Optical time-frequency processor based on atomic quantum memory

Manipulation and detection of photonic spectro-temporal modes enable many quantum information protocols. The standard approach for spectro-temporal processing is to leverage space-time duality by employing electro-optic modulation combined with propagation through highly-dispersive fibers implementing a temporal imaging (TI) setup. A typical example is to perform a frequency-to-time mapping—a Fourier transform—that enables spectral measurements using time-resolving detectors. More advanced combinations of temporal and frequency modulations allow time-frequency mode-sorting that enables optimal filtering and spectral or temporal superresolution measurements. Such state-of-the-art solutions are all well suited for broadband (>100 GHz) systems, such as quantum dots or other solid-state-based setups. Feasible implementations of the protocols merging flexibility of atomic systems and temporal processing capabilities inherently require an ability to manipulate and detect temporal photonic modes with a spectral and temporal resolution matched to the narrowband atomic emission. We demonstrate a novel approach to spectro-temporal processing working in a previously unexplored regime of narrowband atomic emission. Our method is based on atomic gradient-echo quantum memory (GEM) for light that maps incoming light pulses onto atomic coherence—spin waves. We combine the GEM with a spin wave phase modulation caused by a programmable spatially varying light shift of the atomic levels used for the memory. We can imprint almost arbitrary phase profiles onto the coherence and for example, achieve an ultra-large group-delay dispersion for an optical pulse stored in the coherence. Combining this with a simple acousto-optic modulation, we implement far-field temporal imaging with <1 MHz bandwidth and a resolution of <20 kHz. Moreover, with a more advanced protocol, that combines TI with in-memory interference we are able to demonstrate a super-resolution spectrometer performing quantum-optimal measurement of the frequency difference between two emitters, achieving resolution way below the Fourier limit.