Mysteries of the glass transition

Abstract

I do not see any mystery in James Langer’s “mysterious glass transition,” at least with respect to inorganic glasses of one-component systems like silicon dioxide, boron oxide, and so on. To understand the glass transition of inorganic materials, one first has to understand why crystals melt. Near the melting temperature, electrons occupy more and more excited states as temperatures increase. Electrons in excited states possess wavefunctions different from those in their low-energy or ground states. Different wavefunctions mean that the probability distribution of the electrons in space changes. The core ions will be driven to new places as they interact with the excited electrons. However, the electrons will change again and again to other states with different wavefunctions. The arbitrary time series of sufficient electrons in their excited states will cause the core ions to continuously change position. That scenario corresponds to a melt.As the melt cools, the electrons will occupy more and more low-energy states. If the forces of the electrons in their low-energy states are not strong enough to induce a regular order of the core ions, the transition to a glass occurs. Thus melting of chemically bonded solids, and glass formation from their melts, is basically an electronic effect generally neglected in publications dealing with properties of melts and glasses. 1 1. H.-J. Hoffmann, Phys. Chem. Glasses 45, 227 (2004). Glass formation from the melt depends on the strength and sufficiently large number of directed bonds (to stabilize the noncrystalline order) and on the melting entropy ΔSm (that is, melting enthalpy ΔHm , divided by melting temperature Tm ). If ΔSm is small, only a little entropy is released and produced once a bond closes, and the temperature increases locally by just a small amount. This implies that neighboring directed bonds of the undercooled melt can be broken only within a relatively small temperature range below Tm . This interval has to be passed fast enough for glass formation. If ΔSm is large, the temperature interval of recalescence is relatively large to reach Tm and the undercooled melt has enough time to rearrange to crystals during cooling. 2 2. H.-J. Hoffmann, Phys. Chem. Glasses 46, 570 (2005). Now it is easy to understand the “mystery” of the glass transition or what occurs in the glass-transition range. (Imagine that the temperature is rising.) In that range, bonding electrons start to occupy excited states. This causes an additional mechanism for the thermal expansion, an additional contribution of the specific heat capacities (not causing a “jump,” however), and an increase of the damping of resonances of many kinds in glasses. The worldwide standard procedure to determine Tg in glass science is based on the change of the slope of dilatometer curves, not mentioned by Langer. As a consequence of the scenario described here, there is no phase transition at Tg , just an exponential freezing out of electrons from higher to lower energy levels with decreasing temperature.REFERENCESSection:ChooseTop of pageREFERENCES <<CITING ARTICLES1. H.-J. Hoffmann, Phys. Chem. Glasses 45, 227 (2004). Google ScholarCAS2. H.-J. Hoffmann, Phys. Chem. Glasses 46, 570 (2005). Google ScholarCAS© 2008 American Institute of Physics.