Want high-purity single photons, but with control over their temporal shape? A result from our collaboration with Marco Barbieri’s group at University Rome Tre, Heralded generation of high-purity ultrashort single photons in programmable temporal shapes, explains how, and is now published in Optics Express.
One of the bigger limitations of downconverted photon pairs is their tendency to be spectrally entangled, which can be harnessed for quantum advantages in some cases but ruins the interference of photons from distinct sources. By designing specialized sources pumped with broadband laser pulses, it is possible to engineer situations where the time-frequency correlations demanded by the energy conservation of the pump are offset by the conservation of photon momentum in the downconversion material. In these cases though, you are stuck with the photon shape corresponding to the demands on the downconversion medium, limiting their usefulness and adaptability.
In this work, we make use of a phasematching condition known as asymmetric group-velocity matching, where the pump pulse and signal photon travel together, but the idler photon walks off in the medium. This produces a situation where the signal maintains roughly the same temporal width as the pump, but the idler becomes much much longer. In the limit where the walkoff is longer than the pulse widths, the photon pair becomes approximately spectrally separable. Going one step further, the idler photon’s temporal shape is defined entirely by the amount it walks off from the pump, whereas the signal’s shape is defined by the spectral shape of the pump. In this way, custom-shaped photons can be generated by simply shaping the pump pulse, as seen for the Hermite-Gaussian shapes in the figure above.
Of course, there are problems. Due to the sinc shape of the phasematching, the idler is much purer than the signal, as we measure through second-order correlation functions. However, this balances out when the signal photons are heralded by an idler detection in the main peak.
The paper can be found at Optics Express or on the arXiv as 1711.09678. Congrats and thanks to all involved, particularly Vahid Ansari and Emmanuele Roccia, who did the vast majority of the lab work, and the KTP squad at UPaderborn for designing the source.
John M. Donohue 