▎ 摘　　要

The rise of artificial intelligence and Internet of Things has led to an increase in the demand for a large amount of data computing in recent years. To fulfill the requirements of neuromorphic engineering, a promising system composed of artificial neurons and synapses has to be developed. Here, a cheap and mass-productive compacted self-assembly (CSA) graphene with hydrogen (H-2) plasma surface modification is used as the bottom electrode (BE) of gadolinium oxide (GdxOy) memristors to emulate the robust synapses in neuromorphic systems. As the plasma treatment time increases, the increased resistance ratio and reduced operating voltages of Gd(x)O(y)memristors are obtained, which can be attributed to the removal of functional groups on graphene flakes and the enhancement in the redox reaction of CSA graphene during the resistive switching. The Gd(x)O(y)memristors with a 10 min H(2)plasma surface modified CSA graphene BE present outstanding reliabilities of data retention for more than 10(4)s and cycling operation up to 150 times. Additionally, superior bionic characteristics with more adjustable synaptic weight and more harmonious spike-timing-dependent plasticity (STDP) behaviors of Gd(x)O(y)memristors with H(2)plasma surface modified CSA graphene BEs are achieved, providing an opportunity for the applications in future neuromorphic computing systems.