The innovative synthesis technique, glass crystallisation, has proved to be indispensable in the discovery of new metastable oxides with unique properties.  When used in combination with computational prediction of stable compositions within the SrO-SiO2-Al2O3 ternary diagram the discovery of Sr2Si3O8 was made possible.  After discovery of the novel phase variable temperature X-ray diffraction (VTXRD) was used to find the optimum synthesis conditions; glass crystallisation with a heat treatment of 850°C for 12h. Even though the initial precursor glass composition contained aluminium, the crystal structure (solved from PXRD and neutron diffraction) showed it to be Al-free, with the remaining non-crystallised glass being Al-rich, as confirmed by multi-nuclear MAS-NMR and TEM-EDX.  The crystal was found to have a monoclinic, P21 /c, unit cell, with corner sharing SiO4 units forming Zwier ribbon chains in-between 7- or 8- coordinated Sr sites.  Comparison with the known barium analogue Ba2Si3O8, showed the structures to be similar but not isostructural with the difference being a cooperative twisting of the external SiO4 chains creating distorted hexagonal rings.  Although not found in the crystal structure the Al-rich glass matrix was found to play an important role in the crystallisation of Sr2Si3O8.  Attempts at synthesising with the binary precursors, SrO-SiO2, produced a powder with a crystalline phase fraction in excess of 85% but with much broader Bragg peak widths when compared to the aluminium analogue. In an attempt to solve this problem high energy ball milling of the glass was tried to reduce particle size and therefore encourage surface crystallisation, however while it did reduce the optimum synthesis temperature to 800°C it did not reduce Bragg peak widths.  Barium and calcium were also tried as dopants (by substituting for strontium) and the nucleating agents, ZrO2, AgNO3 and Au were independently added to the glass in the hopes of promoting full crystallisation, thus allowing detailed microstructural analysis under TEM measurement conditions.  Differential scanning calorimetry (DSC) analysis of all the samples agreed with the VTXRD results and appeared to show two types of behaviour.  Both the dopants and nucleating agents decreased the crystalline phase fraction, however while the former also increased peak widths the latter decreased it.  This result was unexpected as it was thought that a sharpening of the Bragg peaks (improved crystallinity) would also be associated with an increase in crystalline phase fraction, hence the need for SEM, TEM and NMR microstructure analysis to determine the cause.  While this work focuses on the complex interplay between composition, glass-crystallisation behaviour and the resulting crystal structures and microstructures of Sr2Si3O8-based materials, the discovery of metastable Sr2Si3O8 has proved that when innovative synthesis methods like glass crystallisation are harnessed with computational approaches new phases can be isolated even in well explored phase diagrams.  Future work will explore other complex phase diagrams by these methods, like the CaO-SiO2-Al2O3 system, to discover new compounds with potentially useful properties.