The extraordinary toughness and stiffness of bone are associated with its three main constituents - apatite mineral, collagen protein and water. Variations in composition and organization of these constituents are known to exist as a function of disease and aging. These variations greatly influence bone quality and need to be understood in greater detail. This thesis advances the understanding of molecular organization in bone along three directions:​ quantification of molecular orientation, analysis of mineral deformation in response to hydration changes and loading and investigation of age-dependent bone quality.

First, polarized Raman spectroscopy was adapted for bone tissue applications to quantify molecular organization in non-deproteinated, turbid tissue. This enabled the simultaneous quantitative measurements of altered mineral and collagen orientations in Osteogenesis Imperfecta, a bone disease associated with collagen mutations. Second, the effect of distorting the water environment in bone was investigated by replacing matrix water with deuterium oxide. Changes in hydrogen bonding affected collagen secondary structure, resulting in compression of the mineral lattice as evidenced by changes in peak positions and widths of mineral Raman bands. Further, polarized Raman spectroscopy was used to probe nano-scale deformations due to tensile loading and orientation-dependent strains within the mineral lattice were observed. These results demonstrate the potential of Raman spectroscopy to provide insights on molecular orientation and interaction at the nano-scale.

Third, exploratory data mining tools were employed to identify tissue-level compositional (Raman) and mechanical (nanoindentation) metrics that predict bone quality, instead of the traditionally used linear regressions. The results showed that compositional properties offer only a partial understanding of mechanical properties at the tissue-level and vice versa. Hence, a specific combination of compositional and mechanical metrics was required to reliably classify femoral specimens according to age. These findings suggest that combined metrics will better predict transformations in bone quality than individual metrics and call for novel techniques to explore the complex multi-scale interactions in bone. The multiple lines of evidence presented in this thesis provide an insight into the complex roles that mineral, collagen and water play in governing tissue quality and mechanical properties of bone.