Seznam abstraktů 2025: SEKCE A - NANOMATERIÁLY PRO ELEKTRONICKÉ, MAGNETICKÉ A OPTICKÉ APLIKACE. UHLÍKOVÉ NANOSTRUKTURY, KVANTOVÉ TEČKY

Heterogeneous catalysis plays an indispensable tool in modern industrial processes. When supported nanoparticles are downsized to the single-atom limit, the resulting single-atom catalysts (SACs) have proven special value. These individually dispersed metal sites possess defined coordination structures and enabled reaction pathways that differ from traditional heterogeneous catalysis. While offering high stability even in harsh environments, and maximum utilization of catalytic metals, SACs also satisfy the key goals of a sustainable chemistry. Nevertheless, limited reproducibility and scalability of SAC preparation hinder their further practical applications, and efficient methods of SAC synthesis are urgently needed. A major challenge is to prepare single-atom sites on the catalyst support with high density while avoiding aggregation of metal atoms. This challenge can be traced to the trade-off between diffusion and stabilization of metal atoms during the pyrolysis process.
In this talk, several approaches will be presented. These methods will discuss how the metal resources including ions, organometallic complexes, and particles are transformed into single atoms and their possible mechanisms. This work is a critical step forward toward a scalable synthesis and widespread application of dense SACs and pavs the way for the development of further economically feasible catalyst systems.

Pristine carbon nanotubes have relatively weak electrocatalytic activity. On the other hand, due to the symmetry of the π-electron bands, nanotubes can be doped either donor or acceptor to optimize the kinetics of heterogeneous electron transfer. In this report, we focus on analyzing the effects of acceptor (boron), donor (nitrogen) doping and covalent functionalization of nanotubes with -OH group on the electrocatalysis of cathodic oxygen reduction (ORR) and hydrogen evolution reaction (HER), as well as anodic oxygen evolution reaction (OER). Simulations based on the Gerischer-Marcus model show that acceptor doping shifts the onset of ORR and HER reactions to more negative potentials. At the same time, acceptor doping contributes to increasing the efficiency of the anodic OER by lowering its overpotential. On the other hand, donor doping enhances the kinetics of ORR and HER, while suppressing the OER reaction. We also show that covalent functionalization with an -OH group produces an effect analogous to acceptor doping. We present experimental data obtained in alkaline media, which confirm the predictions of the Gerischer-Marcus model. The results presented outline a strategy for preparing electrode materials containing nanotubes in such applications as water splitting, fuel cells and metal-air batteries.
Acknowledgments: The research was funded by the Ministry of Science and Higher Education under the program ‘Regional Initiative of Excellence’ in years 2024-2027, Project No. RID/SP/0048/2024/01.

Silicon quantum dots (SiQDs) are a promising class of nanomaterials for light-emitting devices and biomedical applications due to their low toxicity and size- and surface-tunable photoluminescence. However, developing precise protocols for their surface termination, ensuring colloidal stability and optical performance, remains challenging.(1) Typical strategies include hydrosilylation, where alkenes covalently attach to SiQD surface via high-temperature or radical initiation. These procedures are highly sensitive to factors such as SiQD quality, size, and surface oxidation, and are limited by a narrow range of usable solvents. Especially in polar, and aqueous, environments, standardized protocols are still underdeveloped.
In this work, we present an optimization study of thermal and radical hydrosilylation in SiQDs synthesized via low-pressure non-thermal plasma in three size fractions (2.5, 4.0, and 5.5 nm) to achieve size-tunable photoluminescence, enhanced photoluminescence quantum yield (PLQY) and colloidal stability in solvents of varying polarity.(2) Specifically, 1-dodecene and undecenoic acid/undecenol were chosen for non-polar and polar terminations, respectively. Reaction parameters such as temperature, duration, and mixture composition were systematically varied. Competitive PLQYs close to 20 % for both non-polar and polar environments were reached.
Thus, surface modification of SiQDs critically depends on the type of ligand and requires precise reaction control, but it enables tailored optical and dispersive properties for targeted applications, particularly in polar solvents, which are crucial for many biological and energy-related uses.
REFERENCES: [1] Matejka, F. et al., Scalable and Bright: Unlocking Functional Silicon Quantum Dots with Near-Unity Internal Quantum Yield through Universal Plasma-Driven Engineering, 2025, submitted; [2] Galar P., et al., Engineering the Thermal and Energy-Storage Properties in Quantum Dots Using Dominant Faceting: The Case Study of Silicon, ACS Nano 2025.

Polarons and excitons belong to a class of extensively studied quasiparticles but their interaction with the local environment has been overlooked so far. We present an unexpected phenomenon of solvent-induced generation of polaronic spin active states in a 2D material fluorographene under UV light. Continuous wave X-band electron paramagnetic resonance (EPR) spectroscopy experiments reveal a massive increase in the EPR signal for fluorographene dispersed in acetone under UV-light irradiation, while the system did not show any significant signal in dark conditions and without the solvent. This phenomenon is solvent-specific, the generation of spin-active states is robust in acetone, moderate in benzene, and absent in cyclohexane.The patterns appear due to the generation of transient magnetic photoexcited states of a polaronic character. Our ab initio calculations demonstrate that polarons are formed at the radical sites in fluorographene and interact strongly with acetone molecules in their vicinity. The solvent-tuneable polaron formation with the use of UV light and an easily accessible 2D nanomaterial open up a wide range of future applications, ranging from molecular sensing to magneto-optical devices.

Additive manufacturing of fiber reinforced PA12 composites offers a path to lightweight, high performance components, but is often hindered by the challenging flow and thermal characteristics of PA12. During filament extrusion, precise control of diameter is achieved by stretching the still molten filament, in which melt strength is of particular importance. We demonstrated that increasing the fiber content form 10 to 20 wt% led to filament breakage due to low melt strength, unless reinforced with at least 0.5 vol% CNT. Integrating CNTs not only prevented streching-induced breakage at higher fiber content but also produced filaments with more uniform cross sections and significantly enhanced printability. The filaments reinforced with 10-20 wt% chopped carbon fiber (CF) achieved exceptional extensional alignment during extrusion—over 80 % of fibers lied within 10° of the flow direction. Introducing 0.5 vol % multi wall carbon nanotubes (CNT) further sharpened this orientation (peak at 0-5°) by increasing extensional viscosity. With this highly aligned structure, the fillers formed continuous, end to end conductive pathways that reduced inter fiber contact resistance. This was reflected in that the electrical conductivity of the CF reinforced composites increased by an order of magnitude with CNT addition. Thermomechanical analysis showed that carbon nanotubes also have a beneficial effect on dimensional stability, which is of particular importance for 3D printing: the addition of CNTs flattened the Tg and minimized both pre- and post-Tg expansion. These results highlight that multiscale reinforcement—combining chopped carbon fibers with CNTs—was key to achieving enhanced orientation, improved conductivity, and filament processability in PA12-based composites for high-performance additive manufacturing.

Now a days studies are focused on the development of conductive polymer composites (CPCs) as well as intrinsically conductive polymers (ICPs) [1]. Certain fillers like carbon fiber, carbon black and graphite may also be added to increase the electrical conductivity of polymers [2]. However, nonflexible and rigid composites have many drawbacks such as limited deformation, brittleness, reduced comfort, limit mobility and difficulty in processing. In our study, we generated polymer aerogels with expanded graphite through freeze-drying and thermal annealing, achieving excellent electromagnetic interference shielding. We used various methods to test and characterize the porous electrode embedded with expanded graphite. A Scanning Electron Microscope (SEM) used to analyze the surface, while Energy-Dispersive X-ray (EDX) and X-ray Diffraction (XRD) determined the elemental composition. The prepared samples of composites were then characterized for the electrical conductivity, EMI shielding and mechanical performance. The conductive expanded graphites having 12% particles showed better results as compared to the others.

Perovskites are optically active materials with high light-absorption coefficients, intrinsic defect tolerance, and size-tunable optoelectronic properties. Zinc oxide (ZnO) is a direct bandgap semiconductor that has gained significant importance in optical, electrical, and optoelectronic applications. Prior works indicate that increased photoluminescence yield and electron transport may be achieved when these material are merged in composite systems [Journal of Luminescence 239 (2021) 118396].
In this work, we studied the structural and electronic properties of heterostructures made of thin layers of CsPbBr3 (CPB) nanocrystals prepared from solutions on different ZnO substrates: 0001-Zn face and 0001-O face with polar surfaces, 1010 and 1120 with non-polar surfaces. The material structure was characterized by scanning electron microscopy (SEM) in the secondary (SE) and backscattered (BSE) electron mode. Nanocomposite work function (WF) and time-resolved surface photovoltage (SPV) were studied by scanning Kelvin Probe (SKP) in the dark and under a solar simulator or UV illumination. Confocal photoluminescence (PL) micro-spectroscopy was used to characterize PL emission spectra and their spatial maps.
Depending on ZnO surface orientation and polarity, SEM shows cubic grains or planar nanoisland of CPB. WF of CPB in the dark also depends on type of ZnO substrate. It is correlated with PL maps and red shift in PL spectra. The PL shift may be caused by quantum confinement effects and/or change of crystalline phase in CPB nanocrystals, which is controlled by ZnO surface arrangement, chemistry, and polarity. SPV is significant under solar simulator illumination and positive for all the samples, where the least pronounced effect is from CPB on the Zn face surface. The results thus provide guidelines for hybrid nanoscintillator designs.
ACKNOWLEDGEMENTS: Authors acknowledge financial support from the Czech Science Foundation project No. 24-12872S and the program “Strategy AV 21” of the Czech Academy of Sciences under the work package VP 27 (Renewable energy resources and distributed energy systems).

Vacuum ultraviolet (VUV) light holds excellent potential for various advanced applications, including VUV lasers for atomic cooling and ultra-precise length measurements using interferometry. However, most conventional materials exhibit strong absorption in the VUV range, which limits their use in optical systems operating at these wavelengths.
In this study, we investigate various fluoride-based thin films deposited via (plasma-assisted) electron-beam vacuum evaporation. These materials exhibit wide band gaps and low absorption in the VUV spectral region, making them excellent candidates for high-reflectivity and high-transparency multilayer optical coatings designed for VUV optics.
The deposited thin films demonstrate surface roughness on the order of a few nanometers and exhibit short autocorrelation lengths, which are beneficial for minimizing scattering and maintaining optical performance in multilayer systems.
ACKNOWLEDGMENT: This work was supported by the Technology Agency of the Czech Republic within the framework of the National Centres of Competence Programme (project No. TN02000020), and by the project Quantum Engineering and Nanotechnology (No. CZ.02.01.01/00/22_008/0004649), co-funded by the European Union and the Czech Academy of Sciences (RVO: 68081731). The CzechNanoLab project LM2023051 funded by MEYS CR is gratefully acknowledged for the financial support of the measurements at CEITEC Nano Research Infrastructure. This research was also supported by the project LM2023039 funded by MEYS CR.

Porous nanomaterials are essential for a wide range of modern technologies due to their high specific surface area. While single-material systems have been extensively studied in the past decades, bi-material nanostructures are nowadays attracting increasing interest, as these materials may offer improved performance or new functionality. This work introduces and investigates a novel and versatile strategy for the solvent-free synthesis of two-component metal/metal-oxide (Me/MeO) nanoparticle-based coatings, which combines a magnetron-based gas aggregation source for nanoparticle synthesis with conventional magnetron sputtering.
Using a Ta2O5/plasmonic metal system as a model, we demonstrate that this technique can produce highly porous Me/MeO nanomaterials with various architectures tailored by the deposition protocol. First, a sequential deposition strategy enables the formation of either MeO nanoparticle films decorated with metal nanostructures or sandwich structures with alternating layers of MeO nanoparticles and plasmonic nanostructures. Alternatively, a simultaneous co-deposition process results in porous films composed of MeO nanoparticles randomly decorated with metal nanostructures. A key advantage of our method is the ability to precisely control the size/number of metal nanostructures in the coatings simply by varying the relative deposition rates of the MeO and Me components. This time- and cost-effective method provides a valuable pathway for fabricating Me/MeO nanomaterials with custom-tailored properties, paving the way for applications in advanced photocatalysis, sensing, biodetection and other emerging fields.
ACKNOWLEDGEMENTS: This work was supported by the grant GAČR 25-14402L from the Grant Agency of the Czech Republic.

The rapid development of artificial intelligence (AI) necessitates new demands on device engineering. Research focuses on methods to enhance hardware energy efficiency and increase computational speed. Current computational resources are far from optimal or natural for effective machine learning (ML), neural networks (NNs), and AI.
Existing concepts include in-memory computing, analog computing, photonic and electro-optic computational cores, and the utilization of phase-change materials (PCMs) for synaptic weights. The application of these concepts promises increased speed, parallelism, and reduced energy consumption. Their combined use is expected to yield order-of-magnitude improvements in both aspects, resulting in positive synergy. The integration of PCM-based components potentially enables the implementation of all aforementioned concepts, individually or concurrently within a single device.
Our work focused on the characterization of selected PCMs in a state close to the material's condition inside an operational component. The switching of active PCMs inside a device is achieved with a short pulse of laser light or an electrical current. It was observed that, in some cases, the properties of the thin material layer significantly differ from those of as-deposited and annealed materials. A critical question is whether and how such differences in physical properties may influence the development and design of specific photonic components. Based on the values obtained from the characterization of PCM thin films prepared via various methods, photonic components were modeled. The impact of deviations in physical properties will be discussed.
ACKNOWLEDGEMENTS: This work was supported by the Czech Science Foundation (23-07574S) and the Ministry of Education, Youth, and Sports (LM2023037).

The behavior of anodic memristors grown on ultra-thin Hf superimposed on Ta films is studied. Concomitant anodization of these metals leads to self-assembled nanoscale oxide columns formation, through an electrical version of Rayleigh-Taylor effect. This is due to ionic currents preferring the less resistive paths, enhancing the growth of the corresponding oxide. Oxide resistivities and structures, transport numbers, Pilling-Bedworth ratios are all considered as determining factors for the anodization process of superimposed systems producing defect engineered oxides. The boundary between Hf and Ta oxides influences the conductive pathways required for the memristive effect, thus being most relevant for fabrication of highly stable and forming-free memristors. The use of superimposed films with gradient but complementary thicknesses allows investigating the ideal Hf/Ta ratios for which the best memristive behavior is obtained. Controlled O vacancies generation is a critical factor in switching uniformity and reproducibility. Therefore, oxide “fingers” formation is a promising electrochemical approach towards defect-engineered memristors. Further investigation of the composite oxide formation, particularly in Hf/Ta superimposed system, is topical. Until now, such systems were not recognized in the literature for the ReRAM applications. This is highly promising since both memory and electrical characteristics are improved by the forming-free nature of the memristors with filaments mediated by oxide nanostructuring.

Electrochromic materials are defined as substances that undergo reversibly change color under an applied voltage and are at the forefront of research in smart technologies, including energy-efficient windows[1], flexible electronics[2], and dynamic military camouflage[3]. One promising direction in this field involves the use of materials that incorporate transition metals, known for their diverse electronic properties and ability to support reversible redox processes. In this study, we present a novel class of electrochromic thin films deposited onto ITO coated electrodes. Obtained materials exhibit clear and reversible color changes in response to electrical signals, with performance characteristics that can be tuned through the selection of molecular components. The most efficient systems demonstrated fast switching behavior, strong optical contrast, and good long-term stability under repeated cycling. Electrochemical and optical measurements confirmed that the color changes arise from coupled electronic changes in both the metal centers and the surrounding molecular framework. The study highlights the potential of transition-metal complexes as versatile components in the design of durable and efficient electrochromic coatings for practical applications.
REFERENCES: [1] Cannavale, Alessandro, et al. "Smart electrochromic windows to enhance building energy efficiency and visual comfort." Energies 13.6 (2020): 1449; [2] Li, Wanzhong, et al. "Progress and challenges in flexible electrochromic devices." Solar Energy Materials and Solar Cells 240 (2022): 111709; [3] Lin, Kaiwen, et al. "A multicolored polymer for dynamic military camouflage electrochromic devices." Solar Energy Materials and Solar Cells 278 (2024): 113180.

The pursuit of technologies aimed at protecting human life has grown substantially in response to increased exposure to a wide range of threats. Among the emerging strategies, the functionalization of natural fibers with nanomaterials (NPs) stands out as a promising approach for developing advanced fibrous structures with enhanced properties. Metal oxide NPs, in particular, are known for their ability to neutralize harmful agents while simultaneously imparting additional functionalities to fibers, all without significantly increasing their weight.
Natural fibers have attracted growing interest as sustainable, biodegradable, and cost-effective alternatives to synthetic materials, due to their abundance and low environmental impact. However, a major challenge in the functionalization of textile substrates with nanomaterials lies in the weak adhesion of nanoparticles to the fiber surfaces. Therefore, effective techniques to improve nanoparticle immobilization are critically needed.
Electrospinning emerges as a powerful method for this purpose. By producing fibers at the nanoscale and depositing them onto textile substrates, electrospinning not only enhances the anchoring of nanoparticles but also increases the specific surface area and filtration efficiency of the final material.
In this study, various natural fibers and biopolymers were functionalized with metal oxide NPs (CaO, MgO, SiO₂, TiO₂, and CeO₂) using straightforward and environmentally friendly methods. The resulting materials were extensively characterized, and their functional properties were systematically assessed.

Nanodiamonds (NDs) represent a versatile class of carbon nanomaterials whose properties are strongly determined by their synthetic origin. In this lecture, I will compare two archetypal types of nanodiamonds: detonation nanodiamonds (DNDs) and nanodiamonds obtained by high-pressure high-temperature synthesis (HPHT NDs). Despite their similar crystalline cores, their history of formation leaves distinct fingerprints in size distribution, defect structure, and ultimately in their physico-chemical behavior.
A particular focus will be devoted to hydrogenated surfaces, which provide a unique platform to study fundamental differences between DNDs and HPHT NDs. Our recent results reveal that hydrogenation leads to strikingly different electronic and interfacial properties depending on the ND origin. These variations manifest in electronic properties, wettability, and adsorption characteristics, underlining the importance of the synthetic route as a hidden variable in surface-related phenomena. As a complementary topic, phonon confinement effects in ultrasmall diamond nanoparticles will be discussed, with emphasis on how vibrational spectra encode particle size and structural order.
Together, these insights highlight that “origin matters”: DNDs and HPHT NDs, while often treated interchangeably, represent distinct materials classes with specific opportunities for tailoring surface reactivity and functionality. Understanding and exploiting these fingerprints is essential for rational design of nanodiamond-based interfaces in sensing, biomedicine, and energy-related applications.

Two material perovskites and aluminum oxide (Al2O3) are considered for integration into a composite structure aimed at enabling efficient and ultrafast radiation conversion in scintillator detectors. Perovskites are optically active compounds known for their high light absorption, intrinsic defect tolerance, and tunable optoelectronic properties that depend on crystal size.
Al2O3, a chemically stable and optically transparent ceramic, is a good electrical insulator with excellent thermal stability, making it an important material in optical, electrical, and optoelectronic applications. Previous studies have shown that using aluminum oxide as an insulating substrate in composite structures can enhance film stability, preserve photoluminescence efficiency, and facilitate charge transport by providing a smooth, inert, and non-conductive foundation [Milan Palei et al., Robust and Bright Photoluminescence from Colloidal Nanocrystal / Al₂O₃ Composite Films via ALD, arXiv:2004.11634 (2020)].
In this study, we investigate the structural and electronic characteristics of nanocomposites composed of CsPbBr₃ (CPB) nanocrystals deposited on Al2O3 substrate. CPB was doped by different concentrations of Mg (0.5-5 mol) and Zn (1 mol) to slow absorption of water. Structural analysis was performed using scanning electron microscopy (SEM) in both secondary electron (SE) and backscattered electron (BSE) modes. The work function (WF) and time-resolved surface and spatial-resolved photovoltage (SPV) of the nanocomposites were examined via scanning Kelvin Probe (SKP), in the dark conditions as well as under illumination with a solar simulator and UV light. Confocal photoluminescence (PL) micro-spectroscopy was employed to analyze PL emission spectra and generate spatially resolved PL maps. SEM images and PL maps revealed Mg and Zn concentration-dependent variations in CPB nanocrystal distribution, morphology, and aggregation. SEM analysis suggests that perovskite layers with higher Mg concentration are thicker than those with lower Mg content, Photoluminescence measurements showed that the highest PL intensity occurred for 1Zn:5Mg, with slight peak shifts and Kelvin probe revealed unstable, sometimes negative apparent work functions for certain Mg/Zn doping levels. These findings highlight how Zn/Mg doping and the use of a nonconductive Al₂O₃ substrate influence the optical performance, stability, and surface potential behavior of CsPbBr₃ nanocrystals, offering insights for optimizing perovskite-based optoelectronic devices.
ACKNOWLEDGEMENTS: Authors gratefully acknowledge the financial support from the Czech Science Foundation Project No. 24-12872S and the program “Strategy AV 21” of the Czech Academy of Sciences, specifically work package VP 27 (Renewable energy resources and distributed energy systems).

Nanodiamonds (NDs) are promising carbon nanomaterials for biomedical, sensing, and adsorption applications, yet their deployment is limited by a poor understanding of their adsorption and colloidal properties and stability in challenging environments.
Here, we systematically investigate the interfacial interactions between four ND types—detonation nanodiamonds (DNDs, nitrogen-rich) and high-pressure high-temperature (HPHT) nanodiamonds (pure), each with positive or negative zeta potential—and a series of cationic and anionic surfactants with increasing alkyl chain length (3 to 16). Zeta potential measurements, Fourier-transform infrared (FTIR) spectroscopy, and transmission electron microscopy (TEM) were employed to monitor surface charge modulation, surfactant adsorption, and aggregation states, respectively.
Our results show that ND composition and surface charge polarity strongly influence surfactant adsorption and colloidal stability of ND-surfactant complexes. Short-chain surfactants interact weakly, whereas longer-chain analogues promote stronger hydrophobic interactions and micelle-related effects, leading to pronounced modifications in zeta potential and dispersion state. Comparative analysis reveals distinct interaction pathways between DNDs and HPHT NDs, with nitrogen content contributing to differences in surfactant binding and aggregation behavior.
This work establishes the combined roles of electrostatic and hydrophobic forces in governing ND-surfactant interfacial assembly. The insights provide a mechanistic basis for designing stable ND dispersions, thereby advancing their rational integration into diverse nanomaterial and biomedical applications.

These days nanotechnology is a crucial research field for all areas of humans’ activity, from medicine and cosmetology through electronics and photonics to catalysis e.g. in chemicals production, environment protection or oil industry [1]. In our research we are focused on the development of the novel route of synthesis of metal nanoparticles in aqueous systems (reffered to as matrixless nanoparticles), numerical modelling of their formation and investigation of their potential applications. Nanoparticles are generated in the process of electro-dissolution of metal electrodes in deionized water without any additives apart from dissolved atmospheric carbon dioxide which acts as a source of counterions for generated metal cations. With this approach we were able to obtain silver [2], copper and gold nanoparticles of a high stability (reaching several years in the case of AgNPs). The synthesis conditions such as duration (from several hours to days) and applied potential were strongly dependent on the chosen metal while the common factors crucial for the synthesis efficiency were: lower temperature ca. 10 - 15 °C due to higher CO2 solubility and presence of the atmospheric molecular oxygen facilitating the initialization of the electrochemical dissolution of metal electrodes. So far the colloids of silver nanoparticles have been examined as growth biostimulants for lettuce (Lactusa sativa L. var. capitata).
References:
1. S. Thota, D. C. Crans; Metal nanoparticles. Synthesis and Applications in Pharmaceutical Sciences, Wiley - VCH, Germany 2018
2. M. Adamowska, et al. Molecules, 27, 2592, (2022)

A titanium-tungsten (Ti-W) thin film library with a compositional spread 3-20 at.% W was fabricated utilizing a combinatorial approach and anodized to investigate volatile analog memristive switching. The combinatorial methodology allows for high-throughput screening across a wide compositional range, which is advantageous for identifying optimal alloy composition, while minimizing variability due to fabrication conditions. This is valuable for the investigation of memristive systems, where even slight compositional differences influence the device´s performance. Tested devices demonstrated consistent analog switching with stable endurance up to 10^4 cycles. HRS/LRS ratio reached values up to 10^7 in certain compositions. Furthermore, the presence of multi-level switching makes them a great candidate for high-density data storage applications. Switching behavior is uniform across the compositional range, showing analog volatile characteristics. The absence of an abrupt SET/RESET transitions suggests and interfacial switching mechanism rather than a filamentary one. Future work will focus on analyzing the structural origins of the observed behavior to optimize device performance and stability.

Block copolymer (BCP) self-assembly offers a flexible means to fabricate nanoscale patterns with potential application to photonics, electronics, and high-performance materials. Nonetheless, unambiguous, rapid, and reconfigurable management of BCP morphology remains a significant challenge due to the complex interplay between thermodynamic and kinetic factors. In this work, we describe a new platform that integrates laser-induced local thermal gradients, machine learning (ML) algorithms, and lab-on-chip microfluidics to tune BCP self-assembly under real-time conditions.
Our method employs a programmable laser array to impose spatially resolved heat distributions within a microfluidic chip that is filled with solutions of BCPs. The local heat inputs are adaptively adjusted real time under feedback provided by a convolutional neural network that is trained to recognize and predict morphologies of BCPs under different conditions. The closed-loop feedback enables adaptive optimization of laser parameters to direct the process of self-assembly to desired structural end-states within minutes rather than hours.
We demonstrate formation of sharp lamellar and cylindrical domains with orientation and periodicity that can be controlled, and that the system can be rendered receptive to disturbances or alterations to the pattern in-situ. The hybrid approach offers a highly flexible, scalable, and rapid method of directing BCP ordering that will allow new paths to fabricating nanostructure materials on demand.

Electrochromic devices, which operate based on materials that change color in response to electrical stimuli, require active layers that are not only reliable but also simple to fabricate and easily scalable. In this study, we report a straightforward on-substrate synthesis of a poly(triphenylamino azomethine) electrochromic polymer directly on a conductive substrate (ITO). The polymer film was obtained via thermal polycondensation of two complementary monomers: an electron-rich diamine and a dialdehyde bearing electroactive triphenylamine units. Both monomers were deposited by spray coating, and the resulting “spray-and-bake” method facilitated the formation of uniform films without the need for solubilizing side chains or complex purification steps. The polymer demonstrated reversible color transitions from yellow to purple and blue upon oxidation, with coloration and bleaching times of approximately 3.3 and 1.2 seconds, respectively. These characteristics were confirmed by spectroelectrochemical analysis and cyclic voltammetry. The polymer layer was successfully integrated into a sandwich-type electrochromic device, which retained stable performance over repeated switching cycles. Importantly, the method proved scalable: devices with an area of up to 100 cm² exhibited similar electrochromic properties, underlining the potential of this approach for use in smart windows and display technologies. Overall, this work demonstrates that in situ polycondensation of azomethines is a viable and efficient strategy for fabricating functional, scalable electrochromic coatings with minimal processing and material waste.

A novel electrochromic monomer based on a naphthalene diimide (NDI) core with polymerizable styrene pendant groups was synthesized and explored using two polymerization strategies: photopolymerization and electropolymerization. In the photopolymerization pathway, UV-induced radical polymerization produced uniform crosslinked films. These materials exhibited reversible electrochromic behavior, transitioning from transparent (monomer) or slightly yellow (polymer) to red-brown and green upon stepwise electrochemical reduction of the NDI unit to its radical anion and dianion states.
An electrochromic device fabricated using the photopolymerized film demonstrated excellent cycling stability over 500 redox cycles, with switching times of 18 seconds for coloration and 6 seconds for bleaching. These results highlight the effectiveness of photopolymerized NDI-based polymers in achieving robust and responsive electrochromic performance.
In parallel, electropolymerization of the same NDI-based monomer was conducted to produce redox-active films via electrochemical deposition. While both methods yielded electrochromically active materials, preliminary observations indicate notable differences in film morphology, electrochemical behavior, and switching dynamics depending on the polymerization route.
This study presents a comparative approach to electrochromic material development, emphasizing how the method of polymerizationcan influence the functional properties of NDI-based systems. The findings support the versatility of the NDI scaffold and offer insights into tuning electrochromic performance via synthetic strategy.

The continuously increasing need to transfer enormous amounts of data forces scientists and engineers to constantly improve multiplexing technologies and develop new types of optical fibers.
One of the promising solutions of great potential are free-form nanostructured fibers. Instead of conventional fabrication methods, such as the Modified Chemical Vapor Deposition (MCVD), one can fabricate a fiber which core consists of thousands of glass rods of nanometric diameters. Two or more types of rods are stacked in a predefined pattern and drawn to obtain a fiber of an arbitrary refractive index distribution. Therefore, one can obtain fibers of optical properties impossible to achieve using other fabrication techniques.
Our project aims to determine the best possible structures of few-mode optical fibers using neural networks of different kinds. Using the Generative Inverse Design Networks (GIDNs) framework [1], we studied optimization of the three-mode fibers composed of two types of glass rods: pure SiO2 and SiO2 with 5% of GeO2. We compared the dense neural network, which usage has been reported recently [2], with a convolutional neural network, which we applied for the first time for the free-form fibers optimization. We observed that convolutional network outperformed the dense one just after first GIDNs cycle, nevertheless both managed to achieve high modes separations above 1.9∙10-3.
ACKNOWLEDGMENTS: This work was supported by the National Science Center, Poland [grant number MAESTRO
14, UMO-2022/46/A/ST7/00238].
REFERENCES: [1] C. Chen and G.X. Gu, Adv. Sci. 7, 1902607 (2020); [2] M. Napiorkowski et al., Eng. Appl. Artif. Intell. 133, 107955 (2024).