Research topics

The theoretical investigation of the interaction of matter with laser or other intense photon pulses by now has reached a level at which fundamental aspects such as the quantum nature of both light and matter, relativity and correlations among the involved particles have become key issues and substantial challenges alike. In this overview, our efforts are summarized towards a detailed understanding of the quantum mechanical interplay of all constituents of atomic, ionic and nuclear systems as well as quantum vacuum and quantum plasma with strong laser fields. Emphasis is placed on various applications with strong connection to experiments at the Institute (see Fig. 1): High-precision calculations involving quantum electrodynamical and nuclear effects are employed for accurate determinations, e.g., of ionic spectra, binding energies and g-factors. Coherent control of atomic and nuclear ensembles is investigated in the light of advanced synchrotron and x-ray free-electron laser sources. And finally our research on extreme field laser physics is directed towards carrying out quantum electrodynamics, nuclear, high-energy and astrophysics with the strongest laser pulses presently available or under construction.

An external static magnetic field gives rise to the Zeeman splitting of atomic energy levels, with a strength characterized by the dimensionless g-factor of the atomic state. The theory of the g-factor of an electron bound in the attractive potential of an atomic nucleus is investigated in the group around Zoltán Harman. These studies are paralleled with a quantum leap in the experimental accuracy in the investigation of the g-factor, especially in Penning trap measurements performed by the division headed by Klaus Blaum. Measuring the electron's g-factor in highly charged ions provide an exciting possibility for testing fundamental theories. Effects of strong-field quantum electrodynamics (QED) are increasingly relevant at higher and higher ionic charges (see Fig. 2). Bound-state QED effects are scrutinized to the highest precision in recent trapped-ion experiments, which have reached the 10^-11 level in terms of relative precision. From a combined analysis of experimental and theoretical data for carbon ions, an improved value of the mass of the electron has been determined to world-leading accuracy. The ALPHATRAP Penning-trap setup also enables the experimental investigation of highly charged ions in virtually arbitrary charge states, which, in collaboration with theoretical calculations, largely enriches the field of applications. Penning-trap measurements have very recently also enabled a novel determination of electronic binding energies. The achieved experimental accuracy on the 1-eV precision level is matched by our large-scale atomic structure calculations. Applications of this collaboration include the discovery of ultranarrow ionic transitions suitable for constructing future atomic clocks, the measurement of nuclear and atomic masses, and the determination of the Q value of the β decay of various atomic isotopes, relevant for constraining the neutrino mass.

The g-factor experiments are currently being extended to a range of ions along the periodic table, including the heaviest stable elements. In a recent study, single-electron tin ions have been experimentally investigated. Due to the high charge of its nucleus, extremely high electrostatic fields of around 10¹⁵ volts per centimetre are present at the location of the bound electron. Our theoretical calculations reach excellent agreement with the measured value after the highly demanding computations of two-loop self-energy corrections, which were evaluated taking into account the nonperturbative nuclear background field. A similar accuracy has been reached for the three-electron tin ion, in which, in addition to the quantum fluctuations of virtual particles created and annihilated from the vacuum, also photons exchanged between the electrons need to be accounted for. In recent studies, we put forward the use of high-precision measurements of the g-factor of such few-electron ions and its isotope shifts as a probe for physics beyond the standard model. The contribution of a hypothetical fifth fundamental force is calculated, and we found that, combining measurements from different ions at accuracy levels projected to be accessible in the near future, one may constrain the new physics coupling constant competitively. This idea was implemented in a recent experiment performed with ALPHATRAP, in which the difference between the g factors of two Ne isotopes was measured with a remarkable precision (see Fig. 3). The combination of theory and experiment provided a novel way of proving the existence of a hypothetical exotic interaction. Another essential motivation of such studies is that g-factor measurements with highly charged ions are anticipated to yield a new value of the fine-structure constant, i.e. the fundamental constant defining the strength of any kind of electromagnetic interactions in the Universe. Furthermore, our theory also enables to extract the magnetic properties of He-3 and Be-9 from high-precision experimental data obtained with the μTEx trap, relevant for the calibration of new nuclear magnetic resonance probes.

The knowledge of the structural and dynamical properties of highly charged ions is also necessary for the modeling and diagnostics of astrophysical and fusion plasmas. The spectra of inner-shell transitions are of great relevance for astrophysical line diagnostics of x-ray binaries and stars, aiming at determining physical parameters of the emitter medium such as element decomposition, density, temperature, and velocity. These transitions are predominantly in the x-ray range, therefore, their efficient radiative excitation has been rather challenging until the advent of modern x-ray sources such as free electron lasers and synchrotrons, providing high intensities in the appropriate wavelength range. Our calculations have contributed to the resolution of a conondrum that has puzzled astrophysicists for decades: The intensity of two important, strong x-ray lines of a certain highly charged iron ion in astrophysical data and in laboratory experiments did not agree with predictions, obscuring the interpretation of astronomical observations (see Fig. 5). A recent electron beam ion trap measurement at the high-brilliance PETRA III synchrotron facility finally agrees with our large-scale calculations and with other modern theoretical results.

In the group of Jörg Evers, currently the emphasis is on atomic Mössbauer nuclei, which have proven successful in ground-breaking experiments on x-ray quantum optics. A central goal is to establish large interacting ensembles of Mössbauer nuclei in solid-state targets as a new platform for quantum dynamics, and to develop novel control schemes for the interaction of the x-rays with the nuclei based on coherence-, interference, nonlinear and quantum effects. Such control schemes open up novel applications for Mössbauer science and powerful tools for x-ray science alike. Over the last years, we developed sophisticated quantum optical frameworks for the description of the nuclear quantum dynamics driven by intense x-ray fields. They allow us to identify and separate all physical processes contributing to the recorded signal, and encompass nonlinear and quantum effects. Based on these models, we engineer advanced quantum optical setups, and we verified a number of predictions of this model experimentally in collaboration with experimental teams around Ralf Röhlsberger (Helmholtz Institute Jena and DESY, Hamburg), Thomas Pfeifer at the Institute, and G. G. Paulus (University and Helmholtz Institute Jena). 

Mössbauer nuclei feature resonances with exceptionally narrow spectral line widths, which form the basis for a broad range of applications across the natural sciences. For x-ray quantum optics, the narrow resonance is favorable, because it translates into long nuclear coherence lifetimes. However, the narrow line width also implies weak driving, since only a tiny fraction of the x-ray pulses is resonant with the nuclei. To address this challenge, a group around Jörg Evers explores the possibility to simulate the effect of strong control fields using mechanical motion of a nuclear target (see Fig. 6). A short x-ray pulse impinging on a nuclear target may either pass without interaction, or interact with the nuclei. By applying a suitable motion to the target immediately after the non-interacting part passed the target, an arbitrary time-dependent phase can be imprinted on the delayed interacting part. In our initial experiments, we used this control to shape the x-ray pulses such that the number of resonant photons is increased – thereby allowing for a stronger driving of nuclei. Further, we demonstrated that the shaped pulses can be used to coherently control nuclear dynamics, by switching between stimulated emission and coherent excitation of the nuclei. Recently, we developed a dynamically tuneable x-ray interferometer for Mössbauer science based on this control (see Fig. 6). Interferometric methods are central to advancing precision measurements, but difficult to realize due to the extreme stability requirements imposed by the short x-ray wavelength. In our setup, two Mössbauer targets act as beam splitters to implement an inline interferometer with spatially overlapping interfering pathways to improve the stability. We engineer the interference such that the transmission vanishes for the empty interferometer (dark-fringe mode) to improve the sensitivity. The relative phase of the two pathways can dynamically be tuned by displacing one of the targets. In the experiment with teams around Ralf Röhlsberger (Helmholtz Institute Jena and DESY) and Thomas Pfeifer (MPIK) at the High Resolution Dynamics Beamline P01 of the synchrotron source PETRA III (DESY), we showcased the new measurement capabilities by interferometrically observing minuscule deformations of a host material due to propagating sound waves. We further demonstrated the control capabilities by switching the interferometer transmission in order to gate the intensity of an x-ray pulse on nanosecond time scales. Our interferometer concept opens avenues towards polarization-sensitive phase measurements, the generation of coherent multi-pulse sequences for controlling nuclear dynamics, and the implementation of feedback loops to adaptively optimize the interferometer, thereby fueling the further development of nuclear quantum optics.

In a collaboration with a team around Ralf Röhlsberger (HI Jena and DESY), we have recently established nuclear resonance scattering on the archetype Mössbauer isotope ⁵⁷Fe at the MID instrument of the European X-ray free electron laser. The high-repetition rate pulse structure together with self-seeding enable a qualitative change in experimental capabilities for Mössbauer science. Most notably, in our experiments, we routinely observed higher excitations of the nuclear ensemble, with up to about 1000 signal photons in the coherent scattering after a single x-ray excitation (see Fig. 7). By contrast, the corresponding average number of signal photons at synchrotron sources is far below one. This progress enables a multitude of new science cases and observables, such as photon correlations or the excitation-dependent study of collective effects. As a first application, we explored Mössbauer spectroscopy on the level of individual x-ray excitations. This addresses the longstanding challenge that in traditional setups, Mössbauer spectroscopy inevitably requires an averaging over extended measurement times. The averaging impedes the study of non-repetitive non-equilibrium phenomena in the nuclear dynamics or in the surrounding host material, e.g., after impulsive stimuli. Previously it had been suggested that an increase in the number of resonant photons delivered by the source could allow for spectroscopy using single x-ray pulses. However, we found that this approach is limited by a finite detector dynamical range, which typically does not suffice to capture the full decay dynamics. It further excludes the vast majority of experimental repetitions with lower signal photon number due to the XFEL photon statistics. To overcome these challenges, we developed a sorting approach, which utilizes the presence of different dynamics classes, i.e. different nuclear evolutions after each excitation. We employ machine learning to identify models for the dynamics classes by a suitable clustering of the data. Interestingly, this approach does not require any theory input and can be used to search for unknown phenomena. Afterwards, also the low-count repetitions can be sorted according to the identified classes, and thereby crucially contribute to the result. The successful demonstration based on data measured at EuXFEL opens up a path towards the future exploration of out-of-equilibrium transient dynamics of the nuclei or their environment. Next to the high number of resonant photons per pulse, EuXFEL also offers an increase in the number of resonant photons per second. In a collaboration with Yuri Shvyd’ko (Argonne National Labs, USA), Olga Kocharovskaya (Texas A&M University, USA; external scientific member at MPIK) and Ralf Röhlsberger (HI Jena and DESY), this allowed us to resonantly excite the ultra-narrow nuclear transition in ⁴⁵Sc at 12.4 keV resonance energy for the first time. The nuclear resonance has a natural lifetime of about half a second (line width 1.4 femto-electronvolt), and its coherent spectroscopy would enable new fundamental tests and potentially a nuclear clock operation at energies of hard x-rays. In the experiment, the nuclear resonance energy was determined 250 times more accurate than before, which sets the stage for the further exploration of the resonance.

Quantum effects such as wave packet interference, entanglement, spin oscillations, and tunneling may be relevant when extremely intense laser pulses are impinging on atomic systems. In particular, a quantum tunneling barrier may be built up in an atom or a highly charged ion by the attractive Coulomb forces that attach the electron to the atomic core and the electric field of a strong laser that pulls the electron away from the core. Moreover, ultra-strong lasers can no longer be treated as pure electric fields; the laser's magnetic field component has to be taken into account, too (nondipole effects). The recent results of the group around Karen Hatsagortsyan contributed to an understanding of the role of sub-barrier as well as nondipole dynamics for shaping subtle features of the photoelectron momentum distribution in tunneling ionization.

One typical topic of interest is the time-resolved understanding of quantum processes during ionization of an atom in a very strong field. A simple model of this process claims that the electron tunnels instantaneously through the laser-generated quantum barrier. However, the accurate description of the sub-barrier dynamics can significantly modify this simple picture. Previously, we have identified a new pathway of strong-laser-field-induced ionization of an atom via recollisions under the tunneling barrier, showing that the interference of the direct and the under-the-barrier recolliding quantum orbits induces tunneling time delay. Recently, we have found that the under-the-barrier-recollision dynamics can induce transient population of the Rydberg states which, counterintuitively, can be much more efficient than the direct multiphoton transitions (see Fig. 9). Accordingly, we have developed the under-the-barrier-recollision theory for Freeman resonances, which predicts distinct features of phenomena in the strong field regime that cannot be explained by the existing direct multiphoton transition scenario. Specifically, it predicts the dominance of high-order Freeman resonances over above-threshold ionization in the photoelectron energy spectra and the flat dependence of the signal of Freeman resonances on the laser intensity, both in the nonadiabatic tunneling regime. We collaborated on this topic with the Korean experimental group around Professor Dong-Eon Kim from POSTECH institute. They were able to experimentally demonstrate these features of Freeman resonances in strong field regime, substantiating the under-the-barrier-recollision dynamics. Thus, the under-the-barrier-recollision is not only responsible for inducing tunneling time delays on the order of attoseconds, but also, with this experimental demonstration, are likely to induce similar substantial amendments in comparable scenarios, modifying, e.g., strong-field molecular, solid-state, and even high-energy tunneling quantum dynamics.

Recently, ion-photoelectron entanglement effects during strong field ionization have experienced special attention of the strong field community. In our recent study, we have shown how nondipole effects can facilitate observation of the ion-photoelectron entanglement. We investigated the three-body dynamics of the photoelectron and nuclei in the dissociative photoionization of diatomic molecules initiated by x-ray photoabsorption. The ion-photoelectron entanglement transfers information of the electronic interference to the ion dynamics. As a consequence, the ion momentum distributions of dissociative molecular photoionization present Young’s double-slit interference when the photoelectron emission angle is fixed. In the ionization of heteronuclear molecules, a nondipole recoil momentum arises due to the different ion masses. Quantifying the nondipole effect via the average of the relative nuclear momentum along the photon propagation direction, we demonstrate that double-slit interference signatures persist in the average longitudinal momentum shift. The interference due to the photoelectron-ion entanglement could serve as an additional leverage for coherent control of molecular dynamics in dissociative photoionization.

Furthermore, the role of the spin may become essential in intense laser-matter interaction, where we have previously investigated its dynamics and back action as well as compared deviating semiclassical models. More recently, we have revealed a counter-intuitive effect of the spin polarization of photoelectrons in tunneling ionization induced by a linearly polarized laser pulse. It was known that ionization can serve as a source of spin-polarized electrons, because spin-orbit coupling in an atomic bound state creates a strong correlation between the electron’s spin and orbital angular momentum. This correlation, combined with the significant angular momentum-dependent ionization probability, results in photoelectron polarization in nonadiabatic tunneling ionization in circularly polarized laser fields. While the conservation law of angular momentum implies that net electron polarization is impossible when ionizing the spinless ground state of rare gas atoms with a linearly polarized laser field, angle-resolved spin polarization can still occur. We demonstrate a nontrivial spin texture of photoelectrons in momentum space, exhibiting a vortex structure relative to the laser polarization axis. In contrast to a circularly polarized pulse, the total polarization of the ionized electron vanishes here; nevertheless, we demonstrate the emergence of significant momentum-resolved spin polarization. We trace the origin of this spin texture to the correlation between spin and the initial transverse velocity of the photoelectron at the tunnel exit, giving rise to spin-dependent quantum orbits. Two distinct mechanisms of photoelectron polarization can exist: one for direct electrons, visible in few-cycle laser pulses, and the other for recolliding electrons, prominent in long pulses. In the latter case, we show that forward rescattering of the spin-dependent quantum orbits is responsible for generating nontrivial momentum-resolved spin texture. Furthermore, the interference between direct and rescattering ionization allows for spin-polarized electron holography (see Fig.10), offering an alternative method to extract atomic fine structural information.

The understanding of the structure of QED in the presence of strong background fields cannot be fully accomplished without a thorough investigation of radiative corrections. Among these we have studied the self-interaction of the electron with its own electromagnetic field (mass operator) and the temporary transformation of a photon into a virtual electron-positron pair (polarization operator). Both analyses have shown qualitatively new effects with respect to the vacuum case. On the one hand, the structure of the mass operator has indicated that, unlike in vacuum, the electron spin dynamics can be substantially altered by the electron self-field. On the other hand, the investigation of the polarization operator has even shown that in the presence of a strong laser beam, virtual electron-positron pairs can propagate along distances which exceed by orders of magnitude the typical microscopic QED lengths of the order of the Compton wavelength λ_C=3.9×10^-11 cm and are comparable with the laser wavelength (of the order of micrometers for optical lasers). As a consequence, the electron and the positron can absorb a very large number of laser photons (of the order of a million for an optical laser of intensity 10²² W/cm²) and the corresponding energy can be potentially exploited to prime high-energy reactions once the virtual pair again annihilates (see Fig. 11). A thorough analysis shows that this process is the analogous in the high-energy domain of the well-known recollision processes in atomic and molecular physics responsible, for example, of high-harmonic generation. Thus, such high-energy recollision processes may represent a basic, key ingredient for a miniaturized version of a high-energy "vacuum" collider.

Coherent extreme ultraviolet (XUV) light sources capable of generating attosecond pulses have become essential tools for probing ultrafast dynamics in atoms, molecules, and solids. Traditionally, such pulses have been produced via high-harmonic generation (HHG), a nonlinear optical process that enables access to attosecond timescales but suffers from low conversion efficiency, particularly at higher photon energies. To overcome these limitations, free-electron lasers (FELs) have emerged as alternative sources offering substantially higher intensities and broader spectral coverage, albeit at the cost of size and complexity. FELs achieve coherence through the formation of microbunches in an electron beam, which emit collectively as they traverse an undulator. Although recent work has demonstrated FEL-based attosecond pulse generation, these facilities remain large and costly. A promising route toward miniaturization is the optical FEL (OFEL), in which an intense laser pulse replaces the undulator. The shorter wavelength of the optical driver offers significant potential for compactness, but the requirement for dense, low-energy electron beams introduces challenges due to emittance growth and Coulomb repulsion. Recent advances in producing collimated and relativistic particle beams, including those composed of both electrons and positrons, open new possibilities for realizing compact coherent light sources in the XUV regime.

QED processes like pair production, cascades and gamma ray emission with radiation reaction are also relevant in quantum plasmas and in astrophysical environments. The formation and dynamics of QED cascades in the magnetospheres of pulsars, particularly above their polar caps, represent a cornerstone of modern pulsar astrophysics. Pulsars are highly magnetized, rapidly rotating neutron stars whose magnetospheres are thought to be filled with dense e^± plasmas generated through QED cascades occurring in vacuum gaps. These cascades are initiated by primary particles accelerated along strong magnetic field lines, which emit high-energy photons via curvature radiation. The curvature radiation photons, in turn, decay into e^±  pairs that subsequently emit synchrotron radiation, driving a self-sustained cascade process. Traditional models and numerical approaches—ranging from magnetohydrodynamic to heuristic particle-in-cell simulations—have provided essential insights into global magnetospheric structure and cascade dynamics, yet often neglect spin and polarization effects intrinsic to QED processes. Recent developments in strong-field QED, propelled by the advent of petawatt-class laser facilities, have revived interest in polarization-resolved dynamics, although most studies focus on terrestrial or laboratory conditions where field configurations differ significantly from those in pulsars. In the pulsar context, where the motion of particles is aligned with the magnetic field rather than perpendicular, polarization and spin effects are expected to play a distinct and potentially nontrivial role. Prior QED-PIC simulations typically adopted unpolarized radiation and pair production rates, thereby omitting the subtle influence of photon polarization on cascade efficiency and structure. This gap underscores the necessity of incorporating polarization-resolved QED processes to accurately capture the microphysics governing cascade initiation and evolution in pulsar polar cap regions.

An international team led by Matteo Tamburini has systematically investigated the role of photon polarization in QED cascades over pulsar polar caps, achieved through the implementation of a Monte Carlo-based QED algorithm that accounts for both e^± spin and γ-photon polarization. By employing both 3D single-particle and 1D QED-PIC simulations, the authors demonstrate that curvature radiation photons exhibit strong linear polarization aligned with the plane of the magnetic field, while synchrotron radiation photons are weakly polarized due to the randomized orientation of e^± acceleration during their gyration (see Fig. 13). As the cascade develops, synchrotron radiation photons become increasingly dominant, leading to a net reduction in average photon polarization. Crucially, the study quantifies the effect of polarization on pair production rates, revealing that the high polarization of curvature radiation photons enhances e^± pair creation by approximately 5% compared to unpolarized models—a finding that contrasts with the suppression typically observed in analogous laser-plasma settings. The simulations further distinguish between “shower-type” cascades, driven by injected electrons without sustained electric fields, and "avalanche-type" cascades that are self-sustained by persistent parallel electric fields. In both cases, polarization-dependent phenomena remain robust, as confirmed by consistency between single-particle and QED-PIC approaches. This work thus constitutes a significant advancement in our understanding of magnetospheric pair creation, providing the first fully self-consistent, polarization-resolved simulation framework for QED cascades in the unique geometrical and dynamical context of pulsar polar caps.

Fast Radio Bursts (FRBs), as another example,  are extremely short, intense flashes of radio waves typically lasting only a few milliseconds. Since their discovery in 2007, they have become one of the most intriguing open problems in contemporary astrophysics. Traditionally, the short observed duration of FRBs has been interpreted as a direct constraint on the size of their central engine. Recent observations of FRBs on the microsecond time-scale challenge this interpretation. In a recent work, we have has put forward a new explanation of the short duration of FRBs: high-energy photons from a FRB’s X-ray/soft gamma-ray counterpart can trigger an electron–positron pair cascade via Compton scattering and the Breit–Wheeler process in the plasma surrounding the source. This cascade rapidly increases the local plasma density, rendering it opaque to the FRB’s radio emission after a light-travel time of up to a few milliseconds, thus truncating the observable burst regardless of the intrinsic emission duration (see Fig. 14). This mechanism, motivated by observations of FRB 200428 from the Galactic magnetar SGR J1935+2154, implies that FRB sources are not necessarily constrained to produce intrinsically millisecond-long radio pulses, and that similar propagation-induced truncation could be common among FRBs, potentially impacting how their duration is interpreted in relation to their origins.

In ultrastrong fields in plasma which can be generated during plasma interaction with petawatt-class lasers, or during different astrophysical instabilities, spin polarized electrons and polarized gamma photons are produced due to radiative polarization. Previously, Karen Hatsagortsyan and co-workers have investigated new methods for the plasma imaging via polarization properties of ejected particles, which will enable in situ probing the plasma dynamical characteristics. In our further study, we established the potential of electron polarization as a source of new information on laboratory and astrophysical plasma instabilities. One of the studies was devoted to the investigation of the polarization properties of electrons during plasma current filamentation of an ultrarelativistic electron beam impinging on an overdense plasma. We showed that electron radiative polarization emerges during the instability along the azimuthal direction in the momentum space. The correlation between the emerging spin polarization and the collective behaviors enables decoding current filamentation instability scenarios via polarization detection.

In another study we address relativistic preturbulent shocks, and examined the mechanisms of electron acceleration. We have revealed a new mechanism of acceleration, dubbed as “Slingshot Acceleration in Relativistic Preturbulent Shocks”, which has been found analyzing the emitted photon polarization. We investigated the transient electron dynamics in the transition to turbulence near the counter-streaming interface of an unmagnetized pair-loaded relativistic collisionless shocks (RCS) precursor, which is potentially associated with the outflow of Gamma-ray bursts (GRBs). We found that the integrated radiation from the RCS exhibits a counterintuitive nonmonotonic dependence of the photon linear polarization degree along the transverse acceleration on the photon energy. The analysis showed that this uncommon polarization behavior of gamma-photons originated from the accelerated electron in the transition region from the preturbulent laminar motion to a chaotic turbulence. This slingshotlike acceleration mechanism is distinct from the well-known stochastic acceleration and is induced by the drifting electric field sustained by the flowing focus of backward-moving electrons (see Fig. 15).  Our results demonstrate the potential of photon polarization as an essential information source in exploring intricate transient dynamics in RCSs with relevance for Earth-based plasma and astrophysical scenarios.

Further, we applied a fully polarization resolved study via particle-in-cell simulations, both for the electron spin and gamma-photon polarization, to radiation reaction dominated magnetic reconnection. We identify a condensation transient plasma evolution of plasmoids accumulated into multiple tiny islands within the reconnection layer, where electrons are strongly polarized while emitting energetic γ-ray photons to undergo radiative spin flips (see Fig. 16). Nonlinear analyses elucidate that the condensation is caused by a spiral attractor appearing in the electron’s phase space due to radiation reaction. The spiral rotation and contraction of the attractor lead to the electrons’ polarization aligned with the magnetic field. In its turn, the spin-flip during a γ-ray emission is responsible for the anomalous linear polarization along the magnetic field.  Thus, from our fully polarization resolved study we conclude that spin-polarized condensed plasmoids may be realized with around 10¹⁰ G magnetic fields in extreme power laser facilities and intrinsically can exist in extreme astrophysical reconnection scenarios, potentially explaining atypical polarization features in observed high-energy cosmic radiation.