Wavefront shaping offers the possibility to compensate for the effect of propagation through heterogeneous media. However, when using a single or a few photons, the feedback signal is typically too weak to allow real-time wavefront shaping applications, which limits applications for quantum communications using entangled photons. In this paper from the team of Yaron Bromberg at the Hebrew University of Jerusalem, the authors overcome this challenge by using as feedback the classical signal of the pump that follows the same path as the entangled photon. It allows adapting in real-time the pump wavefront to compensate for the aberrations/scattering introduced by a heterogeneous dynamic sample.
Scattering is known to destruct the information encoded in photons, this is true for classical and quantum communications alike. With classical light, emerging tools using wavefront shaping techniques allow compensating for the effect of disorder, enabling for instance the focusing of light at given targets, even in the presence of strong scattering. For quantum applications, it is possible to control the wavefront of the profile of the pump beam that stimulates the generation of the single-photons in a non-linear crystal in order to control the correlation of the entangled photons. However, to compensate for dynamical scattering, it requires a feedback signal and a modulation scheme faster than the decorrelation of the medium. Using single photons, one only has access to a weak feedback signal, that requires long acquisition time to cancel the noise, forbidding fast feedback optimizations. One can replace the weak single-photon signal by a bright laser beam at the same wavelength, but it then requires switching between the two sources, again forbidding a fast feedback. The authors discovered that when the pump beam is scattered by the same random sample as the entangled photons, the spatial distribution of its intensity is identical to the spatial correlation map of the entangled photons. One can then use the classical signal from the pump as a feedback for the optimization of the correlation of the entangled photons. The setup used is presented in Fig. 1.
Figure 1. Experimental setup (A), the single count map (intensity) in the absence of diffuser (B) and the coincidence distribution (C). Image from [O. Lib, G. Hasson and Y. Bromberg, Sci. Adv. 6 (2020)].
The similarity of the pump laser intensity field and the correlation map of the entangled photons after propagation through a scattering media can be seen in Fig. 2 A and B. They then use a standard optimization procedure for the pump wavefront to focus the pump light at one given location in the far-field of the scattering medium. It results in focusing both the pump light and the coincidence of the entangled photons (Fig. 2 A and B) while not modifying the single-photon count map (i.e. the intensity pattern of the single photons).
Figure 2. Far-field intensity profile of the pump before (A) and after (B) optimization. Far-field coincidence map and single-photon counts of the entangled photons before (B and C) and after (E and F) optimization. Image from [O. Lib, G. Hasson and Y. Bromberg, Sci. Adv. 6 (2020)].
The authors then compare the efficiency of the pump based feedback scheme to a feedback using directly the coincidence rate of the entangled photons. As expected, due to the week signal to noise ratio, the later fails while the pump based approach is efficient, even with a fast-moving scattering sample (Fig. 3).
Figure 3. Experimental enhancement of pump profile (blue) and coincidence rate (red) as a function of time for a slow-moving scattering sample (A) and a fast-moving one (B) using pump based feedback. In black are shown the results for a feedback based on the coincidence rate. Image from [O. Lib, G. Hasson and Y. Bromberg, Sci. Adv. 6 (2020)].
The technique is efficient as it uses a bright classical light to optimize the properties of the entanglement properties, allowing to cancel scattering of entangled photons from a dynamically moving diffuser in real-time.