Light-Induced Magnetization at the Nanoscale

Triggering and switching magnetic moments is of key importance for applications ranging from spintronics to quantum information. A noninvasive ultrafast control of magnetization at the nanoscale is, however, an open challenge. Here, we propose a novel laser-based scheme for generating atomic-scale charge current loops within femtoseconds. The associated orbital magnetic moments remain ferromagnetically aligned after the laser pulses have ceased and are localized within an area that is tunable via laser parameters and can be chosen to be well below the diffraction limit of the driving laser field. The scheme relies on tuning the phase, polarization, and intensities of two co-propagating Gaussian and vortex laser pulses, allowing us to control the spatial extent, direction, and strength of the atomic-scale charge current loops induced in the irradiated sample upon photon absorption. In the experiment we used He atoms driven by an ultraviolet and infrared vortex-beam laser pulses to generate current-carrying Rydberg states and test for the generated magnetic moments via dichroic effects in photoemission. Ab initio quantum dynamic simulations and analysis confirm the proposed scenario and provide a quantitative estimate of the generated local moments. 
The basic idea is the following: to bypass the optical diffraction limit on the resolution of the volume in which the magnetic moments are generated, we vary the intensity ratio between the Gaussian XUV pulse and the infrared (IR) light vortex that trigger the orbital moments. The XUV pulse is tightly focused and co-propagates with the IR vortex laser [see Fig. 1(a)]. Both lasers intersect the sample in the same focal plane. The IR vortex pulse carries orbital angular momentum (OAM), as well as spin angular momentum (SAM) and has a vanishing intensity in its center, where the XUV intensity peaks. Atoms residing around the optical axis where the XUV (IR) intensity peaks (diminishes), are excited by the XUV laser in the presence of the IR laser. For given laser spots, tuning the intensity ratio between the XUV and the IR pulses controls the cross sectional area where the magnetic moments reside, allowing this way for a nanoscale tuning of the generated magnetic fields with a resolution well below the diffraction limit of the IR laser. The light-induced magnetization follows from a statistical average of the induced atomic magnetic moments. Importantly, the induced magnetic moments and associated magnetic fields are robust and do not average to zero over a few optical cycles of the (petahertz) driving lasers. Our simulations confirm that the maximum magnetic moment is attained by atoms residing around the radial position 2 in Figs. 1(b) and 1(c). The magnetic field associated with the quantum-mechanical induced current is obtained from classical electrodynamics [results in Fig. 1(d)]. 
 

Figure 1.  Experimental setup and temporal buildup of magnetic moments. (a) Sketch of the experimental setup. Collinearly propagating circularly polarized XUV Gaussian laser pulse (violet) and angular-momentum carrying (with a topological charge m), circularly polarized IR vortex laser pulse (red) intersect a gaseous jet of He atoms traversing the focal plane. (b) Depending on the atom positions in the laser spots (depicted cases are for the following radial distances from the propagation direction local orbital magnetic moments are generated in the atom. Theoretical predictions are displayed in (c). The temporal structures of the laser pulses are shown in the inset of (c) (red color indicated the IR vortex pulse, violet color the XUV pulse). (d) shows the quantum mechanically calculated subwavelength confined magnetic field lines for the atom (1) in (b).
 

The experiment was carried out at the FERMI’s Low-Density Matter(LDM) beamline. In order to probe the generated magnetization, we used a method based on dichroism in photoionization. The energy and angle-resolved spectra of the detected photoelectron embody signatures of orbital magnetic moments. Using the velocity map imaging (VMI) spectrometer available at FERMI’s LDM beamline, we measured the differential cross section (DCS) of the photoelectrons, deduced the dichroism obtained by varying both the spin and orbital angular momentum of the IR beam, and compared it with the theory. We performed two sets of measurements. In the first set, the IR laser pulse was a Gaussian-shaped circularly polarized laser pulse, giving rise to circular dichroism only (photon OAM is zero). The DCSs for the first and second sidebands for this case are shown in Figs. 2(a) and 2(f) and compared to the theory. The measured circular dichroism D_Cdistributions are shown in Figs. 2(c) and 2(h), showing a good agreement with the theory and are in line with existing literature. In the second set, we substitute the unstructured IR fields with vortex fields carrying the same SAM as in the first set of measurements but also one quantum of OAM, allowing tuning to total angular momentum (TAM) ±2ℏ. The corresponding results are presented in Figs. 2(b) and 2(g). The TAM dichroism D is shown in Figs. 2(d) and 2(i), while the difference between both dichroism types, i.e., Dc − D is presented in Figs. 2(e) and 2(j). In all cases, we obtained a good agreement with our ab initio modeling. This confirms the femtosecond buildup of atomically confined current densities, leading to the formation of nanometer scale magnetization that lasts for nanoseconds.

 

Figure 2.  Experimental vs theoretical photoelectron spectra and DCSs for different combinations of IR spin and OAM. Experimental verification of angular momentum transfers in first and second sidebands (SB) (upper row SB I, lower row SB II). The first two panels from the left show the individual DCs when the IR field does not carry OAM and is left (CL) or right (CR) circularly polarized (a),(f). Panels (b),(g) correspond to an IR vortex carrying TAM= ±2ℏ. From the measurement we infer the dichroism distributions: circular (c),(h) and TAM dichroism (d),(i). The outer right panels (e), (j) show the difference between both dichroism types.

This research was conducted by the following research team:

Jonas Wätzel¹, Primož Rebernik Ribič², Marcello Coreno^2,3, Miltcho B. Danailov², Christian David⁴, Alexander Demidovich², Michele Di Fraia², Luca Giannessi^2,5, Klavs Hansen⁶, Špela Krušič⁷, Michele Manfredda², Michael Meyer⁸, Andrej Mihelič⁷, Najmeh Mirian², Oksana Plekan², Barbara Ressel⁹, Benedikt Rösner⁴, Alberto Simoncig², Simone Spampinati², Matija Stupar⁹, Matjaž Žitnik⁷, Marco Zangrando^2,10, Carlo Callegari², Jamal Berakdar¹and Giovanni De Ninno^{2, 9,} 

¹ Institut für Physik, Martin-Luther-Universität Halle-Wittenberg, Halle (Saale), Germany
² Elettra-Sincrotrone Trieste S.C.p.A., Trieste, Italy
³ ISM-CNR, Trieste, Italy
⁴ Paul-Scherrer-Institut, Villigen-PSI, Switzerland
⁵ INFN-LNF, Frascati (Rome), Italy
⁶ Center for Joint Quantum Studies and Department of Physics, School of Science, Tianjin University, Tianjin, China
⁷ J. Stefan Institute, Ljubljana, Slovenia
⁸European XFEL, Schenefeld, Germany
⁹ University of Nova Gorica, Nova Gorica, Slovenia
¹⁰Istituto Officina dei Materiali, Consiglio Nazionale delle Ricerche, Basovizza, Italy.



Contact person: 
 
Local contact person: