PHYSICAL CHEMISTRY

E. K. KELER, E. I. KOZLOVSKAYA

ELASTIC PROPERTIES AND CRYSTAL FORMATION IN GLASSES

(Presented by Academician N. N. Semenov, 23 V 1963)

Scientific research and experimental work in recent years in the field of the synthesis and study of the properties of photosensitive glasses has led to the discovery of a method for obtaining glass-crystalline materials of a new type, possessing high mechanical strength, thermal stability, and other valuable properties.

Fig. 1. 1—curve of the dependence of deformation on temperature for the initial glass; 2—thermogram of the initial glass

The combination of interesting technological and technical features of these materials immediately attracted broad attention from researchers.

We have investigated the mechanical properties of the initial glasses and glass-crystalline materials of two systems: 1) alumino-potassium-lithium-silicate (AKLS), 2) alumino-magnesium-titanium-silicate (AMTS).

In principle, the behavior of the materials of these systems proved to be identical; therefore, in what follows the results of their study are generalized.

The deformation–temperature dependence of the initial glasses under a constantly acting load and heating (Fig. 1, curve 1) shows that in the temperature interval \(20^\circ — T_g\) the magnitude of their deformation does not change. At temperatures above \(T_g\) a deformation jump is observed; the point \(T_k\) is revealed quite clearly, above which, in contrast to ordinary glasses (point \(T_3\) \((^1)\)), the curve goes with a noticeable downward slope (temperature interval \(T_k — T_{k1}\)). Then, with a further increase in temperature, a secondary increase in deformation is observed (interval \(T_{k1} — T_{k2}\)). Above the point \(T_{k2}\) the deformation curve runs parallel to the abscissa axis up to the point \(T_f^*\).

The thermogram of the glasses, plotted in the same figure (curve 2), shows that the increase in deformation at temperatures above \(T_g\) corresponds to an endothermic effect; the slope of the deformation curve above \(T_k\) is associated with an exothermic effect and, finally, the changes at the points \(T_k\) and \(T_{k2}\) correspond to the second endothermic and exothermic effects.

In addition to the deformations developing in the interval during the process of continuous—

* \(T_g\)—temperature at which softening of the initial glass begins; \(T_k\)—temperature of effective crystallization; \(T_f\)—temperature of transition to the liquid state; \(T_g — T_k\)—temperature interval of the formation of crystallization centers; interval \(T_k — T_{k1} — T_{k2}\)—temperature interval of crystallization, in which the interval \(T_k — T_{k1}\) is the principal one. \(T_k — T_f\)—temperature interval of the glass-crystalline material.

during heating under a continuously acting load, we investigated the deformation values corresponding to each temperature point separately. It turned out that the greatest deformations during heating of the initial glass develop in the interval \(T_g — T_k\), and the smallest occur on the plateau in the interval \(T_{k2} — T_f\). Intermediate deformation values belong to the intervals \(T_k — T_{k1}\) and \(T_{k1} — T_{k2}\). Unlike ordinary glasses, the magnitude of the elastic deformation in the segment \(T_k — T_f\) is not greater, but less than the deformation of the segment \(20^\circ — T_g\). The changes occurring in the interval \(T_k — T_f\) make the behavior of crystallizing glasses different from that of ordinary glasses \((^1,^2)\). In the final analysis the differences amount to an increase in rigidity (a decrease in elastic deformations) in the interval \(T_k — T_{k1}\), to breaks at the points \(T_k\), \(T_{k1}\), and \(T_{k2}\), and, finally, to a very extended interval \(T_{k2} — T_f\); in this last interval the system is no longer a glass but a glassy-crystalline material with gradually changing phase composition and structure.

To clarify the nature of these changes, we carried out an investigation using the quenching method.

The initial glasses were subjected to various heat treatments. One and the same initial glass was heated to the temperatures: \(T_g\), \(T_g — T_k\), \(T_k\), \(T_k — T_{k1}\), \(T_{k1} — T_{k2}\), and \(T_{k2} — T_f\) (see Fig. 1). At the indicated temperatures, various holding times were applied, from minutes to hours, after which the specimens were quenched in air.

As a result of microstructural analysis of the experimental specimens, it was found that crystallization does not occur up to the temperature \(T_g\). In the interval \(T_g — T_k\), individual crystallized regions were observed in some glasses.

In specimens held at the temperature \(T_k\), against the background of the glassy phase there were found clearly expressed individual crystallized regions (Fig. 2; see insert to p. 1364), the number and volume of which increased as the holding time was increased. The structure of a completely crystallized glass is shown in Fig. 3 (see insert to p. 1364).

The size of the crystals ranges from 2 to \(4\ \mu\). A small amount (approximately 5–7%) of glassy phase is distributed between the crystals. It should be noted that the crystallization process should not be greatly prolonged in time. According to our observations, long holding times are not useful and are even harmful, since they promote the growth of crystals, whose excessive size leads to loosening of the structure and loss of strength. The optimum holding time is established experimentally for each system.

Specimens held in the interval \(T_{k1} — T_{k2}\) and \(T_{k2} — T_f\) for various times showed no essential difference in structure compared with specimens held at the temperature \(T_k — T_{k1}\). The deformation changes and secondary thermal effects (Figs. 1, 2) in the interval \(T_k — T_{k2}\) are apparently associated with crystallization of a second glass phase present in small amounts. It was of great interest to follow the kinetics of crystallization during heat treatment. With the aid of Leonov’s high-temperature microscope \((^3)\), we were able to carry out some experiments in this direction.

Glass specimens were heated in the microscope with a slow rise in temperature and were continuously observed in polarized light. Up to the temperature \(T_g\), no changes were detected in them.

At temperatures in the interval \(T_g — T_k\), and especially near the point \(T_k\), the finest bright points—the centers of crystallization (crystal nuclei)—appeared on the surface of the experimental piece from its various sides. Gradually increasing, as the holding time at the given temperature increased, these points grew into separate, finely crystallized regions. The crystallites were anisotropic; when the microscope stage was rotated, extinction was observed.

With a further increase in temperature, the crystallized regions, expanding, drew closer together and joined. The structure (Fig. 3) shows the final stage of crystal formation, when the scattered crystallized regions have merged into a single fine-crystalline structure.

In the interval \(T_{k^2}—T_f\), no substantial changes in the structure were observed.

Figure 4 gives curves of the elastic moduli of the initial glass and of the glass-crystalline material being formed at different stages of their heating. As can be seen, the value of the shear modulus of the crystallized glass under room-temperature conditions is approximately 1.5–2 times higher than that of the initial glass.

Fig. 4. Dependence of the elastic modulus on temperature. 1 — initial glass, 2 — crystallized glass (glass-crystalline material)

This value remains constant also upon heating up to the point \(T_f\) of glass-crystalline materials, above which the values of the elastic moduli decrease sharply (plastic deformations and flow begin).

In Fig. 4 we already see the “relaxation well” known for ordinary glasses and the sharp break corresponding to the point \(T_k\) \((^{1,2})\). In the temperature interval of crystallization \(T_k—T_{k^2}\), strengthening is observed. In the interval \(T_{k^2}—T_f\) the elastic moduli change hardly at all.

All the samples we obtained with different heat treatments were subjected to bending tests under room-temperature conditions and upon heating.

The greatest strength was shown by samples held in the interval \(T_k—T_{k1}\). The strength (\(\sigma\) in bending) of these samples was \(2400\ \text{kg}/\text{cm}^2\).

From the experimental materials presented, it is evident that changes in the elastic and strength properties of glass-crystalline materials during their crystallization are closely connected with changes in their structure; the most favorable formation of this structure requires, for each system, a definitely established heat-treatment regime.

The temperature range of increased elastic and strength properties of glass-crystalline materials is significantly expanded not only in comparison with the initial glasses, but also in comparison with a number of well-known ceramic materials (porcelain, fireclay, dinas, magnesite, chromomagnesite, etc.), which places them among the most heat-resistant and thermally stable materials.

The study of the elastic and plastic deformations of the initial glasses as a function of temperature and holding time can serve as a reliable and rapid method for identifying the optimum heat-treatment conditions with the aim of obtaining high-quality glass-crystalline materials.

Institute of Silicate Chemistry named after I. V. Grebenshchikov
Academy of Sciences of the USSR

Received
23 I 1963

CITED LITERATURE

1. E. K. Keler, E. I. Kozlovskaya, DAN, 116, No. 2 (1957).
2. E. I. Kozlovskaya, The Glassy State, Publishing House of the Academy of Sciences of the USSR, 1960.
3. A. I. Leonov, Advanced Scientific-Technical and Production Experience, issue 17, 1961.