# Definition of glass transition

I am getting confused about the definition of a "glass transition".

1. I read for example that it can be a transition from a rubber state to a brittle state in polymers. In this case therefore I would consider that as a "amorphous solid $$\rightarrow$$ amorphous solid" transition, since liquid polymers are not involved.

2. I read also that supercooled liquids can undergo a glassy transition, for example water can enter a glassy state. In this sense I would call it a "liquid $$\rightarrow$$ amorphous solid" transition.

Are these both "glass transitions" ?

I read that "glass transitions" are accompanied by small structural changes that cannot easily explain the difference (order of magnitutes) of physical properties between the two states. I can understand this problem in case (1), but in case (2) a liquid and an amorphous material to me are clearly very different (in one atoms diffuse, in the other not). So this makes me wonder if a transition of type (2) is still considered a glass transition or not.

• If the 'amorphous solid' is a configurationally frozen liquid, then it is a glass transition - the viscosity changes smoothly until the material is 'solid'. There are amorphous solids that undergo first-order phase transitions to the liquid (and first-order phase transitions to the crystalline solid). Rubber to brittle is similar - the viscosity changes smoothly from one to the other, and likely from 'rubber' to 'liquid' as well. – Jon Custer Dec 19 '19 at 18:30
• Thanks for your answer. What is a "configurationally frozen liquid" ? For me amorphous solids are materials where atoms individually do not diffuse (like in crystals) but do not possess the long range order typical of crystals. – Thomas Dec 19 '19 at 18:56
• the viscosity changes until the material is "solid" --> for me an "amorphous solid' is already 'solid'. Could you explain better ? – Thomas Dec 19 '19 at 18:57
• pslc.ws/macrog/tg.htm you can read about polymers here. The concepts can be ported to other glasses as well. – Alchimista Dec 20 '19 at 8:38
• @Thomas - those are absolutely correct. the amorphous-to-crystal silicon transition has an associated enthalpy (so first order), and amorphous silicon is not structurally related to the liquid. – Jon Custer Dec 21 '19 at 16:11