Quantum Optical Circuits with Structured Light

Programmable optical circuits are crucial for advancing classical technologies such as optical communication, optical computing, and AI. In parallel, these circuits have pivotal in furthering quantum technologies, enabling quantum measurements, entanglement generation, and photonic quantum computing. Conventional designs for programmable optical circuits use a bottom-up approach, where a desired circuit operation is achieved with a sophisticated mesh of two-mode interferometers. While this approach is widely adopted, the high fabrication tolerance of and precise control required over each circuit element limits its scalability.

In this talk I will discuss an alternative top-down approach to programmable circuits using spatially structured light. Aided by inverse-design techniques, optical circuits are embedded within the higher-dimensional space of large, ambient mode-mixers such as a complex scattering medium and propagation through free-space. First, I will present a method to characterise complex media such as multi-mode fibres (MMF) using multi-plane neural networks (MPNN), showing how it is more accurate and noise-robust than conventional techniques. Following this characterisation, I will show the bespoke unitary and non-unitary optical circuits being implemented within a commercial MMF. As on example, we employe these circuits as quantum gates to manipulate up to seven-dimensional entanglement, facilitating both the transport and certification of entanglement within the transmission channel.

A type of free-space optical circuit known as a multi-plane light converter (MPLC) will be the second focus of this talk. After introducing this platform, I will discuss our work on using it to unambiguously sort seven non-orthogonal modes of light, leading to classification of simple images. Finally, I will discuss our work on multiplexing emission from multiple quantum dots using an MPLC, observing cooperative emission from up to five quantum dots, and measuring two-photon interference from two quantum dots on the same sample. These methods lay the groundwork for optical information processing with structured light, with diverse applications such as optical neural networks, high-dimensional quantum communication, and quantum computing.