Structure and optical band gap study of transparent oxyfluoride glass-ceramics containing CaF2 nanocrystals

    Structure and optical band gap study of transparent oxyfluoride glass-ceramics containing CaF2 nanocrystals M. Rezvani, L. Farahinia PII: DOI: Reference:

S0264-1275(15)30408-1 doi: 10.1016/j.matdes.2015.08.159 JMADE 567

To appear in: Received date: Revised date: Accepted date:

5 July 2015 30 August 2015 31 August 2015

Please cite this article as: M. Rezvani, L. Farahinia, Structure and optical band gap study of transparent oxyfluoride glass-ceramics containing CaF2 nanocrystals, (2015), doi: 10.1016/j.matdes.2015.08.159

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Structure and optical band gap study of transparent oxyfluoride glass- ceramics containing CaF2 nanocryastals

Department of Materials Science and Engineering, University of Tabriz, Tabriz, Iran

RI

1

PT

M. Rezvani1* and L. Farahinia

SC

*Corresponding Author

Email: [email protected]

NU

Tel: +989144159511

MA

Fax: +984133362282

AC CE P

TE

D

Oxyfluoride glass- ceramics containing CaF2 nanocrystals were prepared by one-step crystallization of SiO2- Al2O3- CaO- CaF2 glasses at different temperatures. X- ray diffraction (XRD) results have revealed that CaF2 was the only precipitated crystalline phase in glassceramic samples. Fourier transform infrared spectroscopy (FT-IR) study of the samples showed the effect of crystallization temperature on the structure. According to DTA and XRD results, suitable glass and its associated glass- ceramics were selected as the best sample in order to study the effect of crystallization on the optical properties. UV- Vis spectra of samples were recorded to determine optical properties. Fermi energy level decreased for glass- ceramics due to the increment of semiconducting behavior. Urbach energy, optical band of the specimens have been reduced by increasing of nanocrystals in glassy matrix which could be attributed to increasing of the order of the structural and formation of crystals and dangling bonds, respectively. Keywords oxyfluoride; CaF2 nanoctystal; Band gap; Urbach energy

1. Introduction Fluoride glasses are attractive materials for photonic applications due to their high transparency and low phonon energies [1-2]. However, low thermal, chemical and mechanical properties of them are undeniable [3-4]. Although oxide glasses have high phonon energies, they have more stability and better mechanical properties [5-6]. Hence, there were attempts to combine advantages of both fluoride and oxide glasses. In that sense, glasses containing both fluorine and oxygen, called oxyfluoride glasses, have solved the problem to some extent. In other words, introduction of oxygen to fluoride glass increases the stability and affects the glass formation and the structure of glass networks [7-9]. 1

ACCEPTED MANUSCRIPT

D

MA

NU

SC

RI

PT

Wang and Ohwaki [3] have introduced new transparent oxyfluoride glass- ceramics containing PbxCd1-xF2 crystals. In addition to the favorable stability properties of oxide glasses, these noble glass- ceramics provide a fluoride environment for dopants like rare- earth ions [10]. However, regarding to the environmentally harmful effects of both CdF2 and PbF2, their uses are restricted [11]. LaF3-containing oxyfluoride glass and glass-ceramics are also reported as a more suitable hosts for rare-earth ions (RE) due to the better solubility of LaF3 for RE than PbF2 and CdF2. However, considering the cost of LaF3 materials, their applications are not economical [12-13]. Among the different fluorides, CaF2 is more preferable due to its high transparency to lights with the wavelength within 0.13 to 9.5 µm, refractive index compatibility with the aluminosilicate glass matrix and high solubility of rare-earth ions [14-16]. Therefore, rare-earth ions doped transparent oxyfluoride glass- ceramics containing CaF2 nanocrystals have attracted much attention and become probable candidates for optical applications. A considerable number of studies have been carried out on the effects of rare earth ions on the optical properties of oxyfluoride glass and glass ceramics, like luminescence properties, upconversion behavior and other optical properties [11]. To our knowledge, although effects of crystallization on the optical semiconducting behavior of other glass- ceramic systems have been studied [17-18], there was not any investigation of its potential effects on the oxyfluoride glass- ceramics.

AC CE P

TE

Therefore in the present study, after choosing a suitable composition, we tried to investigate the optical variations arose from the crystallization process. For this purpose, some optical semiconducting properties of oxyfluoride glass- ceramics, like absorption and extinction coefficient, Fermi energy level, direct and indirect optical band gaps and Urbach energy have been calculated.

2. Experimental Procedure

The designated composition of the basic glasses (weight ratio) is presented in Table 1. In fact, we used the most commonly reported composition [4, 14, 19-21]. As it is shown in Table 1, CaO is also added to the three main constituents, i. e., SiO2, Al2O3 and CaF2. Actually, according to reaction 1, replacing some amounts of CaF2 by CaO prevents the loss of F- ions [15-16].

SiO2 + 2CaF2 = SiF4 ↑ + 2CaO

(reaction 1)

Leached SiO2 with high purity, AL2O3 (Merck 101077) and CaF2 (Dae Jung 2508145) are used as the starting materials of the three main components. In order to introduce K2O and CaO, K2CO3 (Sigma- Aldrich P5833) and CaCO3 (Merck 102069) were 2

ACCEPTED MANUSCRIPT

PT

supplied. As2O3 (Merck 1001190100) and Sb2O3 (Acros 213471000) are also added as refining agents to produce bubble-free samples. Since the mentioned composition did not result in suitable melt, applying flux agents seemed inevitable. Therefore, K2O was applied to the batches ax flux agents in 3 different amounts to obtain suitable melts.

AC CE P

3. Results and discussion

TE

D

MA

NU

SC

RI

50 g of batches were mixed and melted in alumina crucibles at 1450 °C for 1 hour in an electric furnace. In order to control the fluoride loss, crucibles were covered by alumina plates and quite high heating rate for melting process (15°C/min) was applied. Then preheated stainless steel molds at 500°C, were used to shape the molten glasses. Finally glassy discs with 0.5 cm thickness were prepared. For releasing the internal stresses of samples, annealing at 500°C for 30 min was carried out. On the basis of differential thermal analysis (DTG- 60AH Shimadzu) results (heating rate of 10°C/min), crystallization temperatures were determined. Glass- ceramic samples were prepared by crystallization of glasses at different temperatures starting from crystallization peak temperature to 850°C. The precipitated crystalline phase in glass- ceramics was identified by X- ray diffraction (XRD, Siemens D- 500) patterns. Optical transmittance measurements were carried out in the UV-Vis spectrum range by means of a UV-Vis Shimadzu 1700 spectrophotometer instrument at room temperature. According to the results obtained from transmittance spectra, absorption and extinction coefficient, Fermi energy level, direct and indirect optical band gaps and Urbach energy of the samples were determined.

3.1. Crystallization of the glass samples 3.1.1. Differential Thermal Analysis (DTA) DTA curves of the basic glasses are presented in Fig. 1. Two distinguishable exothermic peaks are observed at temperatures TP1 and TP2. These two peaks are attributed to the crystallization of CaF2 and anorthite respectively [22]. More K2O have resulted in lower crystallization of CaF2 nanocrystlas. In fact, K2O plays a network modifier role in the glass network. In addition, crystallization of CaF2 crystals in oxyfluoride glasses is a three- dimensional crystal growth process controlled by diffusion [11]. Hence, more non- bridging oxygens (NBOs) created by K2O decrease the viscosity and consequently the mobility of ions gets facilitated. Thus, K2O shifts the crystallization of CaF2 to lower temperatures.

3.1.2. X- ray Diffraction (XRD) patterns and choosing suitable glass ceramic samples XRD patterns are used to characterize the crystalline phases precipitated in the samples. As it is shown in Fig 2, glass sample shows typical amorphous X-ray scattering characteristics; in other words, no unwanted crystallization was occurred during the processing step. After heat treating 3

ACCEPTED MANUSCRIPT

RI

PT

the samples at the first DTA peak temperature for 2 hours (GCH1.5-764, GCK3-717 and GCK4.5-684), CaF2 peaks appeared in the XRD patterns. Crystallized glasses at the second peak temperature of DTA plot show that both CaF2 and CaAl2Si2O8 (anorthite) are precipitated; that is to say, the second exothermic peak is related to crystallization of anorthite. Therefore, in order to obtain oxyfluoride glass- ceramics containing CaF2 nanocrystals, glasses were heat treated at different temperatures ranging from the first peak temperature to 850°C with the steps of 25°C. It is worth to mention that, crystallization at temperatures higher than 850°C did not result in transparent glass- ceramics. Crystallization conditions are summarized in Table. 2.

0.9 BCosB

MA

D

NU

SC

XRD patterns of the glassy samples and the related glass- ceramics are shown in Fig. 2. All peaks in XRD results of crystallized samples are assigned to CaF2 nanocrystals (ICDD: 00-350816). Moreover, the Scherer’s equation (eq. 1) was used to calculate the crystal size of CaF2 crystals embedded in the glassy matrix. (1)

TE

D

Where D is crystal size, λ is the wavelength of X- ray (Cu Kα: 1.541 Å), θB is the diffraction angle and B is the width of the diffraction peak at the half maximum. The peaks get sharper with increase of the crystallization temperature, with indicates the gradual formation of CaF2 nanocrystals in the glass- ceramics.

AC CE P

The calculated crystal size of CaF2 in glass- ceramic samples are presented in Table 2. More K2O additive has resulted in larger crystals in glass- ceramics prepared at the same temperature. This outcome is related to the mechanism suggested by Russel [23] for crystallization of CaF2. Russel pointed to an interface that is formed near the crystals and is enriched of glass formers. Therefore, the mentioned interface hinders the diffusion and prevents the more crystallization of CaF2. Whereas K2O decreases the viscosity, diffusion occurs easier through the interface and consequently CaF2 crystals get larger. As a result, the higher amount of K2O decreases the crystallization temperature and enlarges the crystals in glass- ceramics. In order to choose a suitable glass composition, crystallization temperature and crystals’ size are considered. From this point of view, due to the moderate properties of GK3 glass, it has been chosen as the favorable composition. On the other hand, glass ceramic samples prepared from GK3 glass with mean crystal size about 30 nm and suitable transparency were chosen for further optical studies.

3.2. FT-IR spectroscopy FT-IR spectroscopy was applied with the aim of obtaining structural information of oxyfluoride glass and glass- ceramics. FT-IR spectra of glasses extended from 4000 to 400 cm-1 are presented 4

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

in Fig. 3. Due to the significant amount of SiO2 in the composition of samples, presence of SiO4 tetrahedrons in the glass network was expectable. Bands placed at the ~460, ~700 and ~1170 cm-1 are attributed to the rocking vibrations, symmetric stretching vibrations and asymmetric stretching vibrations of Si-O-Si bonds [24, 25]. Al3+ ions play a network modifier role and replacement of some Si4+ ions by Al3+ ions results in Al-O-Si bonds. Therefore, the absorptions at ~1160 and ~ 560 cm-1 are related to the asymmetric stretching vibration mode of Si-O-Al [26, 27]. Although Ca-F bonds should have resulted in a peak at 443.30 cm-1 [28], overlapping with the rocking vibration mode of Si-O-Si has made this peak somehow difficult to distinguish. In addition, the peaks relating to Ca- F bond get intensified by the increase of crystallization temperature, which is caused by more crystallization of CaF2 crystals in glassceramics. As Hill et. al. [29] have reported, F- ions associate with aluminum atoms in AlO4 tetrahera in oxyfluorige glasses. Therefore, Al-F bonds create a peak at about 1350 cm-1 in FT- IR spectra [30]. On the other hand, due to the replacing oxygens by F-, Ca2+ ions attempt to compensate the negative charge induced in both [AlO4] – and [AlO3F]- tetrahedrons and create a band at ~990 cm-1 related to Si-O-Ca bonds [24]. The small peaks at about 1550 and 1610 cm–1 are related to CO group and molecular water, respectively. The CO peak might be due to the residual carbonates introduced by the raw materials. As it is detectable from Fig. 3, higher crystallization temperature has resulted in more intensified peaks. Since FT-IR is a vibrational technique, the overall spectrum of crystalline material will have much sharper bands than the spectrum of amorphous material.

3.3. Evaluation of optical constants 3.3.1. Absorption, extinction coefficients and determination of Fermi energy level UV- Vis absorbance spectra of oxyfluoride glass and glass- ceramics are presented in Fig. 4. There is no sharp absorption edge in the glassy sample spectrum, which is a characteristic of the glass materials [17]. In contrary to the basic glass, absorption edge becomes more intensive for glass- ceramics which means absorption edge increases with crystallization [31] and obviously shifts to longer wavelength. It may be caused by the reduction number of non-bridging fluorine (NBF) in the residual glassy phase when the CaF2 crystalline phase precipitates [19]. As it is shown in Fig. 4, absorbance is greater for glass- ceramics and the basic glass has the highest transparency. According to the Strong absorption of UV absorption bonds, extinction coefficient (K) obeys from the Fermi- Dirac distribution function, thus, the Fermi energy level can be calculated applying the equation (2): (2)

5

ACCEPTED MANUSCRIPT

PT

Where is Ef the Fermi energy, E is the variable photon energy probing the sample, kB is the Boltzmann constant and T is the ambient temperature (297K) [32-34]. Fig. 5 depicts K vs. hυ plots for different glass samples. The calculated values for Fermi energy of samples are presented in Table 3, which are achieved from least square fittings of Equation (2). Reduction of Fermi energy level with higher crystallization temperatures indicates a change to better semiconducting properties in glass- ceramic samples.

RI

3.4.3. Direct and indirect optical band gap

(3)

MA

NU

SC

For amorphous materials, the absorption coefficient exhibits a sharp increase just before the band gap. The relation between absorption coefficient and the incident photon energy is given by Davis and Mott [35] by the following relation:

AC CE P

TE

D

Where is a constant related to the extent of the band tailing, n is the index which can have different values 2, 3, 1/2 and 1/3 corresponding to indirect allowed, indirect forbidden, direct allowed and direct forbidden transitions, respectively [36-39]. Therefore, by plot vs. Photon energy (Tauc’s plots), the intercept of the obtained line divided by slope, is equal to energy band gap of optical transitions [40]. Fig. 6 illustrates the Tauc’s plots of samples and Table 3 shows direct and indirect allowed optical band gaps of the samples calculated from these plots. Optical band gap decreases for glass- ceramics crystallized at higher temperatures. Actually crystallization and formation of crystals in glassy matrix causes the creation of energy levels in forbidden gap that could reduce band gap of samples. This reduction is due to the decrease in average bond energy and fall in conduction band’s level, which arose from creation of non-bridging oxygens [32]. On the other hand, the reduction in band gap energy could be interpreted by assuming the production of surface dangling bonds around crystallites during the process of crystallization [41]. It has been suggested that when crystallization occurs in amorphous materials, dangling bonds are produced around the surface of crystallites. Thereby the number of surface dangling bonds increase by increasing the crystallization temperature. Increment of dangling bonds leads to the formation of more localized states in the band structure and consequently reduction of the optical energy gap [42]. 3.4.4. Urbach Energy The optical absorption in amorphous semiconductors near the absorption edge is usually characterized by three types of optical transitions corresponding to transitions between tail and tail states, tail and extended states, and extended and extended states. The first two types correspond to and the third one corresponds to . Thus, the plot of absorption coefficient versus photon energy (α vs. hν) has three different regions. In second 6

ACCEPTED MANUSCRIPT

(4)

RI

PT

region, the absorptions is related to transitions from the localized tail states above the valence band edge to extended states in the conduction band and/or from extended states in the valence band to localized tail states below the conduction band. The spectral dependence of absorption coefficient usually follows the so-called Urbach rule (equation (4)).

NU

SC

Therefore, Urbach energy (EU) has been calculated using equation (4) and least square fitting of against hν curves in the tailing part of localized states (Fig. 7) [43-46]. Table 3 represents the Urbach energy values of the samples, which decreases in the case of crystallization at higher temperatures. Since Urbach energy of glassy semiconductors implicitly defines the degree of disorder, crystallization and resulted order of this process decreases the Urbach energy in value.

  

D

TE



Oxyfluoride glasses with different amounts of K2O were prepared by melting process. According to DTA patterns, crystallization temperature of CaF2 was determined and transparent glass ceramic samples containing CaF2 nanocrysyals were obtained. The optimum amount of K2O (3 weight ratio) was chosen considering its effect on the crystallization temperature and CaF2 crystal size. Fermi energy level reduced with higher crystallization temperatures (from 4.29 to 3.07 eV) and consequently samples got more semiconducting properties. Creation of energy level in band gap and presence of dangling bonds are 2 possible reasons for the reduction of band gap energies. Increasing the crystallization temperature decreased Urbach band tailing from 0.29 to 0.21 eV.

AC CE P



MA

4. Conclusions

References [1] V. Nazabal, M. Poulain, M. Olivier, P. Pirasteh, P. Camy, J. L. Doualan, S. Guy, T. Djouama, A. Boutarfaia, J. L. Adam, J. Fluorine Chem. 134 (2012) 18-23. [2] A. J. Stevenson, H. Serier- Brault, P. Gredin, M. Mortier, J. Fluorine Chem. 132 (2011) 11651173. [3] Y. Wang, J. Ohwaki, Appl. Phys. Lett. 63 (1993) 3268-3270. [4] Y. Kishi, S. Tanabe, J. Alloy. Compd. 408- 412 (2006) 842-844. [5] M. C. Goncalves, L. F. Santos, R. M. Almedia, C. R. Chimie 5 (2002) 845-854. [6] X. Zhang, B. Bureau, P. Lucas, C. Boussard- Pedel, J. Lucas, Chem. Eur. J. 14 (2008) 432442. [7] M. Secu , C.E. Secu, S. Polosan, G. Aldica, C. Ghica, J. Non-Cryst. Solids 355 (2009) 1869– 7

ACCEPTED MANUSCRIPT 1872. [8] W. J. Zhang, Q.J. Chen, Q.Y. Zhang, Z.H. Jiang, J. Non-Cryst. Solids 357 (2011) 2278–2281.

RI

PT

[9] S. A. Polishchuk, L. N. Ingat’eva, Y. V. Machenko, V. M. Bouznik, oxyfluoride glasses (a review), Glass Phys. Chem. 37 (2011) 1-20.

D

MA

NU

SC

[10] K. Shinozaki, T. Honma, K. Oh-ishi, T. Komatsu, J. Phys. and Chem. of Solids 73 (2012) 683–687. [11] M. H. Imanieh, B. Eftekhari Yekta, V. Marghussian, S. Shakhesi,I. R. Martin, Solid State Sci. 17 (2013) 76- 82. [12] M. J. Dejneka, J. Non-Cryst. Solids 239(1-3) (1998) 149-155. [13] X. Qiao, X. Fan, J. Wang, M. Wang, J. Non-Cryst. Solids 351 (2005) 357–363. [14] P. Babu, K. H. Jang, E. S. Kim, L. Shi, H. J. Seo, J. Korean Phys. Soc. 54 (2009) 14881491. [15] D. Chen, Y. Wang, E. Ma, Y. Yu, F. Liu, Opt. Mater. 29 (2007) 1693-1699. [16] Zh. Hu, Y. Wang, E. Ma, D. Chen, F. Bao, Mater. Chem. Phys. 101 (2007) 234-237. [17] V. Khani, P. Alizadeh, M. S. Shakeri, Mater. Res. Bull. 48 (2013) 3579-3584.

TE

[18] S. Singh and K. Singh, J. Molecular Structure, 1081 (2015) 211–216.

AC CE P

[19] X. Sun, M. Gu, Sh. Huang, X. Jin, X. Liu, B. Liu, Ch. Ni, J. Lumin. 129 (2009) 773- 777. [20] X. Sun, Sh. Huang, Nucl. Instrum. Meth. A 621 (2010) 322- 325. [21] G. Lakshminarayana, R. Yang, M. Mao, J. Qiu, I. V. Kityk, J. Non- Cryst. Solids 355 (2009) 2668- 2673. [22] L. Farahinia, M. Rezvani, E. Alahgoliyan, Mater. Res. Bull. 70 (2015) 461- 467. [23] C. Russel, Chem. Mater. 17 (2005) 5843- 5847. [24] D. P. Mukherjee, S. Kumar Das, J. Non- Cryst. Solids 368 (2013) 98-104. [25] S. Atalay, H. I. Adiguzel, F. Atalay, Mater. Sci. Eng. A 304-306 (2001) 796-799. [26] U. Grob, S. Rudiger, E. Kemnitz, Solid State Sci. 9 (2007) 838-842. [27] T. Suzuki and Y. Ohishi, Appl. Phys. Lett. 88 (2006) 191912-1. [28]K. Tahvildari, M. Esmaeilipour, Sh. Ghammamy, H. Nabipour, Int. J. Nano Dim. 2 (2012) 269-273. [29] R. Hill, D. Wang, M. Thomas, J. Mater. Sci. 34 (1999) 1767-1774. [30] Y. Zhang, D. Chen, Multilayer integrated film bulk acoustic resonators, Springer, Shanghai, 2013. [31] A. Faeghi Nia, Int. J. Eng. 4 (2014) 609-614. [32] E. El- Diasty, F. A. Abdel Wahab, M. Abdel- Baki, J. Appl. Phys. 100 (2006) 093511. [33] M.A. Khashan, A.M. El-Naggar, Optics Commun., 174 (2000) 445–453. [34] S. U. El-Kameesy, H. M. Eissa, S. A. Eman, Y. A. El-Gamma, Egy. J. Pure & Appl. Sci., 49(1) (2011) 067-070. [35] N. F. Mott, E. A. Davis, Electronic processes in nanocrystalline materials, Clarendon press, 8

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Oxford, 1971. [36] M. S. Shakeri, M. Rezvani, Spectrochim. Acta A 79 (2011) 1920-1925. [37] F. Yakuphanoglu, S. Ilican, M. Caglar, Y. Caglar, J. Optoelect. And Adv. Mater, 9(7) (2007) 2180–2185. [38] H. Ming Oo, H. Mohamed-Kamari, W. M. D. Wan-Yusoff, Int. J. Mol. Sci. 13 (2012) 46234631. [39] Md. R. Sahar, A. S. Budi, Solid State Science and Technology, 14(1) (2006) 115-120. [40] M. Abdel- Baki, F. El- Diasty, F. A. Abdel- Wahab, Opt. Commun. 261 (2006) 65-70. [41] D.K. Dwivedi, H. P. Pathak, N. Shukla, A. Kumar, J. Ovonic Res. 10 (2014) 15- 22. [42] S. Hasegawa, M.Kitagawa, Solid State Commun. 27(1978) 855-858. [43] S. Rani, S. Sanghi, A. Agarwal, V. P. Seth, spectrochim. Acta A 74 (2009) 673-677. [44] S. A. Babanejad, F. Afshari, N. Salarzadeh, E. Ashrafi, Adv. Appl. Sci. Res., 3 (2012) 743748. [45] A.A.Al-Ghamdi, ShamshadA.Khan, Physica B, 404 (2009) 4262–4266. [46] K. Subrahmanyam, M. Salagram, Optical Materials, 15 (2000) 181-186.

9

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

MA

Fig. 1. DTA curves of glassy samples recorded with the heating rate of 10°C/min Fig. 2. XRD pattern of glasses and glasses containing (a) 1.5 (b) 3 (c) 4.5 weight ratio K2O

D

Fig. 3. FT- IR spectra of GK3 and glass- ceramics heat treated at different temperatures

TE

Fig. 4. UV- Vis spectra of GK3 and glass- ceramics heat treated at different temperatures Fig. 5. Extiction coefficient vs. energy plots of glasses heat treated at different temperatures

temperatures

AC CE P

Fig. 6. Tauc’ plots for (a) indirect and (b) direct band gap of glasses heat treated at different

Fig. 7. Determination of Urbach energy of glasses heat treated at different temperatures

10

Fig.1

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

11

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

(a)

12

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

(b) 13

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

(c) Fig. 2

14

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 3

15

Fig. 4

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 5 16

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

(a)

(b) Fig. 6

17

Fig. 7

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

18

ACCEPTED MANUSCRIPT

K2O

37.26 37.26 37.26

28.11 28.11 28.11

1.5 3 4.5

26.89 26.89 26.89

SC

7.73 7.73 7.73

RI

Al2O3 CaO CaF2

AC CE P

TE

D

MA

Sample Code GK1.5 GK3 GK4.5

SiO2

NU

Composition (weight ratio)

PT

Table 1. Compositions of glasses with different amounts of K2O

19

As2O3 Sb2O3 0.2 0.2 0.2

0.2 0.2 0.2

ACCEPTED MANUSCRIPT

GCK1.5-764 GCK1.5-800 GCK1.5-825 GCK1.5-850

764 800 825 850

AC CE P

NU

MA D

TE

GCK3-717 GCK3-725 GCK3-750 GCK3-775 GCK3-800 GCK3-825 GCK3-850

GCK4.5-684 GCK4.5-700 GCK4.5-725 GCK4.5-750 GCK4.5-775 GCK4.5-800 GCK4.5-825 GCK4.5-850

Crystal Size (nm)

SC

Sample Code

Crystallization Temperature (°C)

RI

PT

Table 2. Crystallization condition and crystal size of glass- ceramic samples (heat treated for 2 hours)

16.67 ± 3 39.00 ± 4 42.10 ± 5 43.43 ± 3

717 725 750 775 800 825 850

15.36 ± 4 17.22 ± 5 20.12 ± 3 30.91 ± 6 43.68 ± 3 48.44 ± 8 50.74 ± 5

684 700 725 750 775 800 825 850

19.00 ± 4 19.50 ± 3 22.66 ± 7 24.75 ± 3 33.69 ± 6 45.81 ± 8 48.95 ± 5 51.44 ± 5

20

PT

ACCEPTED MANUSCRIPT

Eindirect

4.29 4.12 3.47 3.07

2.95 2.50 1.43 1.37

Edirect

EU

3.40 3.20 3.02 2.81

0.29 0.27 0.25 0.21

NU

SC

Ef

AC CE P

TE

D

MA

Energy (eV) Sample GK3 GCK3-700 GCK3-725 GCK3-750

RI

Table 3. Optical properties of oxyfluoride glasses heat treated at different temperatures

21

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Graphical abstract

22

ACCEPTED MANUSCRIPT Highlights

TE

D

MA

NU

SC

RI

PT

Glass ceramic samples containing CaF2 nanocrystals were prepared successfully. FT-IR results proved the existence of an oxyfluoride glass structure in samples. Fermi energy and band gap energy are improved for the crystallized glasses. Urbach energy decreased by increasing the crystallization temperature.

AC CE P

   

23