On the Skeptics Guide to the Universe, the hosts noted that the plural of “glass” is “glasses,” not to be confused with a term that refers exclusively to corrective lenses (or many sets of Google’s latest invention).
The news story under discussion was a recent study on the physical properties of a 20 million year old piece of amber. The findings highlight the fact that, even given such a long time period, the amber, which is dried tree resin, hardly flowed at all. This shows the pitfalls in thinking about glasses as “frozen liquids,” as this can lead to some severe misconceptions. For example, as discussed in the same SGU segment, there used to be an urban legend that the “flow” of glass could be observed in the stained-glass windows of old cathedrals, which are supposedly thicker at the bottom than the top. However, this is due to the limitations of Medieval glass-making technology, which could only produce nonuniform slabs, which were logically installed with the thicker end down. (Another good example of accurate observation coupled with a totally off-base interpretation) Glass is a frozen liquid in the sense that if you could take a snapshot of its molecules, they are arranged in a disordered noncrystalline configuration, compared to the regular repeating lattice a solid. You can imagine taking a liquid and instantly freezing all of the molecules in place, and it would look like a glass. However, if you actually try to freeze an ordinary liquid by cooling it, the molecules will try to arrange themselves into their preferred lattice configuration and create a solid. This liquid-solid phase transition is essentially reversible and not path dependent. That is, the system always tries to minimize its free energy, which sets up an epic struggle between the chemical energy (which is minimized by the formation of many regular bonds), vs entropy, which wants to have as much freedom as possible. The key parameter, not surprisingly, is temperature, so there is a well-defined “melting point,” above which entropy wins (liquids are free to flow) and below which chemical energy wins (in a crystalline solid, there are many regular bonds). With glasses, the situation is much more complicated. Instead of a melting point, there is a “glass transition temperature,” which changes the hard glass into a liquid.
As the molten glass is cooled below the glass transition temperature, the molecules try to arrange themselves into a regular lattice, but they move slower and slower and never quite make it. They just get stuck, but the degree to which they remain disordered depends greatly on how quickly the glass was cooled. If it was cooled too fast, it might have to be annealed afterwards in a process that reheats the glass so that the molecules will have enough energy to escape local minima and find more stable configurations. A quick analogy might be the carpentry dictum “don’t tighten the screws too soon” you should keep in mind when assembling furniture from IKEA. You want to allow enough play for the pieces to shift slightly into the “right” positions, then progressively tighten the constraints so they remain there.