Abstract
Nowadays, one of the promising ways for the development of computing systems with high performance and low energy consumption is the creation of artificial synaptic devices that imitate the functions of biological synapses. Such devices have a significant potential for effectively solving problems of pattern recognition, classification, control, and the treatment of diseases of the nervous system. The work demonstrates the imitation of synaptic behavior in a composite multiferroic heterostructure based on the piezoceramics of lead zirconate titanate (PZT) and the amorphous magnetic alloy Metglas. The characteristics of the heterostructure were measured by resonant excitation of the magnetoelectric (ME) effect and applying electric field pulses of various amplitudes and polarities. The ME coefficient αE was considered as a synaptic weight, and the output electrical voltage of the heterostructure as a postsynaptic potential. The study demonstrates the possibility of simulating long-term potentiation (LTP) and depression (LTD) in the ME heterostructure, as well as spike-timing-dependent plasticity (STDP). This work shows promise for creating neuromorphic computing systems based on multiferroic composite heterostructures.
Funder
Russian Science Foundation