Three-dimensional (3D) photonic geometries are attractive for developing novel coupled optical modes that cannot exist in the two-dimensional (2D) nano and microfabrication world. In this thesis, the various optical properties that can be induced as a result of 3D architecture are designed, fabricated, and characterized. Even for the well-established resonance in split-ring resonator-based metamaterials, the addition of the multiple planes of symmetric coupling or decoupling induce isotropic and anisotropic resonances for applications such as ultra-sensitive molecular analysis with two-fold advantage of frequency and amplitude monitoring for small concentrations and low on-chip power inclinometers with nanodegree sensitivity, respectively. The limited spatial coverage of the plasmon-enhanced near-field in 2D graphene ribbons presents a major hurdle in practical applications. The ability to transform 2D materials into 3D structures while preserving their unique inherent properties offers enticing opportunities for the development of diverse applications for next-generation micro/nanodevices. Diverse self-assembled 3D graphene architectures are explored here that induce hybridized plasmon modes by simultaneous in-plane and out-of-plane coupling to overcome the limited coverage in 2D ribbons. While 2D graphene can only demonstrate in-plane bi-directional coupling through the edges, 3D architectures benefit from fully symmetric 360° coupling at the apex of pyramidal graphene, orthogonal four-directional coupling in cubic graphene, and uniform cross-sectional radial coupling in tubular graphene. The 3D coupled vertices, edges, surfaces, and volumes induce corresponding enhancement modes that are highly dependent on their shape and dimensions. While most of this work strives to achieve multiple coupled planes of symmetry, the same ideas are also applied to achieve multiple 3D graphene geometries that break mirror symmetry across multiple planes. The asymmetric graphene induces giant optical activity (chirality) that has remained previously unrealized due to the 2D nature of graphene. The chirality induced within the 3D graphene chiral helixes is also a strong function of the geometrical parameters that are analyzed using a machine-learning-based multivariate regression approach to determine the 3D geometry with the strongest chirality. The hybrid modes introduced through the 3D couplings amplify the limited plasmon response in 2D ribbons to deliver non-diffusion-limited sensors, high-efficiency fuel cells, and extreme propagation length optical interconnects.