Associative learning with Y-shaped floating gate transistors operated in memristive modes

Abstract

We present Y-shaped three-terminal floating gate transistors with positioned quantum dots (QDs) acting as floating gates. The QDs are precisely positioned in the input terminals and the localized charge controls the conductance of the transistors. Connecting two devices enables one to implement associative learning by tuning the QD-charge with two input signals. The number of pulses to develop or to forget the association depends on the widths and amplitudes of the applied voltage pulses. The Y-shaped geometry of the presented device may be considered to implement synaptic functionalities without separating learning and signal transmission in time.