Highlights from 2012
For earlier years see Archive


Nanostructures give directions to efficient laser-proton acceleratorsr

14 March 2017

Nanostructured surfaces have manifold applications. Among others they are used to selectively increase aborption of light. You can find them everywhere, where light harvesting is the key point, e.g. in photovoltaics. But also in laser proton acceleration this approach attracts a lot of attention as nanostructured targets hold the promise to significantly increase maximum proton energies and proton numbers at a given laser energy. As for any other new technology, a high efficiency is a key for a potential future use. Scientists at the Max-Born-Institute (MBI) in Berlin have now investigated, under which conditions the use of nanostructures in laser ion acceleration is beneficial.

If an ultrashort laser pulse (˜30 fs, >1 J) is focused onto a solid target foil, such that relativistic intensities (>1018 W/cm2) are reached, matter is transformed immediately into a plasma by field ionization. Electrons are accelerated to relativistic energies in the laser field. While fastest electrons can leave the target, those with less (but still relativistic) energy are trapped in the coulomb field of the (now) positively charged target and start to oscillate in this field. They form a dynamic sheath that, together with the target surface, generates an electric field of several megavolts per micrometer, in which positive ions (e. g. protons and carbon ions from the surface contamination layer) experience extreme acceleration. This process is called target normal sheath acceleration (TNSA). Fig. 1 shows an image of such a proton bunch.

The idea behind using nanostructured surfaces is now straight forward: Nanostructures increase laser absorption, i. e. more and more energetic electrons are generated which, in turn, can accelerate protons to higher energies.

But there are also alternatives for optimizing the TNSA mechanism - particularly important is the optimization of the plasma gradient, i. e. the density profile of the target. The laser intensities applied are so huge, that ionization of the target does not only happen when the peak of the laser pulse interacts with the target, but already starts during the rising edge of the pulse. The pre-ionized plasma expands, the plasma density decreases. The plasma gradient is therefore essentially determined by the exact temporal pulse structure.

The team of Dr. Matthias Schnuerer from Max-Born-Institute in Berlin has investigated, under which conditions the use of nanostructured targets is beneficial. For this purpose, the physicists have laser structured their targets in-situ. This method of generation of periodic surface structures via a laser (LIPSS) is particularly simple and in principle allows the development of a high repetition rate target system. In a first step, the target surface is nanostructured by applying about 20 strongly attenuated laser pulses. A representative scanning electron microscopy image of such a surface is shown in fig. 2. The structural parameters are similar to those that maximize laser absorption. Structural analysis and simulations show that these structures possess nearly optimal parameters for maximum laser absorption. In the following step, a single fully amplified pulse is focused onto this nanostructured area. Dr. Andrea Lübcke and her co-workers have investigated the influence of those nanostructures on the proton spectrum for different laser intensities. They chose a laser contrast that is optimal at highest intensities. First of all, the scientists could show that nanostructures remain functional even at highest intensities at the present contrast conditions in the sense that they increase the laser absorption as evident from an increase of Kα yield (see. Fig. 2a). For relatively low intensities, nanostructures significantly enhance both the conversion efficiency and proton energies. For example, at 5x1017 W/cm2 the maximum proton energies were increased by a factor of four, the conversion efficiency from laser to proton energy was even enhanced by two orders of magnitude. However, at highest laser intensities with optimal laser plasma parameters no significant benefits from the nanostructures for ion acceleration were measured (Fig. 2b,c). The researchers speculate about fundamental limitations in the energy transfer processes. The scientists were, however, not fully surprised by these results: As in many optimization problems, there are different paths to the optimum and combining them usually does not lead to an even better result. So far, these experiments performed at extreme conditions cannot be theoretically simulated in every respect. It is therefore the merit of this work to have clarified under which conditions the use of nanostructures is beneficial and in which direction new theoretical investigations can be initiiated.


Abb. 1 (click to enlarge)

Fig. 1: Laser accelerated ions, becoming visible in a Wilson chamber.

Abb. 1 (click to enlarge)

Fig. 2: Typical scanning electron microscopy image of nanostructured titanium surface (top). The Kα yield (a) of the nanostructured target is enhanced compared to the plane target over the entire investigated intensity range and indicates that nanostructures are functional even at highest intensity. In contrast, the conversion efficiency (energy transfer into fast protons) (b, logarithmic scale) and the maximum proton energy (c) of the two different targets approach each other at highest intensities.

Original publication: Scientific Reports 7, 44030 (2017) doi:10.1038/srep44030
Prospects of target nanostructuring for laser proton acceleration
Andrea Luebcke, Alexander A. Andreev, Sandra Hoehm, Ruediger Grunwald, Lutz Ehrentraut, Matthias Schnuerer


Dr. Andrea Luebcke Tel. 030 6392 1247
Dr. Matthias Schnuerer Tel. 030 6392 1350


Lattice of nanotraps and line narrowing in Raman gas

8 February 2017

Decreasing the emission linewidth from a molecule is one of the key aims in precision spectroscopy. One approach is based on cooling molecules to near absolute zero. An alternative way is to localize the molecules on subwavelength scale. A novel approach in this direction uses a standing wave in a gas-filled hollow fibre. It creates an array of deep, nanometer-scale traps for Raman-active molecules, resulting inlinewidth narrowing by a factor of 10 000.

The radiation emitted by atoms and molecules is usually spectrally broadened due to the motion of the emitters, which results in the Doppler effect. Overcoming this broadening is a difficult task, in particular for molecules. One possibility to overcome the molecular motion is by building deep potential traps with small dimensions. Previously, this was done e.g. by arranging several counterpropagating beams in a complicated setup, with limited success.

In a cooperation effort of the Max Born Institute (A. Husakou) and Xlim Institute in Limoges, researchers show that subwavelength localization and line narrowing is possible in a very simple arrangement due to self-organization of Raman gas (molecular hydrogen) in a hollow photonic crystal fibre. Due to Raman scattering, the continuous-wave pump light transforms into the so-called Stokes sideband, which travels back and forth in the fibre due to reflections from fibre ends and forms a stationary interference pattern - a standing wave with interchanging regions of high and low field [Fig. 1]. In the high-field regions, the Raman transition is saturated and is not active, and the molecules have high potential energy since they are partially in the excited state. In the low-field region, the molecules are Raman-active, and they have low potential energy since they are close to the ground state. These low-field regions form an array of roughly 40 000 narrow, strong traps, which contain localized Raman-active molecules. The size of these traps is around 100 nm (1 nm = 10-9m), which is much smaller than the light wavelength of 1130 nm. Therefore the emitted Stokes sidebands have a very narrow spectral width of only 15 kHz - this is 10 000 times narrower than the Doppler-broadened sidebands for the same conditions!

The self-organization of the gas manifests also on the macroscopic scale. First, the calculations show that the Raman process mainly happens exactly in the fibre section where the standing wave is formed, as shown in the top panel of Fig. 1. Second, the macroscopic gradient of the potential leads to the gas flow towards the fibre end, which is observed by eye in the experiment. This strong localization and the linewidth narrowing can find various uses, e.g. in spectroscopy. However, it can also be used as well as a method to periodically modulate the density of the gas, which is naturally suited for developing quasi-phase-matching schemes for other nonlinear processes, such as effective generation of high harmonics.


Fig. 1 (click to enlarge)

Fig. 1: On the macroscopic scale, the pump light transforms into forward-propagating Stokes (FS) radiation, which is partially reflected from the fibre end and becomes backward-propagating Stokes radiation (BS) which is also amplified by the pump. In the region where both FS and BS are strong, they form interference pattern of standing wave, which is shown on the microscopic scale. In the low-field regions (denoted by red-color molecules) the molecules are in the ground state and strongly trapped, as shown by the potential in the bottom panel. Exactly these trapped molecules are Raman-active, leading to line narrowing.

Original publication: Nature Communications 7, 12779 (2016) doi:10.1038/ncomms12779
"Raman gas self-organizing into deep nano-trap lattice"
M. Alharbi, A. Husakou, M. Chafer, B. Debord, F. Gérôme and F. Benabid


Dr. A. Husakou Tel. 030 6392 1280


Ultrasmall atom motions recorded with ultrashort x-ray pulses

1st Februar 2017

Periodic motions of atoms over a length of a billionth of a millionth of a meter (10-15 m) are mapped by ultrashort x-ray pulses. In a novel type of experiment, regularly arranged atoms in a crystal are set into vibration by a laser pulse and a sequence of snapshots is generated via changes of x-ray absorption.

A crystal represents a regular and periodic spatial arrangement of atoms or ions which is held together by forces between their electrons. The atomic nuclei in this array can undergo different types of oscillations around their equilibrium positions, the so-called lattice vibrations or phonons. The spatial elongation of nuclei in a vibration is much smaller than the distance between atoms, the latter being determined by the distribution of electrons. Nevertheless, the vibrational motions act back on the electrons, modulate their spatial distribution and change the electric and optical properties of the crystal on a time scale which is shorter than 1 ps (10-12 s). To understand these effects and exploit them for novel, e.g., acoustooptical, devices, one needs to image the delicate interplay of nuclear and electronic motions on a time scale much shorter than 1 ps.

In a recent Rapid Communication in Physical Review B, researchers from the Max Born Institute in Berlin (Germany), the Swiss Federal Laboratories for Materials Science and Technology in Dübendorf (Switzerland), and the National Institute of Standards and Technology, Gaithersburg (USA) apply a novel method of optical pump - soft x-ray probe spectroscopy for generating coherent atomic vibrations in small LiBH4 crystals, and reading them out via changes of x-ray absorption. In their experiments, an optical pump pulse centered at 800 nm excites via impulsive Raman scattering a coherent optical phonon with Ag symmetry [movie]. The atomic motions change the distances between the Li+ und (BH4)- ions. The change in distance modulates the electron distribution in the crystal and, thus, the x-ray absorption spectrum of the Li+ ions. In this way, the atomic motions a mapped into a modulation of soft x-ray absorption on the so-called Li K-edge around 60 eV. Ultrashort x-ray pulses measure the x-ray absorption change at different times. From this series of snapshots the atomic motions are reconstructed.

This novel experimental scheme is highly sensitive and allows for the first time to kick off and detect extremely small amplitudes of atomic vibrations. In our case, the Li+ ions move over a distance of only 3 femtometers = 3 x 10-15 m which is comparable to the diameter of the Li+ nucleus and 100000 times smaller than a distance between the ions in the crystal. The experimental observations are in excellent agreement with in-depth theoretical calculations of transient x-ray absorption. This new type of optical pump-soft x-ray probe spectroscopy on a femtosecond time scale holds strong potential for measuring and understanding the interplay of nuclear and electronic motions in liquid and solid matter, a major prerequisite for theoretical simulations and applications in technology.


Abb. 1 (click to enlarge)

Abb. 1: In an x-ray absorption experiment light excites a strongly bound core electron into a conduction band state. On the left of the figure such a transition is shown. An electron which is strongly bound to a Lithium nucleus (green) is excited into a conduction band state (red) that interacts with both the Lithium nucleus and Borohydride group. This conduction band state is therefore sensitive to a modulation of the distance Q between Lithium nucleus and Borohydride group and as a result the x-ray absorption process is sensitive to such a modulation (cf. Figs. 2(b) and 3(d) in the main article). On the right side of the figure the Lithium K-edge x-ray absorption spectrum for different strongly exaggerated displacements is shown.
Movie Movie: What happens in the unit cell of crystalline LiBH4 after impulsive Raman excitation with a femtosecond laser pulse? Upper panel: measured transient absorption change Δ A(t) (symbols) as we vary the time delay between infrared pump pulses and soft x-ray probe pulses at photon energy of ħω = 61.5 eV [cf. Fig. 3(a) in the main article. The lower box shows the atoms in the unit cell of LiBH4 with red boron atoms, gray hydrogen atoms, and green Li atoms. The moving blue circle in the upper panel is synchronized with the moving atoms in the lower panel. The amplitude of the motion is strongly exaggerated (i.e. times 30000) to visualize the pattern of the motion. The reddish color of the unit cell indicates the intensity of the infrared pump pulse.

Original publication: Physical Review B 95, 081101 (R) (2017)
Ultrafast modulation of electronic structure by coherent phonon excitations
J. Weisshaupt, A. Rouzée, M. Woerner, M. J. J. Vrakking, T. Elsaesser, E. L. Shirley, and A. Borgschulte


Dr. Michael Woerner Tel. 030 6392 1470
Jannick Weisshaupt Tel. 030 6392 1471
Dr. Arnaud Rouzée Tel. 030 6392 1240
Prof. Dr. Marc Vrakking Tel. 030 6392 1200
Prof. Dr. Thomas Elsaesser Tel. 030 6392 1400


Unified time and frequency picture of ultrafast atomic excitation in strong fields

5 January 2017

The insight that light sometimes needs to be treated as an electromagnetic wave and sometimes as a stream of energy quanta called photons is as old as quantum physics. In the case of interaction of strong laser fields with atoms the dualism finds its analogue in the intuitive pictures used to explain ionization and excitation: The multiphoton picture and the tunneling picture. In a combined experimental and theoretical study on ultrafast excitation of atoms in intense short pulse laser fields scientists of the Max Born Institute succeeded to show that the prevailing and seemingly disparate intuitive pictures usually used to describe interaction of atoms with intense laser fields can be ascribed to a single nonlinear process. Moreover, they show how the two pictures can be united. The work appeared in the journal Physical Review Letters and has been chosen to be an Editors' suggestion for its particular importance, innovation and broad appeal. Beside the fundamental aspects the work opens new pathways to determine laser intensities with high precision and to control coherent Rydberg population by the laser intensity.

Although the Keldysh parameter, introduced in the 1960's by the eponymous Russian physicist, clearly distinguishes the multiphoton picture and the tunneling picture, it has remained an open question, particularly in the field of strong field excitation, how to reconcile the two seemingly opposing approaches.

In the multiphoton picture the photon character shines through as resonant enhancement in the excitation yield whenever an integer multiple of the photon energy matches the excitation energy of atomic states. However, the energy of atomic states is shifted upwards with increasing laser intensity. This results in resonant-like enhancements in the excitation yield, even at fixed laser frequency (photon energy). In fact, the enhancement occurs periodically, whenever the energy shift corresponds to an additional photon energy (channel closing).

In the tunneling picture the laser field is considered as an electromagnetic wave, where only the oscillating electric field is retained. Excitation can be viewed as a process, where initially the bound electron is liberated by a tunneling process, when the laser field reaches a cycle maximum. In many cases the electron does not gain enough drift energy from the laser field to escape the Coulomb potential of the parent ion by the end of the laser pulse, which would lead to ionization of the atom. Instead, it remains bound in an excited Rydberg state. In the tunneling picture there is no room for resonances in the excitation since tunneling proceeds in a quasi-static electric field, where the laser frequency is irrelevant.

In the study the excitation yield of Ar and Ne atoms as a function of the laser intensity has been directly measured for the first time, covering both the multiphoton and tunneling regimes. In the multiphoton regime pronounced resonant enhancements in the yield have been observed, particularly in the vicinity of the channel closings, while in the tunneling regime no such resonances appeared. However, here excitation has been observed even in an intensity regime which lies above the threshold for expected complete ionization.

The numerical solution of the time dependent Schrödinger equation for the investigated atoms in a strong laser field provided excellent agreement of the theory with the experimental data in both regimes. A more detailed analysis revealed that both pictures represent a complementary description in the time and frequency domain of the same nonlinear process. If one considers excitation in the time domain one can assume that electron wave packets are created periodically at the field cycle maxima. In the multiphoton regime it can be shown that the wave packets are created predominantly close to the maximum intensity of the pulse and thus interfere constructively only if the intensity is close to a channel closing. With this, regular enhancement in the excitation spectrum results effectively only at the photon energy separation. In the tunneling regime the wavepackets are also created periodically at the field cycle maxima, however, predominantly at the rising edge of the laser pulse which, in turn, leads to an irregular interference pattern and consequently, to irregular variations in the excitation spectrum. These rapid variations are not resolved in the experiment and the detected excitation spectrum is smooth.


Fig. 1 (click to enlarge)

Fig. 1: Yield of excited atoms as a function of the laser intensity. At a laser intensity of 200TW/cm2, in the vicinity of a 6 photon channel closing, a strong resonant enhancement of a factor 100 is visible. For the argon data, the theoretical curve is also displayed (red dashed curve), which is in excellent agreement with the experimental data.

Original publication: Phys. Rev. Lett. 118, 013003 (2017) doi:10.1103/PhysRevLett.118.013003
"Unified Time and Frequency Picture of Ultrafast Atomic Excitation in Strong Laser Fields"
H. Zimmermann, S. Patchkovskii, M. Ivanov, and U. Eichmann


Dr. S. Patchkovskii Tel. 030 6392 1241
Prof. Dr. U. Eichmann Tel. 030 6392 1371


Amplification of relativistic Electron Pulses by Direct Laser Field Acceleration

5 January 2017

Controlled direct acceleration of electrons in very strong laser fields can offer a path towards ultra-compact accelerators. Such a direct acceleration requires rectification and decoupling of the oscillating electromagnetic laser field from the electrons in a suitable way. Researchers worldwide try to tackle this challenge. In experiments at the Max Born Institute, direct laser acceleration of electrons could now be demonstrated and understood in detail theoretically. This concept is an important step towards the creation of relativistic and ultra-short electron pulses within very short acceleration distances below one millimeter. Resulting compact electron and related x-ray sources have a broad spectrum of applications in spectroscopy, structural analysis, biomedical sciences and for nanotechnology.

The way electrons can be accelerated up to relativistic kinetic energies in strong laser fields is a fundamental issue in the physics of light-matter interaction. Although the electromagnetic fields of a laser pulse force a free electron previously at rest to oscillations with extremely high velocities, these oscillations cease again when the light pulse has passed by. A net energy transfer by such a direct acceleration of a charged particle in the laser field cannot take place. This fundamental principle - often discussed in physics exams - is valid for certain boundary conditions of the spatial extent and intensity of the laser pulse. Only for particular, different boundary conditions, electrons can indeed receive a net energy transfer via acceleration from the strong laser field. These conditions can be set e.g. by focusing of the laser pulse or the presence of strong electrostatic fields in a plasma.

Worldwide, scientists are looking for solutions how fast electrons can be extracted from extremely strong laser fields and how one can obtain short electron pulses with a high charge density via ultra-short laser pulses.

In light fields of relativistic intensity (I> 1018 W/cm2) electrons oscillate with velocities close to the speed of light. The corresponding kinetic energy reaches values from MeV to GeV (at I> 1022 W/cm2. Strong light fields are realized by focusing ultra-short laser pulses with high energy down to areas of few micrometers. The resulting spatial intensity distribution does already enable the acceleration of the electrons up to high kinetic energies. This process is known as ponderomotive acceleration. It is an essential process for the interaction between strong light fields and matter. Various theoretical studies, however, have predicted that the number of electrons and their kinetic energy can be further significantly increased by a direct acceleration in the laser field, but only if the electron-light interaction is interrupted in a properly tailored way. These considerations were the starting point for the experiments by Julia Braenzel and her colleagues at the Max Born Institute.

In the experiments at MBI, the electrons were decoupled from the light pulse at a particular moment in time, using a separator foil that is opaque for the laser light but can transmit fast electrons. We could show that this method leads to an increase of the number of electrons with high velocities. At first, a 70 TW Ti:Sapphire laser pulse (2 J @ 35 fs) irradiates an 30 - 100 nm thin target foil consisting of a PVF-polymer. In the laser propagation direction, about 109electrons are accelerated up to several MeV energy via the ponderomotive force. During this interaction the foil is almost fully ionized and transformed into plasma.

For sufficiently thin target foil thicknesses below 100nm a fraction of the incident laser light can be transmitted through the plasma. The transmitted light starts to overtake the electrons already emitted in this direction. This corresponds to a quasi-intrinsically synchronized injection of slow electrons into the transmitted, but still relativistic laser field (<8 x 1018W/cm2). If a second thin separator foil is placed at the correct distance behind the first one, amplification in the electron signal for a particular energy interval is observed. Fig. 1a) shows a schematic of the temporal evolution in the experiment and Fig. 1b) presents a direct comparison of the detected electron spectral distribution for a single foil and a double foil configuration, where the second foil acts as a separator. This foil is opaque for the laser light but is transparent for the fast electrons and hence enables a decoupling of both. The time at which the interaction between electrons and transmitted light is interrupted depends on the distance between the two foils.

The experiments carried out in the group of Matthias Schnürer demonstrate that an amplification of the electron signal can obtained and is maximized for a particular distance. The amplification vanishes for very big distances. Numerous measurements as well as numeric simulations confirmed the hypothesis that electrons with high kinetic energy can indeed be extracted out of the light field if they are decoupled appropriately. If the separator foils is located at an optimized position, slow electrons with kinetic energies below 100keV are accelerated to about ten times higher kinetic energies. This effect leads to a concentration of electrons in a narrow energy interval. In contrast to experiments using the different mechanism of laser wake field acceleration, where the production of GeV electrons has already been demonstrated, the direct laser acceleration demonstrated here can be scaled up to high laser intensities and high plasma densities. Beyond the fundamental insight in laser-matter interactions, the direct laser acceleration demonstrated in this work holds promise for the future realization of compact sources of relativistic electrons.


Fig. 1a (click to enlarge)

Fig. 1a: Schematic of the direct electron acceleration in a laser field and its realization in the experiment.
Steinmeyer Fig. 1b: Detected electrons in the laser propagation direction from a single (F1) and double foil (F1F2) target configuration, where the soncond foils acts as a speparator. The plastic foils used were about F1=35nm and F2=85 nm thick. N e values represent the integrated electron numbers for the whole detection range (0,2-7,5 MeV) with respect to the spectrometer aperture.
Fig. 1b (click to enlarge)  

Original Publication: Phys. Rev. Lett. 118, 014801 (2017) doi:10.1103/PhysRevLett.118.014801
"Amplification of Relativistic Electron Bunches by Acceleration in Laser Fields"
J. Braenzel, A.A. Andreev, F. Abicht, L. Ehrentraut, K. Platonov, and M. Schnürer


Julia Bränzel Tel. 030 6392 1338
Dr. Matthias Schnuerer Tel. 030 6392 1315


Archive .... here