We report on the fabrication of ferroelectric tunnel junctions of BiFe0.45Cr0.55O3 as a tunneling barrier, Nb-doped (111) SrTiO3 as a bottom electrode, and platinum as a top electrode. BiFeO3 is a generic multiferroic material with perspectives for multiferroic tunnel junctions, with chromium being introduced to shift and enhance the magnetic ordering from canted magnetization to ferrimagnetism. We deposit the ferroelectric films by radio frequency magnetron sputtering, an industry-compatible synthesis method. After confirming the BiFe1–x Cr x O3 film composition and its partial crystallinity, we found possible indications of ferroelectricity through piezoresponse force microscopy. X-ray photoelectron spectroscopy together with optical band measurements provide the electronic band profile of the Nb:SrTiO3/BiFe1–x Cr x O3/Pt structures. Pulsed electrical characterization reveals resistive switching with very high fatigue resistance (>106 cycles) consistent with direct tunneling across a trapezoidal barrier for a surface fraction of the film. These results make BiFe1–x Cr x O3 a promising candidate for ferroelectric tunnel junctions in particular, as they are able to operate as artificial synapses for neuromorphic circuit tiles as evidenced by spike-timing-dependent plasticity.

While machine learning algorithms are becoming more and more elaborate, their underlying artificial neural networks most often still rely on the binary von Neumann computer architecture. However, artificial neural networks access their full potential when combined with gradually switchable artificial synapses. Herein, complementary metal oxide semiconductor-compatible Hf 0.5 Zr 0.5 O 2 ferroelectric tunnel junctions fabricated by radio-frequency magnetron sputtering are used as artificial synapses. On a single synapse level, their neuromorphic behavior is quantitatively investigated with spike-timing-dependent plasticity. It is found that the learning rate of the synapses mainly depends on the voltage amplitude of the applied stimulus. The experimental findings are well reproduced with simulations based on the nucleation-limited-switching model.

No abstract