PBS Space Time

Whenever we open a new window on the universe we discover something new.

Whether it's figuring out how to see to greater distances like with telescopes, or down to smaller size-scales like with microscopes, or perhaps expanding our vision to new wavelengths of light or via exotic means such as in neutrinos or gravitational waves.

Well, the 2023 Nobel prize in physics has been awarded to three physicists for opening just such a new windowbut it's not a window to a new size scale or a new mode of seeing-its for a new window in time.

Its for attosecond physicsthe billionth of a billionth of a second that represents the timescale of the insides of atoms.

This years Nobel in physics is for a microscope in time Every 230 million years the solar system completes one orbit around the Milky Way.

Every 243 years Venus passes between the Sunand the Earth.

Every year monarch butterflies migrate betweenthe United States and Mexico.

Every 3seconds a Kinesin protein travels down one of your cytoskeletal filaments.

As we look to smaller and smaller scales, we find faster and faster processes.

Makes intuitive sensebig things surely have a hard time moving faster than their constituents.

The trend continues to the smallest scalesthe motion of individual atoms during chemical reactions, or of individual electrons moving in atoms.

The timescale of their motion is not measured in microseconds nor nanoseconds, they're measured in attoseconds.

And this is a very small amount of time indeed.

There are as many attoseconds in a second as there are seconds in the entire history of the universe.

Probably there are some things to be discovered buried in that universe of phenomena that play out in every single second.

And this is why the Swedish Academy of Sciences made a very reasonable choice in awarding Anne L'Hullier, Pierre Agostini and Ferenc Krausz the 2023 Nobel Prize in Physics for bringing into existence the field of attosecond physics.

Before we get to how our newly minted laureates managed this, lets talk about why it was so difficult in the first place.

Think about what it takes to observe something.

Observing means watching things bounce off other things, whether its light bouncing off an apple into our eyes or electrons bouncing off a tardigrade into our electron microscope.

The smaller the object you want to observe, the smaller the particle you need to fling at it.

For example, you could get a rough idea of my shape by hurling beach balls at me and seeing how they scattered.

But youd get a better picture pelting me with ping pong balls.

Or even better, photons.

Resolving small distances requires scattering many particles with small spatial separation, into a well-defined spatial location.

Resolving very small times requires scattering many particles with small separations in time and well-defined temporal locations.

Imagine you want to take a video of a fast processsay, a hovering hummingbird.

If your camera aperture stays open too long youll only capture a blur over the span of the wings motion.

The aperture opening needs to have a well-defined temporal locationin other words, it needs to be quick.

But quick exposures arent going to give you a very good movie of the humminbird if they arent separated by less than a wing beat.

Then youll only get a random seeming collection of stills.

You need short temporal separation to get a meaningful sequence.

So how do you get tight temporal location and short temporal separation down to the attometer?

Lets start by thinking about the temporal location of a photon.

Thats limited by its periodyou cant clock a photon to any time smaller than a single up-down cycle of its electromagnetic wave Light with a period of attoseconds is in the X-ray part of the electromagnetic spectrum.

We routinely use X-rays in imaging, so we should be fine there, right?

Not quite.

A typical laser fires one photon every few femtoseconds - thats 1000 times longer than an attosecond.

So even if you make a movie of attosecond motion with a conventional laser, the abysmal frame rate would turn the motion into a hopeless blur.

It is possible to build extremely fast lasersfor example, Free-Electron lasers But they require an electron accelerator to work, and they are honestly too powerful and way too dangerous to give to the average physicist to play around with, and besides, blasting your sample with an X-ray laser of the power required would be like taking an x-ray using a nuclear explosion, it's just not practical.

Enter our first Nobel laureate.

Back in the 1980s Anne LHuillier and her colleagues were playing around with irradiating Argon gas with infrared lasers.

They observed something pretty weird.

Normally, when you irradiate a gas youll get different types of light out.

That includes some of the light you put in, some light corresponding to the electron transitions of the atom, as well as some random thermal light from the jiggling of the atoms.

But LHuilliers team noticed that the outgoing beam consisted of the original in-going frequency plus some higher frequencies that didnt correspond to any known process.

So where did these other frequencies come from?

Its pretty cool actually.

As a laser pulse passes by a gas atom, it can nudge the electromagnetic field holding the electrons in place in such a way that electron escapes by quantum tunneling.

The electron can then be pulled straight back toits atom and will release the energy it gained in a single photon, think of it like a stretching a spring and letting it go.

That one photon will have a higher energy than each of themany photons that made up the laser pulse.

In fact itll have a frequency thats some integer multiple of the laser photons, and a cool detail is that it'll be an odd multiple, for complex reasons regarding symmetry This process is called high harmonic generation, and it basically adds overtones to the laser beam.

Play a note on any musical instrument and youll get the fundamental frequency plus higher frequency overtones.

The fundamental corresponds to the longest wavelength standing wave that can fit along the resonating string or air column, while the overtones correspond to all shorter wavelength standing waves that also fit.

The balance of the strength of the overtones determines the precise character of the sound of an instrumentits timbre.

So, the timbre of the laser was altered by the argon cloud.

Thats a cute effect, but were not quite at Nobel-level science just yet.

To get to attosecond physics, we need one more analogy with acoustics.

Listen to this.

Its a very low G note.

And this is a very low A.

Individually theres nothing unusual, but if we play them together we hear a sort of "wawawa".

This is known as a "beat" in acoustics, and its due to the fact that the sine waves of the G and the A line up perfectly at certain points, making them louder with constructive interference, while theyre perfectly out of alignment elsewhere, and so cancel out with destructive interference.

For a pair of waves the beats are quite spread out, with a lower frequency than both of its parents.

But if you add more and more waves, its possible to narrow the width of the beats so you end up with sharp pulses and very little in between.

For that you need a large number of frequencies of similar intensityfor example, the rich spectrum of overtones produced in Anne LHuilliers experiment.

In fact, it proved possible to get pulses that were mere hundreds of attoseconds in their temporal width.

Our new attosecond-resolved pulses werent ready for application just yet.

The potential value is that they would allow us to measure attosecond-scale events.

But if the pulses themselves are attosecond-scale events, how can we calibrate them to start with?

It's like trying to measure your height with a series of rules whose individual length you dont know.

And here we get to our second Nobel Laureate, Pierre Agostini.

Agostini was able to calibrate the pulse train by again using constructive and destructive interferencethis time in reference to the ingoing laser beam.

He did this by deflecting part of the beam and adding a delay to it before recombining it with the now-frequency-multiplied beam.

In this way he could measure the width of the pulsesclocking them at 250 attoseconds.

He also found that the pulses were what we call phase locked, which means the beats were nice and consistent and just what we needed for attosecond measurements.

These pulse trains provided pulses with attosecond temporal locality and separation, just as we required.

But for some applications it would be preferable to have single, isolated attosecond pulses.

And that effort is thanks to our third Laureate, Ferenc Krausz.

I wont go into the gory details, but for your viewing pleasure heres their Rube-Goldberg experimental setup.

With intricate phase and amplitude manipulation, they were able to create isolated pulses of 650 attoseconds, whose width was known to 150 attosecond precision.

OK, attosecond resolution achieved.

So whatcan we do with it?

As Krausz himself states in an interview, they invented this technology because looking at Nature in a new way is wonderful, literally, it fills you with wonder.

But we did just give ourselves a new superpower, so itd be a shame not to use it.

The first application of attosecond pulses was to look at electron motion in atoms and molecules.

Electrons travel the breadth of their orbitals in a handful of attoseconds, or rather the fuzzy quantum clouds that define the electron in an atom change on that timescale.

By hitting these clouds with attosecond pulses we can study the shapes and dynamics of these electrons.

Attosecond pulses can also be used to manipulate electrons on tiny timescales, which has a number of powerful applications.

One effort that Ferenc Krauszs team is working on at the Max Planck institute for Quantum Optics is in molecular fingerprinting, in which attosecond pulses are frequency-tuned to cause vibrations in specific molecules.

In this way the detailed molecular composition of a sample can potentially be cataloged.

The Krausz team are using this to develop molecular fingerprinting devices for medical diagnosis.

Another very exciting possibility is the creation of ultrafast electronics.

If we have two metal plates with opposite electric charge and we shoot an isolated attosecond pulse at one of the plates it can be absorbed by an electron, which hops to the second plate.

This is the photoelectric effect, and weve known about it for 120 years.

But in the configuration I described its also a transistor, a key component in most of our technology.

Regular transistors control the flow of electricity between two charged plates by changing the charge in a third plate, but this new type of transistor controls the flow using light itself, which can in principle be much, much fasterespecially if the pulse width is measured in attoseconds.

Krausz claims that it may be possible to increase the power of computers by a factor of 100,000 this way.

That sounds optimistic.

But even if we get a fraction of that its an incredible achievement, and may save Moore's law for a few more decades.

Those are just a couple of applications in medicine and electronics, but as we look at the universe with this new tool we are bound to find new applications and to make new discoveries.

After all, every time we open a new window to the universe new mysteries are revealedwhether that window is to the largest, the smallest, and now the fastest phenomena in Space Time