Transparent, insulating coatings can be applied to windows to increase the energy efficiency of buildings. Amorphous material like silica make good thermal insulators due to their local atomic disorder that impedes heat conduction. Pores can additionally be added to the material to further reduce heat conduction by decreasing the material density while adding interfaces that scatter heat carriers. This concept has been extensively used in highly porous silica aerogels, which are valued for their ultra-low thermal conductivities. However, these aerogels significantly scatter light, and cannot be used for applications that require high optical transparency. This thesis examines four different nanoporous silica networks, synthesized using a combination of template-assisted and template-free methods, to understand the relationship between the structure of each network and its thermal conductivity, and to explore effective, scalable synthetic methods for producing nanoporous materials with optimized porosity. In the first part of this dissertation (Chapters 2 to 5), we explore the effect of nanoscale architecture on polymer-templated silica-based networks. Silica-based thin films with various types of precursors, pore sizes, particle sizes and dopants in the network are studied. We found that although silica is amorphous, the change in the nanoscale architecture at fixed porosity and changes in the chemical composition of the walls can both be used to tune the thermal conductivity. In the second part (Chapter 6), we combined the knowledge gained from our thin film studies to synthesize hollow silica shells that can be assembled to produce transparent, thermally insulating monoliths. We optimize the synthesis of hollow silica shells to reduce their sizes from c.a. 30 nm to below 15 nm, and then demonstrate a method to assemble those shells into mechanically robust, monoliths. In the final part of this work (Chapter 7 and 8), we use small angle X-ray scattering to understand the structural changes that occur when silica precursors react to form wet gels and when those gels are dried to produce monoliths. Insights into the changes in the nanoscale architecture allow us to further optimize the optical transparency and minimize the thermal conductivity of the final monoliths.