In the current built environment, the three main glass elements are glass window blocks, (structural) float glass and (structural) glass bricks. Float glass is the most used type of glass element and it is evolving from a skin material to a structural material. Glass hollow blocks became popular in the 1930s, but have lost their appeal due to little structural capacity and outdated aesthetics. A new and upcoming technique is the usage of cast glass bricks, which provides high structural performance without the use of substructures. The 21st century is unmistakably characterized by climate change and the effort being put into the battle against it. It is the duty of engineers to design in a sustainable way. ‘Sustainable’ can be interpreted in many ways; the focus within this paper lies on the usage of recycled materials, the recyclability of the final product and the energy usage during the lifespan of the façade (linked to thermal performance). When looking at the material itself, glass is a very recyclable material. This is widely done for glass bottles. However, the recyclability of float glass is very limited. Windowpanes and structural glass panes both consists of several layers of glass, combined with a PVB interlayer. Mostly, these combined panes are themselves processed further into insulated glass units, or IGUs. This end product consists of many different materials, that are hard to separate and for this reason, IGUs are seldom recycled at this moment. Additionally, not only the end product is rarely recycled, little recycled material is used for the production of float glass. The tolerances for using waste cullet in the production of float glass is very limited. Since the lifecycle cannot easily be changed for float glass, it is interesting to look into cast glass. The production technique is different from the production of float glass. Glass is melted into a mould, rather than spread along a bath of tin. The latest research shows the possibilities of using glass cullet and thus casting waste glass. One opportunity that arises from this, is the development of waste glass bricks. One main problem of cast glass bricks is its thermal performance. Due to the lack of cavities, as present in IGUs or hollow glass bricks the thermal performance is not optimal, especially with the increasing need for high-performing facades to meet sustainability demands. There are several strategies to improve the thermal performance of glass structures. The most simple solution for reducing the thermal transmittance is the introduction of air pockets or cavities. Air is a relatively good insulator, especially in small confined geometries. Convection reduces the thermal resistance of air and the bigger the air pocket, the more convection can occur. Not only air can provide insulation, this can also be done with other materials. Inert gasses, like argon and krypton, have a lower thermal conductivity than glass, but are more costly and would require a more complex production process to fill the cavity and to ensure complete sealing. Translucent solutions can be achieved with materials like aerogel and glass fibres. Light transmittance is still present, but the result is not transparent. If transparency or translucency is completely ignored, traditional insulation materials, like rock wool and polystyrene, could increase the thermal resistance too. Improvements can also be made with either reducing the amount of as long-wave, infrared energy going through the glass utilizing coatings or by reducing the conductivity of the glass itself by changing the molecular build-up. The glass recipe influences the thermal conductivity. This thesis focuses on the thermal performance of cast glass bricks and in specifically the investigation of the impact of cavities on the thermal performance and the development of a relevant production method for cavity cast glass bricks. Air cavities can be added on either system-level or element-level. On a system-level, the design of a singular brick does not change, but the typology in which it is used does. Cavity walls with either a double brick wall or with an additional float glass wall will improve the thermal performance but will utilize a lot more material. When changing the geometry of the brick itself, one or more cavities can be introduced to improve the performance. Since smaller cavities have a relatively better performance, multiple small cavities will automatically perform better than a singular larger cavity. Several designs are drawn, varying from a single or double cavity to an arrangement of glass shards, providing lots of encapsulated air pockets. Thermal bridges within a design influence the overall thermal performance. The heat will choose the path with the least resistance, which is via the glass. It is possible to lengthen the path that the heat must undertake, by introducing multiple air chambers and increasing the overall thickness of the brick. This increased thickness will also aid the structural performance. Creating multiple air pockets is directly linked to many glass-air transitions and thus a more complex geometry and production process. Producing such a brick with current standard manufacturing techniques would require glue. Glueing is not very environmentally friendly. The glue is not easily (if not impossible) to remove and therefore the final product is not recyclable. With experimental research, the potential of tack fusing glass is highlighted. This ‘tack fusing’ is done at a temperature below the actual fusing temperature, where the glass still creates a permanent bond, without relaxation and shape loss. This temperature lies around the dilatometric softening point of the specific glass type. The potential of the tack fusing method is explored employing shear tests of fused specimens. Three types of specimens are tested: two different series of tack fused specimens at 650˚C (1hr, 3hr dwell at top temperature) and one series of glued samples (DELO 4468) to be used as a reference. DELO4468 is a high-performance glue, which is mostly used for its high strength. Nine design concepts are drawn, based on a solid brick of 200x200x100 mm. Seven element-based designs are limited within these boundaries since increased thickness (even with only glass) would lead to higher thermal performance. Two system-level alternatives are not in-depth researched but are still examined as reference. These two alternatives cannot be kept within these boundaries and are thus at increased depth of the finished wall. All nine designs are scored in a multi-criteria analysis. The main focus is the thermal performance and therefore the thermal transmittance of each design needs to be analysed. This is for eight out of nine designs done with TRISCO, a steady-state 3D thermal analysis software. The last design, a fused shard brick, consisting of a random arrangement of random-shaped shards and is not easily modelled in 3D software. For this reason, it is not analysed in TRISCO but analysed through an experimental analysis. This experiment replicates a situation in which one side of the brick is warmer than the other while measuring the temperatures and heat fluxes. With these measurements, the thermal transmittance can be estimated. The last part of this thesis is the multi-criteria analysis, or MCA, in which all alternatives are scored against several criteria. One of these criteria is the thermal performance, others are sustainability, producibility, aesthetical potential and transparency. The sustainability is scored against several checks, related to the recyclability and the required temperature for production. The producibility and the aesthetical potential is scored utilizing a small expert survey. Both criteria are subjective of nature, but by including the opinion of several experts, the outcome is inter-subjective. The transparency is scored regarding the number of refractive surfaces (because with each air-glass transition visibility is lost due to reflections) and the percentage of non-transparent materials in the cross-section. The result of this thesis is not a single design, but insights in the potential of each concept. All designs are not ready for direct application but would require further structural verification and thermal optimisation, depending on the governing thermal requirements. There is not one perfect solution since each façade has different performance requirements, but the results do provide insights for an improvement from the current technologies.