| ||||||||
28 November (2024)
26 November (2024)
Nice work from Camilo of pulsed quantum excitation.
21 November (2024)
After the Chaos, the harmony of the mountains.
laussy.org/wiki/Hiking_in…
laussy.org/wiki/Hiking_in…
Miguel's Excited state quantum phase transitions spilling all over the place.
Alternative Theory Seminar @icmm_csic in 35 minutes. You can already challenge the speaker @MBastarrachea in the Seminar room.
icmm.csic.es/en/alternative…
icmm.csic.es/en/alternative…
20 November (2024)
The C of BCS and of BCM... heisingberg.eu/2024/11/18/the…
18 November (2024)
🔑«Unlocking multiphoton emission from a single-photon source through mean-field engineering» or playing hide and seek with photons in resonance fluorescence. A thread on Sang Kyu Kim et al.'s work over the last years, now in arXiv:2411.10441.
If you like to think of a single-photon source as a two-level system (2LS) that gets saturated and whose reloading time prevents from emitting more than one photon at a time, then I refer you to nature.com/articles/s4159… & pubs.aip.org/aip/apq/articl… for a better view of this paradigm.
Here we discuss instead the mind-boggling physics that occurs when the two-level system is coherently driven, but so weakly that the saturation of the two-level system σ plays no role. This is known as the Heitler regime. The light is coherently scattered but remains antibunched.
Since it's not the saturation that spreads apart the photons, what is? An interference! Decompose σ into its mean field <σ> and its fluctuations ς (σ=<σ>+ς), and you'll see that intensity correlations involve interferences between the classical |<σ>|² and the quantum <ς⁺ς>.
<σ> is boring, just a weak coherent state. But ς, by Jove, will trick the most seasoned experts. This quantum field boils with virtual photons, whose quantum amplitudes interfere out to result in single photons. That's how you get quantum light from scattering a laser off a 2LS.
Now the step from Maths to Magic. Remove <σ> (it's easy since it's classical; destructive interferences will do the job) and the virtual photons spill out from the quantum field: remove the coherent fraction and your single-photon source becomes a multiphoton one!
We're looking into an abyss here and one needs special vision to see naked quantum fields. This special vision is Glauber's multiphoton correlators g⁽ⁿ⁾ that tell you how much more (>1) or less (<1) n-photon coincidences you get compared to a classical field of same intensity.
That's what bathophobia looks like. ℱ is the external laser disrupting the interference. At ℱ=1—with only quantum fluctuations—we dive into the abyss: the intensity〈s⁺s〉vanishes as it comes with arbitrarily more multiphotons than if it was classical light: g⁽ⁿ⁾→∞∀n.
You don't have to touch ∞, just get close enough asymptotically. Here are some numbers: for Ω≡Ωₛ/γₛ—ratio of driving to emission rate—of 0.1, you expect ≈17k 2-photon/s for γₛ=1/216ps but instead get 3 millions/s. You expect only ≈16 3-photon/s but get ≈0.1 million/s.
Go deeper into the abyss with Ω=0.01. Your signal dims considerably, but its multiphoton deviations increase: instead of ≈2×10⁻⁴ 2-photon/s you now get 370/s and instead of 2×10⁻¹¹ 3-photon/s you get 0.2/s! A huge amount of multiphotons in comparison to expectations!
If you understand signal not as a mere intensity, but as this anomaly, this observation of multiphotons that were not expected, as clicks on photon-coincidence units, you can swim in the quantum vacuum and, with ℱ as a helm, navigate quantum fluctuations by tuning the phase.
Such deep immersion into extreme quantum depth has been realized in Kai Müller (left)'s group at @TU_Muenchen as part of Elena del Valle (right)'s Hans Fischer fellowship there, in collaboration with Sang Kyu Kim (experimentalist), Eduardo Zubizarreta (theorist) et al. inter me.
And after about a year of hard work, this was successful. This picture is from the early days with Sang Kyu still healthy-looking and smiling as he's insulating his setup to be stable enough to wipe out the huge mean field without also blowing out the small quantum fluctuations.
This is right after glorious days in Madrid (including Días del Arroz at the cafeteria) where he perfectioned phase stabilization with Carlos Antón @AntnSo @UAM_Madrid to the extreme point needed to undress the quantum dot down to its full frontal ς, with not even its socks on.
And back to München, this is Lukas Hanschke, whose seminal experimental demonstration of the origin of antibunching as an interference [https://t.co/8YCtJwciX8] started the quest to seize control of what Lukas had instead disrupted more curtly and more bluntly (with filters).
This is what they saw after weeks of immersion in this soundproof cubicle: 3-photon coincidence maps with, at ℱ=0 (no external field), standard single-photon emission carving antibunching in the random background and, at ℱ=1, its negative where multiphotons now stand on vacuum!
Let me zoom on this picture of the vacuum fluctuations: there are no clicks (blue=0) as most of the intensity is gone with the coherent fraction so only random events provide a noisy background, except on diagonals (Δ)τᵢ=0. At τ₁=τ₂=τ₃, however, BOOM, a huge multiphoton peak!
Here is how those multiphoton peaks evolved as they tuned the laser ℱ to control the interference, compared to the theory expectation for their 0.15 driving. Black is intensity: you see it drop considerably at ℱ=1 as two and three photons start to arrive in anomalous amounts.
Agreement is not perfect but for such low driving, they had to integrate over long times and the setup started to drift. At higher driving, the agreement becomes perfect but correlations are less strong: farther from the bottom of the trench, quantum deviations are smaller.
We're speaking of days of integrations, and 0.15 is not even ≪1. There's a counter-intuitive quirk that the smaller the driving, the smaller the ratio of quantum over coherent signal yet the stronger the quantum deviations, like M&W's homeopathic lagger: youtube.com/watch?v=QAboul…
I focused on ℱ=1 where the system reduces to its quantum fluctuations, but at 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 values of the 𝑐𝑜𝑛𝑡𝑖𝑛𝑢𝑜𝑢𝑠 field ℱ, even fishier effects take place! If you scroll up again, you can see that instead of superbunching ∀n, we have antibunching for selective n.
This, they also observed (cf. inset above) and the physics is also crazy, first confirming that multiphoton observables are independent, e.g., you can have joint g⁽²⁾≪1 (2-photon suppression) and g⁽³⁾≫1 (3-photon enhancement) or vice-versa, or together. They saw everything.
I won't discuss now even more shocking results, like multiphoton amplification or high-intensity external fields never killing the effect but transferring it to arbitrary-high Fock |n〉instead. Diving that deep into quantum fields must be done with decompression steps on Twitter.
So I refer the experts to the paper instead arxiv.org/abs/2411.10441 and will come back to the really dizzying rapture of the deep in a subsequent post (and theory paper).
heisingberg.eu/2024/11/18/the…
A blog post we wrote with Natalia at the occasion of her visiting the @icmm_csic on this year's Physics @NobelPrize by @HopfieldJohn and @geoffreyhinton.
A blog post we wrote with Natalia at the occasion of her visiting the @icmm_csic on this year's Physics @NobelPrize by @HopfieldJohn and @geoffreyhinton.
Note that, as discovered by Hofpield, your own neural network will be able to retrieve blog from bolg!
17 November (2024)
Replying to @aandr314
You'll see that we have a peculiar vision of what "signal" is.
16 November (2024)
Replying to @CristianoCiuti
Gorgeous. Looks like a renaissance painting.
15 November (2024)
Replying to @david__co
surtout depuis l'hypothèse du continu...
Replying to @AntnSo and @UAM_Madrid
doesn't look great for setup stabilization...
Replying to @GDNonline
Restoring dignity to parliamentarism. Vonnegut would have loved that.
14 November (2024)
Replying to @polyquantique
arXiv time arxiv.org/localtime — Sunday 20:00 EST (that is Mon, 18 Nov 2024 01:00 UTC).
On Monday I'll tell you about our good old problem of single-photon emission, which never stops to surprise us.
inspiring wisdom: «Have confidence in your ability to change the world - but never lose sight of the privilege you have.» Applies to all of us scientists, not only Royal fellows.
13 November (2024)
Back to school! (but next year) qoqm.nanotec.cnr.it
11 November (2024)
Replying to @pmcouteaux and @radiocourtoisie
Condoléances à la France.
9 November (2024)
Visit of a @_HEISINGBERG_ superstar to our beautiful Cantoblanco campus (her first time here, can you believe it?) She'll be back. Will tell you more about it later.
8 November (2024)
Replying to @EmmanuelMacron
La diplomatie de Oui Oui.
7 November (2024)
La France est tombée bien bas. La plus ancienne nation occidentale, réduite à l'état de petite pute de parking. Merci à ce monsieur d'avoir tenté un peu de dignité. On a connu des chevaliers Bayard plus tenaces mais après tout, pourquoi se sacrifier pour un pays vidé de son âme.
6 November (2024)
Again, and particularly in this case, success is responsibilities and duties. The world is on fire and he promised to stop it overnight. With this incredible comeback, he also showed that nothing is impossible. I wish him to make the peace Nobel prize great again.
4 November (2024)
Prof. Natalia Berloff @Cambridge_Uni will visit our @icmm_csic this Thursday to tell us about computing with non-Hermitian driven-dissipative systems:
wp.icmm.csic.es/seminars/2024/… @_HEISINGBERG_
wp.icmm.csic.es/seminars/2024/… @_HEISINGBERG_
3 November (2024)
Replying to @InstantVintage_
Et Langlois, dans «un Roi sans divertissement». Quel acteur, tellement au dessus de lui même qu'il semblait toujours être quelqu'un d'autre.
1 November (2024)
and one of 1985 per accompanying adult.
Don't boycott culture. Don't boycott science. Boycott war, arms deals, the military. Not the artists, not the scientists, not the people. Even a war criminal, if he also is an artist or a scientist, this side of him, it cannot be condemned, ignored or erased. The evil sides must.
Replying to @SNQSFinleyGroup
Prof. Finley knows a lot of the scariest stuff. Always good company, even at Halloween.