## SciPost Submission Page

# Engineering Gaussian states of light from a planar microcavity

### by Mathias Van Regemortel, Sylvain Ravets, Atac Imamoglu, Iacopo Carusotto, Michiel Wouters

#### This is not the current version.

### Submission summary

As Contributors: | Mathias Van Regemortel |

Arxiv Link: | http://arxiv.org/abs/1712.08012v2 |

Date submitted: | 2018-04-16 |

Submitted by: | Van Regemortel, Mathias |

Submitted to: | SciPost Physics |

Domain(s): | Theoretical |

Subject area: | Quantum Physics |

### Abstract

Quantum fluids of light in a nonlinear planar microcavity can exhibit antibunched photon statistics at short distances due to repulsive polariton interactions. We show that, despite the weakness of the nonlinearity, the antibunching signal can be amplified orders of magnitude with an appropriate free-space optics scheme to select and interfere output modes. Our results are understood from the unconventional photon blockade perspective by analyzing the approximate Gaussian output state of the microcavity. In a second part, we illustrate how the temporal and spatial profile of the density-density correlation function of a fluid of light can be reconstructed with free-space optics. Also here the nontrivial (anti)bunching signal can be amplified significantly by shaping the light emitted by the microcavity.

###### Current status:

### Ontology / Topics

See full Ontology or Topics database.### Author comments upon resubmission

We have adapted the manuscript according to the suggestions of both referees. In particular, a study has been added where we discuss the possible impact of various noise sources upon our proposed optical scheme.

### List of changes

- We have added a paragraph in the introduction to motivate our work better. It is outlined why photons generated with the unconventional photon blockade can still bear interesting nonclassical features. Furthermore, we explain why the scheme that we propose may have more flexibility than other setups proposed in this context.

- We have added an Appendix (C), with references in the main text, where we discuss the effect of possible noise sources on the setup.

- The order of Figs. 1 and 2 was changed

- Some typo's where corrected

### Submission & Refereeing History

## Reports on this Submission

Show/hide Reports view### Anonymous Report 1 on 2018-4-20 Invited Report

### Strengths

1. All remarks from my previous report have been addressed satisfactorily, except one.

### Weaknesses

1. The remark on pure dephasing has not been addressed satisfactorily

### Report

I acknowledge that the Authors have addressed all my original points. I think however that the criticism on the effect of pure dephasing has not been satisfactorily addressed. In Section C.2 the Authors compare the additional broadening induced by pure dephasing to the overall broadening, concluding that the former is negligible in typical cases. This is however not correct for unconventional blockade. In several of the original works, including Ref. 20 of the current manuscript, it is shown that UPB is destroyed when the pure dephasing rate becomes comparable to the nonlinear Kerr energy (which in the present case should be g*n_0^2). This is the result that one obtains when modeling the driven-dissipative process leading to UPB. The Authors study the occurrence of UPB in terms of an optimal squeezing analysis, which assumes a given form of the density matrix of the system, with optimized squeezing and displacement features. The point is that, this optimal state will never be achieved as the steady state of a driven-dissipative setup if the pure dephasing rate exceeds the nonlinear energy. The Authors should carry out a simulation of the driven-dissipative process (possibly restricted to few modes, for simplicity) including typical polariton dephasing rates, and hopefully show that the proposed mechanism is robust to pure dephasing in typical experimental conditions.

I recommend that the manuscript is accepted only upon this analysis, as I am afraid that the current considerations in Section C.2 are technically inaccurate.

### Requested changes

Solve the quantum master equation (within a few-mode approximation) in presence of drive, dissipation and pure dephasing, to conclusively assess the effect of pure dephasing on the proposed scheme. For this, parameters relevant to state-of-the-art polariton system should be used.

We admit that our previous analysis of dephasing was inaccurate and we would like to thank the referee for noticing this. However, we do not see the purpose of performing a hard-core simulation of the master equation for this model, which is continuous and contains a high number of particles in the condensate mode. In the new analysis, we point out that the primary effect of dephasing would be a scattering of condensate particles with phonons to nonzero momentum modes. We obtain this by evaluating the effect of the dephasing jump operator upon the system with Lindbladian dynamics. We also estimate the impact of this spurious effect, but we were unable to find accurate measures of the dephasing rate of a polariton system in the literature. In any case, a pumped excitation scheme is a way to circumvent this problem.

We hope that the referee is now willing to accept our work for publication in SciPost with this new analysis.