Journal of Non-Crystalline Solids 284 (2001) 128±133
www.elsevier.com/locate/jnoncrysol
High-pressure densi®cation of ¯uoride, GeS2 and silicate glasses Yoji Kawamoto a,*, Koichi Miyauchi a, Masanori Shojiya a, Shinichi Sakida b, Naoyuki Kitamura c a
Department of Chemistry, Faculty of Science, Graduate School of Science and Technology, Kobe University, Nada, Kobe 657±8501, Japan b Venture Business Laboratory, Kobe University, Nada, Kobe 657±8501, Japan c Department of Optical Materials, Osaka National Research Institute, AIST, Ikeda, Osaka 563-8577, Japan
Abstract Densi®cation of glasses having dierent chemical-bond properties and also dierent structures have been examined by treating ionic ZrF4 -based ¯uoride glasses, mixed ionic and covalent SiO2 -based oxide glasses, and a covalent GeS2 sul®de glass at temperatures of about three-quarters of their respective glass-transition temperatures under pressures up to 9 GPa. The densities of GeS2 glass and silicate glasses of SiO2 contents >50 mol% increased with increasing applied pressure. In ZrF4 -based glasses and silicate glasses with SiO2 contents <50 mol%, on the other hand, the densities increased until an applied pressure around 3 GPa, the maximum densities were measured at pressures around 3 GPa, and then decreased with increasing pressure. The pressure dependences of permanent densi®cation observed in the latter glasses are a permanent densi®cation that has not been reported to our knowledge. Ó 2001 Elsevier Science B.V. All rights reserved. PACS: 81.40.Vw; 81.05.Kf; 61.43.Fs; 83.20.Hn
1. Introduction Glasses subjected to high-pressure treatments increase in density. The eect persists even after removal of the pressure [1±4]. The seemingly permanent densi®cation of glass is of interest to us from the viewpoint of glass science and technology because the changes in glass structure are retained and consequently the optical, electrical, mechanical and magnetic properties of the glass
* Corresponding author. Tel.: +81-78 803 5679; fax: +81-78 803 5679. E-mail address:
[email protected] (Y. Kawamoto).
may be changed without changing glass composition. Although permanent densi®cation experiments have been performed on silicate and sul®de glasses [5,6], there are no experiments of ¯uorozirconate glasses and no studies exist from the standpoint of the chemical-bond properties and packing density (de®ned as the fractional molar volume occupied by the ions assuming them to be perfect spheres) to our knowledge. In ¯uorozirconate glasses the chemical bonds between cations and ¯uoride anions are ionic [7] and consequently the packing densities of the ions are 660% [8]. On the other hand, the Ge and S of GeS2 glass, a typical sul®de glass, are covalently bonded. The packing density
0022-3093/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 2 2 - 3 0 9 3 ( 0 1 ) 0 0 3 9 1 - X
Y. Kawamoto et al. / Journal of Non-Crystalline Solids 284 (2001) 128±133
of Ge and S in the glass is about 20%. 1 Silicate glasses have both ionic and covalent bonding and the packing densities of glass constituents range 20±50%, depending on the glass composition. 2 In the present study the permanent densi®cation of ionic ZrF4 -based ¯uoride glasses, mixed ionic and covalent SiO2 -based oxide glasses and a covalent GeS2 sul®de glass are examined and discussed based on the chemical-bond properties and packing densities.
2. Experimental procedure 2.1. Glass samples Fluorozirconate glasses of the 60ZrF4 30BaF2 10EuF3 (ZBE) and 55ZrF4 17BaF2 5EuF3 23NaF (ZBEN) compositions, silicate glasses of the xSiO2 0:2
100 xK2 O 0:4
100 xCaO 0:4
100 xSrO composition (x 42:1, 44.4, 47.1, 50.0, 57.1 and 66.7) and a sul®de glass of the GeS2 composition were prepared and annealed at the respective glass-transition temperatures. The preparation procedures of the respective samples have been described elsewhere [9±12]. 2.2. Permanent densi®cation treatments Densi®cation of the samples was carried out with a 6±8 multi-anvil high-pressure apparatus [13]. Specimens of 8 8 2 mm3 in size were submitted to pressures of 1.5, 3.0, 4.5, 6.0 and 9.0 GPa with a 0:033 GPa min 1 ascending rate at room temperature. Then those were heated to temperatures of 270°C and 250°C (ZBE and ZBNE ¯uoride glasses, respectively), 355°C, 340°C, 355°C, 370°C, 386°C and 422°C (silicate glasses with 42.1, 44.4, 47.1, 50.0, 57.1 and 66.7 mol% SiO2 , respectively) and 270°C (GeS2 glass) within 5 min at each pressure and were kept at these temperatures for 30 min. After the densi®1 Value calculated from the measured density and the covalent radii in [19]. 2 Values calculated from the measured density and the ionic radii in [18,19].
129
cation treatments the temperatures were decreased to room temperature and then the applied pressures were released with a 0:033 GPa min 1 descending rate. The experimental errors in the pressure and temperature regulation in densi®cation treatments are 2:23 10 8 GPa and 0.3°C, respectively. 2.3. Density measurements Densities of the undensi®ed and densi®ed specimens were measured by the Archimedes method, using CCl4 as an immersion liquid. The experimental error in the density measurements is about 0:005 g cm 3 .
3. Result In the present silicate glasses the average number of bridging-oxygen ions per SiO4 tetrahedron, Y, can be calculated from an equation Y 6 200=x where x is SiO2 mol% [12]. Therefore, glasses with x: 42.1, 44.4, 47.1, 50.0, 57.1 and 66.7 mol% have Ys of 1.25, 1.5, 1.75, 2.0, 2.5 and 3.0, respectively. In the glasses with smaller Ys the connectivity of SiO4 tetrahedron units is less. Hereafter, glasses with x: 42.1, 44.4, 47.1, 50.0, 57.1 and 66.7 are labeled as the 1.25, 1.5, 1.75, 2.0, 2.5 and 3.0 silicate samples, respectively.
Fig. 1. Pressure dependence of densities of undensi®ed and densi®ed ZBE ¯uoride samples. The line is drawn between data symbols.
130
Y. Kawamoto et al. / Journal of Non-Crystalline Solids 284 (2001) 128±133
Fig. 2. Pressure dependence of densities of undensi®ed and densi®ed ZBEN ¯uoride samples. The line is drawn between data symbols.
The densities of the ZBE and ZBEN ¯uoride samples before and after densi®cation are plotted against applied pressure in Figs. 1 and 2, respectively. The densities of the 1.25, 1.5, 1.75, 2.0, 2.5
Fig. 4. Pressure dependence of densities of undensi®ed and densi®ed GeS2 samples. The line is drawn between data symbols.
and 3.0 silicate samples before and after densi®cation are plotted against applied pressure in Fig. 3. Densities of the GeS2 glass upon pressurization are shown in Fig. 4. In the ®gure the densities of crystalline b-GeS2 (low-temperature modi®cation) [14], a-GeS2 (high-temperature modi®cation) [15] and II-GeS2 (high-pressure modi®cation) [16] are plotted for reference.
4. Discussion 4.1. Densi®cation of the respective glass systems
Fig. 3. Pressure dependences of densities of undensi®ed and densi®ed silicate samples with x: 1.25, 1.5, 1.75, 2.0, 2.5 and 3.0. The lines are drawn between data symbols of each type.
The following observations may be made from Figs. 1±3 and 5 in relation to the permanent densi®cation of the glasses studied. The densities of the GeS2 sample and the 2.0, 2.5 and 3.0 silicate samples with SiO2 -contents P50 mol% increased with increasing pressure. In the two ¯uoride samples and the 1.25, 1.5 and 1.75 silicate samples with SiO2 contents <50 mol%, on the other hand, the densities increased until a pressure 3 GPa, and then decreased with increasing pressure. Here it is noted that the pressure dependencies of densities observed in the latter samples are a permanent densi®cation that has not been reported so far to our knowledge.
Y. Kawamoto et al. / Journal of Non-Crystalline Solids 284 (2001) 128±133
131
Fig. 5. Relationship between ionicity and packing density in ZrF4 -based, SiO2 -based and GeS2 samples.
Fig. 6. Relationship between connectivity of tetrahedron and packing density in ZrF4 -based, SiO2 -based and GeS2 samples.
Fig. 5 shows the relationship between ionicity and packing density in the ZBE and ZBEN ¯uoride samples, the 1.25, 1.5, 1.75, 2.0, 2.5 and 3.0 silicate samples and the GeS2 sample. In the ®gure a SiO2 sample is also shown for reference. The ionicity in the respective samples were calculated from a difference in electronegativity between two elements using the Pauling equation [17]. On the other hand, the packing densities, obtained from the molar volumes calculated from the densities of glasses and the ionic [18] and/or covalent [19] radii of elements in ¯uoride, silicate and GeS2 samples, are given in Table 1. From the ionicity and packing density relationship in Fig. 5 we deduced that glasses of ionicity >64% and packing densities >39% (a hatched area in the ®gure) have a dierent permanent densi®cation dependence on pressure. Fig. 6 shows the relationship between connectivity of glass-network forming tetrahedra and
packing density in the 1.25, 1.5, 1.75, 2.0, 2.5 and 3.0 silicate samples and the GeS2 sample. It is noted here that the ZBE and ZBEN ¯uoride samples are composed of the ZrF7 and/or ZrF8 polyhedral units [20±22] so that this relationship is inapplicable. However these polyhedra have 100% ionicity. Tentatively, therefore, the connectivity of these polyhedra was assumed to be zero and plotted in Fig. 6. The ®gure indicates that samples with smaller connectivity of glass-network forming polyhedra have larger packing densities and that glasses with Y < 2:0 and packing densities >about 39% (a hatched area in the ®gure) have a dierent permanent densi®cation dependence on pressure. 4.2. Structures of permanently densi®ed samples Changes in the F coordination environments around Zr4 and Eu3 of the ZBE ¯uoride glass
Table 1 Ionic and/or covalent radii of elements in ¯uoride, silicate and GeS2 glasses [18,19] Ionic radius (nm) Fluoride glass 4
Zr Ba2 Eu3 Na F a b
Covalent radius (nm) Silicate glass
0.084 0.142 0.107 0.116 0.133
O in SiO2 . O in K2 O, CaO and SrO.
K Ca2 Sr2 O2 b
Silicate glass 0.137 0.100 0.118 0.140
Si Oa
GeS2 glass 0.104 0.066
Ge S
0.122 0.104
132
Y. Kawamoto et al. / Journal of Non-Crystalline Solids 284 (2001) 128±133
with permanent densi®cation have been examined by Raman and ¯uorescence spectroscopies [23]. In the 1.25, 1.75 and 2.5 silicate samples, changes in silicon±oxygen structure with permanent densi®cation have been examined by 29 Si magic-anglespinning nuclear magnetic resonance spectroscopy [10]. The structural studies of permanently densi®ed GeS2 sul®de glass have been performed by means of extended X-ray absorption ®ne structure (Ge±K and S±K) and Raman spectroscopies and an X-ray radial distribution analysis [23]. On the other hand, the reason for the pressure dependence of densi®cation of the ZBEN ¯uoride sample has been analyzed using a molecular dynamics (MD) simulation technique [24±26]. The MD simulations have revealed a mechanism for the densi®cation as follows: at pressures >3 GPa the connectivity of the network-forming ZrFn and EuFn polyhedra partially transforms from corner-sharing type to edge-sharing one. The change in the connectivity causes a contraction of interstices around the network-modifying Ba2 and Na ions, around which the coordination numbers increase to those in the high-pressure polymorphs of the BaF2 and NaF crystals [27,28]. After decompression the structural relaxation occurs in the ZBEN glass. The extent of the relaxation depends on temperature and pressure because of the `fragility' of this glass [29]. The high-pressure network structures are fully or partially preserved after decompression at temperature
(x: 42.1±66.7) and a covalent GeS2 sample by treating them at temperatures about three-quarters of their respective glass-transition temperatures and pressures to 9 GPa. In the GeS2 sample and the SiO2 -based samples with x P 50, which have covalent-bonds and smaller packing densities, the densities increased with increasing pressure. This pressure dependence of densi®cation has been observed for a variety of glasses so far studied. In the ZrF4 -based samples and the SiO2 -based samples with x < 50, which have a larger ionicity and larger packing densities, the densities showed the maximum values at a pressure 3 GPa. Such a pressure dependence of permanent densi®cation is a permanent densi®cation that has not been reported to our knowledge. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
5. Conclusion Permanent densi®cations were examined on ionic 60ZrF4 30BaF2 10EuF3 and 55ZrF4 17BaF2 5EuF3 23NaF ¯uoride samples, mixed ionic and covalent xSiO2 0:2
100 xK2 O 0:4
100 xCaO 0:4
100 xSrO silicate samples
[16] [17] [18] [19] [20]
P.W. Bridgman, I. Simon, J. Appl. Phys. 24 (1953) 405. J.D. Mackenzie, J. Am. Ceram. Soc. 46 (1963) 470. H. Cohen, R. Roy, Phys. Chem. Glasses 6 (1965) 149. N. Mizouchi, A. Cooper Jr., J. Am. Ceram. Soc. 56 (1973) 320. S. Sakka, J.D. Mackenzie, J. Non-Cryst. Solids 1 (1969) 107. G. Parthasarathy, E.S.R. Gopal, Bull. Mater. Sci. 3&4 (1985) 271. Y. Kawamoto, I. Nohara, Solid State Ionics 22 (1987) 207. A. Konishi, H. Izumi, R. Kanno, Y. Kawamoto, J. Mater. Sci. 29 (1994) 1584. K. Miyauchi, J. Qiu, M. Shojiya, Y. Kawamoto, N. Kitamura, Mater. Res. Bull. 34 (1999) 1383. S. Sakida, K. Miyauchi, Y. Kawamoto, N. Kitamura, J. Non-Cryst. Solids 271 (2000) 64. Y. Kawamoto, S. Tsuchihashi, J. Am. Ceram. Soc. 52 (1969) 626. H.J.L. Trap, J.M. Stevels, Glastech. Ber. K 32 (6) (1959) 31. N. Kawai, M. Togaya, A. Onodera, Proc. Jpn. Acad. 49 (1973) 623. G. Dittmar, H. Schafer, Acta Crystallogr. B 32 (1976) 1188. G. Dittmar, H. Schafer, Acta Crystallogr. B 31 (1975) 2060. C.T. Prewitt, H.S. Young, Science 149 (1965) 535. L. Pauling, in: The Nature of the Chemical Bond, 3rd Ed., Cornell University, New York, 1960. R.D. Shannon, C.T. Prewitt, Acta Crystallogr. B 25 (1969) 925. L. Pauling, M.L. Huggins, Z. Krist. 87 (1934) 205. Y. Kawamoto, T. Horisaka, J. Non-Cryst. Solids 56 (1983) 39.
Y. Kawamoto et al. / Journal of Non-Crystalline Solids 284 (2001) 128±133 [21] Y. Kawamoto, T. Horisaka, K. Hirao, N. Soga, J. Chem. Phys. 83 (1985) 2398. [22] Y. Kawamoto, in: Materials Science Forum, vol. 6, TransTech, Aedermannsdorf, Switzerland, 1985, p. 767. [23] K. Miyauchi, J. Qiu, M. Shojiya, Y. Kawamoto, N. Kitamura, J. Non-Cryst. Solids 279 (2001) 186. [24] Y. Tamai, Y. Kawamoto, Chem. Phys. Lett. 302 (1999) 15.
133
[25] Y. Tamai, Y. Kawamoto, J. Chem. Phys. 112 (2000) 3875. [26] Y. Tamai, Y. Kawamoto, Phys. Rev. B. 62 (2000) 865. [27] D.P. Dandekar, J.C. Jamieson, Trans. Am. Crystallogr. Assoc. 5 (1969) 19. [28] Y. Sato-Sorensen, J. Geophys. Res. 88 (1983) 3543. [29] C.A. Angell, J. Non-Cryst. Solids 73 (1985) 1.