Sodium borosilicate glasses are well known for their immiscibility in the liquid state, which makes them suitable as stating materials for porous glasses. Background for that application is the spinodal decomposition to at least two glass phases with an interpenetrating microstructure. One of the phases is removed by leaching. Pore sizes between 2 and 100 nm are state of the art. To widen the application range microstructure with wider pore sizes might be useful. To achieve that goal the following strategies were followed:
Shirazu porous glass, basically a calcium alumoborosilicate glass, can be used to develop porous glasses with pore sizes of about 5 to 6 µm. To achieve that the composition was varied by increasing the B2O3 content. The addition of other network modifying components to sodium borosilicate glasses compositions reduced the phase separation tendency of the glasses und was thus not a successful strategy to increase pore sizes.
Prolonged heat treatment was tested with a sodium borosilicate glass with (in mole %) 8 Na2O, 26 B2O3, 66 SiO2. In literature reports two types of microstructure development have been expected: the approximation of a long term limit of pore sizes or their infinite growth with a growth rate reduced with time. In the experiments described here times up to 64 d and temperatures between 630 °C and 750 °C were applied for heat treatment. The resulting materials were leached in dilute acid and in NaOH to generate porosity. As cast and heat treated glasses as well as porous materials were analyzed by thermal analysis, mercury intrusion porosimetry, nitrogen sorption, X-ray diffraction, X-ray fluorescence spectroscopy, helium pycnometry and scanning electron microscopy, if applicable. The development of microstructure parameters with treatment time was evaluated to characterize the kinetics of microstructure formation. Since the experiments were focused on long term heat treatment the initial growth stages described in literature were not observed. Here, two subsequent growth stages were identified. In both stages the pore sizes follow a pore size time law dp µ tb. Ostwald ripening might be responsible for the first stage with b » 0.33. In the second stage with b » 0.5, diffusion possibly supported by viscous flow might be responsible for structure formation. The observed pore size time laws do not limit the pore sizes and mean pore diameters of more than 13 µm were generated. Parallels to ceramic sintering processes will be discussed. The calculated activation energy of the second investigated spinodal decomposition in the investigated material is 220 ± 10 kJ mole-1.