This work reports Si and SiGe(C) nanoparticle formation on high-k dielectrics to work as charge storage nodes in NC Flash memories. When applying high fields for programming or erase operations the bands for the high-k dielectric bend sufficiently to cause Fowler Nordheim (FN) tunneling through a narrow triangular barrier, while higher bandgap SiO 2 remains in direct tunneling regime. In our proposed high-k based cells the memory operations are achieved at lower voltages and reduced charge loss. SiGe NCs possess a lower bandgap than Si NCs, and deeper potential wells when embedded in the gate dielectric and hence longer retention times. A different approach involves demonstrating that a chaperonin protein lattice can be used as a template to assemble nanocrystal (NC) arrays for Flash memory fabrication. Different types of NCs from colloidal suspensions can be incorporated into Flash memory fabrication. The demand for high-density, low-cost, low-power/voltage, and high speed (programming, erase and read operations) semiconductor memory have led to the current baseline memory technologies including non-volatile flash memories to proliferate into several revolutionary and evolutionary approaches such as nanoparticle floating gate and single/few electron non-volatile memories [1] . They each have unique strengths and challenges in the aforementioned areas (Fig.1]. Flash memory 1-T cells are more compact than DRAMs and do not require refresh. However, in the current embodiment, with a continuous polysilicon floating gate, SiO 2 or oxide-nitride-oxide (ONO) tunneling and control gate dielectrics, where the cells are programmed with channel hot electrons (CHE) and block erased by Fowler-Nordheim (FN) or direct tunneling from the floating gate into the source, there are severe scaling problems. These cells have relatively slow write/erase (~ms) and access (~80 ns) times. Shorter channel lengths and lower power Figure.1. Memory options from ITRS 2003.