Historical research guide · 0630 · 2022
Realizing a rapidly switched Unruh-DeWitt detector through electro-optic sampling of the electromagnetic vacuum
Sho Onoe · Thiago L. M. Guedes · Andrey S. Moskalenko · Alfred Leitenstorfer · Guido Burkard · Timothy C. Ralph
In one minute
Electro-optic sampling is the ultrafast-optics trick of firing a light pulse shorter than one oscillation of the field you want to look at, then reading that field off the pulse’s polarisation. Sho Onoe and Timothy Ralph in Queensland, with Thiago Guedes, Alfred Leitenstorfer and Guido Burkard at Konstanz and Andrey Moskalenko at KAIST, show that this laboratory technique is the same thing as an Unruh-DeWitt detector — the idealised particle detector theorists use to ask what an accelerating observer sees in empty space. The strong probe pulse plays the part of switching the detector on and off inside a fraction of one cycle, and switched that fast the detector registers real photons out of the vacuum: the virtual particles the uncertainty principle allows for very short intervals are handed over as countable excitations of the probe. The authors locate the crossover — which detected frequencies give genuine detector behaviour and which give ordinary squeezing — and specify the crystal, the pulse and the optics that would put it on a bench.
Connection
Chapter 2 argues the vacuum is a structured medium you can measure rather than a bookkeeping convention, and this paper turns the most sensitive vacuum probe in ultrafast optics into a detector whose readings are stated in the language of quantum field theory on a spacetime metric. It belongs with chapter 5 too, because it is a laboratory system standing in for curved-spacetime physics: the fast switch acts as a horizon, and how sharp that horizon is decides whether vacuum entanglement can be harvested.
From the source
The nonlinear interaction behind electro-optic sampling is rewritten as an Unruh-DeWitt detector. Driving a zincblende-type crystal with a strong coherent pulse shorter than one cycle of the sampled field gives an interaction unitary of beam-splitter form between the probe mode and a Gaussian subcycle mode of the vacuum, the same form the idealised detector takes when its coupling is switched on and off inside a fraction of a cycle. The strong optical pump plays exactly the role the switching function plays in the detector model.Section IV B and Section VI A; equations 30, 31 and 36 to 40
A field mode shorter than its own cycle is not empty even in the Minkowski vacuum. Because such a mode is broad enough in frequency to contain both positive- and negative-frequency components, the expectation value of its photon number is nonzero and grows as the pulse gets shorter; the authors attribute those photons to virtual particles, which the uncertainty principle permits for correspondingly short intervals. In their numerical plot the two quadrature variances begin to depart from shot noise once the Gaussian envelope reaches roughly 0.45 of the inverse carrier frequency.Section III B, equations 21 to 24 and Figure 2 with its caption
Two regimes are separated, and the boundary between them is calculated. Using what the authors call the first-order unitary evolution, the electro-optic interaction is either a beam-splitter exchange or a two-mode squeezing. Above the probe carrier frequency it is the beam-splitter case, and that is the genuine detector regime in which vacuum virtual particles are promoted to real near-infrared excitations; below it the process is ordinary squeezing, and the counted photons are one half of a created pair. Deviations of the quadrature variances from unity in the detector regime are the signal of sampled virtual particles.Section VI B and VI D, Figure 5; Conclusion, second paragraph
The measurement is specified with laboratory numbers. The probe is centred at 255 terahertz with a 5.8 femtosecond intensity full width at half maximum, focused to a 3 micrometre waist radius and carrying about 5 times ten to the ninth photons; the crystal is 7 micrometres of zinc telluride with an electro-optic coefficient of 4 picometres per volt; the detection band is 1 terahertz wide. The readout is ellipsometry: a band-pass filter, a phase-shifting waveplate, a half-wave plate at one eighth of a right angle, a Wollaston prism splitting the two polarisations, and a photon counter on each output.Section VI C and VI D; Figure 4 with its caption
The first-order unitary treatment predicts quadrature variances below the shot-noise level at some detected frequencies, where ordinary first-order perturbation theory does not. Sub-shot-noise variance is one of the trademarks of quantumness, so the authors put it forward as the working regime in which the existence of virtual photons carrying quantum correlations in the electromagnetic ground state can be verified experimentally. That measurement is the one to watch.Section VI D, final paragraph; Abstract, final sentence
Fast switching acts as a horizon, and the shape of the switch decides what can be harvested. In Unruh-Davies and Hawking radiation a horizon separates the two members of a virtual pair and leaves them delocalised and entangled; switching the detector on and off quickly does the same job. With a Gaussian pump the horizon is soft, so the probe picks up both partners in the same mode and the authors see no strong decoupling — the squeezing visible in their two figures is the signature of that. Searching for pump profiles sharp enough to observe vacuum entanglement between regions of spacetime is the research direction they name.Conclusion, final paragraph
The way in
https://doi.org/10.1103/PhysRevD.105.056023How to cite it
Sho Onoe, Thiago L. M. Guedes, Andrey S. Moskalenko, Alfred Leitenstorfer, Guido Burkard, Timothy C. Ralph (2022) Realizing a rapidly switched Unruh-DeWitt detector through electro-optic sampling of the electromagnetic vacuum doi:10.1103/PhysRevD.105.056023