Nobel Prize lecture: Ferenc Krausz, Nobel Prize in Physics 2023

Scientists won the Nobel Prize
12 followers
3 Video Views·Dec 1, 2025
Now we come to the creation of individual short landed passes and we are going to welcome
professor and director Ferenc Kraus. So please join me in welcoming him. Thank you.
Good morning ladies and gentlemen. It is a tremendous honor to be here and I thank the
Swedish Academy for bestowing it on me. It's still hard to believe. I would also like to thank
all persons in my life who supported me, helped me learn, grow and achieve. My family,
my teachers in Hungary, my co-workers in Austria and in Germany and our collaborators from all
over the world. Some of them will show up on my slides and I already now cordially thank
them as well as the many more who are not going to show up for their invaluable
contributions without which I would certainly not stand here today.
These two gentlemen played a very special role. I owe them very very special thanks and gratitude.
Arnold Schmidt's guidance and Paul Corkum's deep insight into electron phenomena have
profoundly and decisively influenced my path. This lecture is going to be devoted to how
attosecond physics opens the door to the world of subatomic motions of electrons
which were inaccessible to human observation until the turn of the new millennium.
For decades after following its discovery, this elementary particle was thought when bound
in atoms to be revolving around the nucleus. Eventually quantum mechanics revealed that
subatomic motion can only emerge by exciting an electron from its ground quantum state
to a superposition of stationary states and the resultant temporal variation of the electron's
probability density constitutes the closest analogy to classical motion exemplified here
by the dipole oscillations in the 1s2p superposition state in hydrogen.
One of the messages that I would like you to take home from my lecture is that control is
absolutely central to time-resolved observation and exploration. For example,
the control of electrons current by transistors in macroscopic
nanoscale but yet macroscopic circuits with microwave electric fields underlies
the capturing of fast varying electronic signals in macroscopic circuits. Analogously,
the control of electrons current by the electric field of light opens the door to observing
and eventually even controlling electron phenomena in atomic and even subatomic dimensions.
But how actually to control the electric field of light in order to be able to control
with this controlled force that it exerts on these electrically charged particles to control
the atomic scale electron current? Four Nobel Prize winning concepts paved the way.
Lasers utilizing non-linear optics and chirp pulse amplification can
yield ultra-short, very powerful ultra-short pulses. Whereas the control of the laser
frequency comb can lead to pulses with a controlled waveform. Controlling the waveform of light
actually means that we have to shape the electric field within its oscillation period.
What does this mean? Well, what it technically requires is an octave spanning bandwidth.
Lacking lasers with such a gain bandwidth, we have no other choice but try to broaden the spectrum
of the pulses outside the laser. And that's exactly what we can do in a wonderful device
invented in Milano by Orazio Svelto, Santo de Silvestri, Mauro Nizzoli.
We can send a short laser pulse through a gas-filled hollow fiber where it is subjected
to self-phase modulation. And this non-linear process broadens its frequency spectrum.
The result is a pulse that carries a frequency sweep with a strongly broadened spectrum.
And these pulses can subsequently be compressed by another wonderful device,
chirped multi-layer mirrors, invented in Budapest by Robert Zippert and Carpath Ferenc.
These two devices together are able to actually advance femtosecond technology to its limits,
to the limit that is imposed by nature, to the limit of a single wave cycle.
So thanks to these two technical developments,
powerful pulses consisting of a single intense oscillation cycle became available
with controlled carrier envelope phase since 2003. And with this later development,
they actually opened the door to precision attosecond metrology. But before I come to that,
I just would step back a couple of years to briefly review the first attosecond generation
and measurement experiment in Vienna. The multi-cycle driving laser pulses used in the
experiments of N and Pierre were replaced by pulses comprising a single wave cycle,
single very intense wave cycle. When focused on atoms, the extremely rapid turn-on
in these pulses implies an unprecedented rate and temporal confinement of ionization.
And as a direct consequence, a very short, very brief
emission of XUV light from the higher order harmonic generation process, particularly if we
filter the highest order harmonics, in our case with a bandpass multilayer, which also focuses
the beam to a second jet of neon gas, where the XUV pulse knocks electrons free in the presence
of the few-cycle laser field, and actually provided the first very strong experimental
indication of the existence of an isolated sub femtosecond pulse in a series of experiments
performed by Michael Henschel and Reinhard Kimberger in our laboratory, at that time
at the Technical University of Vienna back in 2001. A couple of years later, laser pulses
with controlled waveform enabled precision attosecond metrology by implementing an ingenious
concept of Paul Corkum, the concept of a light field-driven stray camera. The XUV pulse again hits
neon atoms, again in the presence of the laser field, but this time we measure the time of flight
of those electrons which are catapulted from the atoms parallel to the direction of the laser field.
Depending on the moment of release, the laser field accelerates or decelerates them,
and you see here from this equation that now physics further simplified after the electron
is liberated, the laser field is so strong that actually we can describe the motion
in a very good approximation classically. Just Newton's equation is being solved here
to calculate this delta P, which is the kick in the back of the electron to move it faster or slower,
and this very kick is depending on the moment of release, and this actually maps the temporal
profile of the XUV pulse to a final momentum distribution of photoelectrons.
Recording this XUV-induced laser field-streaked photoelectron spectra as a function of delay
results in a spectrogram which allows an ambiguous complete retrieval
of both the laser waveform as well as the complex amplitude of the attosecond XUV pulse.
So this experiment actually demonstrates how control enables observation,
this time on an attosecond timescale, and actually renders the basic tools,
the attosecond varying force that this light electric field exerts on electrons,
and the attosecond duration XUV emission available for real-time observation and control of electron
light phenomena. With these tools in place, we can set out to perform
attosecond time-resolved spectroscopy. You have learned from the wonderful talk of Anne
how actually such spectroscopy can be done by using XUV pulses to kind of trigger the motion,
like ionization, and use the basic laser field to capture the motion. Here I turn the game around
and I will show you an example where we use actually the attosecond varying and attosecond
controlled force that our single cycle laser pulse exerts on electrons to trigger the motion,
and use the attosecond XUV pulse to actually capture it, to take the snapshots with a delay
and try to reconstruct the motion from this series of snapshots. So basically we just use
this single cycle laser field to ionize an atom, actually in the same focus which we used before
to characterize the tools by attosecond streaking, and this time we put some other gas there,
krypton gas, and the single cycle laser field actually ionizes this gas, and we send the XUV
pulse with some delay through the ionizing medium and analyze its spectrum, the energy
distribution of the photons contained in this attosecond XUV probe, and when the system starts
ionizing then we see some changes in the spectrum, and that's what, with the help of some theory,
as we will see in a moment, gives us information about how the electron is actually liberated and
is coming out, is catapulted out of the atom. So obviously this kind of measurement is able to
provide direct time domain insight into what I think can be considered arguably the most
fundamental strong field phenomenon, optical field-induced electron tunneling. Here we can
observe how this process, in this case in krypton atoms, is substantially confined to the
central wave cycle of a cosine-shaped near-infrared waveform, which we have measured simultaneously
by attosecond streaking. So remember two attosecond techniques are being employed in this
very experiment, attosecond transient absorption spectroscopy, which leads to the transient
absorption spectra I'm going to show you in a moment, and simultaneously attosecond streaking
to learn precisely at which moment actually, or what is the reference on the time scale,
what is the time zero with respect to which we delay our attosecond XUV part.
One of the exciting discoveries that we learned from this experiment was that the about
700 attosecond rise time of the ion population turned out to be significantly slower than
predicted by the quasi-static ADK theory. So this kind of non-adiabatic tunneling
could not have been discovered by any previous spectroscopy. When subjecting these krypton atoms
to the strong field, actually the electrons can be liberated from two different
so-called spin-orbit quantum states of the 4p subshell of krypton,
with comparable probability. This means that the hole in the electron cloud of this krypton atom
is created in a superposition state, because the hole can be either in one or in the other
spin-orbit state. Remember we learned at the beginning of this lecture that putting the
hole in a superposition state means motion, subatomic motion. Indeed, the sequence of
transient absorption spectra that is shown here in a false color representation does exhibit
pronounced temporal variation, periodic temporal variation for increasing delay of the
XUV probe. The 6 femtosecond delay period actually precisely corresponds to the 0.7
electron volt energy separation between the two spin-orbit levels. To develop the film
from this transient absorption spectra
captured with 150 attosecond exposure, the photographer's job here is taken over by
theorists, and I would very much like to join previous speakers Anne and Pierre to emphasize
how incredibly important the role of theorists was in the emergence of attosecond science.
Here you see the result of that modeling, and the excellent agreement between theory and experiment
allows actually to reconstruct the attosecond snapshots of the hole density distribution
versus time delay, and even the initial quantum phase between the two components
of the wave packet of this hole that we create. From all this information we can completely
reconstruct the subatomic motion following tunneling ionization. Here in slow motion
we play with a temporal magnification of 10 to the 15. When rendering these motions accessible
to human observation, attosecond spectroscopy furnishes us with a temporal magnifying power
that is equivalent to the combined magnifying power of a space telescope and an electron microscope.
In the words of Eva Olsson, we can now open the door to the world of electrons.
Attosecond physics gives us the opportunity to understand mechanisms that are governed by
electrons. The next step will be utilizing them. But which applications are likely to yield,
in the words of Max Planck, the richest and most lasting gains that we can possibly draw
from this new perception? If you accept that this may be a right question, we have already
mastered what Ernie Rubik considers to be the most important and difficult task,
namely finding the right question. My tentative answers appear to be deep dives
into the delicate details of light-electron energy exchange at the nanoscale and into the
composition of functional constituents, biofluids, body fluids, and tissues of living organisms
for tackling contemporary grand challenges. Ever since the beginning of the new millennium,
advancement of process or performance was driven by miniaturization,
whilst the clock rate actually determining the speed of processing was stagnating at about 10
gigahertz. Can attosecond physics possibly contribute to exploiting the fourth dimension
for further advancing electron-based signal processing? Another contemporary
grand question relates to the millions of premature deaths on this planet year by year,
mostly in low- and middle-income countries. Most of these deaths could be prevented by
early intervention following, of course, early detection of the condition, urging the quest for
affordable health monitoring. Can the molecular composition of human blood provide sufficiently
comprehensive information to serve as a basis for population health screening, and can
attosecond physics possibly contribute to that? In order to be able to tackle such questions,
attosecond technology, based on complex vacuum systems, has to become simpler and more powerful.
Few cycles, or actually single-cycle light, recently enabled this, owing to its ability
to realize the same process that we've just studied in atoms, actually in a wide-gap solid,
namely optical field-induced electron tunneling. But why is this extreme temporal confinement so
essential here? Because no solid matter could survive, could otherwise survive, the required
critical field strength of about two volts per angstrom. Only by confining this extreme field
strength to a single oscillation cycle can irreversible breakdown be avoided. Tunneling
then occurs within a fraction of the central field cycle in a cosine-shaped waveform. This
means actually again less than one femtosecond for visible or near-infrared light.
Sub-femtosecond carrier injection replaces sub-femtosecond XUV light and thereby obviates
the need for vacuum systems for a class of applications extraordinarily important for
advancing electronics, namely the attosecond control and probing of valence and conduction
band carrier dynamics. In a primary step, sub-femtosecond carrier injection is used for
optical field sampling by implementing it in a small volume where the test field can separate
the injected electrons and holes towards two electrodes and drive actually a current in an
circuit connecting these two electrodes. Here you see this current as a function of delay with
respect to the red infrared test field which is obviously measured with a good fidelity as
confirmed by attosecond streaking. Now this ability to precisely measure the oscillating
field of light offers a new probe for attosecond phenomena in the form of the electric field of
light all the way from the infrared meanwhile to the ultraviolet. It can actually probe
the arguably most fundamental consequence of any light-matter interaction,
the displacement of electronic charge density, briefly polarization. It is embedded, it is
inscribed in the electric field of the transmitted wave. Referencing this transmitted field to the
input field actually allows us to unambiguously retrieve, at least for a thin medium, the induced
polarization. Together with its driving field it actually directly yields by means of this
simple formula the rate of energy exchange between the field and the electronic system.
If you now integrate this rate from a certain time before the electric field is coming in
to a certain time point then we get basically the transferred energy and that's what I'm
showing you here for an insulator and the semiconductor subjected to strong fields.
For the highest field strength the energy transfer appears to be completely irreversible.
So basically the transferred energy just stays there after the laser pulse is over.
But for reduced field strength what we see is that actually the field can completely reverse
this energy transfer, this is the yellow curve, at least for the six-photon bandgap. This is a
very important regime because such a reversible process is really ideal for dissipation-free
ultra-fast signal processing. This regime is completely missing in the two-photon
bandgap medium where reactive and dissipative nonlinearities always appear in combination.
These studies actually showcase how real-time access to the light electron energy transfer
explores optimum conditions for ultra-fast optoelectronic signal manipulation.
We have now seen how light field sampling and the oscillating light as a probe
may help advance electronics. In the remaining few minutes I would like to address how the same
technical capability provides access to minuscule changes in the molecular composition of human
blood plasma which is a sensitive indicator of health disease transitions. To sense these
transitions we propose, instead of searching for individual biomarkers, actually accessing the
global molecular landscape via field-dissolved spectroscopy. To this end we expose human blood
plasma to a single-cycle infrared laser pulse which coherently excites molecular vibrations
within its spectral band. The vibrating molecules in turn radiate coherent infrared waves in the
excitation. The resultant signal, which you see here, is characteristic of the molecular
composition of the sample. We capture this electric field molecular fingerprint with
highly sensitive attosecond metrology. This fingerprint signal is expected to change whenever
emerging abnormalities change the sample's molecular composition which we recently
verified for several chronic conditions such as cancer, cardiovascular and metabolic disorders.
As a prominent example, we compared 470 samples of lung cancer patients with the same number
of samples of reference individuals free from cancer. Briefly, controls. With the two cohorts
being carefully matched regarding the distribution of the subject's age and other parameters.
These are the electric field molecular fingerprints of all those samples. Clear differences come to
light between the two cohorts some thousand femtoseconds after the excitation peak,
where molecular emission starts dominating over the excitation. We can now quantitatively
analyze these differences in terms of the gaps between the adjacent zero crossings of the
oscillatory signal. Briefly, half cycle gaps. Here we show the mean and the standard deviation
of the half cycle gap, what I would call spectroscopic markers. More than 350 of them
evaluated from the fingerprints of stage 1 to stage 4 lung cancer. From the fingerprints of
the controls and most importantly the differences between the two cohorts. These cancer-induced
signals, cancer-induced changes in the half cycle gaps range from tens of attoseconds to a couple of
femtoseconds. Here we plotted these lung cancer-induced changes in half cycle gaps for
different stages in comparison with the spread of controls. The efficiency with which a well-trained
algorithm is capable of distinguishing cases from controls critically depends on this comparison.
The classification efficiency can be improved either by increasing the disease-induced signal
or by reducing the spread of controls. The cancer-induced signal increases for progressing
stages and so does the classification efficiency, the so-called area under the curve. These AUC
values are promising for non-metastasized lung cancer. However, they could be even further
improved by reducing the spread of controls even at the current stage of development of
infrared fingerprinting. That's exactly what our computer modeling predicts for referencing each
individual cancer fingerprint to a set of controls collected from the same individual before
contracting the disease. The spread of these controls is much reduced thanks to a strongly
reduced within-person variability of the infrared fingerprints over time as compared to the spread
in a healthy control group. We pursue proving the predicted performance
in a longitudinal follow-up study in Hungary. The predicted above 90% detection efficiencies
constitute unprecedented values for stage 1 to stage 4 lung cancer. The price to be paid
for this early detection and thereby protection against being diagnosed with the deadly stage 4
will be a yearly blood test. The study in Hungary may reveal that infrared fingerprinting
is able to screen other severe chronic conditions at an early stage and consequently it will become
affordable for screening whole populations and may save millions of lives year by year.
I take the Academy's decision as a mandate for the pursuit of this very goal with utmost
determination and as a mandate for exploiting our increased visibility for the benefit of the
children in Ukraine who are already dreaming of joining in like Vladislav from Lyceum number 10
in Marhanets near the front line. Just as all other children anywhere else, they deserve a
chance to realize their dreams and we badly need all of them, we badly need all of them
to carry on once our own time comes to an end. Thank you.
