Publications Repository - Helmholtz-Zentrum Dresden-Rossendorf

Hyperspectral imaging methods allow extensive information on rock properties to be captured over large areas (e.g., entire drill cores). While typically interpreted in terms of mineralogy, these data are also sensitive to textural properties like grain size and porosity. In this study, we explore possible links between hyperspectral data and physical rock properties, using deep learning to predict multiple petrophysical properties, including slowness (the reciprocal of P-Wave Velocity), density, and gamma ray readings. Our dataset consists of three boreholes drilled in the Spremberg region of Germany. Our deep learning model achieves high predictive performance, with test R2 scores of 0.889 for slowness, 0.949 for density, and 0.659 for gamma-ray readings. Shapley value analyses reveal that the hyperspectral bands used by these models coincide with known mineralogical absorption features, especially in the long- and mid-wave infrared range. These results suggest that hyperspectral data might, with appropriate training data and validation, be used as a reliable proxy for important physical rock properties, facilitating non-invasive geological assessments. Most importantly, this methodology could be scaled up to allow rock-property prediction over large areas (e.g., large drill core collections or, potentially, outcrops or mine-faces). Our research thus advances the application of hyperspectral data in geoscience, paving the way for more accurate and efficient subsurface characterizations with significant implications for resource exploration.

Hyperspectral data provides rich quantitative information on both the mineralogical and fine-scale textural properties of rocks, which, in turn, largely control their petrophysical characteristics. We therefore developed a deep learning model to predict petrophysical properties directly from hyperspectral drill core data. Our model learns relevant features from high-dimensional hyperspectral data and co-registered sonic, gamma-gamma density and gamma-ray logs to infer slowness, density, and gamma-ray counts. We demonstrated the performance of this approach on data acquired in the Spremberg region of Germany. Our results demonstrate that with meticulous pre-processing steps and thorough data cleaning, one can overcome the difference in capturing resolution and learn the relationship between hyperspectral data and petrophysics. Using a test dataset from a spatially independent borehole, we generate a pixel-resolution (≈ 1 mm2) model of the petrophysical properties and resample it to match the measured logs. This test indicates substantial accuracy, with R2 scores and root-mean-squared errors (RMSE) of 0.7 and 16.55 μs.m-1, 0.86 and 0.06 g.cm-3 and 0.90 and 15.29 API for the slowness, density and gamma-ray readings respectively. Overall, our findings lay the groundwork for building deep learning models that can learn to predict physical and mechanical rock properties from hyperspectral data. Such models could provide the high-resolution but large-extent data needed to bridge the different scales of mechanical and petrophysical characterisation.

In this study, we investigated a novel anti-cancer drug design approach by revisiting diclofenac-based carborane-substituted prodrugs. The redesigned compounds combine the robust carborane scaffold with the oxindole framework, resulting in four carborane-derivatized oxindoles and a unique zwitterionic amidine featuring a nido-cluster. We tested the anti-cancer potential of these prodrugs against murine colon adenocarcinoma (MC38), human colorectal carcinoma (HCT116), and human colorectal adenocarcinoma (HT29). The tests showed that diclofenac and the carborane-substituted oxindoles exhibited no cytotoxicity, the dichlorophenyl-substituted oxindole had moderate anti-cancer activity, while with the amidine this effect was strongly potentiated with activity mapping within low micromolar range. Compound 3 abolished the viability of selected colon cancer cell line MC38 preferentially through strong inhibition of cell division and moderate apoptosis accompanied by ROS/RNS depletion. Our findings suggest that carborane-based prodrugs could be a promising direction for new anti-cancer therapies. Inhibition assays for COX-1 and COX-2 revealed that while diclofenac had strong COX inhibition, the re-engineered carborane compounds demonstrated a varied range of anti-cancer effects, probably owing to both COX-inhibition and COX-independent pathways.

The Virtual Geoscience Conference (VGC) provides a vital interdisciplinary platform for researchers in geoscience, geomatics, and related fields to exchange insights on the latest methodological advancements and applications. The conference brings together experts focused on the development and application of geomatic techniques—such as LiDAR, photogrammetry, immersive visualization, computer vision, computer graphics, and terrestrial InSAR—acrossEarth and environmental sciences. The increasing adoption of 3D measurement technologies, UAV platforms, and innovative 2D imaging approaches such as hyperspectral and thermal imaging, is evident across diverse sub-disciplines within geoscience. These sophisticated, multi-sensor toolkits enable unprecedentedly detailed quantitative mapping, facilitating analyses that enhance our understanding of the natural environment and address societal challenges, including resource and energy security, geohazards, disaster management, and climate/environmental change.

The key strategy to increase recycling efficiency is to identify components as early as possible before shredding . We propose a Smart sEnsor nEtworK (SEEK) for improved identification of materials in complex waste streams. It relies on automated data acquisition from multiple sensors, with real-time processing driven by AI-based algorithms.

Digital outcrop models have become a powerful tool for detailed geological mapping, allowing geological exposures to be characterized in unprecedented detail, while simultaneously mitigating access limitations that hinder conventional mapping approaches. Here we present an emerging workflow that fuses digital outcrop data with high resolution ground- and UAV- based hyperspectral imaging products to better discriminate lithological units, marker horizons and mineralogical trends.
These hyperspectral data provide valuable information on gangue and alteration minerals, which can help quantify important spatial variations in mineral systems. In some settings, bulk mineralogy (or even grade information) can be directly estimated from hyperspectral data, allowing continuous mm-cm resolution mapping across large areas (meters to kilometers). Otherwise cryptic changes in e.g., clay or white-mica mineralogy and mineral chemistry can be identified and used to distinguish e.g., paleotemperature zonations within epithermal or massive sulfide deposits.
In this contribution, the strengths and weaknesses of the hyperspectral method will be discussed in the context of ore geology and mineral system research, and illustrated using using various examples from Germany, Spain, Greenland, Morocco and Italy, where hyperspectral data has helped constrain the geometry of different geo-bodies across a range of tectonic environments and the associated diagenetic and mineral systems.

Rare earth elements (REE) are key constituents in electronic devices (e.g. smartphones, batteries), being present in both end-user and industrial applications. The rapid innovation cycles of electronic devices, combined with the increasing demand for new technological applications (e.g. mobility and e-cars) pose a challenge for the supply of REE, which are considered as Critical Raw Materials (CRM). This scenario calls for rapid, non-invasive methods that enable the identification of new REE-rich mining resources. Furthermore, the high supply risks associated with CRM such as REE drive technological developments to compensate and overcome market fluctuations by turning previously not mined co-resources into valuable and economic modalities, such as re-mining materials.

We present an investigation focused on the identification of REE in waste rocks and tailing materials from the mine of Siilinjärvi (Finland). The deposit in the area consists of alkaline-carbonatite rocks, with the most important REE-bearing minerals being apatite (average REE concentration: 0.4% (wt%)) and monazite (REE concentration: up to 67% (wt%)). Mining activities focus on extraction of phosphate from fluorapatite, and the chemical reactions involved in this extraction generate phosphogypsum (PG) as a by-product. Literature reports indicate that REE can be incorporated to the PG matrix in the crystallisation process, with the most relevant examples including Nd, Ce, La, Sm, Gd, Tb, Dy, and Eu.

Our goal is to highlight how the sequential acquisition by multiple optical methods (multi-sensor approach) can trace REE contents for individually identified REE from pristine rocks to processing waste dumped in tailings. Each material type was scanned by two fast hyperspectral imaging (HSI) sensors integrated in a conveyor-belt system: a reflectance-based HSI sensor operating in the visible to near-infrared and short-wave infrared (Specim AisaFenix); and an innovative laser-induced fluorescence line scan sensor (HSI-LiF, Freiberg Instruments). The optical sensing results were validated by mineralogical methods (mineral liberation analysis (MLA)). MLA results for PG indicate the presence of REE-bearing minerals including gypsum, apatite, and monazite (respective abundances (wt%): 97.4, 0.6, and 0.08).

Optical features characteristic of Nd were identified on rocks and tailings samples by both HSI-reflectance and HSI-LiF sensors. Spectral signatures were detected in HSI-LiF spectra for an additional REE group including Sm, Er, and Pm.

We highlight that efficient, non-invasive optical sensing can detect and re-evaluate tailing materials as a baseline for economic considerations according to market needs. The results confirm that REE detected on the pristine rocks of the mine can be traced through the mineral processing route to be found again in the tailings material. The multi-sensor optical detection based on HSI-reflectance and HSI-LiF, accordingly, provides an efficient non-invasive tool for exploring both mining and re-mining potential by providing immediate results on REE types and their spatial abundance, when employed as scanning techniques.

Hyperspectral imaging is gaining widespread use in the resource sector, with applications in mineral exploration, geometallurgy, and mine mapping. However, the sheer size of many hyperspectral datasets (>1 Tb), and associated data correction and analysis challenges, limit the integration of this technique into time-critical exploration and mining workflows. We present an overview of several newly developed real-time processing capabilities to mitigate these challenges, and so provide hyperspectral data and derived products (e.g., mineral abundance estimates) in near real-time. This allows for efficient, timely, and automated delivery of hyperspectral data to enhance geological activities.
Hyperspectral data generally needs to be corrected, coregistered, cropped and masked, before derivative results can be generated, visualized and stored. To achieve real-time processing, each of these steps, which can involve the computationally intense manipulation of several Gb worth of spectral data, need to be completed within the 1-3 minutes a typical instrument or scanner takes to capture a data cube. To help with this, we have developed a python-based asynchronous processing pipeline, crunchy, that uses a file-discovery-based triggering mechanism to spawn parallel processing workflows that automatically perform these tasks. Coregistered and radiometrically corrected results are then stored using a directory-based data structure managed by a second python utility, hycore, that facilitates (1) consistent data storage, (2) file-based out-of-core processing, and (3) management of the various metadata required to localize and give meaning to hyperspectral drill core data. We have also developed a third python tool, hywiz, to enable an easy browser-based interaction with hycore databases. This includes the visualisation of sensor results and analysis products for individual trays and drillhole mosaics. Additional data such as assays, logging notes or downhole geophysical data can be overlain on these to enable integrated interpretation of otherwise disparate datasets.
We hope that these tools will enable greater use of hyperspectral data in research and industry, and facilitate e.g., hyperspectrally enhanced core-logging, sample selection, vectoring and, potentially, realize self-updating 3-D geological models.

Hyperspectral imaging is gaining widespread use in the resource sector, with applications in mineral exploration, geometallurgy, and mine mapping. However, the sheer size of many hyperspectral datasets (>1 Tb), and associated data correction and analysis challenges, limit the integration of this technique into time-critical exploration and mining workflows. Here we present an overview of several real-time processing capabilities we have developed to mitigate these challenges, and so provide hyperspectral data and derived products (e.g., mineral abundance estimates) in near real-time. This allows for efficient, timely, and automated delivery of hyperspectral data to enhance geological activities.

The unique optical features of gold nanoparticles (AuNPs), such as localized surface plasmon resonance (LSPR), have attracted substantial attention in biosensing applications. In closely spaced nanoparticle assemblies, the electromagnetic fields of neighboring particles interact strongly, leading to significant near-field coupling which influence the LSPR signature. Therefore the collective plasmonic characteristics in ordered arrangements of AuNPs are more sensitive to variations in the local refractive index (RI) than in individual particles. Such local RI modifications also take place when biomolecules bind to the AuNP surfaces. Therefore, by increasing RI sensitivity, AuNP assemblies have the potential to detect biomolecules of interest with much higher accuracy.
In this study, a comprehensive investigation is presented comparing the plasmonic spectra of linear periodic assemblies of AuNPs against individual particles (50 nm diameter) for biosensing applications. Simulations using the Finite-Difference Time-Domain (FTDT) method suggested that longitudinal coupling along the AuNP lines were more sensitive to RI changes than transversal coupling. Practical experiments supported the simulation results through an exemplary attachment of the biomolecules to AuNP assemblies. A pro-inflammatory cytokine- Tumor Necrosis Factor Alpha (TNF-α), which is an important marker in cancer research influencing various aspects of tumorigenesis, tumor progression, and therapeutic response was chosen for the bio-functionalization process. Polydimethylsiloxane (PDMS) templates were used to confine large arrays (cm²) of amine-functionalized AuNPs into linear assemblies on glass substrates. RI sensitivity of the AuNP assemblies during various steps of the functionalization process were investigated using UV-Vis spectrometry. Promising experimental results exhibited enhanced RI sensitivity of the linear assemblies as compared to individual or randomly ordered AuNPs, offering a favorable approach towards plasmonic biosensing applications.

Existing diagnostic methods for cancer, for example- imaging and tissue biopsies, are expensive, invasive and impractical for repeated examination. While soluble biomarkers can be quantified quickly and non-invasively, their detection in ultra-low concentrations remains a challenge. Sensors based on Field Effect Transistors (bioFETs) with silicon channels are highly sensitive [1] and offer an ideal platform. Plasmonic gold nanoparticles (AuNPs) have tunable optical properties and their free electron clouds undergo collective oscillation upon interaction with light of specific wavelengths, exhibiting Localized Surface Plasmon Resonance (LSPR) [2]. Furthermore, ordered arrays of NPs exhibit coupled plasmonic properties which are highly susceptible to variations in the local refractive index- a change which also occurs when biomolecules attach to NP surface.
Our presented work investigates template-assisted assembly of AuNP chains and biofunctionalization with exemplary cytokine: Tumor Necrosis Factor Alpha (TNF-α). The research aims towards utilization of the coupled plasmonic properties of the AuNP chains to contribute to the change in interfacial charges at the bioFET channel and thereby improve the performance of bioFETs through optical gating.

[1] Trang Anh Nguyen-Le et al., Biosensors & Bioelectronics (2022); 206:114124
[2] Christoph Hanske et al., Nano Letters 2014 14 (12), 6863-6871

Gold nanoparticles (AuNPs) attract substantial interest in biosensing applications due to their unique optical properties, such as localized surface plasmon resonance (LSPR). The near field interactions between individual NPs, e.g. in assemblies or arrays, have a significant impact on LSPR signature. Compared to individual particles, ordered arrangements of AuNPs offer collective plasmonic properties which are more susceptible to variations in the local refractive index (RI). Such local changes in RI also occur when biomolecules attach to the surface of AuNPs. Hence, AuNP assemblies could potentially enable the detection of biomolecules of interest with unprecedented accuracy due an enhancement of RI sensitivity. This study presents a comprehensive investigation of the plasmonic spectra of linear periodic assemblies of AuNPs (590 nm periodicity) against individual particles (50 nm diameter) for biosensing applications. Simulations using Finite-Difference Time-Domain (FTDT) method showed that longitudinal coupling along the AuNP lines is more sensitive to RI changes than transversal coupling. This was further investigated experimentally through an exemplary attachment of the biomolecules to the nanoparticle lines. Tumor Necrosis Factor Alpha (TNF-α), a pro-inflammatory cytokine which plays a multifaceted role in cancer research, influencing various aspects of tumorigenesis, tumor progression, and therapeutic response was chosen for the bio-functionalization experiments. Wrinkled polydimethylsiloxane (PDMS) templates were used to confine the cm² large arrays of amine-functionalized AuNPs into lines on glass substrates. RI sensitivity of the AuNP assemblies during various steps of the functionalization were investigated using UV-Vis spectrometry. Promising experimental results demonstrate enhanced RI sensitivity of the linear assemblies, compared to single NPs, offering a new approach towards plasmonic biosensing applications.

Understanding charge transport properties of large-area single-layer 2D materials is crucial for the future development of novel optoelectronic devices. In this work, the synthesis and electrical characterization of large-area single-layers of Cu3BHT 2D conjugated coordination polymers are reported. The Cu3BHT are synthesized on the water surface by the Langmuir-Blodgett method and then transferred to SiO2/Si substrates with pre-patterned electrical contacts. Electrical measurements revealed ohmic responses across areas up to ≈1 cm2, with a mean resistance of approximately 53 ± 3 kΩ at a probe separation of 50 µm. Cooling and heating cycles show hysteresis in the electrical response, suggesting different current pathways are formed as the samples underwent structural-chemical changes during temperature sweeps. This hysteresis vanished after several cycles and the conductivity shows a stable exponential behavior as a function of temperature, suggesting that a temperature-dependent tunneling process is governing the conduction mechanism in the analyzed polycrystalline single-layer Cu3BHT samples. These results, together with density functional theory calculations and valence band X-ray photoelectron spectroscopy data suggest that the single-layer samples exhibit a semiconducting rather than a metallic behavior.

In this study, we unveil a conceptual technology for fabricating artificial metalloenzymes (ArMs) by deeply integrating hemin into protein scaffolds via a protein refolding process, a method that transcends the conventional scope of surface-level modifications. Our approach involves denaturing proteins, such as benzaldehyde lyase, green fluorescent protein, and Candida antarctica lipase B, to expose extensive reactive amino acid residues, which are then intricately linked with hemin using orthogonal click reactions, followed by protein refolding. This process not only retains the proteins’ structural integrity but expands proteins’ functionality. The most notable outcome of this methodology is the hemin@BAL variant, which demonstrated a remarkable 83.7% conversion rate in cyclopropanation reactions, far surpassing the capabilities of traditional hemin-based catalysis in water. This success highlights the significant role of protein structure in the ArMs’ activity and marks a substantial leap forward in chemical modification of proteins. Our findings suggest vast potentials of protein refolding approaches for ArMs across various catalytic applications, paving the way for future advancements in synthetic biology and synthetic chemistry.

We combine 2 m resolution airborne (HySpex) and 30 m resolution satellite (EnMAP) hyperspectral data to address the challenge of mixed pixels in satellite imagery. Endmembers are manually selected from HySpex data, and Non-negative Least Squares (NNLS) spectral unmixing is applied to generate high-resolution spectral abundance maps. These maps are then resampled to match EnMAP’s spatial resolution and used to predict an endmember library from the EnMAP scene. This predicted library is then used for unmixing the EnMAP data over a broader area. When compared to spectral abundance maps generated from direct endmember selection from EnMAP alone, the unmixing results using the predicted library closely align with the high-resolution output, despite some land cover changes over time. In contrast, the spectral abundance maps from low-resolution endmembers lack detail. We discuss the implications of our approach for improved spatial and temporal mapping.

Spectral indices use band division to target specific absorption features or reflectance changes (e.g., the red-edge in vegetation spectra). The resultant intensities help detect targeted objects across a scene. However, most standardized spectral indices formulas are designed for multispectral sensors (e.g., Landsat/ASTER), with hyperspectral data typically downsampled spectrally to fit these formulas.

Band averaging can work well for heterogeneous scenes, but we have found that it can suppress subtle variations over relatively homogeneous scenes, such as dense vegetation or extensive carbonate-rich zones. We postulate that this is because the higher spectral resolution is not being fully leveraged. For example, with NDVI, a slight shift in the red-edge can indicate the state of plant health. However, averaging bands in the NIR and Red range to fit the standard NDVI equation will easily overlook this indication. Similarly, in carbonate-rich zones, where calcite and dolomite may be present in close mixtures, averaging many spectral bands may also miss the necessary shifts to differentiate between these minerals.

In this contribution, we propose a Tree-structured Parzen Estimator (TPE) algorithm that can help to optimize spectral band selection for spectral index analyses with hyperspectral sensors, while retaining compatibility with well-established multispectral ones. TPE, a Bayesian hyperparameter optimization technique, improves selection based on previous trials. It treats the continuous range of hyperspectral band wavelengths as a search space and evaluates initial samples with a Gaussian mixture model. The algorithm iteratively generates new candidate band combinations by exploring areas of the search space that yield better performance.

We tested this approach for NDVI, focusing on the NIR and Red bands at the Hohes Holz, (Germany) research site, and for calcite with a signature absorption at 2337 nm in the Marinkas carbonatite region (Namibia). These applications assessed the suitability for both VNIR and SWIR spectral regions of operational hyperspectral sensors - PRISMA, EnMAP, and EMIT. The results show that for the NDVI index (NIR-RedNIR+Red), the maximally correlated band equations for PRISMA and EMIT are 913.45-664.89913.45+664.89 and 902.37-671.09902.37+671.09 respectively. For the calcite spectra index (2190:22242293:23452375:24302293:2345), the optimized band equations are 2199.452322.132400.002322.13 for EnMAP and 2204.502330.332396.882330.33 for EMIT. The resultant hyperspectral indices highlight more subtle variations than their multispectral counterparts, facilitating comparison across sensors.

Well-exposed sites for mineral exploration are scarce, and presently potential mineral-rich areas globally are mostly covered with vegetation and topsoil, which are suboptimal for direct remote sensing-based exploration. Arsenic (As) is known to be a pathfinder element for gold mineralisation, thus mapping its anomaly across a terrain can be of very high value. Our study area is located in the Rise and Shine Shear Zone (RSSZ) in South Island, New Zealand and is part of the Otago schist. Gold mineralisation here is hosted in the upper greenschist facies rocks within the RSSZ and is separated from lower greenschist facies rocks by the post-mineralisation Thomson’s Gorge Fault. Previous geological studies have recognised and mapped the general geology of the area and carried out geophysical mapping, however, hyperspectral remote sensing has never been applied to be used for gold exploration in a similar setting. This study aims to explore relevant information on concealed subsurface geology using its surface manifestations via airborne high-resolution hyperspectral imaging.
Initially, we performed band selection employing recursive feature elimination using field data and a mineralogical understanding of the area. Subsequently, an orthogonal total variation component analysis (OTVCA) was conducted on the resultant 85 spectral bands to consolidate the information in 10 spectral bands. The OTVCA results were finally classified into three levels of soil As concentration; <20 ppm, 20-100 ppm, and >100 ppm using a random forest classifier.
We found an inherent connection between geology and exposed soil, which contributes extensively to the classification accuracy but introduces challenges in analysis (e.g., miss-classifications). Despite these complexities, the delineated high As concentration zones are a good match with potential gold mineralisation zones in the RSSZ area. This research adds valuable insights to gold exploration in such a challenging setting.

Discovery of the mineral resources fuelling the green energy transition requires innovative, robust, and accurate remote sensing datasets for regional-scale mineralogical assessment. New satellite-based hyperspectral data could help provide this by allowing the identification of subtle mineralogical changes over large areas along with repeated temporal data for efficient monitoring. Although mineral systems usually have large spatial footprints, they are often difficult to detect from satellites due to their subtle spectral manifestation (e.g., less spectral dominance, or minor shifts in absorption characteristics), and/or influence of vegetation, soil, etc, reducing the pixel area of the actual mineral outcrop. Airborne hyperspectral data, however, can provide finer spatial resolution, potentially capturing more spectrally complex information than satellite-based hyperspectral data, but over smaller areas and at a relatively higher expense. A balanced integration of the airborne and the satellite-based hyperspectral data can thus address many current gaps in the mineral prospecting domain.

Das Verständnis der hydrogeochemischen Prozesse und deren numerische Modellierung sind für die Beurteilung der Schadstoffmigration in Grundwassersystemen, einschließlich der Anwendungen bei der Entsorgung radioaktiver Abfälle, von entscheidender Bedeutung. Das Projekt SANGUR (Systematische Sensitivitätsanalyse für mechanistische geochemische Modelle unter Verwendung von Felddaten aus kristallinem Gestein) befasst sich mit der Frage, welche Methoden und Parameter für den reaktiven Transport unter besonderer Berücksichtigung von Heterogenitäten in kristallinem Gestein relevant sind. Es wurde ein Workflow entwickelt, der die Datenerfassung mit geostatistischen Ansätzen und einer Modellreduktion basierend auf einer Sensitivitätsanalysen kombiniert. Ziel ist die Verbesserung der Vorhersage der Radionuklidrückhaltung im Fernfeld eines Endlagers.

Die Gesteinssimulation wird in unserem Ansatz mit Hilfe von Multinary Random Fields realisiert.[1] Als Trainingsdatensatz dienen Kristallinproben aus der Lausitz, deren mineralogische Zusammensetzung mittels MLA (Mineral Liberation Analysis) bestimmt wurde. Der gewählte Simulationsansatz erlaubt es, eine Vielzahl von Realisierungen zu berechnen und damit die mineralogische Zusammensetzung und deren Variabilität entlang der Migrationspfade unter Berücksichtigung von Unsicherheiten zu beschreiben. Dies ist eine Voraussetzung für die Auswahl realistischer Oberflächenkomplexierungsmodelle und –parameter, die wiederum die Berechnung smarter Kd-Matrizen zur Beschreibung der Radionuklidmigrationsmuster ermöglichen.[2] Sowohl die Eingabeparameter als auch die Smart-Kd-Matrix werden in die anschließende Sensitivitätsanalyse einbezogen, um die Relevanz einzelner Parameter, aber auch deren Abhängigkeiten für die Simulation der Radionuklidrückhaltung zu ermitteln. Dadurch kann einerseits mehr Aufwand in die Bestimmung der wichtigsten Parameter und ihrer Unsicherheiten investiert werden, andererseits können Parameter mit geringerem Einfluss als Konstanten gesetzt werden, was zu weniger komplexen und rechenzeitintensiven Modellen führt.

Wir stellen die Arbeitsschritte sowie die Ergebnisse des gesamten Workflows vor und können so erste Aussagen zur Frage der Relevanz von Modellparametern wie Mineralzusammensetzung der Festphase, Zusammensetzung der Fluidphase, pH-Wert und Simulationsskala präsentieren.

Referenzen:

[1] Menzel, P. et al. (2020) Math. Geosci. 52, 731 – 757. [2] Stockmann, M. et al. (2017) Chemosphere 187, 277 – 285.

Efficient and sustainable production, recovery and recycling phases of semiconductors (SC) life cycles require non-invasive, inline methods able to identify their composition in material streams. Ideally, the sensor system should be fast and incorporated into conveyor-belt operations. Rapid identification as well as spatial distribution maps would allow for real-time monitoring and quality control of the material stream. Considering these requirements, we suggest the sequential use of fast hyperspectral reflectance imaging (HSI) and Raman spectroscopic sensors for the identification of SC types in a sensor network configuration. We propose spectral proxies based on electronic properties derived from HSI-reflectance (i.e. absorption edge linked to the band gap values) and Raman sensors (i.e. Raman-active phonon modes) for SC identification. We identify potential limitations of each proxy on identifying undoped/doped SC materials, and discuss which process workflows enable optimized SC classification. We demonstrate the multi-sensor approach with SC standards (GaAs, GaSb, InP, 4H-SiC, and Borosilicate) which are relevant for both opto- and power-electronic devices, and showcase the potential of sequential data acquisition by fast HSI-reflectance sensors in the visible to shortwave-infrared (integration times: (4.5–18) ms) and Raman scattering (excitation laser: 532 nm, acquisition times: (0.5–10) s).

The high electrolytic currents in aluminum reduction cells can provoke magnetohydrodynamic wave
instabilities in the liquid-liquid cryolite-Al interface. Most critical for the safe operation of aluminum
smelters is the metal pad roll (MPR) instability, which sets the interface into self-growing rotational
motions. Such interface instabilities are commonly averted by keeping the poorly conducting cryolite
layer sufficiently thick, but at the detriment of the energy efficiency. Mohammad et al. [1, 2] have
recently demonstrated that the cryolite layer can be markedly reduced when adding an oscillating
component to the electrolytic current, inhibiting exponential growth of the MPR instability. We
dedicate this paper to the investigation of the underlying physics behind this new wave suppression
technique. We analyze the MPR stability using a simplified mechanical model, which reduces the
mathematical problem to a set of two coupled Mathieu’s differential equations. The state of stability is
calculated both numerically using Floquet theory and analytically by applying the complexification-
averaging method. Our analysis reveals that observed stability patterns can essentially be attributed
to a simultaneous occurrence of parametric resonance and anti-resonance. We identify ideal system
parameters and show ways to verify the rather exotic phenomenon of parametric anti-resonance in
MPR experiments.

Living organisms establish interaction through the exchange of physicochemical signals. The cumulative effect of these exchanges cascades across scales controlling the emergence and maintenance of home-ranges and territories. Therefore, a theoretical framework aiming to elucidate the role of behavior in how animals partition the use of space must adopt a bottom-up approach, incorporating individual-level interactions. During this presentation we will discuss a potential data-driven structure for this framework. We will start by analyzing animal tracking data, looking for signatures of interactions, specifically focusing on contact-type interactions. Then, having characterized interactions, we move towards a connection between a quantified behavioral trait (e.g. level of aggressiveness among individuals) and how space use is partitioned.

In recent works [1,2], collaborators and I, have worked a general framework to coarse-grain the dy-
namics of physicochemical substances that mediate the interaction of a given population. We generally
note that the dynamics timescale, by itself, can qualitatively change the effective description that emerge
from the coarse-graining procedure. Interestingly, at the slow limit we recover Turing-like models, with
two coupled partial differential equations, and, on the fast limit, we obtain a single equation which is
non-local (or kernel-based), accounting for spatially-extended interactions. Therefore, our work bridge
two class of models that are used to study pattern formation and reveal how the dynamics of mediating
substances induce distance-dependent interaction between individual of the focal population.
The crucial and fundamental point of our results is that pattern formation is not only controlled by the
operator dynamics of the mediators (activator-inhibitors). Just by changing the the timescale (keeping
same dynamics), the system detour from standard pattern formation criteria (Turing’s three criteria).
We show that this occur due to how nonlinearities associated to the mediators propagation cascade to
large-scales. This general findings are then concretely applied to: i) a population where individuals are
constantly releasing a toxic substances that diffuse and decay [1]; ii) a population which release substances
that can increase or decrease the motility of individuals [2].
Lastly, we will discuss the challenging in accessing the dynamics of mediators from data and how
theoretical developments could assist on this initiative. Furthermore, a model that helps bridge scale
could help infer from top-down (from pattern to mediators) features of the mediators dynamics [3].

Interactions between organisms are mediated by an intricate network of physico-chemical substances and other organisms. Understanding the dynamics of mediators and how they shape the population spatial distribution is key to predict ecological outcomes and how they would be transformed by changes in environmental constraints. However, due to the inherent complexity involved, this task is often unfeasible, from the empirical and theoretical perspectives. In this paper, we make progress in addressing this central issue, creating a bridge that provides a two-way connection between the features of the ensemble of underlying mediators and the wrinkles in the population density induced by a landscape defect (or spatial perturbation). The bridge is constructed by applying the Feynman-Vernon decomposition, which disentangles the influences among the focal population and the mediators in a compact way. This is achieved though an interaction kernel, which effectively incorporates the mediators' degrees of freedom, explaining the emergence of nonlocal influence between individuals, an ad hoc assumption in modeling population dynamics. Concrete examples are worked out and reveal the complexity behind a possible top-down inference procedure.

The data publication contains the primary data used for the publication. There are three groups of data: Raman data, optical data, and DFT data. The two former are in txt format, the DFT partially as Excel and partially as txt file.

The FLUKA Monte Carlo Radiation Transport and Interaction code package is widely used to simulate the interaction of particles with matter in a variety of fields, including high energy physics, space radiation, medical applications, radiation protection and shielding assessments, accelerator studies, astrophysical studies and well logging. This paper gives a brief overview of the FLUKA program and describes recent developments, in particular, improvements in the modelling of particle interactions and transport are described in detail. In addition, an overview of selected applications is given.

Magnetoelectric uniaxial antiferromagnet Cr2O3 (chromia) is a prospective material for fundamental research with the recent demonstrations of spin superfluidity [1] and flexomagnetism [2], as well as for spintronics applications [3].

We present a theoretical study of the properties of such magnetic topological solitons in chromia as domain walls and compare them with experimental observations. Structural defects like grain boundaries which are commonly present in thin films, act as the pinning sites of domain walls. The energy landscape formed by the boundaries of small grains enables the critical size of the chromia bit below which they mainly tend to be in the single-domain magnetic state even with zero-field cooling state preparation [4].

The magnetic symmetry of chromia is characterized by the center of anti-inversion, which leads to the coupling between gradients of magnetic texture and external field. We describe the coupling between the domain wall in chromia and the magnetic field and show the presence of a new field-induced spin-reorientation phase transition below the spin-flop phase. The theoretical conclusions are confirmed by the scanning nitrogen-vacancy magnetometry and X-ray magnetic linear dichroism (XMLD) measurements.

[1] W. Yuan et al., Sci. Adv. 4 (2018) eaat1098
[2] P. Makushko et al., Nat. Commun. 13 (2022) 6745.
[3] J. Han et al., Nat. Mater. 22 (2023) 684; H. Meer et al., Appl. Phys. Lett. 122 (2023) 080502.
[4] Rickhaus, Pylypovskyi, Seniutinas, Borras, Lehmann, Wagner, Žaper, Prusik et al., Nano Letters (2024, in press) arXiv:2406.19085v1

Magnetoelectric uniaxial antiferromagnet Cr2O3 (chromia) is a promising material for applications in spintronics [1] which exhibits rarely observed physical phenomena such as spin superfluidity or flexomagnetism [2]. Here we present a systematic study of the chromia’s σ-model derived using the spin Hamiltonian for chromia. We focus on the properties of the domain walls and finite-size samples, confirming the analytical predictions with spin-lattice simulations and X-ray magnetic linear dichroism (XMLD) measurements.

While the commonly used σ-model of a bipartite antiferromagnet has a quadratic coupling between the Neel vector (antiferromagnetic order parameter) n and magnetic field B, in chromia there is a linear coupling between B and the gradient of n, previously predicted from the symmetry considerations [3]. We quantify this coupling in terms of spin lattice parameters including exchange integrals and interatomic distances. The resulting energy term modifies the boundary conditions for n and is responsible for the appearance of a new field-driven spin-reorientation phase.

The boundary conditions for n at a c-plane cut are modified by an additional term that mimics unidirectional surface anisotropy of the easy-axis type, with the axis aligned with the magnetic field. This term becomes active in the spin-flop state, leading to the near-surface tilt of n from the in-plane to the out-of-plane direction.

The domain wall (DW) lying in the basal plane of chromia possesses a finite magnetic moment MDW(p, B) whose direction is determined by the DW polarity p. We found that in magnetic fields starting from of about 55% of the spin-flop field, the Zeeman energy associated with MDW can lower the total energy below the energy of the collinear state. This is manifested as an additional field-driven phase transition preceding the spin-flop transition. While the contribution of MDW(p, B) to the total magnetic moment of large single-crystal samples is negligibly small, it is comparable with the total moment of chromia films with a thickness below several hundredths of nanometers in the spin-flop state. This new phase in chromia effectively lowers the critical spin-reorientation field by almost a factor of two. Using XMLD measurements in thin films and single crystals of chromia in a wide range of temperatures, we show that thin-film samples possess a sizeable finite magnetization in magnetic fields about twice smaller compared to single-crystal samples, in accordance with theoretical predictions.

[1] J. Han et al., Nat. Mater. 22 (2023) 684; H. Meer et al., Appl. Phys. Lett. 122 (2023) 080502.
[2] W. Yuan et al., Sci. Adv. 4 (2018) eaat1098; P. Makushko et al., Nat. Commun. 13 (2022) 6745.
[3] A. F. Andreev, J. Exp. Theor. Phys. 63 (1996) 15062.

The transition towards a green economy has led to an increased demand for raw materials, which are mainly sourced by mining. Mining activities generate residues such as rock wastes, tailings and stockpiles. These materials are associated with environmental and safety risks that need to be carefully managed throughout their life cycle, with an emphasis on stability and the prevention of water and soil pollution. Earth-observation (EO)-based techniques are seldom used for monitoring these deposits, and multi-sensor field data is commonly not integrated despite recent technological advances. We will develop holistic, full-site services for the geotechnical and environmental monitoring as well as valorisation of mining-related deposits based on a combination of EO and in situ geophysical data. The work will be accomplished under the “Multiscale Observation Services for Mining related deposits” project (MOSMIN for short), and funded by the European Union Agency for the Space Programme (EUSPA) with project number 101131740. MOSMIN services will use Copernicus EO data for time-resolved, spatially extensive, remote monitoring of ground deformation and surface composition. Innovative change detection algorithms will highlight displacements and identify environmental hazards. Satellite data will be integrated with real-time, high-resolution data obtained from unoccupied aerial vehicles and sensors installed at the site, leveraging the power of machine learning for fusion and resolution enhancement of multi-scale, multi-source data. Novel, non-invasive geophysical techniques such as distributed fibre-optic sensing will provide subsurface information to identify potential risks such as internal deformation and seepage. In collaboration with international mining companies, MOSMIN will use pilot sites in the EU, Chile and Zambia to develop and trial comprehensive monitoring services, which are calculated to have a Total Available Market of €1.2bn and expect to be commercialised shortly after project completion by three industry partners. The MOSMIN integrative service and tools will improve the efficiency and reliability of monitoring, maximise resource utilisation and help mitigate environmental risks and the impact of mining operations. - On behalf of the MOSMIN consortium

The transition towards a net zero economy has led to an increased demand for the critical raw materials required for green technologies. Recycling alone is not capable of compensating the requirements for the foreseeable future. At the same time, the extractive sector is facing increasing difficulties in getting stakeholder support to develop new projects. Therefore, our goal is to adopt exploration techniques that minimize environmental impact and prioritize non-invasive methods. For this, we use innovative remote sensing methods to not only improve mineral detection and mapping, but also foster social acceptability for the mining and exploration industry. Hyperspectral imaging (HSI) is a rapidly developing technology that allows for fast and systematic identification of key minerals at the Earth’s surface and provides information about mineral abundances and associations. Several recently launched satellites and the rapid rise of NewSpace (commercial providers) are also opening new opportunities. In this contribution, we illustrate a process including joint drone-borne HSI and satellite-based HSI. We leverage the potential of the different platforms and imaging systems, taking in account their respective advantages and disadvantages. We suggest a vertical integration of drone-borne high spatial and spectral resolution but with limited coverage, and large scale imaging with 30 m ground sampling provided by satellites such as EnMAP, PRISMA and soon to be launched Planet (Tanager) and CHIME. We argue that a combination of machine learning and spectroscopy, accompanied by a structural analysis provides an ideal solution to map potential targets accurately. ​​The processing chain includes radiometric and geometric corrections, co- registration, spectral and structural mapping. We showcase this approach at two study sites in Namibia: The Marinkas Quellen Carbonatite Complex and the Uis pegmatite-hosted tin mine. Both of these deposits host CRMs used in today’s green technology i.e., REEs and lithium respectively.

The shift towards a sustainable economy has led to a heightened demand for raw materials, typically obtained through mining operations. Mining operations produce byproducts like rock waste, tailings, and stockpiles, which pose environmental and safety hazards. It's crucial to effectively manage these materials throughout their lifespan, prioritizing stability and the prevention of water and soil contamination. Earth- observation (EO)-based techniques are rarely used for monitoring these deposits, and multi-sensor field data is commonly not integrated despite recent technological advancements. We aim to establish holistic, full-site services for geotechnical and environmental monitoring, along with the valorization of mining- derived deposits. This will be achieved through an integrated approach utilizing both EO data and in situ geophysical data. The work will be accomplished under the MOSMIN project: “Multiscale Observation Services for Mining related deposits”, and funded by the European Union Agency for the Space Programme (EUSPA) with project number 101131740. MOSMIN services will use Copernicus EO data for time-resolved, spatially extensive, remote monitoring of ground deformation and surface composition. Cutting-edge algorithms for change detection will pinpoint displacements and identify environmental hazards. Satellite data will be integrated with real-time, high-resolution data obtained from drones and sensors installed on site, leveraging the power of machine learning for fusion and resolution enhancement of multi-scale, multi- source data. Novel geophysical techniques such as distributed fibre-optic sensing will provide subsurface information to identify potential risks such as internal deformation and seepage. MOSMIN will collaborate with international partners and mining companies to leverage pilot sites located in the EU, Chile, and Zambia. These sites will serve as testing grounds for the development of comprehensive monitoring services. The MOSMIN integrative services and tools will improve the efficiency and reliability of monitoring, maximise resource utilisation and help mitigate environmental risks and the impact of mining operations. - On behalf of the MOSMIN Consortium.

Electrostatically driven carrier magnetic separation of the fluorescent powder Y₂O₃:Eu and process upscaling with high-gradient magnetic separation.

"Peptide-based Separation System for the Recovery of Palladium from the Chemical-Pharmaceutical Industry"
Presented by Dr. Nora Schönberger at BioKreativ - Forum, Berlin, 8 November 2024.

Palladium is a critical raw material essential for pharmaceutical synthesis, yet its limited availability and high cost present significant challenges for the industry. Current methods for palladium recovery are inefficient, resource-intensive, and environmentally harmful, exacerbating the dependency on politically sensitive supply regions and increasing production expenses. This presentation introduces a novel bio-based separation system leveraging palladium-binding peptides to efficiently recycle this valuable metal.

The innovative approach combines biotechnological peptide development with functionalized membranes, enabling selective palladium recovery from catalytic processes. Through rational design, phage surface display, and AI-assisted optimization, peptides are tailored for high binding affinity and stability. An interdisciplinary roadmap ensures scalability and integration into industrial systems, incorporating life cycle assessments and eco-efficiency analyses to align with sustainable chemistry principles.

This cutting-edge technology not only enhances resource security and lowers production costs but also minimizes toxic waste and environmental impact, advancing the circular economy and promoting sustainable industrial practices.

We experimentally demonstrate that the coacervation of a biopolymer can trigger a hydrodynamic instability when a coacervate is formed upon injection of a Xanthan Gum dispersion into a cationic surfactant (C14TAB) solution. The local increase of the viscosity due to the coacervate formation induces a viscous fingering instability. Three characteristic displacement regimes were observed: a viscous fingering dominated regime, a buoyancy-controlled ”volcano” regime and a ”fan”-like regime determined by the coacervate membrane dynamics. The dependence of the spatial properties of the viscous fingering pattern on the Péclet and Rayleigh numbers is investigated.

Printed circuit boards represent an extraordinarily challenging fraction for the recycling of waste electric and electronic equipment. Due to the closely interlinked structure of the composing materials, the selective recycling of copper and closely associated precious metals from this composite material is compromised by losses during mechanical pre-processing. This problem could partially be overcome by a better understanding of the influence of particle size and shape on the recovery of finely comminuted and well-liberated metal particles during mechanical separation. Here, we propose a workflow to quantify the role of the size and shape of such particles in various separation processes. As a case study, we compare an analytical heavy liquid separation to a new type of eddy current separator. Using X-ray computed tomography, we were able to distinguish metallic and non-metallic phases and determine the size and 3D microstructure of individual particles. For both separation processes, we trained a particle-based separation model that predicts the probability of individual particles to end up in the processing products. In particular, elongated particles were found to display a negative correlation between particle size and sphericity of metallic particles. In line with this correlation, the predicted metal recoveries are positively correlated with particle size but negatively correlation with sphericity in both separation processes. The suggested workflow is easily transferred to other recycling material systems. It allows to quantify the role of 3D geometrical particle properties in separation processes and provide robust predictions for the recoverability of different raw materials in complex recycling streams.

The use of actinium-based radiopharmaceuticals is on the rise, but the coordination chemistry of trivalent actinium remains poorly understood. The most stable isotope of Ac (227Ac) has a short half-life of 21.77 years, making experiments with this element quite ambi-tious. Computational chemistry is the way forward for exploring actinium chemistry. There have been several attempts to apply combined experimental and theoretical approaches for designing suitable chelators for Ac3+-radiopharmaceuticals, including 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetramethylene phosphonic acid (DOTP),[1] and 3,4,3,3-(LI-1,2-HOPO).[2]
Here, DFT calculations were first performed to compare the binding affinity of Ac3+ with different chelators. It was revealed that diethylenetriaminepentaacetic acid (DTPA) and “HOPO-O8” are slightly more effective chelators than the widely-used DOTA. DOTP demonstrated the most optimal performance. It is important to pursue the specific high affinity of the chelator to-wards Ac3+ to deliver the radionuclide to the target cells. However, it is also imperative to have molecular interactions with the receptor for the recognition of the radiopharmaceuticals. In the next step, molecular dynamics (MD) simulations of somatostatin receptor 2 (SSTR2) overex-pressed in neuroendocrine tumors in complex with several different radiopharmaceutical com-pounds have been performed. Bonding and non-bonding parameters involving actinium have been developed using Metal Center Parameter Builder implemented in Amber 20 as well as using DFT calculations with B3LYP functionals using Gaussian16. MD simulations performed using GROMACS program package as depicted in Figure 1.
DOTP, which has been suggested in previous study as an excellent alternative to DOTA,[1] was found to perform better than DOTA not only because of better affinity to Ac3+ but also in terms of the overall higher affinity of Ac3+-DOTP-TATE to the receptor compared to the corre-sponding DOTA complex. Detailed energetic analysis revealed that this is primarily due to elec-trostatic interactions stemming from high negative charge of DOTP. Further analysis of the sec-ondary structure of the receptor revealed that Ac3+-DOTP-TATE perform excellently also in terms of ligand recognition and affect the “toggle switch” for the activation of somatostatin recep-tor. Furthermore, effect of adding “linker” between the chelator and the peptide part of the radio-pharmaceutical have been investigated and it has been revealed that the addition of linker indeed increases the ligand affinity to the receptor.

Derzeit sind etwa 5 Milliarden Leuchtstofflampen in der Europäischen Union in Gebrauch. Diese Lampen funktionieren, nach dem Prinzip, dass ein gemischtes Pulver aus rotem, grünem und blauem Leuchtstoffpulver durch UV-Licht angeregt wird, um sichtbares Licht mit einem gewünschten Spektrum zu erzeugen (Abbildung 1). Diese Leuchtstoffpulver enthalten hohe Mengen an fluoreszierenden Seltenen Erden, kostbaren Metallen mit hohem Versorgungsrisiko, die in den meisten technologischen Anwendungen verwendet werden und eine sehr wichtige Rolle in der Energiewende erfüllen.
Momentan werden Leuchtstoffpulver überall in Europa deponiert (Abbildung 2). Die Aufreinigung dieser Abfallpulver würde es ermöglichen, sie in neuen Produkten zu recyceln. Allerdings ist es mit herkömmlichen Verfahren sehr aufwendig, die einzelnen Pulver voneinander zu trennen. Wir haben eine Lösung gefunden, um dieses Material mithilfe von Biotechnologie effizient und kostengünstig aufzureinigen und zu recyceln. Dabei verwenden wir Biomoleküle, um die Oberfläche von magnetischen Partikeln so zu verändern, dass sie selektiv an eines der Leuchtstoffpulver binden. Durch das Hinzufügen eines Magneten werden die magnetischen Partikel angezogen, und das gebundene Leuchtstoffpulver wird mitgezogen und gereinigt (Abbildung 3).
Unsere ersten Experimente mit diesem neuen Trennverfahren, die wir mit dem roten Leuchtstoffpulver durchgeführt haben, zeigen, dass wir eine Rückgewinnung von über 80 % und eine Reinheit von über 90 % erreichen können. Wir haben ein Patent angemeldet und eine Finanzierung für ein Validierungsprojekt namens Magnetische Aufbereitung zur Gewinnung Seltener Erden aus Leuchtstoffpulvern (MAGSEL) erhalten. Wir hoffen, dass die Biotechnologie mit MAGSEL dazu beitragen kann, eine Kreislaufwirtschaft für die wichtigen Metalle zu schaffen, die wir für die Energiewende benötigen.

Abbildung 1 Funktionsprinzip von Leuchtstofflampen: Wenn die Lampe unter Strom steht, fliegen Elektronen von der Kathode zur Anode. Sie regen gasförmige Quecksilberatome an, die daraufhin UV-Licht emittieren. Dieses UV-Licht regt eine Mischung aus roten (typischerweise Y₂O₃: Eu³⁺), grünen (typischerweise LaPO₄: Ce³⁺, Tb³⁺) und blauen (typischerweise BaMgAl₁₀O₁₇: Eu²⁺) Leuchtstoffpulvern an.

Abbildung 2 Deponie von Leuchtstoffpulvermischungen, die aufgrund eines Mangels an geeigneten Trennverfahren derzeit nicht recycelt werden. Quelle: https://indaver.com/expertise/materials-recovery-from-waste/lamps

Abbildung 3 Darstellung des Verfahrens zur Reinigung von Leuchtstoffpulvern. Biomoleküle werden verwendet, um die Oberfläche von magnetischen Partikeln so zu modifizieren, dass sie selektiv an eines der Leuchtstoffpulver binden. Nachdem jedes Abfallpulver in einem Magnetfeld gereinigt wurde, werden die Leuchtstoffpulver in neuen elektronischen Geräten recycelt.

Presentation of the MAGSEL validation project for Rare Earth Element Recovery from Fluorescent Powder to representatives of LAREC.

Presentation of the MAGSEL validation project for Rare Earth Element Recovery from Fluorescent Powder to representatives of INDAVER NV.

Modern electronic devices play a crucial role for evolving the technological landscape within the European Union and to facilitate the transition into a future energy system based on renewables. However, such devices typically incorporate 20-60 different raw materials, including many that face significant supply risks and that have been categorized either as “critical” or even “strategic”. Moreover, the extraction of these much needed raw materials from geogenic ore deposits is typically energy intensive and results in significant environmental impacts. Although current flows of waste electric and electronic equipment (WEEE) contain greatly elevated concentrations of many of these raw materials – often exceeding concentrations in geogenic ore deposits - only a very small number of them are typically recovered as secondary raw materials. The development of concepts and technologies required for a more comprehensive recycling typically faces practical challenges, mainly due to the complex composition of WEEE and the minute scale of its components.

To overcome the challenges of recovering multiple metals from WEEEs, this study proposes a workflow to evaluate the recyclability of state-of-the-art electronic devices by detailed characterization of components and a particle-based evaluation of the separation efficiency of target components with several separation technologies. A case study is used to illustrate the intended workflow. The investigated flow sheet comprises comminution of WEEE to obtain particle sizes in the scale of individual electronic components, followed by subsequent physical separation processes, including size, density, magnetic and eddy current separation. The particles present in the various streams of each processing step are characterized by X-ray computed tomography (CT) to obtain their 3D geometrical properties and composition in metallic and polymeric phases. These particle datasets are then used for particle-based separation modelling, to quantify the influence of particle size, shape, liberation, and association in their recovery. In future work, this approach will be used to evaluate recyclability already during the design of electronic devices, also considering exergy and life-cycle assessment perspectives.

Electronic waste contains high amounts of valuable metals in the form of ultrafine (<10 μm) inorganic powders. Currently, only a minor fraction of these metals is economically recycled, whereas the vast majority ends up in landfill. Separation of the inorganic powders would significantly enhance the recyclability of these secondary resources. However, the most prominent particle separation (froth flotation, gravity, magnetic and electric separation) processes were developed by the mining industry for primary resources. These processes are only partially suitable for electronic waste recycling because they face challenges related to the ultrafine particle sizes and the complex waste composition (typically >60 elements in electronic waste).
In a novel approach, we propose the use of magnetic nanoparticles (MNPs) as carriers for the magnetic separation of critical raw materials from electronic waste. MNPs can be synthesized costeffectively with a broad variety of surface functionalization possibilities and exhibit unique superparamagnetic properties. We present a case study for the recycling of rare-earth elements from ultrafine fluorescent lamp powders by separation based on the selective attachment of MNPs.
First, we obtained a Massart ferrofluid with monodisperse maghemite nanoparticles, electrostatically stabilized with a negatively charged citric acid coating. These MNPs form an electrostatically driven selective monolayer on the surface of the red fluorescent powder Y2O3:Eu (YOX). Subsequently, a gradient magnetic field is used to selectively purify YOX from other fluorescent powders. After magnetic separation, the pH is increased beyond the isoelectric point of YOX, the MNPs detach from the surface, the two types of particles are then separated based on their size difference and the MNPs are successfully reused in new rounds of magnetic carrier separation. The presented study represents a significant advancement in the utilization of MNPs for the recycling of ultrafine inorganic powders from electronic waste and has been submitted for a European patent application. In coming work, we will collaborate with a lamp recycling company to scale up this process by means of high-gradient magnetic separation.
[1] Acknowledgements: The MAGSEL project is co-financed by tax revenue on the basis of the budget adopted by the Saxon state parliament and the European Union.

Electronic waste contains high amounts of valuable metals in the form of ultrafine (<10 µm) inorganic powders [1]. Currently,
only a minor fraction of these metals is recycled economically. Separation of the inorganic powders would strongly enhance
the recyclability of these secondary resources. However, the most prominent particle separation (froth flotation, gravity,
magnetic and electric separation) processes were developed by the mining industry for primary particles [2,3]. These
processes are only partially suitable for secondary resources and face challenges with regards to the ultrafine particle sizes
and the high complexity (typically, >60 elements are present in electronic waste).
In a novel approach, we propose the use of magnetic carriers derived from various life science applications (such as magnetic
drug delivery, purification, hyperthermia, imaging, etc. [4]) for the magnetic separation of critical raw materials from
electronic waste. Magnetic nanoparticles (MNPs) exhibit excellent properties and can be synthesized cost-effectively. The
small size and high specific surface area of ultrafine powders provide benefits for the attachment of MNPs, as opposed to
their hindrance of conventional separation processes. Achieving attachment selectivity of MNPs to the desired target
powders is crucial for the selectivity of the separation process. This draws inspiration from the common practice of MNP
functionalization with biomolecules in the aforementioned fields of life science[5].
In this presentation, we discuss a case study involving biotechnologically functionalized MNPs for the carrier magnetic
separation of rare-earth element-containing phosphors from fluorescent lamps Figure 1 [6,7]. We provide a comprehensive
overview of MNP synthesis and functionalization, determination of their interaction affinity with various phosphors,
application in magnetic separation, as well as post-separation detachment and MNP reuse. Special emphasis is placed on
MNP colloidal stability and magnetic field gradient.
Our work presents a novel approach to recycling rare-earth elements from fluorescent lamps. More broadly, it represents a
significant advancement in the utilization of biotechnologically functionalized MNPs for the recycling of ultrafine inorganic
powders from electronic waste.
Figure 1 Overview of a case study involving biotechnologically functionalized MNPs for the carrier magnetic separation of rare-earth element-containing
phosphors from fluorescent lamps. [A] The blue (BaMgAl10O17: Eu2+), green (LaPO4: Ce3+, Tb3+ or CeMgAl11O19: Tb3+) and red (Y2O3:Eu3+) phosphors coated as ultrafine particles on the inner surface of a glass tube. [B] Sequential separation of the phosphors after grinding of the lamps by utilizing selective magnetic
carriers. [C] Low carbon-footprint reuse of the critical raw materials in new electronic devices.
1. Rudolph, M. A Aufbereitungs-Technik/Mineral Processing 2018, 59, 65-73.
2. Eckert, K.; Schach, E.; Gerbeth, G.; Rudolph, M. Materials Science Forum 2019, 959, 125-133
3. Luo, L.; Nguyen, A.V. Separation and Purification Technology 2017, 172, 85-99
4. Schwaminger, S.P.; Bauer, D.; Fraga-García, P.; Wagner, F.E.; Berensmeier, S. CrystEngComm 2017, 19, 246-255.
5. Le Jeune, M.; Secret, E.; Trichet, M.; Michel, A.; Ravault, D.; Illien, F.; Siaugue, J.-M.; Sagan, S.; Burlina, F.; Ménager, C. ACS Applied Materials &
Interfaces 2022, 14, 15021-15034
6. Boelens, P.; Lei, Z.; Drobot, B.; Rudolph, M.; Li, Z.; Franzreb, M.; Eckert, K.; Lederer, F. Minerals 2021, 11
7. Boelens, P.; Bobeth, C.; Hinman, N.; Weiss, S.; Zhou, S.; Vogel, M.; Drobot, B.; Azzam, S.S.A.; Pollmann, K.; Lederer, F. Journal of Magnetism and
Magnetic Materials 2022, 563, 169956

Progress in our knowledge of tumor mechanisms and complexity led to the understanding of the physical parameters of cancer cells and their microenvironment, including the mechanical, thermal, and electrical properties, solid stress, and liquid pressure, as critical regulators of tumor progression and potential prognostic traits associated with clinical outcomes. The biological hallmarks of cancer and physical abnormalities of tumors are mutually reinforced, promoting a vicious cycle of tumor progression. A comprehensive analysis of the biological and physical tumor parameters is critical for developing more robust prognostic and diagnostic markers and improving treatment efficiency. Like the biological tumor traits, physical tumor features are characterized by inter and intratumoral heterogeneity. The dynamic changes of physical tumor traits during tumor progression and as a result of tumor treatment highlight the necessity of their spatial and temporal analysis in clinical settings. This review focuses on the biological basis of the tumor specific physical traits, the state of the art methods of their analyses, and the perspective of clinical translation. The importance of tumor physical parameters for disease progression and therapy resistance, as well as current treatment strategies to monitor and target tumor physical traits in clinics, is highlighted.

Bioassays are versatile bioanalytical methods, based on the use of well plates for analytical research and clinical diagnostic testing.
Seamless integration of flexible, multimodal actuator/ sensors into microplates:
Thermal interface
Electrochemical impedance spectroscopy (EIS)
Extended Gate Field Effect Transistor (EGFET)-based biosensing

Bioassays are versatile bioanalytical methods, based on the use of microplates for analytical research and clinical diagnostic testing. Our vision is to
seamlessly integrate multimodal actuators/sensors for label-free, low-cost, and automated real-time monitoring of bioassays. This will be achieved by a thermal, electrical and EGFET interface.

Open set recognition (OSR) constitutes a critical endeavor within the domain of computer vision, frequently deployed in applications, such as autonomous driving and medical imaging recognition. Existing OSR methodologies predominantly center on the acquisition of a profound association between image data and corresponding labels, facilitating the extraction of discriminative features instrumental for distinguishing novel categories. Nevertheless, real-world scenarios often introduce not only novel classes (referred to semantic shift) but also intricate environmental modifications that engender alterations in the distribution of established classes (termed as covariate shift). The latter phenomenon has the potential to undermine the robust correlation between images and labels established by conventional statistical correlation modeling approaches, consequently resulting in significant degradation of OSR performance. Causal correlation stands as the fundamental linkage between entities, routinely harnessed by humans to enhance their cognitive capacities for a more profound comprehension of the intricate world. With inspiration drawn from this perspective, our work herein introduces the causal inference-inspired open set recognition (CISOR) approach tailored for real-world OSR (RWOSR). CISOR represents the pioneering initiative to leverage the stability inherent in causal correlation to construct two pivotal modules: the covariate causal independence (CCI) module and the semantic causal uniqueness (SCU) module, both instrumental in addressing the RWOSR problem. The CCI module adeptly confronts the challenge of covariate shift by imposing constraints on the correlations between inter-class causal features. This strategy effectively mitigates the impact of spurious correlations between distinct categories on the generalization capacity of discriminative features. Furthermore, in order to counteract the issue of semantic shift, the SCU module harnesses correlations between causal features within the same class as constraints, thereby facilitating the extraction of resilient causal features endowed with superior discriminative capabilities. Empirical findings substantiate the superior efficacy of the proposed CIOSR method when compared to state-of-the-art approaches across diverse RWOSR benchmark datasets. The source code of this article will be available at https://github.com/yangzhen1252/RWOSR1.

In recent years, remote sensing recognition technology has found extensive applications in diverse fields, such as modern agriculture, forest management, urban planning, natural resource management, and disaster monitoring. However, the existing remote sensing recognition tasks face significant challenges because of the complex and ever-changing observation environment and the rapid growth of observation classes. The detection performance of existing closed-set recognition methods (where the test set does not contain unknown classes) is greatly limited. Therefore, numerous remote sensing open-set recognition (RSOSR) methods have been proposed to cope with more demanding but practical scenarios in the open world, including scenes or targets with unknown classes. Despite this, there is still a lack of comprehensive work on RSOSR technology. This article presents a comprehensive review of existing RSOSR technologies, covering relevant definitions, model principles, evaluation standards, and method comparisons. We then identify and discuss the limitations of current RSOSR technologies while highlighting promising research directions.

The increasing severity, duration, and frequency of destructive floods can be attributed to shifts in climate, infrastructure, land use, and population demographics. Obtaining precise and timely data about the extent of floodwaters is crucial for effective emergency preparedness and mitigation efforts. Deep convolutional neural networks (CNNs) have shown astonishing effectiveness in various remote sensing applications, including flood mapping. One of the key limitations of CNNs is that they can only predict whether a desired feature will appear in an image, not where it can be recognized. To address this limitation, the incorporation of self-attention mechanisms deployed in vision transformers (ViTs) can be particularly effective. However, the self-attention modules in the ViTs are complex and computationally expensive, and they require a wealth of ground data to attain their full capability in image classification/segmentation. Thus, in this paper, we develop the Residual Wave Vision U-Net (WVResU-Net), a deep learning segmentation architecture that utilizes advanced Vision Multi-Layer Perceptrons (MLPs) and ResU-Net for accurate and reliable flood mapping using Sentinel-1 SAR’s dual polarization data. Results showed the significant superiority of the developed WVResU-Net algorithms over several well-known CNN and ViT deep learning models, including Swin U-Net, U-Net+++, Attention U-Net, R2U-Net, ResU-Net, TransU-Net and TransU-Net++. For example, the segmentation accuracy of TransU-Net++, SwinU-Net, ResU-Net, R2U-Net, Attention U-Net, TransU-Net, and U-Net+++, was significantly improved by approximately 5, 12, 13, 13, 16, 19, and 23 percentage points, respectively in terms of recall obtained by the WVResU-Net with a recall value of about 69.67%. The code will be made publicly available at https://github.com/aj1365/RWVUNet

Convolutional neural networks (CNNs) are used extensively in remote sensing due to their capacity to capture intricate features from a broad range of object patterns, irrespective of object size, shape, or color. These networks excel at extracting high-frequency spectral information such as angles, edges, and outlines. The classification boundary zone, however, becomes hazy for CNNs because they learn characteristics by means of a fixed shape kernel concentrated on the central pixel and can perform poorly in image classification at class boundaries. In addition, CNNs are not designed to capture global relationships. Thus, in this letter, we propose an attention graph convolutional network (Attention-GCN) as a solution to the aforementioned shortcomings. The developed model illustrated a high level of superiority over several CNN and vision transformer (ViT)-based models. For example, in the Augsburg data benchmark, the developed algorithm exhibited an average accuracy of 61.11%, substantially outperforming other models such as HybridSN, iFormer, EfficientFormer, graph convolutional network (GCN), CoAtNet, 2D-CNN, 3D-CNN, and ResNet by approximately 9, 13, 14, 15, 18, 24, 25, and 29 percentage points, respectively. The code will be made publicly available at https://github.com/aj1365/AGCN

Long-term safety assessments for nuclear waste disposal face considerable challenges due to uncertainties resulting from the complex geological, geochemical and environmental processes. This work focuses on enhancing the predictive capability of reactive transport models (RTM) for radionuclide migration in fluids within repositories in crystalline host rock. In particular, the work is focused on investigating the influence of uncertain parameters on radionuclide sorption behavior in crystalline rocks. This is achieved by means of systematic Global Sensitivity analysis (GSA) techniques. The distribution coefficient (Kd) is a key parameter quantifying sorption behavior, obtained by means of geochemistry databases. A Quasi Monte Carlo sampling of input parameters, including mineral composition, pH/Eh, and Uranyl concentrations, was employed to study their effects on Kd values. GSA identifies the important variables affecting the uncertainty in the assessment results. Two GSA methodologies where utilized in this work, namely CUSUNORO and High Dimensional Model Representation (HDMR). By performing CUSUNORO and HDMR together, we capture first-order non-linear and second-order effects, respectively, revealing interaction effects between input parameters on the distribution coefficient. Moreover, the compositional data sampling poses a challenge due to the interdependencies which can alter the results of sensitivity analysis. To address this, we implemented transformation techniques to mitigate the interdependency problem. Our findings contribute to a deeper understanding of these processes, providing valuable insights for enhancing the reliability and robustness of long-term safety assessments for nuclear waste disposal sites.

Predictive models are increasingly used for managerial and operational decision-making. The use of complex machine learning algorithms, the growth in computing power, and the increase in data acquisitions have amplified the black-box effects in data science. Consequently, a growing body of literature is investigating methods for interpretability and explainability. We focus on methods based on Shapley values, which are gaining attention as measures of feature importance for explaining black-box predictions. Our analysis follows a hierarchy of value functions, and proves several theoretical properties that connect the indices at the alternative levels. We bridge the notions of totally monotone games and Shapley values, and introduce new interaction indices based on the Shapley-Owen values. The hierarchy evidences synergies that emerge when combining Shapley effects computed at different levels. We then propose a novel sensitivity analysis setting that combines the benefits of both local and global Shapley explanations, which we refer to as the “glocal” approach. We illustrate our integrated approach and discuss the managerial insights it provides in the context of a data-science problem related to health insurance policy-making.

Code for reproducibility

We examine the construction of variable importance measures for multivariate responses using the theory of optimal transport. We start with the classical optimal transport formulation. We show that the resulting sensitivity indices are well-defined under input dependence, are equal to zero under statistical independence, and are maximal under fully functional dependence. Also, they satisfy a continuity property for information refinements. We show that the new indices encompass Wagner’s variance-based sensitivity measures. Moreover, they provide deeper insights into the effect of an input’s uncertainty, quantifying its impact on the output mean, variance, and higher-order moments. We then consider the entropic formulation of the optimal transport problem and show that the resulting global sensitivity measures satisfy the same properties, with the exception that, under statistical independence, they are minimal, but not necessarily equal to zero. We prove the consistency of a given-data estimation strategy and test the feasibility of algorithmic implementations based on alternative optimal transport solvers. Application to the assemble-to-order simulator reveals a significant difference in the key drivers of uncertainty between the case in which the quantity of interest is profit (univariate) or inventory (multivariate). The new importance measures contribute to meeting the increasing demand for methods that make black-box models more transparent to analysts and decision makers.

Cubic zirconia (c-ZrO2) is considered a highly radiation-tolerant material. It is also capable of incorporating a variety of large cations within its crystal structure, making it a promising material as a waste matrix for actinide immobilization. In this study, various syntheses of cerium(IV)-doped zirconia co-doped with Gd(III)/Y(III) were conducted to identify compositions exhibiting a pure cubic structure, with cerium serving as a plutonium analogue. Four compositions were chosen for the production of dense ceramics. The ceramic production of ZrO2 was conducted with a constant Ce(IV) concentration of 18 mol% and varying Gd/Y concentrations. Purely cubic solid solutions phases were obtained for compositions where the trivalent dopant concentrations exceeded 15 mol% (Fig. 1). The full width at half maximum (FWHM) of the XRD peaks in the dense ceramics increased by a factor of 2 in relation to the starting powder material. Their radiation tolerance was assessed through external ion irradiation experiments. In preparation for these experiments, the ceramic surfaces was polished, and half of the pellet was masked using Al-foil. The non-masked part of the pellet was irradiated with 14 MeV Au4+ ions to simulate the recoil of daughter products from alpha decay. Samples were irradiated at two different fluences, 1014 ions/cm2 (A1) and 1015 ions/cm2 (A2). Subsequent to irradiation, analyses were conducted with scanning electron microscopy (SEM) and synchrotron X-ray diffraction in grazing incidence mode (GI-XRD).
The cubic ceramic phases demonstrated excellent radiation tolerance, displaying no significant radiation damage of the structure and maintaining their cubic crystal structure even after irradiation at the highest fluence, A2 (Fig.2). However, diffraction peak broadening following irradiation is visible, suggesting that irradiation has induced microstructural changes to the samples (Fig. 2, right). A non-systematic shift of the Bragg peaks towards lower angles is observed in the irradiated part, particularly pronounced for fluence A2, indicating an expansion of the lattice. No amorphous contributions could be observed in the diffractograms. These observations demonstrate the high radiation tolerance of the ZrO2 crystal structure, and corroborate their use as waste forms for high-level actinide-bearing waste.

Zirconia (ZrO2), exhibits several advantageous properties, including high thermal
stability, chemical inertness, and the capacity to incorporate substantial quantities of
actinides and lanthanides into its host crystal structure [1]. These characteristics make
zirconia a promising candidate for the immobilization of radionuclides from spent nuclear
fuel [2][3]. In the present study, the chemical durability and radiation resistance of doped
zirconia materials has been investigated. Cerium (Ce) has been used as a plutonium
(Pu) analogue. To stabilize the cubic ZrO2 phase at low tetravalent Ce doping
concentrations, trivalent yttrium (Y) was incorporated as a co-dopant during synthesis.
Both powder samples and dense ceramic pellets were produced for solubility and
irradiation investigations, respectively. For the irradiation investigations, half of the pellet
surface was masked with aluminum foil to protect the pristine side, and the other half
was irradiated with 14 Mev Au4+ ions applying two fluences: 1x1014 ions/cm2 (A1), and
1x1015 ions/cm2 (A2). The pellets were then analyzed using scanning electron
microscopy (SEM), vertical scanning interferometry (VSI), and synchrotron powder x-ray
diffraction (SPXRD) in gracing incidence mode. The results (Fig.1) showed no significant
difference between the pristine and the irradiated side, indicating a high radiation
tolerance of these pellets. Solubility studies of powder samples with identical composition
to the irradiated pellets were conducted in a low-pH environment (0 ≤ pHm ≤ 0.8).
Additional solubility investigations for selected U-doped zirconia samples, under both
oxidizing and reducing conditions were performed in parallel. After 6 months, the yttriumstabilized
samples with cubic structure exhibited slightly lower solubility compared to
those without yttrium (monoclinic or tetragonal structure). These findings speak for an
enhanced chemical stability in addition to the exceptional radiation tolerance of especially
the cubic zirconia modifications.

We present the Materials Learning Algorithms (MALA) package, a scalable machine learning framework designed to accelerate density functional theory (DFT) calculations suitable for large-scale atomistic simulations. Using local descriptors of the atomic environment, MALA models efficiently predict key electronic observables, including local density of states, electronic density, density of states, and total energy. The package integrates data sampling, model training and scalable inference into a unified library, while ensuring compatibility with standard DFT and molecular dynamics codes. We demonstrate MALA's capabilities with examples including boron clusters, aluminum across its solid-liquid phase boundary, and predicting the electronic structure of a stacking fault in a large beryllium slab. Scaling analyses reveal MALA's computational efficiency and identify bottlenecks for future optimization. With its ability to model electronic structures at scales far beyond standard DFT, MALA is well suited for modeling complex material systems, making it a versatile tool for advanced materials research.

Zirconia (ZrO2), exhibits several advantageous properties, including high thermal stability,
chemical inertness, and the capacity to incorporate substantial quantities of actinides and
lanthanides into its host crystal structure [1]. These characteristics make zirconia a promising
candidate for the immobilization of radionuclides from spent nuclear fuel [2][3]. In the present
study, the chemical durability and radiation resistance of doped zirconia materials have been
investigated. Cerium (Ce) has been used as a plutonium (Pu) analog. To stabilize the cubic ZrO2
phase at low tetravalent Ce doping concentrations, trivalent yttrium (Y) was incorporated as a codopant
during synthesis. Both powder samples and dense ceramic pellets were produced for
solubility and irradiation investigations, respectively. For the irradiation investigations, half of
the pellet surface was masked with aluminum foil to protect the pristine side, and the other half
was irradiated with 14 Mev Au4+ ions applying two fluences: 1x1014 ions/cm2 (A1), and 1x1015
ions/cm2 (A2). The pellets were then analyzed using scanning electron microscopy (SEM),
vertical scanning interferometry (VSI), the powder diffraction in Bragg Brentano mode and
grazing incidence diffraction. The results (Fig.1) showed no significant difference between the
pristine and the irradiated side, indicating a high radiation tolerance of these pellets. Solubility
studies of powder samples with identical composition to the irradiated pellets were conducted in
a low-pH environment (0 ≤ pHm ≤ 0.8). Additional solubility investigations for selected U-doped
zirconia samples, under both oxidizing and reducing conditions were performed in parallel. After
6 months, the yttrium-stabilized samples with cubic structure exhibited slightly lower solubility
compared to those without yttrium (monoclinic or tetragonal structure). These findings speak for
enhanced chemical stability in addition to the exceptional radiation tolerance, especially the cubic
zirconia modifications.

This study aimed to enhance the accuracy of forest fire susceptibility mapping (FSM) by innovatively applying recursive feature elimination (RFE) with an ensemble of machine learning models, specifically Support Vector Machine (SVM) and Random Forest (RF), to identify key fire factors. The fire zones were derived from MODIS satellite imagery from 2012 to 2017. Further validation of these data has been provided by field surveys and reviews of land records in rangelands and forests; a total of 326 fire points were determined in this study. Seventeen factors involving topography, geomorphology, meteorology, hydrology, and human factors were identified as being effective primary factors in triggering and spreading fires in the selected mountainous case study area. As a first step, the RFE models RF, Extra Trees, Gradient Boosting, and AdaBoost were used to identify important fire factors among all selected primary factors. The SVM and RF models were applied once on all factors and secondly on those derived from the RFE model as the key factors in FSM. Training and testing data were divided tenfold, and the model’s performance was evaluated using cross-validation. Various metrics, including recall, precision, F1 score, accuracy, area under the curve (AUC), Matthew’s correlation coefficient (MCC), and Kappa, were employed to measure the performance of the models. The assessments demonstrate that leveraging RFE models enhances the FSM results by identifying key factors and excluding unnecessary ones. Notably, the SVM model exhibits significant improvement, achieving an increase of over 10.97% in accuracy and 8.61% in AUC metrics. This improvement underscores the effectiveness of the RFE approach in enhancing the predictive performance of the SVM model.

Convolutional neural networks (CNNs) and vision transformers (ViTs) have shown excellent capability in complex hyperspectral image (HSI) classification. However, these models require a significant number of training data and are computational resources. On the other hand, modern Multi-Layer Perceptrons (MLPs) have demonstrated a great classification capability. These modern MLP-based models require significantly less training data compared with CNNs and ViTs, achieving state-of-the-art classification accuracy. Recently, Kolmogorov–Arnold networks (KANs) were proposed as viable alternatives for MLPs. Because of their internal similarity to splines and their external similarity to MLPs, KANs are able to optimize learned features with remarkable accuracy, in addition to being able to learn new features. Thus, in this study, we assessed the effectiveness of KANs for complex HSI data classification. Moreover, to enhance the HSI classification accuracy obtained by the KANs, we developed and proposed a hybrid architecture utilizing 1D, 2D, and 3D KANs. To demonstrate the effectiveness of the proposed KAN architecture, we conducted extensive experiments on three newly created HSI benchmark datasets: QUH-Pingan, QUH-Tangdaowan, and QUH-Qingyun. The results underscored the competitive or better capability of the developed hybrid KAN-based model across these benchmark datasets over several other CNN- and ViT-based algorithms, including 1D-CNN, 2DCNN, 3D CNN, VGG-16, ResNet-50, EfficientNet, RNN, and ViT.

The Fibroblast Activation Protein (FAP) plays a pivotal role, particularly in cancer, being overexpressed in the microenvironment of solid tumors, rendering it an attractive target. Based on the UniCAR platform technology, UniCAR target modules (TMs) have been engineered to specifically address this antigen. These TMs, comprising either a single-chain variable fragment (ScFv) or immunoglobulin G (IgG) format, coupled with the UniCAR peptide epitope E5B9, act as a bridge between universal CAR-T cells and target cells, enhancing safety, and efficiency [1]. This study explores gold nanoparticles (AuNPs), both spherical and branched, as nanocarriers for anti-FAP TMs. Branched AuNPs with NIR absorbance extend beyond conventional targeting, holding potential as photothermal agents for localized therapy. This multifaceted approach aims for enhanced cell labeling, photothermal effects, and cytokine activation, advancing the therapeutic capabilities of anti-FAP-targeted immunotherapy. Surface biofunctionalization of particles was achieved through site-directed immobilization of biomolecule-peptide epitope conjugates, utilizing the cysteine terminus at the peptide epitope, to facilitate the formation of a protein monolayer, allowing precise and stable functionalization. Incubation of the FAP-expressing cell line (HT1080 hFAP) with anti-FAP TM coated NPs, monitored via surface plasmon resonance Scattering (SPRS) imaging, indicated successful cell labeling without inducing toxicity at an optical density of 0.1 OD (~272 pM). Viability assessments conducted on all treated cells demonstrated no toxicity concerns. Specificity testing conducted on PC3 cells, employed as a negative control, revealed no discernible increase in scattering intensity. Ongoing investigations are dedicated to optimizing parameters, including concentration and incubation time, to maximize therapeutic potential, aiming to optimize FAP-targeted nanoparticles for advanced therapeutic and diagnostic applications.

The interplay of electronic charge, spin, and orbital currents, coherently driven by picosecond long oscillations of light fields in spin-orbit coupled systems, is the foundation of emerging terahertz lightwave spintronics and orbitronics. The essential rules for how terahertz fields interact with these systems in a nonlinear way are still not understood. In this work, we demonstrate a universally applicable electronic nonlinearity originating from spin-orbit interactions in conducting materials, wherein the interplay of light-induced spin and orbital textures manifests. We utilized terahertz harmonic generation spectroscopy to investigate the nonlinear dynamics over picosecond timescales in various transition metal films. We found that the terahertz harmonic generation efficiency scales with the spin Hall conductivity in the studied films, while the phase takes two possible values (shifted by π), depending on the d-shell filling. These findings elucidate the fundamental mechanisms governing nonequilibrium spin and orbital polarization dynamics at terahertz frequencies, which is relevant for potential applications of terahertz spin- and orbital-based devices.

Movement behavior is critical in shaping population dynamics, yet theoretical frameworks linking individual movement to demographic outcomes remain limited. Here, we introduce the range-resident logistic model, an extension of the spatial logistic model that incorporates realistic range-resident movement. Through individual-based simulations and analytical approximations, we show that population carrying capacity strongly depends on home range size, with traditional models often both overestimating and underestimating abundances. To bridge movement and demography, we introduce a simple crowding index, calculated solely based on spatial scales measured in the simulations, which accurately predicts the population’s carrying capacity across a wide parameter range. By incorporating movement as a parametric stochastic process, our model bridges existing frameworks for sessile and freely moving organisms, correctly recovering their results as limiting cases. The range-resident logistic model provides a unified perspective on how movement behavior shapes population growth and spatial distribution, offering insights into upscaling individual movement to population-level consequences.

In this lecture, I will present an overview of existing models to describe animal movement and revisit techniques to upscale the population-level consequences of different movement patterns.

In this lecture, I will present an overview of existing mathematical models to describe patterns of animal movement.

Animals use space non-uniformly and occupy home ranges that are significantly smaller than the population range. Modern tracking technologies and the development of new statistical tools have made it possible to quantify these features of individual movement behavior with unprecedented precision. Yet, how range-resident movement impacts animal encounters and interactions still needs to be better understood, preventing us from properly scaling up the consequences of individual movement behavior to the population level and investigating how movement behavior drives larger-scale ecological processes. In this presentation, I will first introduce a framework that incorporates range-resident movement in different encounter metrics, including pairwise encounter rates, spatial distributions of encounters, and encounter probability. I will discuss how this refined approach deviates from previous theories lacking range-resident movement and finally lay the ground for developing new statistical estimators that will enable the application of this new theory to tracking data.

Below-ground plant interactions are a key driver of vegetation pattern formation. However, many existing models rely on functional forms for these interactions that frequently lack empirical support. This gap between models and data stems primarily from our limited understanding of below-ground plant growth processes. Unlike aboveground shoot competition, the study of below-ground plant growth is hampered by our inability to observe roots. We have few observations of intact root systems in soil and lack a comprehensive theory for root system responses to their environment. In this presentation, I will first review previous theoretical efforts to explain plant below-ground competition and discuss how they lead to seemingly contradictory predictions. Then, I will introduce our recent theoretical and experimental work and show how it resolves existing controversy and provides a unifying framework to study below-ground plant interactions, both competitive and facilitative. I will conclude by discussing future research lines that depart from our results, including extensions to larger spatial scales
and applying this new modeling approach to vegetation patterns.

Collective behaviors, in which many individuals exhibit some degree of behavioral coordination, are frequent in nature and observed across a continuum of scales, from microbial aggregates to ungulate migrations. Intriguingly, however, such coordination is sometimes imperfect, and “out-of-sync” individuals exist in many of these systems. The roots of such imperfect coordination, and hence the mechanisms underlying the emergence of out-of-sync individuals, will undoubtedly differ across systems. Nevertheless, the occurrence of imperfect coordination across such different systems and scales raises fundamental questions about its causes and consequences. Are “out-of-sync” individuals merely inevitable byproducts of large-scale coordination attempts, or can they, at least in some systems, be a variable trait that selection can shape with potential ecological consequences?

I will address this question by combining empirical data on slime-mold imperfect aggregation and observed patterns of partial migration observed within three ungulate specie. In each of these systems, we find that the number of individuals that do not engage in the collective behavior is unrelated to the total population size, suggesting that a complex individual decision-making process underlies the onset of the collective behavior. Using a minimalistic modeling framework, we propose that imperfectly synchronized collective behaviors are, in fact, a dynamic population partition process that originates from each individual making a stochastic signal-based decision. The parallelisms between these two seemingly different systems suggest that imperfectly synchronized collective behaviors could be critical to understanding social behaviors and ecological dynamics across scales.

Theoretical models in ecology often assume well-mixed populations and thus that individuals interact with one another with the same probability regardless of their spatial location. This strong assumption results in mathematically very tractable models, often based on ordinary differential equations for the population size, that have taught us a lot about how populations change over time. However, these models do not account for the fact that populations are spatially structured, which favors interactions between nearby individuals. In this presentation, I will present some of our recent work investigating the effect of space in shaping population dynamics. First, I will very briefly discuss how stochastic spatial models, from lattice models to systems of interacting particles, can provide more accurate descriptions of ecological populations and how to analyze them computationally and analytically. Then, I will use a specific example based on an asymmetric voter model to show how modeling spatial processes can reverse the outcome of species competition predicted by a well-mixed model.

A large body of existing ecological theory, from species interaction to disease transmission, relies on very strong and unrealistic assumptions about the way individuals move and get to interact with each other and with the environment. Specifically, several models assume that individuals behave like the molecules of an ideal gas: following completely random trajectories through the entire area occupied by the population and only interacting with each other when their trajectories intersect. In this presentation, I will first discuss how traditional population dynamics models emerge from ideal gas assumptions. Then, I will present our ongoing research to refine those models so they incorporate movement features observed in GPS tracking data. I will discuss examples covering both the development of new theory (based on random walk models, spatially-extended nonlinear dynamical systems, and stochastic calculus) and its application to ecological data.

Populations very often self-organize into regular spatial patterns with important ecological and evolutionary consequences. Yet, most existing models neglect the effect that external biophysical drivers might have both on pattern formation and the spatiotemporal population dynamics once patterns form. Here, we investigate the effect of environmental flows on pattern formation and population dynamics using a spatially nonlocal logistic model (or Fisher-Kolmogorov equation) coupled to a simple shear and a Rankine vortex flow. We find that, whereas population abundance generally decreases with increasing flow intensity, the effect of the flow on the pattern instability depends on the spatial structure of the flow velocity field. This result shows that the velocity field interacts with the spatial feedbacks responsible for pattern formation in non-trivial ways, leading to a variety of spatiotemporal population dynamics regimes in which the total population abundance can exhibit either regular oscillations with a characteristic frequency or more erratic dynamics without a well-defined period. More generally, the diversity of spatiotemporal population dynamics caused by the interplay between self-organizing feedbacks and environmental flows highlights the importance of incorporating environmental and biophysical processes when studying both ecological pattern formation and its consequences.

Many natural phenomena are quantified by counts of observable events, from the annihilation of quasiparticles in a lattice to predator-prey encounters on a landscape to spikes in a neural network. These events are triggered at random intervals when an underlying dynamical system occupies a set of reactive states in its phase space. We derive a general expression for the distribution of times between events in such counting processes assuming the underlying triggering dynamics is a stochastic process that converges to a stationary distribution. Our results contribute to resolving a long-standing dichotomy in the study of reaction-diffusion processes, showing the inter-reaction point process interpolates between a reaction- and a diffusion-limited regime. At low reaction rates, the inter-reaction process is Poisson with a rate depending on stationary properties of the event-triggering stochastic process. At high reaction rates, inter-reaction times are dominated by the hitting times to the reactive states. To further illustrate the power of this approach we apply our framework to obtain the counting statistics of two counting processes appearing in several biophysical scenarios. First, we study the common situation of estimating an animal's activity level by how often it crosses a detector, showing that the mean number of crossing events can decrease monotonically with the hitting rate, a seemingly 'paradoxical' result that could possibly lead to misinterpretation of experimental count data. Second, we derive the ensemble statistics for the detection of many particles, recovering and generalizing known results in the biophysics of chemosensation. Overall, we develop a unifying theoretical framework to quantify inter-event time distributions in reaction-diffusion systems that clarifies existing debates in the literature and provide examples of application to real-world scenarios.

Spot patterns, in which vegetation patches form a hexagonal lattice, are frequent in nature and could serve as an early-warning indicator of abrupt vegetation collapses. Consequently, they have been intensively studied using both individual-based models and density-based field equations. Yet, the relationship between these two approaches remains unclear, particularly in scenarios where vegetation dynamics exhibit strong long-range spatial correlations and traditional mean-field approximations fail. To solve this issue, we develop a new method that refines mean-field approximations by describing both the dynamics of the biomass density field and its spatial correlations. This new approach harnesses the strengths of both individual and density-based mdoels, treating spatial correlations explicitly and allowing for the identification of spatial instabilities resulting in periodic patterns. Our results indicate that this new approximation predicts the parameter regimes where regular periodic patterns emerge more accurately than mean-field models, suggesting that it could provide a more robust framework to perform further nonlinear analysis.

Ecologists have long investigated how demographic and movement parameters determine the spatial distribution and critical habitat size of a population. However, most models oversimplify movement behaviour, neglecting how landscape heterogeneity influences individual movement. We relax this assumption and introduce a reaction–advection–diffusion equation that describes population dynamics when individuals exhibit space-dependent movement bias toward preferred regions. Our model incorporates two types of these preferred regions: a high-quality habitat patch, termed ‘habitat’, which is included to model avoidance of degraded habitats like deforested regions; and a preferred location, such as a chemoattractant source or a watering hole, that we allow to be asymmetrically located with respect to habitat edges. In this scenario, the critical habitat size depends on both the relative position of the preferred location and the movement bias intensities. When preferred locations are near habitat edges, the critical habitat size can decrease when diffusion increases, a phenomenon called the drift paradox. Also, ecological traps arise when the habitat overcrowds due to excessive attractiveness or the preferred location is near a low-quality region. Our results highlight the importance of species-specific movement behaviour and habitat preference as drivers of population dynamics in fragmented landscapes and, therefore, in the design of protected areas.

Direct encounters, in which two or more individuals are physically close to one another, are a topic of increasing interest as more and better movement data become available. Recent progress, including the development of statistical tools for estimating robust measures of changes in animals’ space use over time, facilitates opportunities to link direct encounters between individuals with the long-term consequences of those encounters. Working with movement data for coyotes (Canis latrans) and grizzly bears (Ursus arctos horribilis), we investigate whether close intraspecific encounters were associated with spatial shifts in the animals’ range distributions, as might be expected if one or both of the individuals involved in an encounter were seeking to reduce or avoid conflict over space. We analyze the movement data of a pair of coyotes in detail, identifying how a change in home range overlap resulting from altered movement behavior was apparently a consequence of a close intraspecific encounter. With grizzly bear movement data, we approach the problem as population-level hypothesis tests of the spatial consequences of encounters. We find support for the hypotheses that (1) close intraspecific encounters between bears are, under certain circumstances, associated with subsequent changes in overlap between range distributions and (2) encounters defined at finer spatial scales are followed by greater changes in space use. Our results suggest that animals can undertake long-term, large-scale spatial changes in response to close intraspecific encounters that have the potential for conflict. Overall, we find that analyses of movement data in a pairwise context can (1) identify distances at which individuals’ proximity to one another may alter behavior and (2) facilitate testing of population-level hypotheses concerning the potential for direct encounters to alter individuals’ space use.

Germanium (Ge), the next-in-line group-IV material, bears great potential to add functionality and performance to next-generation nanoelectronics and solid-state quantum transport based on silicon (Si) technology. However, most research has been directed towards strain-relaxed and defective (Si)Ge buffers to overcome the ~4.2 % lattice mismatch between Si and Ge. Here, we investigate the direct implementation of two-dimensional high-quality crystalline Ge layers on Si. Ultra-low growth temperatures (TGe = 100°C - 350°C) and pristine growth pressures (≲1e10 mbar) are necessary to obtain a substantial Ge layer supersaturation and crystalline growth down to the lowest TGes. Under the employed growth conditions and strain-free growth of Ge on Ge(001), positron annihilation lifetime spectroscopy demonstrates that TGe does not influence the concentration of point defects. Therefore, a systematic investigation of the Ge growth on Si(001) was conducted, varying the Ge coverage (1, 2, 4, 8, 12, and 16 nm) and TGe (100°C to 300°C, in increments of 50°C) to assess the influence of these parameters on the layer’s structural quality. Atomic force microscopy revealed a rippled surface topography with superimposed grainy features and the absence of quantum dots. Transmission electron microscopy unveiled pseudomorphic, highly crystalline growth within the grains with defective domains separating them. X-ray diffraction confirmed the presence of both pseudomorphic areas and regions containing defects. Spatially resolved strain fluctuations were confirmed by nanobeam x-ray diffraction measurements. Therefore, strain contributes to the formation of the ripples, which originate from the kinetic limitations of the ultra-low temperatures. The excellent crystalline quality of Ge layers grown at TGes as low as 100°C can significantly impact applications based on optoelectronics and nanoelectronics.

High-content image-based screens are a well-established technique relying on the perturbation of biological or biochemical systems and the subsequent phenotypic readout with the aim of identifying e.g. genes or chemical compounds modulating a process of interest. Noteworthy examples are RNAi screens, which leverage the gene silencing mechanism of RNA interference to interrogate the role of individual genes in specific processes, or small molecule screens employed by pharmaceutical companies as a pivotal step of the drug discovery process. Regardless of the precise type of biological perturbation method employed, off-target effects are an inevitable nuisance requiring special processing to filter them out. Furthermore, high-content screens harbour invaluable information on biological interactions potentially benefiting other research endeavours in systems biology. In a bid to both reliably identify off-target effects and unearth buried phenotypes, we conducted an image-based human genome-wide RNAi screen involving infection with vaccinia virus and subject the results to computational methodologies. In detail, we perform an enrichment via databases and apply XGBoost to the obtained tabular data. In order to harness recent advances in the realm of natural language processing and its applications to biological sequences, we also incorporate the sequence information of proteins identified as hits. The resulting model thus represents an instance of multimodal AI. The proposed method is also applicable to other screening techniques, such as CRISPR-based screening.

ML and DL are revolutionising our abilities to analyse biomedical images. Among other host-pathogen interactions may be readily deciphered from microscopy data using convolutional neural networks (CNN). We demonstrate in several studies how the definition of novel ML/DL tasks may aid in studying infection and disease phenotypes. Specifically, ML/DL algorithms may allow unambiguous scoring of virus-infected and uninfected cells in the absence of specific labelling. Accompanied by interpretability approaches, the ability of CNN to learn representations, without explicit feature engineering, may allow for uncovering yet unknown phenotypes in microscopy. Furthermore, we demonstrate novel ML/DL approaches to simplified 3D microscopy acquisition using conventional 2D hardware. Finally, we exemplify how generative AI can be applied to tasks like image denoising, reconstruction and resolution enhancement in fluorescence and brightfield microscopy. Taken together, we show novel approaches to established algorithms in Computer Vision and Data Science. Applied to microscopy data, these approaches allow for the extraction of observations from datasets large enough to not be suitable for manual analysis. We argue that this shows that reformulating conventional ML/DL tasks to answer biological questions may facilitate novel discoveries in Infection and Disease Biology.

Axonal injury is a key determinant of long-term outcomes after traumatic brain injury (TBI) but has been difficult to measure clinically. Fluid biomarker assays can now sensitively quantify neuronal proteins in blood. Axonal components such as neurofilament light (NfL) potentially provide a diagnostic measure of injury. In the multicenter BIO-AX-TBI study of moderate-severe TBI, we investigated relationships between fluid biomarkers, advanced neuroimaging, and clinical outcomes. Cerebral microdialysis was used to assess biomarker concentrations in brain extracellular fluid aligned with plasma measurement. An experimental injury model was used to validate biomarkers against histopathology. Plasma NfL increased after TBI, peaking at 10 days to 6 weeks but remaining abnormal at 1 year. Concentrations were around 10 times higher early after TBI than in controls (patients with extracranial injuries). NfL concentrations correlated with diffusion MRI measures of axonal injury and predicted white matter neurodegeneration. Plasma TAU predicted early gray matter atrophy. NfL was the strongest predictor of functional outcomes at 1 year. Cerebral microdialysis showed that NfL concentrations in plasma and brain extracellular fluid were highly correlated. An experimental injury model confirmed a dose-response relationship of histopathologically defined axonal injury to plasma NfL. In conclusion, plasma NfL provides a sensitive and clinically meaningful measure of axonal injury produced by TBI. This reflects the extent of underlying damage, validated using advanced MRI, cerebral microdialysis, and an experimental model. The results support the incorporation of NfL sampling subacutely after injury into clinical practice to assist with the diagnosis of axonal injury and to improve prognostication.

Three-dimensional (3D) in vitro cancer models gains increasingly popularity as pre-clinical platforms for evaluating the efficacy of existing anti-cancer drugs and for discovering innovative therapeutic approaches.[1, 2] These models aim to recreate the multicellular compact structures and spatial architecture observed in human solid tumors.[3] The efficiency of immunotherapy stays limited for solid tumors. Tumor heterogeneity and the complex physical and biochemical conditions of the microenvironment hinder immune cells from effectively infiltrating malignant tissues, which can significantly impact the overall therapeutic performance.[4] In our research, we have successfully established a 3D prostate cancer model by co-culturing PC3 and HT1080 cells within 3D micro hydrogel beads which were generated using a custom-designed, high-throughput droplet-based microfluidic platform, coupled with a UV gelation system. We conducted a comparative analysis of PC3 and HT1080 spheroid growth and structures in both monoculture and co-culture conditions. Importantly, our study validated the synergistic efficacy of a dual-targeting molecular approach, utilizing UniCAR-T cell therapy, which simultaneously targeted the tumor microenvironment and the cancer cells. This dual-targeting strategy was found to be more effective when compared to mono-targeting approaches. Our developed 3D prostate cancer model holds significant potential for advancing cancer research, particularly in understanding the critical role of the tumor microenvironment in tumor development, prognosis, and therapy. It provides a more comprehensive platform for testing novel therapeutic interventions and evaluating their impact on the complex interactions within the tumor microenvironment.

Detecting virus-infected cells in light microscopy requires a reporter signal commonly achieved by immunohistochemistry or genetic engineering. While classification-based machine learning approaches to the detection of virus-infected cells have been proposed, their results lack the nuance of a continuous signal. Such a signal can be achieved by virtual staining. Yet, while this technique has been rapidly growing in importance, the virtual staining of virus-infected cells remains largely uncharted. In this work, we propose a benchmark and datasets to address this. We collate microscopy datasets, containing a panel of viruses of diverse biology and reporters obtained with a variety of magnifications and imaging modalities. Next, we explore the virus infection reporter virtual staining (VIRVS) task employing U-Net and pix2pix architectures as prototypical regressive and generative models. Together our work provides a comprehensive benchmark for VIRVS, as well as defines a new challenge at the interface of Data Science and Virology.

Computational image-quality enhancement for microscopy (deblurring, denoising, and optical deconvolution) provides researchers with detailed information on samples. Recent general-purpose deep learning solutions advanced in this task. Yet, without consideration of the underlying physics, they may yield unrealistic and non-existent details and distortions during image restoration, requiring domain expertise to discern true features from artifacts. Furthermore, the large expressive capacity of general-purpose deep learning models requires more resources to train and use in applications. We introduce DeBCR, a physics-informed deep learning model based on wavelet theory to enhance microscopy images. DeBCR is a light model with a fast runtime and without hallucinations. We evaluated the image restoration performance of DeBCR and 12 current state-of-the-art models over 6 datasets spanning crucial modalities in advanced light microscopy and cryo-electron tomography. Leveraging optic models, DeBCR demonstrates superior performance in denoising, optical deconvolution, and deblurring tasks across both LM and cryo-ET modalities.

Geometric effects in curvilinear nanomagnets can enable chiral, anisotropic and even magnetoelectric responses. Here, we study the effects of magnetic frustration in curvilinear (quasi-)1D magnets represented by spin chains arranged along closed space curves of constant torsion. Considering the cases of easy- and hard-axis anisotropy in ferro- and antiferromagnetic samples, we determine their ground states and analyze the related magnetoelectric multipoles. A constant torsion along the chain results in alternating regions of high and low curvature, facilitating the spin spiral state perturbed by the (anti)periodic boundary conditions on the magnetic order parameter. While easy-axis ferromagnetic chains develop a purely toroidal configuration with the magnetic toroidal moment oriented along the geometry symmetry axis, hard-axis antiferromagnetic chains support multiple magnetic toroidal domains. Our findings suggest that tailoring curvature and torsion of a spin chain enables a new physical mechanism for magnetic frustration, which can be observed in the inhomogeneity of the magnetic order parameter and in the local ferrotoroidal responses.

Phase imaging is gaining importance due to its applications in fields like biomedical imaging and material characterization. In biomedical applications, it can provide quantitative information missing in label-free microscopy modalities. One of the most prominent methods in phase quantification is the Transport-of-Intensity Equation (TIE). TIE often requires multiple acquisitions at different defocus distances, which is not always feasible in a clinical setting. To address this issue, we propose to use chromatic aberrations to induce the required through-focus images with a single exposure, effectively generating a through-focus stack. Since the defocus distance induced by the aberrations is small, conventional TIE solvers are insufficient to address the resulting artifacts. We propose Zero-Mean Diffusion, a modified version of diffusion models designed for quantitative image prediction, and train it with synthetic data to ensure robust phase retrieval. Our contributions offer an alternative TIE approach that leverages chromatic aberrations, achieving accurate single-exposure phase measurement with white light and thus improving the efficiency of phase imaging. Moreover, we present a new class of diffusion models that are well-suited for quantitative data and have a sound theoretical basis. To validate our approach, we employ a widespread brightfield microscope equipped with a commercially available color camera. We apply our model to clinical microscopy of patients' urine, obtaining accurate phase measurements.

Antiferromagnets (AFMs) represent a wide class of magnetically ordered materials. While they offer a lot of technologically promising properties and rich physics like efficient spin torques or ultrafast dynamics, a high degree of their magnetic compensation challenges the readout of their state [1]. Here, using the magnetoelectric two-sublattice AFM Cr2O3 with the easy axis of anisotropy as the case study, we show that the difference between the bulk and surface magnetic point symmetry groups (MPSG) can result in a sizeable Dzyaloshinskii-Moriya interaction (DMI) driven by the surface symmetry [2].
The bulk Cr2O3 crystals characterized by MPSG have two nominally compensated, high-symmetry crystallographic cuts: m plane and a plane . Their MPSGs correspond to and , respectively, are determined as subsets of the bulk MPSG by the requirement to leave the polar vector of the surface normal unchanged [3].
MPSG on the surface is lower than in the bulk. This is a factor of modification of the surface energy density w depending on magnetization and the Neel vector (the AFM order parameter). For the m-plane crystallographic cut, MPSG allows w to have the homogeneous DMI terms and even a weakly ferromagnetic response. The microscopic origin of this DMI is the combination of the single-, inter-ion anisotropies and antisymmetric exchange driven by the break of the spatial inversion at the interface. By building the phenomenological energy functional and by means of ab initio calculations, we show that m-plane Cr2O3 is the canted ferrimagnet with finite magnetization components out-of-plane and along the c axis. Furthermore, the magnetization direction is switched with the change of a bulk antiferromagnetic domain. At 0 K, the equilibrium out-of-plane magnetization is 0.1 Bohr magnetons per unit cell due to the spin canting by 0.5° from the interface. This corresponds to the surface-symmetry-driven DMI of the homogeneous type coefficient being about 1 mJ/m2. We performed zero-offset Hall measurements on the m-plane single crystal Cr2O3 to detect the out-of-plane interfacial magnetization. Using the field-cooling procedure to set the preferential AFM domain, we found a sizeable transversal resistance ρ which changes the sign with the AFM domain at room temperature. The temperature dependence of ρ indicates the different temperature dependencies of the anisotropy and antisymmetric exchange contributing to the surface-symmetry-driven DMI. The same procedure for a-plane Cr2O3 indicates the presence of interfacial magnetization related to the weak ferromagnetism of this crystallographic cut in Cr2O3 [2].
Our findings contribute to the understanding of the available magnetic responses of nominally compensated AFM surfaces and interfacial phenomena like spin pumping.

The exploration of the coordination chemistry of actinides significantly lags behind that of transition metals as well as their lanthanide homologues. As such, a fundamental understanding of the binding properties in actinide compounds is still leaving many open questions. Therefore, systematic investigation of various coordination motives around an actinide center can be used as benchmark to evaluate what analytic techniques can reveal about novel actinide-ligand bonding. In this study, we focus on a square antiprism coordination of only oxygen donor atoms in an actinide series ranging from thorium to plutonium. Installing either one or two transition metals in close proximity to the actinide, leads to an 8+2 coordination at the actinide center. These heterobi- and trimetallic complexes have been investigated using single-crystal X-ray diffraction, NMR, HERFD-XANES, and SQUID magnetometry. The experimental findings were further analyzed with quantum chemical calculations. A comparison with their monometallic counterparts gives new insight into actinide-transition metal bonding.

Overview of new research questions emerging from the RadoNorm project on the characterization of exposure to NORM

One of the aims of the Work Package 2 “Exposure” of the European project RadoNorm (2020-2025) is to develop methodologies and tools applicable at European level to identify and quantify exposure of population and environment due to Naturally Occurring Radioactive Material (NORM). In particular, focus is on adapting and optimizing current approaches for identification and evaluation of exposure in light of international and national requirements for handling NORM more compelling than in the past and based on new available scientific evidence. The project is in an advanced stage of progress and several results have been obtained. However, these also let emerge research questions, which may be worth addressing in the future. For example, a methodology to establish a NORM inventory has been developed and applied to gather systematic information on NORM from European countries. Information gained, abundant for naturally occurring radionuclides (NOR), indicated that additional and more systematic information on amounts and handling approaches of other contaminants is needed. This as a basis for the establishment of a more efficient optimized and integrated approach for evaluating NORM involving situations.Impact of most recent ICRP dose coefficients for intake of radionuclides by workers and for external radiation has been analyzed for several generic NORM scenarios. Moreover, with the aim to help stakeholders for a practicable implementation of the radiation protection requirements, screening values (defined in terms of activity concentration corresponding to annual effective dose of 1 mSv/year) have been derived for NORM residues disposable in conventional landfills, and for the reuse of NORM sludge as fertilizer in agriculture. For obtaining these screening values, however, generic consideration of groundwater pathway proved difficult and a systematic analysis of types of landfills and typical hydrological characteristics of these landfills at EU level (i.e. a kind of mapping) would be necessary to use water flow and solute transport models with optimized and less conservative parameters.
In addition, for the groundwater pathway, in order to better define the sorption of NOR to soil, e.g. via the Kd parameter, sorption/desorption properties of uranium, radium and polonium have been investigated experimentally or via geochemical models (e.g. smart-Kd approach). To improve the predictive capability of these models, gaps for thermodynamic NOR data have to be filled, especially for radium and polonium. In addition, NOR interactions with organic matter and quantification of the microbial influence on NOR migration in soil are needed to better predict the radionuclide mobility over space and time, which is needed for realistic dose calculations and evaluation of remediation activities.

Urinary tract infection (UTI) is a common disorder. Its diagnosis can be made by microscopic examination of voided urine for markers of infection. This manual technique is technically difficult, time-consuming and prone to inter-observer errors. The application of computer vision to this domain has been slow due to the lack of a clinical image dataset from UTI patients. We present an open dataset containing 300 images and 3,562 manually annotated urinary cells labelled into seven classes of clinically significant cell types. It is an enriched dataset acquired from the unstained and untreated urine of patients with symptomatic UTI using a simple imaging system. We demonstrate that this dataset can be used to train a Patch U-Net, a novel deep learning architecture with a random patch generator to recognise urinary cells. Our hope is, with this dataset, UTI diagnosis will be made possible in nearly all clinical settings by using a simple imaging system which leverages advanced machine learning techniques.

The Helmholtz-Zentrum Dresden-Rossendorf (HZDR) is advancing towards an integrated data management strategy that spans the entire data lifecycle, embracing the needs of diverse stakeholders such as Helmholtz, the National Research Data Infrastructure (NFDI), and the European Open Science Cloud (EOSC). This talk will address how the HELPMI initiative, a metadata framework designed for laser and plasma research (HELPMI), aligns with and supports HZDR’s comprehensive data management objectives. We will explore HELPMI’s contributions to data standardization, sharing, and FAIR principles (Findable, Accessible, Interoperable, Reusable) within HZDR, discussing its synergy with the broader data management lifecycle and strategic goals. Attendees will gain insights into HELPMI’s role in enhancing data transparency and collaboration, laying a foundation for sustainable research data practices that resonate with national and European frameworks. Finally, we will highlight the roadmap for deeper integration, challenges, and potential areas of development that will help HZDR achieve a unified, robust and sustainable data strategy.

Autophagy is an essential degradation program required for cell homeostasis. Among its functions is the engulfment and destruction of cytosolic pathogens, termed xenophagy. Not surprisingly, many pathogens use various strategies to circumvent or co-opt autophagic degradation. For poxviruses, it is known that infection activates autophagy, which however is not required for successful replication. Even though these complex viruses replicate exclusively in the cytoplasm, autophagy-mediated control of poxvirus infection has not been extensively explored. Using the prototypic poxvirus, vaccinia virus (VACV), we show that overexpression of the xenophagy receptors p62, NDP52, and Tax1Bp1 restricts poxvirus infection. While NDP52 and Tax1Bp1 were degraded, p62 initially targeted cytoplasmic virions before being shunted to the nucleus. Nuclear translocation of p62 was dependent upon p62 NLS2 and correlated with VACV kinase mediated phosphorylation of p62 T269/S272. This suggests that VACV targets p62 during the early stages of infection to avoid destruction and further implies that poxviruses exhibit multi-layered control of autophagy to facilitate cytoplasmic replication.

We previously identified the bisbenzimide Hoechst 33342 (H42) as a potent multi-stage inhibitor of the prototypic poxvirus, the vaccinia virus (VACV), and several parapoxviruses. A recent report showed that novel bisbenzimide compounds similar in structure to H42 could prevent human cytomegalovirus replication. Here, we assessed whether these compounds could also serve as poxvirus inhibitors. Using virological assays, we show that these bisbenzimide compounds inhibit VACV spread, plaque formation, and the production of infectious progeny VACV with relatively low cell toxicity. Further analysis of the VACV lifecycle indicated that the effective bisbenzimide compounds had little impact on VACV early gene expression but inhibited VACV late gene expression and truncated the formation of VACV replication sites. Additionally, we found that bisbenzimide compounds, including H42, can inhibit both monkeypox and a VACV mutant resistant to the widely used anti-poxvirus drug TPOXX (Tecovirimat). Therefore, the tested bisbenzimide compounds were inhibitors of both prototypic and pandemic potential poxviruses and could be developed for use in situations where anti-poxvirus drug resistance may occur. Additionally, these data suggest that bisbenzimide compounds may serve as broad-activity antiviral compounds, targeting diverse DNA viruses such as poxviruses and betaherpesviruses.

Gallium is classified as a technology metal as it is important for technological innovations. It is also referred to as a strategic metal, which emphasizes its economic relevance. In addition, gallium is a critical raw material that is strategically important but only available in limited quantities. However, recycling dissolved gallium from lowconcentration
wastewater is often not done due to the lack of suitable technologies.
This research presents a membrane-based approach using the siderophore Desferrioxamine B for the recycling of gallium. Nanofiltration membranes were used to separate gallium from other metal impurities (such as arsenic). The membranes recovered about 70% of gallium from low-concentrated synthetic wastewater.
Afterward, the membranes were tested using industrial wastewater, and a similar recovery rate was observed. A model was developed to predict operation parameters that would lead to the highest recovery rate of gallium with the minimum impurities. The model showed that recycling more than 90% of gallium from wastewater is possible using this approach. Therefore, the siderophore-assisted nanofiltration approach demonstrated in this research showed great potential for the sustainable recycling of gallium from industrial wastewater.

Large Language Models (LLMs) have gained significant popularity in recent years for their ability to answer questions in various fields. However, these models have a tendency to "hallucinate" their responses, making it challenging to evaluate their performance. A major challenge is determining how to assess the certainty of a model's predictions and how it correlates with accuracy. In this work, we introduce an analysis for evaluating the performance of popular open-source LLMs, as well as gpt-3.5 Turbo, on multiple choice physics questionnaires. We focus on the relationship between answer accuracy and variability in topics related to physics. Our findings suggest that most models provide accurate replies in cases where they are certain, but this is by far not a general behavior. The relationship between accuracy and uncertainty exposes a broad horizontal bell-shaped distribution. We report how the asymmetry between accuracy and uncertainty intensifies as the questions demand more logical reasoning of the LLM agent, while the same relationship remains sharp for knowledge retrieval tasks.

Scientists and engineers use simulators to model empirically observed phenomena. However, tuning the parameters of a simulator to ensure its outputs match observed data presents a significant challenge. Simulation-based inference (SBI) addresses this by enabling Bayesian inference for simulators, identifying parameters that match observed data and align with prior knowledge. Unlike traditional Bayesian inference, SBI only needs access to simulations from the model and does not require evaluations of the likelihood-function. In addition, SBI algorithms do not require gradients through the simulator, allow for massive parallelization of simulations, and can perform inference for different observations without further simulations or training, thereby amortizing inference. Over the past years, we have developed, maintained, and extended sbi, a PyTorch-based package that implements Bayesian SBI algorithms based on neural networks. The sbi toolkit implements a wide range of inference methods, neural network architectures, sampling methods, and diagnostic tools. In addition, it provides well-tested default settings but also offers flexibility to fully customize every step of the simulation-based inference workflow. Taken together, the sbi toolkit enables scientists and engineers to apply state-of-the-art SBI methods to black-box simulators, opening up new possibilities for aligning simulations with empirically observed data.