This study investigates the limited selectivity of the Cu111 surface for C-C bond formation during CO2 reduction and explores the factors influencing selectivity using Cu nanoparticles smaller than 2 nm. The optimal nanoparticle size for C-C bond formation on the 111 facet with minimal overpotential is determined using density functional theory. A suitable supporting surface to enhance the stability and catalytic performance of the Cu-based nanoparticles is identified. Various Cu catalyst geometries, including planar surfaces and cuboctahedral, icosahedral, and truncated octahedral Cu nanoparticles, are considered. Size-dependent effects on the binding energies of reaction intermediates and hydrogen atoms are examined. Carbon-based surfaces, particularly 2SO2-doped graphene nanoribbons, are stable hosts for the Cu nanoparticles and help in retaining the activity for CO2 reduction. Scaling relations between the binding energies of the intermediates suggest COOH binding energy as an energy descriptor. Through multiparameter optimization and with the help of parity line and graphical construction, Cu38 and Cu79 are found to be the most promising surface for C2 product generation. This study provides insights into the factors influencing the selectivity and catalytic performance of Cu nanoparticles, aiding the development of efficient catalysts for CO2 reduction.