Control and manipulation of the pulse-mode structure of light in the classical domain often works on principles of amplitude filtering and gain, which do not preserve the fragile quantum features of non-classical states. For example, using a spectral filter to generate narrow bandwidth pulses from initially broader pulses works well with classical lasers, since photon statistics of laser light are robust to loss. However, generating a single-photon pulse with a narrowed spectrum from an initially broader single photon by spectral filtering does not work; it introduces vacuum and leads to a mixed output state. This example highlights the key characteristics required to manipulate the spectral-temporal mode structure of quantum states – lossless unitary transformations of the field. Here we are developing various approaches to coherently control the pulse-mode structure of light that preserve its quantum nature.
Frequency Conversion in Nonlinear Waveguides
Since the colour of beam of light tells us the energy of the photons in that beam, in order to manipulate the information encoded in quantum rainbows
, we need tools and techniques which change the energy of photons. One of the ways we accomplish this is using a nonlinear process called sum-frequency generation
. In this process, a single photon mixes with a strong laser pulse in a special crystal where they may combine to form a photon at a higher energy; for example, a green (medium-energy) photon might combine with a red (low-energy) laser beam, and we'd detect their combination by measuring a violet (high-energy) photon after the crystal.
By engineering the properties of the crystal where this interaction takes place, we can create devices which allow us to not only change the colour of photons, but also their shape in time and colour. As part of the QCUMbER research programme, we have developed techniques in long Lithium Niobate waveguides, increasing the interaction between the photons and the control laser pulses. With this technology, called the quantum pulse gate, we can select and reshape complex pulse shapes with features faster one trillionth of a second. The technique is easily reconfigurable to custom pulse shapes by changing the shape of the strong laser pulse. This enables us to construct and access a large quantum alphabet for communication and computation, and phase-sensitively probe ultrafast quantum states.
If you have questions, please feel free to contact John M. Donohue from the University of Paderborn.
Quantum memories are first and foremost thought of as devices that store a quantum state of light, in our case a quantum rainbow, just as a classical memory stores classical information. It turns out however, that quantum memories can do more: they can actually store a specific quantum rainbow and return it in a different shape. This capability has many applications. You can use a quantum memory to spell a message, by imprinting “letters” on your quantum states. Alternatively, you could use them as a translator for different alphabets, allowing different devices to talk to each other.
The quantum memory we have developed during QCUMbER is based on a hot gas of Caesium atoms. This has several benefits: first, we do not need a complex setup (e.g. a “super fridge” or cryostat) to operate our memory; second, we can use broadband photons, which makes our memory compatible with high clock-rate networks; finally, we can operate our memory with simple lasers, allowing for a low-cost and compact packaging in the future.
There is still work to be done before our memory will be ready for commercialisation, but we have taken the important first step. We have demonstrated that our memory operates on only one user-chosen quantum rainbow and we have shown that we can interface different quantum rainbows. Currently, we are taking the next step by reducing the quantum noise level of our memory to a point where we can interface it with genuine single photons.
Question? Ask Benni Brecht from the University of Oxford.