ABSTRACT
The rapid development of artificial neural networks and applied artificial intelligence has led to many applications. However, current software implementation of neural networks is severely limited in terms of performance and energy efficiency. It is believed that further progress requires the development of neuromorphic systems, in which hardware directly mimics the neuronal network structure of a human brain. Here, we propose theoretically and realize experimentally an optical network of nodes performing binary operations. The nonlinearity required for efficient computation is provided by semiconductor microcavities in the strong quantum light-matter coupling regime, which exhibit exciton-polariton interactions. We demonstrate the system performance against a pattern recognition task, obtaining accuracy on a par with state-of-the-art hardware implementations. Our work opens the way to ultrafast and energy-efficient neuromorphic systems taking advantage of ultrastrong optical nonlinearity of polaritons.
Subject(s)
Artificial Intelligence , Neural Networks, Computer , Brain , Humans , Neurons , SemiconductorsABSTRACT
Thin layers of transition metal dichalcogenides have been intensively studied over the last few years due to their novel physical phenomena and potential applications. One of the biggest problems in laboratory handling and moving on to application-ready devices lies in the high sensitivity of their physicochemical properties to ambient conditions. We demonstrate that novel, in situ capping with an ultra-thin, aluminum film efficiently protects thin MoTe2 layers stabilizing their electronic transport properties after exposure to ambient conditions. The experiments have been performed on bilayers of 2H-MoTe2 grown by molecular beam epitaxy on large area GaAs(111)B substrates. The crystal structure, surface morphology and thickness of the deposited MoTe2 layers have been precisely controlled in situ with a reflection high energy electron diffraction system. As evidenced by high resolution transmission electron microscopy, MoTe2 films exhibit perfect arrangement in the 2H phase and the epitaxial relation to the GaAs(111)B substrates. After the growth, the samples were in situ capped with a thin (3 nm) film of aluminum, which oxidizes after exposure to ambient conditions. This oxide serves as a protective layer to the underlying MoTe2. Resistivity measurements of the MoTe2 layers with and without the cap, exposed to low vacuum, nitrogen and air, revealed a huge difference in their stability. The significant rise of resistance is observed for the unprotected sample while the resistance of the protected one is constant. Wide range temperature resistivity studies showed that charge transport in MoTe2 is realized by hopping with an anomalous hopping exponent of x ≃ 0.66, reported also previously for ultra-thin, metallic layers.
ABSTRACT
Monolayer transition-metal dichalcogenides (TMDs) manifest exceptional optical properties related to narrow excitonic resonances. However, these properties have been so far explored only for structures produced by techniques inducing considerable large-scale inhomogeneity. In contrast, techniques which are essentially free from this disadvantage, such as molecular beam epitaxy (MBE), have to date yielded only structures characterized by considerable spectral broadening, which hinders most of the interesting optical effects. Here, we report for the first time on the MBE-grown TMD exhibiting narrow and resolved spectral lines of neutral and charged exciton. Moreover, our material exhibits unprecedented high homogeneity of optical properties, with variation of the exciton energy as small as ±0.16 meV over a distance of tens of micrometers. Our recipe for MBE growth is presented for MoSe2 and includes the use of atomically flat hexagonal boron nitride substrate. This recipe opens a possibility of producing TMD heterostructures with optical quality, dimensions, and homogeneity required for optoelectronic applications.
