperature and viscosity of silicates [15], while it can improve the microstructure [16] and promote or inhibit the formation of crystalline phases [17, 18], including nanocrystals [19].
To address all the questions mentioned above, this work proposes using CGs from end-of-life solar panels as raw materials to produce oxyfluoride sodium calcium silicate glasses, achieved by incorporating . Our methodology is simple; the obtained samples have high-optical quality and allow a recycled fraction of raw materials as high as 80% of the final weight of the glass samples. Using CG as a raw material may contribute to the sustainability of the glass industry and positively impact the environment by reducing energy demand and the extraction of new raw materials from nature, ultimately reducing overall emissions and environmental footprint. Additionally, as we will show, these CG-derived materials may provide a platform to research and development of new glasses and glass-ceramics for several applications.
Materials and methods
The weight of the compounds mixed to produce each sample is shown in Table 1, where the sample names refer to the mol% of . The CG, sourced from photovoltaic panels previously studied [20], was fragmented into particles approximately 5–10 mm in length. The reagents added to the mix were and , both from Sigma-Aldrich with 99.0% purity. To keep the ratio approximately constant through all the samples, concentration was adjusted accordingly.
| Sample | Cover-glass (g) | CaF (g) | NaCO (g) | Total (g) |
| CgCAF05 | 13.00 | 1.00 | 1.94 | 15.94 |
| CgCAF12 | 12.25 | 2.45 | 1.83 | 16.53 |
| CgCAF15 | 11.80 | 3.00 | 1.76 | 16.56 |
| CgCAF18 | 11.30 | 3.50 | 1.69 | 16.49 |
| CgCAF20 | 10.90 | 4.00 | 1.63 | 16.53 |
Table 1. Mass of each sample prepared in this work
Reagents were mixed with CG in a platinum crucible and heated in a furnace, using a 10 °C/min heating rate, with holding steps of 40 minutes at 200°C, 60 minutes at 500°C, 90 minutes at 800°C, and 120 minutes at 1200°C. This thermal profile was chosen to ensure a homogeneous fusion. of the mixture and to avoid foaming or spillage during melting. After this stage, the molten glass was poured into a stainless steel mold and rapidly transferred to a second furnace maintained at , where it was kept for approximately 12 hours to relieve internal stresses. Upon cooling to room temperature, the samples were removed from the mold. Figure 1 shows a photo of the obtained samples, which were cut, polished, or milled for the analyses.
Figure 1. Samples investigated in this work, as prepared for analyses. The diameter of the discs is about 1.0 cm.
The analytical techniques employed in this study encompass compositional, thermal, structural, and optical analyses. X-ray fluorescence (XRF) was used to determine the chemical composition of the samples. The analysis was performed on fused beads at approximately (Panalytical – Axios). The beads were prepared from powdered material, previously ground to pass through a 325-mesh sieve, using lithium tetraborate as the flux. Additionally, loss-on-ignition analysis was carried out to account for volatile content up to . Differential thermal analysis (DTA) was performed in the temperature range of to , using a heating rate of /min under a synthetic air atmosphere with a flow rate of 50 mL/min (TA Instruments – SDT Q600). The density of the samples was obtained using Archimedes’ principle with distilled water at room temperature. Mass measurements were performed with an analytical balance with a precision of 0.1 mg (Kern ABT - 120-5DM). Structural characterization was performed by X-ray powder diffraction (XRPD) on powdered samples. The analysis used CuK radiation ( Å) at 40 kV and 15 mA, with a scanning rate of 5°/min, step size of 0.02°, and a range of 5 to 80° (Rigaku - Miniflex 600). Also, in situ XRPD analysis was performed on powdered material to evaluate the temperature-dependent structural evolution. The measurements were conducted under air in the Bragg–Brentano – geometry, using Ni-filtered CuK radiation ( Å), over a range of 15 to 60°, with a step size of 0.016° and a counting time of 0.5 s/step. A temperature step of 25°C was applied from 500 to 950°C. Data were collected on a Bruker-AXS D8 Advance diffractometer (Karlsruhe, Germany) equipped with an Anton Paar oven chamber (model HTK 1200N, Graz, Austria), which allows measurements up to 1473 K. The powdered samples were placed in a platinum-lined corundum sample holder, and the diffracted intensities were recorded using a LynxEye detector.
Optical characterization involved Raman spectroscopy, UV-Vis-NIR and spectroscopic ellipsometry analysis. Raman spectra were collected on a Bruker-Senterra Raman microscope using a 532 nm laser at 20 mW, in the spectral range of 50 to 1542 cm. A ×20 objective lens was used for laser focusing, and the signal was integrated for 2 seconds. UV-Vis-NIR transmittance was measured in the spectral range of 190 to 3600 nm (Shimadzu - UV-3600i Plus), and the refractive index of the samples was determined by spectroscopic ellipsometry at an incidence angle of 55° (HORIBA - UVISEL Plus).
Results and discussion
Although the CG is expected to be low iron soda lime glass, and the reagents added are quite pure, the samples were submitted to XRF analysis to confirm their composition. These results are shown in Table 2.
| Compound | CG | CgCAF05 | CgCAF12 | CgCAF15 | CgCAF18 | CgCAF20 |
| SiO | 72.62 | 62.90 | 58.22 | 56.62 | 54.18 | 52.51 |
| CaO | 9.35 | 7.63 | 7.59 | 7.35 | 7.11 | 6.88 |
| NaO | 12.48 | 17.62 | 16.12 | 15.28 | 15.19 | 15.15 |
| MgO | 4.46 | 3.81 | 3.55 | 3.44 | 3.29 | 3.19 |
| AlO | 0.48 | 0.37 | 0.36 | 0.33 | 0.31 | 0.32 |
| KO | 0.32 | 0.26 | 0.26 | 0.22 | 0.24 | 0.24 |
| CaF* | - | 5.20 | 12.40 | 15.25 | 17.98 | 20.61 |
Table 2. Composition obtained by XRF expressed as a percentage (mol%). *CaF concentration was estimated assuming that Ca concentration obtained by XRF is the sum of CaO (from CG) plus CaF (added).
Assuming the CG composition obtained from the XRF analysis and the weights shown in Table 1, we can calculate the stoichiometric relations for CgCAF samples, which are very similar to those obtained experimentally by XRF. In this way, though fluorine is well known to be volatile, these results corroborate the assumption that no significant mass was lost during the melting process. As previously stated, we also aimed to keep the ratio SiO/NaO constant, and the XRF analysis confirms this it lies in the range of ∼3.5 - 3.7 for the samples.
Figure 2 presents the XRPD patterns for all glass samples. The absence of sharp peaks in all patterns confirms the amorphous nature of the samples, consistent with a glassy structural framework. These results are in agreement with previously reported diffractograms for glasses and glass-ceramics [13, 21, 22]. Notably, the diffraction halos exhibit two overlapping broad bands, which can be more clearly distinguished by fitting the patterns with two Lorentzian curves, as shown in 2, represented by the filled green and blue curves corresponding to bands centered at approximately and , respectively. The area ratio between the green and blue bands decreases from 1.16 for the CG sample to 0.69 to the CgCAF20 sample. In the CG sample, the most intense band is centered around , whereas in samples with increasing CaF content, the band at becomes more prominent. This behavior is likely associated with structural rearrangements driven by the incorporation of fluoride ions. As reported by Iwamoto and Makino [23], fluoride ions in calcium fluorosilicate systems can engage in Si-F or Ca-F bonding. At lower CaF concentrations, the formation of Si-F bonds is favored, whereas at higher concentrations, Ca-F interactions become more prominent. This transition is accompanied by a change in the fluoride coordination environment, from four-fold coordination, characteristic of a CaF-type quasi-lattice, to six-fold coordination typical of a NaCl-type quasi-lattice. Notably, the NaCl-type structure possesses smaller average lattice parameters than the CaF-type configuration. According to Bragg’s law, a reduction in the lattice spacing results in a shift of diffraction features towards higher angles. Nonetheless, confirmation of this hypothesis requires more comprehensive structural analyses, including advanced modeling approaches.
Figure 2. X-ray powder diffraction patterns.
As demonstrated in Table 2, CgCAF samples have a significantly lower proportion of compared to the CG as concentration increases. Figure 3, presents the mass density and refractive index at 628 nm as a function of concentration. The mass density of the samples increases linearly with the addition of , with an experimental uncertainty of g/cm, which renders the associated error bars negligible in the graphical representation. The linear increase in density observed with increasing content in the multicomponent soda-lime glass matrix arises from the combined effects of compositional modifications and structural rearrangements. Specifically, the gradual replacement of the lighter network-forming, by , along with concurrent variations in the concentrations of other constituents such as , , and , introduces heavier cations (e.g., , ) and larger anions () into the glass structure. These changes lead to an increase in the average molar mass and promote a partial depolymerization of the silicate network, thereby enhancing atomic packing efficiency. The resulting structural compaction is corroborated by the shift of the amorphous halo to higher values in the XRPD patterns, indicating a reduction in the mean interatomic distances. At the same time, the refractive index remained constant (), within the error margins. Muniz et al. [13] have investigated sodium calcium silicate glasses containing amounts of CaF₂ similar to those in the present work, showing about the same mass density and a smaller refractive index (1.47).
Figure 3. Mass density and refractive index at 628 nm of the samples as a function of CaF₂ concentration. Error bars are negligible for the density in this range.
Figure 4 presents the thermograms of all samples, including CG, where the key events (glass transition), (onset of crystallization), and (crystallization peak) are also indicated. A summary of the characteristic temperatures identified for all samples is shown in Table 3. However, in these experiments, could not be identified for the CG, CgCAF05, and CgCAF12 samples. As can be seen, in all CgCAF samples, it is possible to
Figure 4. Differential thermal analysis of CG and CgCAF samples.
| Sample | (°C) | (°C) | (°C) | (°C) | (°C) |
| CG | 560 | - | - | - | - |
| CgCAF05 | 491 | - | 636 | 812 | - |
| CgCAF12 | 507 | - | 637 | 751 | - |
| CgCAF15 | 475 | 636 | 664 | 695 | 161 |
| CgCAF18 | 469 | 644 | 673 | 697 | 175 |
| CgCAF20 | 491 | 626 | 646 | 694 | 135 |
Table 3. Thermal parameters of glass samples.
identify two crystallization peaks, a result widely observed in the literature on silicates containing CaF [13, 16], as well as a significant reduction in the glass transition temperature (50–90°C). It is essential to note the satisfactory glass stability (-) observed in these samples, suggesting a suitable working temperature that may enable these materials to be produced in different shapes and sizes. Although could not be identified.
An in situ high-temperature XRPD study was performed on the CgCAF12 sample to investigate its crystallization characteristics under thermal treatment. The corresponding diffraction patterns are shown in Figure 5.
Figure 5. In situ XRPD performed on the CgCAF12 samples.
Consistent with DTA measurements, a crystalline phase appears at 600°C. Nevertheless, we cannot identify this phase using the ICDD PDF5 database. Additional work will be necessary to elucidate this point. The proportion of this unknown phase increases until 800°C and after decreases but is present in small amounts at 950°C. At 650°C we observe the crystallization of C3S ( JCPDS file 00-017-0445), which disappears at 800°C. The Combeite phase ( JCPDS file 04-008-0810) crystallizes around 700°C, and its proportion increases until 950°C to be finally the main phase at high temperature. After this experiment, the PXRD diagram collected at room temperature shows the presence of Combeite and a minimal amount of Akermanite (JCPDS file 01-079-2425). Some of these phases have been linked to intriguing features, such as antimicrobial activity [24]. This indicates that CG-derived materials may also result in glass-ceramics with unique properties, which we will explore in future work.
Figure 6 shows the Raman spectra recorded for all samples, along with deconvolution using a multipeak Gaussian fit applied to the CgCAF05 sample (including the individual Gaussian components) as a representative example. Each Raman spectrum was fitted using 11 Gaussian components labeled (a to k) in the figure. This number was chosen based on the stability of the fitting process, as additional components did not significantly improve the quality of the fit. A weak band (g), located at approximately 870 , could not be fitted for the CgCAF05 sample. The Raman scattering of the studied glasses is dominated by bands in the 550–750 range (c, d and e) and in the 900–1200 range (h, i, j and k). Raman spectroscopy has proven to be a powerful technique for the structural investigation of silicate glasses [25, 26, 27], although multiple overlapping bands pose a significant challenge for complete spectral interpretation.
Figure 6. Raman spectra of CgCAF samples.
The bands within the 900–1200 range (h, i, j and k) are attributed to symmetric stretching vibrations of Si–O bonds in various tetrahedral units, commonly referred to as bands (where to 4), with $n denoting the number of bridging oxygens per tetrahedron [28, 27]. It is well established that the band near , associated with the - vibration involving bridging oxygens, shifts to lower frequencies - or new bands emerge in lower frequency regions - due to the disruption of -- linkages [29]. Furthermore, the primary determinant of Raman band frequency within this range is the force constant of the - (non-bridging oxygen) bond. In silicate glasses containing , the substitution of oxygen atoms by fluorine distorts the silicon’s local electronic environment due to the higher electronegativity of fluorine. This distortion weakens the remaining - bonds within the tetrahedra, leading to lower bond force constants and, consequently, lower associated vibrational frequencies [30]. This effect explains the slight “k” band shift with increasing concentration; this behavior was also observed by Muniz et al. [31] for a similar glass composition.
On the other hand, the - bond behaves similarly to the - bond, and the stretching vibration of the Si-F bond in tetrahedra typically appears near 950 cm in fluorine-containing silica glasses [30, 32]. Therefore, the formation of Si-F bonds may influence the frequencies of Si-O related vibrations and could account for the systematic increase in intensity observed for band “h” (960 cm) as content increases.
The bands in the 600 cm region (c, d, and e) are assigned to symmetric vibrations of bridging oxygens (BO) in Si-O-Si linkages [28, 31]. In conventional soda-lime glasses, bands near 500 and 560 cm are attributed to bending motions of Si-O-Si in highly polymerized tetrahedral structures ( and species), while bands above 600 cm are typically associated with Si-O bending in depolymerized tetrahedra ( species) [28, 33]. Although the Raman spectra of the intermediate samples (with contents ranging from 12.15 to 18 mol%) exhibit negligible variation among themselves, a notable shift of approximately 20 cm is observed in the position of bands “e” and “k” when comparing the initial (CgCAF05) and final (CgCAF20) samples of the series. This spectral evolution indicates an increase in the fraction of NBO relative to total oxygens, suggesting a progressive depolymerization of the silica network with increasing CaF content [13, 34, 35]. Such behavior implies significant structural modifications, including the possible formation of new Q species or phases, or a substantial reorganization within the silicate network.
The observed low-intensity Raman bands can also be identified. Band “a”, located near 350 cm, is attributed to the vibrations of the network-modifying cations [31]. Although weak, its integrated area increases linearly with CaF concentration. Band “f”, observed at 800 cm and invariant concerning CaF content, is assigned to the Si-O stretching with dominant Si atom motion [36]. Symmetric stretching vibrations of Ca-F bonds are expected to appear around 485 cm (band “b”) [30]. However, due to the highly ionic character of the Ca-F bond, Raman bands associated with these vibrations exhibit low intensity, and barely to no variation is observed as a function of CaF content in the glass compositions studied.
Finally, samples of about 0.6 mm were used to measure the light transmittance in the range 190-3500 nm. These spectra are shown in Figure 7, demonstrating the overall high transmittance of the CgCAF samples. As shown in Figure 3, the refractive index of these samples is , which translates to reflection losses of about 4% at each air-glass interface. As can be seen, the transmittance remains near 90% from UV to NIR, evidencing the high optical quality of the CgCAF samples in this range. As transmittance (∼90%) plus reflection (∼8%) accounts for about 98% of the incoming light, it can be concluded that the absorption and scattering losses are pretty low in these samples. The inset in Figure 7 also highlights the wider transmittance window in the UV range of CgCAF samples compared to the original soda-lime glass, which may be associated with changes in the glass structure and could point to potential applications in UV-related technologies.
Figure 7. UV-VIS spectra.
Conclusions
In this work, we have demonstrated the utilization of end-of-life cover glass from commercial solar panels to produce oxyfluoride glasses by incorporating CaF. XRF analyses have proven that the final samples are free of contaminants, such as iron, that could introduce color to the material, reducing its value and limiting its potential applications. This confirms that, due to the high purity and transparency of the soda-lime cover glass, we could use it to make up to 80% of the total weight of CgCAF glasses. In contrast, conventional flat glass production allows only a small proportion of the material to be recycled. Furthermore, the melting temperature of the samples was about 1200°C, significantly lower than the melting temperature of soda-lime glass, which favors a reduced energy consumption and carbon emissions to produce CgCAF samples. Thermal analysis showed a 50–90°C reduction in temperature relative to the soda-lime, the presence of at least two crystallization processes and a pretty good stability of the glass phase, demonstrated by values in the range 100–180°C. XRPD confirmed the glassy nature of the samples and indicated a depolymerization process as the CaF₂ concentration increases, which is corroborated by mass density and Raman data. In situ XRPD measurements in the CgCAF12 sample demonstrated the formation of combeite, diopside, and Ca₃SiO₅ triclinic
phases, besides some unidentified peaks that require further studies. Finally, the UV-VIS-NIR transmittance demonstrated the good optical quality of all samples and the improved UV transmittance window compared to the original cover glass employed in this work. The ensemble of results shows that this family of materials can be explored as glass or be crystallized to produce glass-ceramics that may be tailored for different applications. Further studies should focus on understanding the crystallization process of these materials and incorporating optically active ions. If scaled for practical applications, these oxyfluoride glasses could provide a viable route for the reintegration of end-of-life cover glass into the industrial value chain. Such an approach would not only enhance the economic value of this waste material but also promote circularity and sustainability within the glass and photovoltaic industries.
Acknowledgement
The authors thank CNPq (grants 409475/2021-1, 402473/2023-0 and 304060/2023-2) for the financial support.
References
- [1]D. D. F. D. Rio, B. K. Sovacool, A. M. Foley, S. Griffiths, M. Bazilian, J. Kim, D. Rooney, Decarbonizing the glass industry: A critical and systematic review of developments, sociotechnical systems and policy options (3 2022). doi:10.1016/j.rser.2021.111885.
- [2]M. Zier, P. Stenzel, L. Kotzur, D. Stolten, A review of decarbonization options for the glass industry (6 2021). doi:10.1016/j.cemx.2021.100083.
- [3]J. H. Butler, P. D. Hooper, Glass Waste, 2nd Edition, Elsevier, 2019, pp. 307–322. doi:10.1016/B978-0-12-815060-3.00015-3. URL https://linkinghub.elsevier.com/retrieve/pii/B9780128150603000153
- [4]B. Caudle, S. Taniguchi, T. T. Nguyen, S. Kataoka, Integrating carbon capture and utilization into the glass industry: Economic analysis of emissions reduction through CO₂ mineralization, Journal of Cleaner Production 416 (2023) 137846. doi:10.1016/j.jclepro.2023.137846. URL https://linkinghub.elsevier.com/retrieve/pii/S0959652623020048DOI
- [5]C. D. Westbroek, J. Bitting, M. Craglia, J. M. C. Azevedo, J. M. Cullen, Global material flow analysis of glass: From raw materials to end of life, Journal of Industrial Ecology 25 (2021) 333–343. doi:10.1111/jiec.13112. URL https://onlinelibrary.wiley.com/doi/10.1111/jiec.13112
- [6]T. Bristogianni, F. Oikonomopoulou, Glass up-casting: a review on the current challenges in glass recycling and a novel approach for recycling “as-is” glass waste into volumetric glass components, Glass Structures & Engineering 8 (2023) 255–302. doi:10.1007/s40940-022-00206-9. URL https://link.springer.com/10.1007/s40940-022-00206-9DOI
- [7]M. P. Belançon, M. Sandrini, V. S. Zanuto, R. F. Muniz, Glassy materials for silicon-based solar panels: Present and future, Journal of Non-Crystalline Solids 619 (2023) 122548. doi:10.1016/j.jnoncrysol.2023.122548. URL https://linkinghub.elsevier.com/retrieve/pii/S0022309323004143DOI
- [8]X. Li, Q. Gu, Q. Wang, J. Luo, D. Liu, Y. Chang, Renewable energy in the mining industry: Status, opportunities and challenges, Energy Strategy Reviews 56 (2024) 101597. doi:https://doi.org/10.1016/j.esr.2024.101597. URL https://www.sciencedirect.com/science/article/pii/S2211467X24003067
- [9]G. Gaustad, M. Krystofik, M. Bustamante, K. Badami, Circular economy strategies for mitigating critical material supply issues, Resources, Conservation and Recycling 135 (2018) 24–33. doi:10.1016/j.resconrec.2017.08.002.DOI
- [10]D. Sica, O. Malandrino, S. Supino, M. Testa, M. C. Lucchetti, Management of end-of-life photovoltaic panels as a step towards a circular economy, Renewable and Sustainable Energy Reviews 82 (2018) 2934–2945. doi:10.1016/j.rser.2017.10.039. URL https://doi.org/10.1016/j.rser.2017.10.039
- [11]T. Zink, R. Geyer, Material recycling and the myth of landfill diversion, Journal of Industrial Ecology 23 (2019) 541–548. doi:10.1111/jiec.12808.DOI
- [12]R. F. Muniz, B. S. Greatti, M. Sandrini, M. P. Belançon, R. Valcarenghi, R. D. Bini, V. S. Zanuto, Glass Application in Solar Energy Technology, IntechOpen, 2025. doi:10.5772/intechopen.1010177. URL http://dx.doi.org/10.5772/intechopen.1010177 https://www.intechopen.com/online-first/1218426
- [13]R. F. Muniz, V. O. Soares, G. H. Montagnini, A. N. Medina, M. L. Baesso, Thermal, optical and structural properties of relatively depolymerized sodium calcium silicate glass and glass-ceramic containing CaF₂, Ceramics International 47 (2021) 24966–24972. doi:10.1016/j.ceramint.2021.05.224. URL https://doi.org/10.1016/j.ceramint.2021.05.224
- [14]R. F. Muniz, A. Steimacher, F. Pedrochi, V. S. Zanuto, L. M. Azevedo, J. H. Rohling, M. L. Baesso, A. N. Medina, Eu²⁺-Nd³⁺ co-doped glasses for solar spectrum modification via nuv/visible to nir down-conversion, Journal of Alloys and Compounds 888 (2021) 161484. doi:10.1016/j.jallcom.2021.161484. URL https://doi.org/10.1016/j.jallcom.2021.161484
- [15]J. Zheng, X. Xing, W. Liu, Z. Pang, R. Hu, Q. Xue, J. Wang, H. Zuo, An integrate study of the effects of CaF₂ on the viscous behavior and structure of CaO-SiO₂-MgO-Al₂O₃-CaF₂ blast-furnace slag, Journal of Non-Crystalline Solids 648 (2025) 123310. doi:10.1016/j.jnoncrysol.2024.123310.DOI
- [16]M. Riaz, R. Zia, A. Mirza, T. Hussain, F. Bashir, S. Anjum, Synthesis, characterization of CaF₂ doped silicate glass-ceramics, Materials Science and Engineering: C 75 (2017) 872–876. doi:10.1016/j.msec.2017.02.141. URL https://linkinghub.elsevier.com/retrieve/pii/S0928493116319798DOI
- [17]J. Xu, P. Chen, Effect of crystallization behavior and phase evolution on glass-ceramics derived from alumina silicate solid waste with addition high content CaF₂, Chemical Engineering Journal 506 (2025) 159998. doi:10.1016/j.cej.2025.159998.DOI
- [18]J. Laonamsai, P. Tasi, P. Wiwatrojanagul, M. Sriondee, T. Suriwong, P. Julphunthong, Synergistic effects of CuO and CaF₂ additives in facilitating low-temperature tricalcium silicate formation and stabilization, Results in Engineering 26 (2025) 104620. doi:10.1016/j.rineng.2025.104620.DOI
- [19]N. Pawlik, B. Szpikowska-Sroka, T. Gorczyńska, J. Pisarska, W. A. Pisarski, Structural and photoluminescence investigations of Tb³⁺/Eu³⁺ co-doped silicate sol-gel glass-ceramics containing CaF₂ nanocrystals, Materials 14 (2021) 754. doi:10.3390/ma14040754.
- [20]M. P. Belançon, M. Sandrini, F. Tonholi, L. S. Herulcano, G. S. Dias, Towards long term sustainability of c-Si solar panels: The environmental benefits of glass sheet recovery, Renewable Energy Focus 42 (2022) 206–210. doi:10.1016/j.ref.2022.06.009. URL https://linkinghub.elsevier.com/retrieve/pii/S1755008422000515
- [21]R. P. F. de Almeida, C. Bocker, C. Rüssel, Size of CaF₂ crystals precipitated from glasses in the Na₂O/K₂O/CaO/CaF₂/Al₂O₃/SiO₂ system and percolation theory, Chemistry of Materials 20 (2008) 5916–5921. doi:10.1021/cm801426u. URL https://pubs.acs.org/doi/10.1021/cm801426u
- [22]C. Rüssel, Nanocrystallization of CaF₂ from Na₂O/K₂O/CaO/CaF₂/Al₂O₃/SiO₂ glasses, Chemistry of Materials 17 (2005) 5843–5847. doi:10.1021/cm051430x. URL https://pubs.acs.org/doi/10.1021/cm051430x
- [23]N. Iwamoto, Y. Makino, A structural investigation of calcium fluorsilicate glasses, Journal of Non-Crystalline Solids 46 (1) (1981) 81–94. doi:10.1016/0022-3093(81)90076-4.
- [24]J. P. Caland, J. Baptista, G. C. Peiter, K. M. de Aguiar, H. Coelho-Júnior, J. P. Sinnecker, J. F. Felix, R. Schneider, Nanostructured glass-ceramic materials from glass waste with antimicrobial activity, Molecules 29 (7 2024). doi:10.3390/molecules29133212.
- [25]L. Vidal, E. Joussein, M. Colas, J. Cornette, J. Sanz, I. Sobrados, J. L. Gelet, J. Absi, S. Rossignol, Controlling the reactivity of silicate solutions: A FTIR, Raman and NMR study, Colloids and Surfaces A: Physicochemical and Engineering Aspects 503 (2016) 101–109. doi:10.1016/j.colsurfa.2016.05.039.
- [26]J. Gao, G. Wen, T. Huang, B. Bai, P. Tang, Q. Liu, Effect of slag-steel reaction on the structure and viscosity of CaO-SiO₂-based mold flux during high-Al steel casting, Journal of Non-Crystalline Solids 452 (2016) 119–124. doi:10.1016/j.jnoncrysol.2016.08.036. URL https://linkinghub.elsevier.com/retrieve/pii/S0022309316303611
- [27]H. Tian, Z. Wang, T. Zhao, C. Wang, A raman and multinuclear 29Si, 27Al, and 19F NMR study on the structural roles of CaF2 in SiO2–CaO–Al2O3-based welding fluxes, Metallurgical and Materials Transactions B 53 (2022) 232–241. doi:10.1007/s11663-021-02359-4. URL https://link.springer.com/10.1007/s11663-021-02359-4DOI
- [28]A. K. Yadav, P. Singh, A review of the structures of oxide glasses by Raman spectroscopy, RSC Advances 5 (2015) 67583–67609. doi:10.1039/c5ra13043c.DOI
- [29]Y. Tsunawaki, N. Iwamoto, T. Hattori, A. Mitsuishi, Analysis of CaO2 SiO2 and CaO2 SiO2 CaF2 glasses by Raman spectroscopy, Journal of Non-Crystalline Solids 44 (2-3) (1981) 369–378. doi:10.1016/0022-3093(81)90039-9.
- [30]R. W. Luth, Raman spectroscopic study of the solubility mechanisms of f in glasses in the system CaO-CaF2-SiO2, American Mineralogist 73 (3-4) (1988) 297–305.
- [31]R. Muniz, M. Baesso, F. Sato, A. Bento, J. Rohling, A. Medina, High pressure effect on the short-and intermediate-range structure of depolymerized soda lime silicate glass: Insights from micro-Raman spectroscopy, Vibrational Spectroscopy 110 (2020) 103113. doi:10.1016/j.vibspec.2020.103113.DOI
- [32]P. Dumas, J. Corset, Y. Levy, V. Neuman, Raman spectral characterization of pure and fluorine-doped vitreous silica material, Journal of Raman spectroscopy 13 (2) (1982) 134–138. doi:10.1002/jrs.1250130207.DOI
- [33]M. Wang, J. CHeng, M. Li, F. He, Raman spectra of soda–lime–silicate glass doped with rare earth, Physica B: Condensed Matter 406 (20) (2011) 3865–3869. doi:10.1016/j.physb.2011.07.014.DOI
- [34]A. Novatski, A. Steimacher, A. N. Medina, A. C. Bento, M. L. Baesso, L. H. C. Andrade, S. M. Lima, Y. Guyot, G. Boulon, Relations among nonbridging oxygen, optical properties, optical basicity, and color center formation in CaO–MgO aluminosilicate glasses, Journal of Applied Physics 104 (9) (nov 2008), doi:10.1063/1.3010306. URL https://pubs.aip.org/jap/article/104/9/094910/389590/Relations-among-nonb
- [35]C. O’Shaughnessy, G. S. Henderson, H. W. Nesbitt, G. M. Bancroft, D. R. Neuville, The influence of modifier cations on the raman stretching modes of Qn species in alkali silicate glasses, Journal of the American Ceramic Society 103 (2020) 3991–4001. doi:10.1111/jace.17081.DOI
- [36]B. Boizot, S. Agnello, B. Reynard, R. Boscaino, G. Petite, Raman spectroscopy study of β-irradiated silica glass, Journal of non-crystalline solids 325 (1-3) (2003) 22–28. doi:10.1016/S0022-3093(03)00334-X.DOI