How Physicists Created the Double Slit Experiment In Time

The famous double-slit experiment, which demonstrated that light is both a wave and a particle, has been performed using “slits in time”.

0:00 Light Can Interfere in Time as well as Space
0:52 The Classic Double Slit Experiment
2:21 The Double Slit Experiment in Time
5:05 Ad Segment
6:15 The Experimental Results

This video was sponsored by NordVPN

Where to find me:

A few people have asked so I've added the info below. Some of these are affiliate links. If you make a purchase it doesn't cost you anything extra, but a percentage of the sale will help support this channel and my work to bringing entrepreneurship into science.

[donseesyourshaydim]

If you get 200 more views today, 2/3 of it is me rewatching

[ScienceDiscussed]

Great explaination of this really cool experiment. Also I like the addition of the motion graphics. Your videos are getting better and better. Great Work!

[tylershepard]

The interference pattern from a double slit is (essentially) the Fourier transform of the spatial pattern of the slits (integrating over a complex exponential in space), so it would make sense that it would be true in time as well. But I also am coming at this from an electrical engineering perspective where we love to apply the Fourier transform.

[davidgreenwitch]

Maybe the situation can be explained easier.
What is important - and when thinking about it, should wonder more people - is the question why light is actually bending at the two splits in the classical experiment. Because this makes the result clearer.

The reason why a laser beam can behave like a ray until it hits the split and afterwards like a wave, is the Heisenberg uncertainty principle.
The moment you define the location with high precision, the more you mess up the impulse. Thus, with the light passing a very tiny (certain) split, you create an impulse pretty random (uncertain).
That is why suddenly the light can change the direction and move to the "side" instead of continuing its former trajectory. It is then when it can interact with another wave (even with its own) and get "diverted" (simplified speaking).

With the temporal splits, it is the same. The shorter the impulse of the photon is, the more specific its location is defined. However, that gives a more random impulse.
This is already well documented for single impulses. A "pure" red light (or laser of a single wavelength) in super short impulses (e.g. photons of synchrotron emitters) is located at a certain location and thus changes colors - since a different (uncertain) impulse means an uncertain energy. Another energy in light means another color. So if your light pulse is just short enough, it may change from red to green or so.
This is not new, but might help understanding the next step.

Like with the two waves in the spatial experiment, the two waves in short temporal distance can interfere with each other. This interferance is the pattern we see.

Maybe that helps a bit.

[theoreticaltherapy]

"Light travels through all possible paths at the same time" --- that is deep

[kasuha]

There's no surprise about broadening the spectrum on short pulses of light, it's similar result to when you do a fourier analysis on short pulse of sound - even if the sound itself is constant pitch at single frequency, the transition from "no sound" to "sound" and back introduces whole spectrum of sound frequencies that need to add up to form such a pulse. I'm somewhat curious how they arranged things for the two pulses to interfere, though. So It could be equally interesting to see the description of the instrument they used to detect the pulses and different frequencies of the result.

[MathIndy]

The Heisenberg uncertainty equation is usually written as (delta-x)(delta-p)>h/2*pi but you can also easily rearrange the Heisenberg equation so that, rather than position and momentum, it instead refers to energy and time. That is (delta-E)(delta-t) on the left side but remember that a photon's frequency is directly proportional to its energy (E=hf). So, in the traditional double slit experiment the delta-x is confined to one of two slits so the uncertainty in the lateral momentum must increase (two probability waves spread out and form a spacial interference pattern). From the (delta-E)(delta-t) point of view if you confine the (delta-t) to two time slits, then a similar thing must happen except now the two uncertainty "waves" are in the E=hf frequency. This creates two interfering frequencies and the associated beat pattern that is observed.

[TerryBollinger]

This is a fascinating and well-done explanation, but my brain keeps yelling, "Aren't these just Fourier transforms stretching wave packets? And similarly, doesn't the frequency spread for the same reason that sharply banging a gong produces loads of frequencies?" Yes, there is time uncertainty, but doesn't that also stretch the wave packets in length, allowing a more mundane explanation? I loved the potential for that femtosecond cutoff; that's one of those nifty new-tech enables that could go in very unexpected directions. Very cool.

[paulweiler]

A trillion times faster than the blink of an eye WHAT

[ThiloDiver-rkpz]

Completely missed the point...
This is just equivalent to the creation of spectral "sidebands" on a radio frequency carrier in case of amplitude modulation at frequencies not very much smaller than the carrier frequency.
The probe laser represents the carrier with long coherence length and the femtosecond double pulses create an amplitude modulation which creates the optical sidebands in the spectral domain. And because of the longer coherence length they of course(!) overlap to create the optical spectrum of a double pulse.
This, in fact is trivial, just check the well-established "pulse shaping in frequency domain" technologies. There even are holograms with spectral properties to create pulse sequences from a single readout pulse or vice versa. This was already demonstrated back in the 90ties, when femtosecond lasers became available...
Just a consequence of the optical Fourier transform from time to frequency and back.

[GoingToBeWild]

Its unclear if this interference is a byproduct of the probabilistic absorption and emission of light by electrons in the reflective material or the light waves over time intervals in space 😵‍💫

[_vizec]

What about the reaction time of the sensor that “reads” the incoming light? Could it just be that the material the sensor is made out of is entering a harmonic with the light and “re-releasing” it at a different frequency, only to get trapped and “re-read”? Like how atoms release infrared to describe heat? Maybe this is just the “heat equivalent” of a harmonic. This could just describe the interaction between light and matter, not what light actually is (particle, wave, wave-particle). The regular double slit experiment’s harmonic function would just be based on the size and distance the slits are from each other. Either way this interaction could be very valuable if understood fully. I can imagine a “light calculator” by varying the size (or duration of pulses for this experiment) of each slit against the other and summing or running other algorithms with the results.

[Hansulf]

Wait... Couldn't you make an optical transitor like this?

[anywallsocket]

The ‘photon’ doesn’t go through both slits (in space or time), the ‘photon’ is the absorbed (detected) state. The wavefunction is what goes through both slits, either because it is dynamic while spacetime is fixed, or (more likely) because it is fixed while spacetime is actually dynamic.

[HuygensOptics]

I'm not even sure what it means, interference in the frequency domain. Or how that could come about from 2 pulses that do not overlap in time. If you want to study these phenomena, you need some kind of grating to separate the different frequencies, which introduces path length differences. In 2.3 ps, light only travels 0.7mm, so part of these pulses might actually overlap temporally at the detector if you don't take extreme care to avoid this.

[williambunting]

That is very interesting. I am very interested in the progress towards optical computers. I see a future where a computer is a specially grown crystal where the processing path bounces back and forth through the crystal between switching devices, and through cavities to allow for processing time variations. This will give better protection against “jamming” of computer operations. More funding to your field!

[aaronsmith]

What I'd like to know is if the pulse laser just shot a single photon at a time, would you still see the quantum probabilistic interference pattern? Like in the classical two slit experiment. Of course, I'm guessing you would.

[haros]

Well the word time traveling is kind of clickbaity. It shows wace particle duality behaviour in the temporal dimension, which obviously isn't quantified, the planck time isn't the world is in frames per second, just current physics break beyond it. Anyways this does NOT imply retrocausality. The delayed quantum eraser would, if it wasn't Debunked. And how retrocausality would even be possible if quantum decoherence and fluctuations make the future fundumentally unpredictable. It going back in time, even in theory (aside the infinite energy needed) is impossible.

I see commenters and fans of superdetermenism get annoyed by quantum indeterminenism and it "violates causality" while accepting science fiction like retrocausality. Indeterminenism doesn't violate causality, retro does

[earthone]

Also time is emergent from events. How do you factor this into the experiment?

[earychow]

Let's not confuse the term color with spectrum, the term color specifically is always about human visual perception. When you say color you are talking about 1. Stimulus (CIE Standard Human Observer model) or 2. Cognitive color perception (Abney Effect. etc.), and both has NOTHING to do with the time slit experiment.

For example, when you said broadening of colors at 8:00, I assume that you're actually talking about broadening of spectrum, this is different because generally a broader, less-pointy-curve-graph spectrum produces a less chroma (or less saturated) color when it comes to human perception, If you broaden it to the extreme where it's a flat line then you get illuminant E, which is one of the standard "white"s.

So when you're talking about spectrum just call it spectrum or spectral distribution. Don't talk about color when you don't mean human perception.