Accelerated crystal growth in a lithia aluminosilicate glass

No viscosity change in the vicinity of the crystal/melt interface is to be expected for crystal growth in silicate systems at constant temperature in a deeply undercooled isochemical melt. In most experimental studies, however, absolute chemical conformity is not realized and the formation of an interfacial zone with presumably strongly altered viscosity will occur. In the present investigation isolated Li2O⋅Al2O3⋅4SiO2 (LAS4) single crystals grew in a slightly alumina deficient base glass at temperatures above glass transition. Due to the depletion of alumina and the enrichment of excess components and impurities of the continuously growing interfacial zone (“shell”), its viscosity will change. Dedicated experiments showed that the viscosity of the interfacial zone is up to 7 orders of magnitude lower than the viscosity of the base glass. The decreased viscosity at the crystal growth front is assumed to cause the experimentally observed increase of the crystallisation velocity with the crystal radius. This behaviour can be quantitatively modelled introducing a linear decrease of the activation enthalpy with the crystal radius. Obviously, the proposed model can be likewise used in the opposite case, i.e. for an increasing viscosity of the interfacial zone during crystal growth. According to the standard concept of a fractional viscosity relation, a coupling factor can formally be extracted from the comparison between the experimentally determined activation enthalpies of the crystal growth velocity and of the base glass viscosity. The resulting unrealistically low numerical value indicates that it is not the high viscosity of the base glass, but the lower viscosity of the interfacial zone that must be considered for the determination of a physically meaningful coupling factor.