The glass transition temperature of silicate and borate glasses

The dependence of the glass transition temperature on silicate composition as well as on borate composition is discussed. Some new experimental results on borosilicate glass-enamels are put together with the existing literature data. It is demonstrated that in the glass transition interval the activation energy E(Tg) for viscous flow is related to the glass transition temperature. The activation energy for viscous flow is given by E(Tg) = 260Tg ± 10% [J/mol]. In the case of silicates, the glass transition temperature decreases almost linearly with the molar fraction x of network modifier, NM. In the interval 0.01 < x < 0.6 the glass transition temperature is: T g ( x ) = ( 1080 - 626 x ) ± 5 % [ K ] for ( Na 2 O + MgO ) x ( Al 2 O 3 + SiO 2 ) 1 - x T g ( x ) = ( 967 - 626 x ) ± 5 % [ K ] for ( PbO ) x ( SiO 2 ) 1 - x T g ( x ) = ( 895 - 626 x ) ± 5 % [ K ] for ( Na 2 O ) x ( SiO 2 ) 1 - x . This is easily explained assuming that viscous flow is controlled by the motion of SiO4 tetrahedra with one oxygen bridge already broken. In the same time the alkaline cations impose some spatial hindrance, therefore, larger is the cation the higher is the hindrance. The composition, x, dependence of the glass transition temperature of borates is more complicated. It has a maximum at about 30% of network modifier. This happens because introduction of NM changes borate structure from one consisting of BO3 triangles sharing edges to BO4 tetrahedra similar to SiO4 tetrahedra in silicates.