Accepted
, , Received Date: 2025-07-04
Abstract +
Adhesion at the nanowire-substrate interface plays a critical role in determining the performance, integration density, and long-term reliability of micro/nano devices. However, existing measurement techniques, such as peeling tests based on atomic force microscopy and in-situ electron microscopy techniques, often suffer from operational complexity, limited environmental applicability, and large measurement uncertainties. To solve these problems, this study proposes a cross-stacked bridge testing method based on optical microscopy nanomanipulation (OMNM), which can directly and quantitatively measure nanowire–substrate interfacial adhesion energy under ambient conditions. In this method, nanowires are precisely stacked on the target substrate to form a grid structure, where miniature bridges spontaneously appear at the intersections. The bridge geometry is governed by the mechanical balance between nanowire bending deformation and interfacial adhesion. By combining Euler–Bernoulli beam theory with the principle of energy conservation, a quantitative model is established to correlate arch geometry with adhesion energy, thereby realizing reliable measurement. Using this method, we measure the adhesion energies of SiC, ZnO, and ZnS nanowires on Si substrates. The SiC/Si system yields an adhesion energy of (0.154 ± 0.030) J/m2, which is in excellent agreement with the van der Waals (vdW) theoretical value (~0.148 J/m2), confirming that its interfacial behavior is dominated by vdW forces. In contrast, the measured adhesion energies for ZnO/Si ((0.120 ± 0.034) J/m2) and ZnS/Si ((0.192 ± 0.043) J/m2) are significantly higher than their corresponding vdW predictions (0.090 J/m2 and 0.122 J/m2, respectively). This discrepancy is attributed to surface polarization in ZnO and ZnS nanowires, which induces additional electrostatic attraction and thus enhances interfacial adhesion. These findings not only reveal the coupling mechanism between vdW forces and electrostatic interactions in polar nanowire systems but also provide new experimental evidence for understanding complex interfacial phenomena. The proposed OMNM-based cross-stacked bridge testing method offers advantages of operational simplicity, high accuracy, and broad applicability. In addition to nanowires, it can be extended to other low-dimensional nanostructures, such as nanotubes and two-dimensional materials. Looking forward, this approach holds promise as an efficient platform for building adhesion energy databases of realistic systems and for advancing mechanistic insights into interfacial adhesion. Furthermore, it can provide valuable guidance for the design, optimization, and reliability evaluation of next-generation nanoelectronic and optoelectronic devices, thereby contributing to micro/nano fabrication and functional device engineering.
, , Received Date: 2025-06-25
Abstract +
To address the frequency sweeping nonlinearity of frequency-modulated continuous-wave signals generated by a current-modulated distributed feedback laser diode, we propose and experimentally demonstrate a pre-distortion method based on a feedforward neural network. For this method, the beat frequency signals of the distributed feedback laser diode under a sawtooth-waveform current modulation are first experimentally obtained, and then the time-frequency curves of the distributed feedback laser diode output are obtained by performing a Hilbert transform on the beat signals. Subsequently, three-layer feedforward neural networks with 10, 5, and 3 hidden-layer neurons are constructed, respectively. By taking the driving current and the time-frequency curves as the input and output of the feedforward neural network, respectively, the nonlinear mapping relationship between them is established. Finally, a backpropagation algorithm is utilized to obtain the pre-distortion modulation current. Taking this current under the modulation frequency from 1 kHz to 10 kHz to drive the distributed feedback semiconductor laser (DFB-LD), the performance of the generated frequency-modulated continuous-wave (FMCW) signals is analyzed. We use nonlinear regression coefficients and residual root mean square values to characterize the performance. For the modulation frequency set at 4 kHz, the frequency sweeping nonlinearity and the residual root mean square value are reduced from 5.29 × 10–3 and 281 MHz to 1.77 × 10–5 and 15.15 MHz, respectively. With the modulation frequency fixed at 6 kHz, the frequency sweeping nonlinearity decreases from 5.58 × 10–3 to 1.52 × 10–5 and the residual root mean square declines from 251.98 MHz to 12.17 MHz in the proposed scheme. Across the entire tested frequency range from 1 kHz to 10 kHz, the nonlinearity remains stable at ~10–5 after adopting the pre-distortion scheme, with RMS values consistently below 20 MHz. The proposed method is expected to provide a new scheme for the linearization technology of the sweep signal in high-precision frequency-modulated continuous-wave light detection and ranging systems.
, , Received Date: 2025-07-24
Abstract +
In this work, a special striped water electrode dielectric barrier discharge device is designed. Through numerical solutions of the Laplace equation, the spatial distribution of the applied electric field is revealed to exhibit a strip-shaped nonuniform distribution featuring the alternating regions of enhanced and weakened field intensity. These field gradients play a pivotal role in governing the plasma, for the intensified regions act as preferential sites for discharge onset, directly shaping the formation and evolution of plasma structures. Using this device, a series of novel striped patterns is observed in the discharge of a mixed gas of air and argon, marking a significant advancement in pattern formation studies. Notably, four striped superlattice patterns are obtained for the first time, each displaying intricate structural hierarchies. Among them, the large and small dot honeycomb striped superlattice pattern featuring structural complexity is chosen to investigate the formation mechanisms. The pattern is composed of three substructures: small dots, large dots, and a honeycomb framework. In the experiment, the emission spectra of different substructures are measured using a spectrograph, revealing that they are in different plasma states. The spatiotemporal dynamic behaviors of the pattern are observed using a high-speed camera and two photomultiplier tubes. It is found that the discharge sequence is small dots → large dots → honeycomb framework, where the honeycomb framework is formed by the superposition of random discharge filaments. The electric field distributions at different times are simulated by solving the Poisson equation, and the result well explains the formation mechanism of the above-mentioned patterns.
, , Received Date: 2025-07-17
Abstract +
Rydberg atoms possess a large electric dipole moment and exhibit high sensitivity to electromagnetic signals. Receivers based on Rydberg atoms represent a novel reception mechanism, demonstrating broad application prospects in the field of communication. Current research has not addressed the influence of the operating parameters of Rydberg atomic receiver on channel capacity. This study establishes a channel capacity model for Rydberg atomic receiver based on Shannon's formula and the response mechanism of the electromagnetically induced transparency (EIT) effect. Using this model, the influences of atomic number density, laser beam waist, and coupling laser Rabi frequency on the channel capacity of Rydberg atomic receiver are analyzed. A strategy for optimizing channel capacity by adjusting the coupling laser Rabi frequency is proposed, and an analytical solution for the Rabi frequency that maximizes channel capacity is derived. The accuracy of this analytical solution is then verified through numerical simulations. The channel capacity corresponding to the analytical solution in this study is similar to the optimal channel capacity obtained using the one-dimensional optimization method (Newton’s method) and is superior to the results obtained by the quadratic interpolation method, demonstrating the effectiveness of the proposed analytical solution in optimizing the channel capacity of Rydberg atomic receiver. This research provides theoretical guidance for designing high-performance Rydberg atomic receiver and optimizing channel capacity.
,
Abstract +
The neutron total cross section is one of the most fundamental nuclear data. In the low-energy resonance region, discrepancies among measurements are caused by experimental backgrounds, sample self-shielding, and instrumental broadening, complicating the precise determination of cross sections. In this work, the neutron total cross sections of 169Tm were measured at the Back-n facility of the China Spallation Neutron Source (CSNS). Wing-shaped lithium glass detectors were employed to record the transmitted neutron signals from samples with thicknesses of 0.5 mm and 4.5 mm. The raw data were processed, including time-of-flight correction and normalization to the incident proton number, to account for fluctuations in the proton beam intensity. The in-beam γ background was quantified using the saturated resonance absorption technique with glass scintillators. Following background subtraction, transmission and total cross-section data were obtained over the energy range of 1–100 eV. The measured spectra were analyzed using the SAMMY code within the framework of the R-matrix. Capture data from Ren et al. were incorporated to improve the reliability of the extracted resonance parameters. For the resonance near 8 eV, the Γγ was set to the library average value of 86 meV. The resonance energy was determined to be 8.037 eV, consistent with the recently reported resonance by I. Knapová et al., and the neutron widths for both total spin states were evaluated. Based on the extracted parameters, the neutron total cross section of 169Tm was reconstructed using the Reich–Moore approximation. The reconstructed cross section shows good agreement with those recommended by ENDF/B-VIII.1 library, confirming the reliability of the resonance parameters extracted from the capture–transmission measurements. Overall, the present measurements and analysis provide a set of resonance parameters for 169Tm, enhancing the experimental foundation for both nuclear physics research and nuclear industry.
The datasets presented in this paper, including the neutron transmission, neutron total cross section, and resonance parameters are openly available at https://www.doi.org/10.57760/sciencedb.j00213.00192 (Please use the private access link https://www.scidb.cn/s/MrUVry to access the dataset during the peer review process)
The datasets presented in this paper, including the neutron transmission, neutron total cross section, and resonance parameters are openly available at https://www.doi.org/10.57760/sciencedb.j00213.00192 (Please use the private access link https://www.scidb.cn/s/MrUVry to access the dataset during the peer review process)
,
Abstract +
The synthesis of superheavy nuclei (SHN) is a leading research frontier in nuclear physics today. In the experiments for synthesizing SHN via fusion-evaporation reactions, the appropriate choice of projectile-target combination and determination of the optimal incident energy are crucial. The number of SHN that can be synthesized with stable projectiles is very small. The fusion-evaporation reaction with radioactive projectile is one of the promising ways for SHN synthesis and it is of great significance to investigate this kind of reactions deeply. In this work a systematic study has been carried out on the fusion-evaporation reactions with radioactive projectiles. The capture cross section is calculated with the empirical coupled channel model, the fusion probability is computed by the dinuclear system model with a dynamical potential energy surface (DNS-DyPES model) and the survival probability is determined through the statistical model.
In the systematic study, 11 actinide isotopes with $Z=90$--100 are used as targets which are $^{232}$Th, $^{231}$Pa, $^{238}$U, $^{237}$Np, $^{244}$Pu, $^{243}$Am, $^{248}$Cm, $^{249}$Bk, $^{251}$Cf, $^{254}$Es and $^{257}$Fm. Projectiles are isotopes between proton and neutron drip lines for elements $Z=4$--32 and most of these projectiles are radioactive. By combining these projectiles and targets, 4969 reaction systems are proposed for synthesizing isotopes of superheavy elements Z=104-122. Through large-scale calculations, the excitation functions for $2n$--$5n$ evaporation channels of each reaction system are obtained. With the results of these reaction systems, we establish a synthesis cross section dataset for superheavy nuclei. For each reaction system, the dataset includes the identities of the synthesized SHN, the optimal incident energies and the maximal evaporation residue cross sections in $2n$-$5n$ evaporation channels. This dataset may serve as a theoretical support for synthesizing new superheavy nuclides and elements.
Additionally, taking the reactions with $^{232}$Th target as examples, we discuss systematic trends in the results and explore the underlying SHN synthesis mechanism. The synthesis cross sections of these reactions, shown in Fig. 1, are in vastly differences. We find that inner fusion barrier of the compound system formed after the projectile touches the target and fission barrier of the compound nucleus are key factors that influence the synthesis cross section. Qualitatively, the projectile-target combinations with relatively large synthesis cross sections are featured by a lower inner fusion barrier in the compound system formed upon contact which favors fusion and a higher fission barrier in the compound nucleus which enhances survival probability. These conclusions may provide valuable references for the theoretical research related to superheavy nuclei synthesis. The dataset presented in this paper are available at the Science Data Bank at http://www.doi.org/10.57760/sciencedb.27854 (Please use the private access link https://www.scidb.cn/s/bimY7j to access the dataset during the peer review process).
In the systematic study, 11 actinide isotopes with $Z=90$--100 are used as targets which are $^{232}$Th, $^{231}$Pa, $^{238}$U, $^{237}$Np, $^{244}$Pu, $^{243}$Am, $^{248}$Cm, $^{249}$Bk, $^{251}$Cf, $^{254}$Es and $^{257}$Fm. Projectiles are isotopes between proton and neutron drip lines for elements $Z=4$--32 and most of these projectiles are radioactive. By combining these projectiles and targets, 4969 reaction systems are proposed for synthesizing isotopes of superheavy elements Z=104-122. Through large-scale calculations, the excitation functions for $2n$--$5n$ evaporation channels of each reaction system are obtained. With the results of these reaction systems, we establish a synthesis cross section dataset for superheavy nuclei. For each reaction system, the dataset includes the identities of the synthesized SHN, the optimal incident energies and the maximal evaporation residue cross sections in $2n$-$5n$ evaporation channels. This dataset may serve as a theoretical support for synthesizing new superheavy nuclides and elements.
Additionally, taking the reactions with $^{232}$Th target as examples, we discuss systematic trends in the results and explore the underlying SHN synthesis mechanism. The synthesis cross sections of these reactions, shown in Fig. 1, are in vastly differences. We find that inner fusion barrier of the compound system formed after the projectile touches the target and fission barrier of the compound nucleus are key factors that influence the synthesis cross section. Qualitatively, the projectile-target combinations with relatively large synthesis cross sections are featured by a lower inner fusion barrier in the compound system formed upon contact which favors fusion and a higher fission barrier in the compound nucleus which enhances survival probability. These conclusions may provide valuable references for the theoretical research related to superheavy nuclei synthesis. The dataset presented in this paper are available at the Science Data Bank at http://www.doi.org/10.57760/sciencedb.27854 (Please use the private access link https://www.scidb.cn/s/bimY7j to access the dataset during the peer review process).
,
Abstract +
With the technological advantages of high thrust, high specific impulse, long life, variable specific impulse and high efficiency, the variable specific impulse magnetoplasma rocket engine has become the essential advanced propulsion system for the deep space exploration and manned space flight in the future. In the variable specific impulse magnetoplasma rocket engine, the ion cyclotron resonance heating stage is linked with the helicon plasma source. The operation status of the helicon plasma source has a direct influence on the ion heating process in the ion cyclotron resonance heating stage. It is of great significance for the testing and the optimization of the engine performance to reveal the influence of the ionization process on the ion heating process. In this paper, a multi-fluid model in which the ion cyclotron resonance heating stage is linked with the helicon plasma source was developed. The numerical simulation with different input current of helicon plasma source and different pressure was performed to analyze the effect of the operation status in the helicon plasma source on the ion energy density in the ion cyclotron resonance heating stage. The results show that the discharge mode of the helicon plasma source gradually changes with the increase of the input current and the plasma density jump appears while the ion temperature basically remains unchanged. With the plasma density jump and nearly identical ion temperature the ion energy density jump also appears in the simulation domain. Similar with the results of the simulation under different input current of the helicon plasma source, the plasma density and the ion energy density also jump when the pressure increases. However, the ion temperature decreases due to the deviation between the input frequency and the resonance frequency.With the numerical model and the input conditions of this paper, the ionization process in the helicon plasma source is decoupled with the ion heating process in the ion cyclotron resonance heating stage. The energy gain of a single ion in the ion cyclotron resonance heating stage does not change with the operation status of the helicon plasma source which account for the ability of the engine to work in multi mode.
,
Abstract +
The manufacturing process of superconducting quantum processor chips faces unique metal contamination challenges, with significant differences in material systems and process characteristics compared to traditional semiconductor chips. This study focuses on the issue of metal contamination in the fabrication process of quantum chips, systematically analyzing the sources, diffusion mechanisms, and prevention strategies of metal contamination in quantum chips. It particularly emphasizes the bulk diffusion and surface migration behaviors of superconducting materials (such as Ta, Nb, Al, TiN) on sapphire and silicon substrates. The aim is to provide theoretical basis and technical references for process optimization, and to promote the industrialization process of quantum computing technology in our country.
The metal contamination in the fabrication of quantum chips is mainly caused by the metal film materials used in the process, the external environment, or the unintended metal impurity atoms introduced during the manufacturing process. Among them, some quantum chip components directly use superconducting metal materials. Unlike semiconductor chips, they cannot achieve front and back stage isolation, resulting in the continuous presence of metal surface migration channels, and there are exposed metal structures on the chip surface. Metal contamination often leads to two fundamental failure problems: circuit short circuits and leakage currents. These problems mainly result from the bulk diffusion of metal impurities in the dielectric layer and the migration behavior on the sample surface. The diffusion and migration rates of metals are affected by temperature, interface reactions, defects, and grain boundaries. The results show that the sapphire substrate, due to its dense lattice structure, exhibits excellent anti-diffusion performance, reducing the risk of contamination and providing a stable interface environment for superconducting quantum chips. For silicon substrates, special attention needs to be paid to the contamination risks posed by high-mobility metals such as Au, In, and Sn. Experimental verification shows that Ti/Au under bump metallization structures on silicon substrates are prone to Au penetration diffusion, and increasing Ti thickness does not significantly improve the blocking effect. The low-temperature process (<250 °C) and ultra-low-temperature operating environment (mK level) of quantum chips effectively suppress metal diffusion, but exposed metal surfaces and material diversity still pose unique challenges.
The study recommends establishing a dedicated metal contamination prevention system for quantum chips and proposes future research directions, including evaluation of novel materials, surface state regulation, and long-term reliability studies. This work provides important theoretical support and technical guidance for process optimization and performance enhancement of superconducting quantum chips.
The metal contamination in the fabrication of quantum chips is mainly caused by the metal film materials used in the process, the external environment, or the unintended metal impurity atoms introduced during the manufacturing process. Among them, some quantum chip components directly use superconducting metal materials. Unlike semiconductor chips, they cannot achieve front and back stage isolation, resulting in the continuous presence of metal surface migration channels, and there are exposed metal structures on the chip surface. Metal contamination often leads to two fundamental failure problems: circuit short circuits and leakage currents. These problems mainly result from the bulk diffusion of metal impurities in the dielectric layer and the migration behavior on the sample surface. The diffusion and migration rates of metals are affected by temperature, interface reactions, defects, and grain boundaries. The results show that the sapphire substrate, due to its dense lattice structure, exhibits excellent anti-diffusion performance, reducing the risk of contamination and providing a stable interface environment for superconducting quantum chips. For silicon substrates, special attention needs to be paid to the contamination risks posed by high-mobility metals such as Au, In, and Sn. Experimental verification shows that Ti/Au under bump metallization structures on silicon substrates are prone to Au penetration diffusion, and increasing Ti thickness does not significantly improve the blocking effect. The low-temperature process (<250 °C) and ultra-low-temperature operating environment (mK level) of quantum chips effectively suppress metal diffusion, but exposed metal surfaces and material diversity still pose unique challenges.
The study recommends establishing a dedicated metal contamination prevention system for quantum chips and proposes future research directions, including evaluation of novel materials, surface state regulation, and long-term reliability studies. This work provides important theoretical support and technical guidance for process optimization and performance enhancement of superconducting quantum chips.
, , Received Date: 2025-07-08
Abstract +
Silicon photomultipliers (SiPMs) have been widely used in the field of weak light detection. However, SiPMs utilizing small-sized Geiger-mode avalanche photodiode (G-APD) cells face the limitations due to a restricted effective geometric fill sactor (GFF), which leads to relatively low photon detection efficiency (PDE), and additionally, constrained by the intrinsic properties of silicon materials, their PDE in the near-infrared band is also relatively insufficient. To address the above issues, this work proposes a regional optical field modulation approach based on topological photonic crystals (TPCs), aiming to improve the PDE of SiPMs without modifying their internal structure. Through COMSOL electromagnetic wave frequency-domain simulation, the multi-band synergistic mechanism of dead-zone topological edge state guidance, photosensitive region slow-light effect, and Bragg scattering is revealed. In the 460–700 nm band, the honeycomb lattice in the dead zone induces topological edge states via Floquet periodic analysis, while the periodic dielectric distribution of the lattice excites Bragg scattering to reduce photon reflection loss at the metal surface and precisely couples photons to the photosensitive region, leading to an increase in effective GFF from 46.4% to 63.1% at 621 nm. In the 700–1100 nm band, the periodic dielectric distribution of the honeycomb lattice further excites Bragg resonance to reduce metal surface reflection loss, and simultaneously, the multiple scattering mechanism substantially extends the propagation path of photons in the dead zone to improve the coupling probability with the photosensitive region. The designed periodic silicon pillar structure in the photosensitive region effectively extends the lateral propagation path of photons through the slow-light effect, while Bragg scattering reduces reflection loss, resulting in a significant increase in absorption efficiency from 41.19% to 51.37% at 900 nm. Simulation results show that this design scheme increases the average PDE of SiPMs by 50% in the 460–1100 nm band (with a peak value of 81%) and can be implemented via mainstream etching processes (electron beam lithography + reactive ion etching). Compared with traditional microlens and plasmonic structures, TPCs exhibit significant advantages in broad-spectrum response and process simplification. This work provides a new topological photonics approach for photon recycling and PDE enhancement of SiPMs.
, , Received Date: 2025-06-20
Abstract +
Complex networks are powerful tools for characterizing and analyzing complex systems, with wide applications in fields such as physics, sociology, technology, biology, and port terminal management. One of the core issues in complex networks is the mechanism behind the emergence of scaling laws. In real-world networks, the mechanisms underlying the emergence of scaling laws may be highly complex, making it difficult to design network evolution mechanisms that fully align with reality. Explaining real networks through simple mechanisms is a meaningful research topic. Since Barabási and Albert discovered that growth and linear preferential attachment are mechanisms that generate power-law distributions, scholars have identified various forms of preferential attachment that produce power-law degree distributions. However, the most famous and useful one remains the linear preferential attachment in the BA model. Can scale-free behavior also emerge from random attachment and growth? In traditional network analysis, nodes are assumed to join the system at discrete, equally spaced time intervals, often based on the unfounded assumption that interarrival times follow a uniform distribution. In reality, nodes arrive randomly, and their interarrival times do not necessarily follow a uniform distribution. Although complex networks have flourished over the past two decades, they still cannot fully describe real systems with multiple interactions. Hypernetworks, which capture interactions involving more than two nodes, have become an important subject of study, and the mechanisms underlying the emergence of scaling in hypernetworks are a key research focus. The paper first introduces the concept of cliques in hypernetworks. A 1-element clique is a node, a 2-element clique is an edge in a complex network, a 3-element clique represents a triangle in higher-order networks, and a 4-element clique corresponds to a tetrahedron in higher-order networks. Secondly, we propose a clique-driven random hypernetwork evolution model. By combining stochastic processes, nodes arrive in continuous time, which better reflects real-world scenarios and provides a justified distribution for node interarrival times. Using Poisson process theory, we analyze the clique-driven random hypernetwork evolution model, avoiding arbitrary assumptions about node interarrival time distributions commonly made in traditional network analysis, thereby making the network analysis more rigorous. We derive an analytical expression for the cumulative degree distribution and the power-law exponent of the node degree distribution. Finally, we validate the theoretical predictions through computer simulations and empirical analysis of collected real-world data. The results show that the clique-driven random hypernetwork evolution model employs a simple connection mechanism, and that scale-free behavior emerges from growth and random attachment in higher-order structural networks. In our model, not only do nodes join the network in continuous time, but new nodes also randomly select d-element cliques, resulting in a power-law degree distribution. When d = 2, the power-law exponent of the node degree distribution in our model matches that of the BA model. When d > 2, the power-law exponent of the degree distribution depends on the number of elements of the driving clique (simplex dimension). We can directly estimate the power-law exponent of the model's degree distribution using the number of elements of the driving clique.
, , Received Date: 2025-07-09
Abstract +
Infinite-layer nickelates, obtained by removing the apical oxygen from perovskite precursors, are the first nickelate system to exhibit superconductivity and provide a platform for exploring nontraditional superconductivity. Although the traditional CaH2 sealed-tube reduction method is simple and effective, it is an ex-situ process that tends to cause surface contamination or degradation, making it unsuitable for surface-sensitive measurements like angle resolved photoemission spectroscopy (ARPES). To address this issue, we establish three different in-situ atomic hydrogen reduction methods in an ultrahigh vacuum chamber—namely, a lab-based RF plasma cracker, an industrial RF plasma cracker, and a thermal gas cracker. The key parameters including hydrogen flow, RF power or filament temperature, reduction temperature, and timeare comprehensively optimized using each of the above methods. Structural evolution is monitored by X-ray diffraction (XRD), surface morphology is characterized by atomic force microscopy (AFM), and superconducting properties are examined through electrical transport measurements. The results show that all three in-situ methods can achieve reduction and superconducting properties comparable to or better than CaH2 reduction. Moreover, all atomic hydrogen approaches yield lower surface roughness than CaH2 from the same precursor, highlighting their clear advantage in enhancing surface flatness. Notably, the industrial RF plasma source, due to its higher hydrogen production efficiency, enables sufficient reduction under milder conditions, resulting in even smoother surfaces. This study also provides a detailed summary of the parameter optimization for each method, providing valuable guidance for the controlled reduction of high-quality infinite-layer nickelate thin films.
, , Received Date: 2025-07-09
Abstract +
Freezing of multicomponent droplets and thin films is ubiquitous in natural environments and engineered settings. Previous studies on multicomponent droplets, including Marangoni-driven self-lifting droplets and soap-bubble freezing, have identified the roles of interfacial flow and solute redistribution, often exhibiting a snow-globe effect of migrating ice particles. Curvature and field-of-view constraints in droplet systems hinder continuous observation of a single object. Here, utilizing the comparability of interfacial heat and mass transfer between droplets and films, we employ a flat isopropanol-water binary film on a cooled substrate to achieve high-resolution, time-resolved in-situ microscopy observation of individual separated ice flakes within a supercooling (ΔT) range of the substrate. Experiments show that with the increase of ΔT, the external shape of ice flakes evolves from hexagonal pyramid to dodecagonal pyramid and ultimately to a nearly-conical form, accompanied by the decrease of transparency. We quantify morphological evolution by using a shape factor β and qualitatively distinguish crystal-structure differences by combining bright-field and dark-field microscopy. A minimal model that couples solute and thermal diffusion with Marangoni stress rationalizes the observations: solute-concentration gradients primarily drive structural evolution, while the competition between advection and diffusion governs anisotropic growth. These results provide mechanistic insight into interfacial freezing dynamics of multi-component liquid films and establish flat-film microscopy as a platform for single-flake kinetics.
, , Received Date: 2025-08-08
Abstract +
Two-dimensional ferroelectric α-In2Se3 possesses many fascinating physical properties. However, chemical-vapor-deposited ferroelectric α-In2Se3 typically requires high temperatures (>650 ℃). In this work, α-In2Se3 is synthesized at 400 to 460 ℃ by introducing a KCl/LiCl/NH4Cl ternary catalyst, resulting in a 200 ℃ reduction in growth temperature compared with ferroelectric α-In2Se3 synthesized by the traditional chemical vapor deposition (CVD) method. The surface morphology of the as-prepared material is controlled by temperature and gas flow rate. As the growth temperature increases from 400 to 460 ℃, the synthesized α-In2Se3 is changed from discrete hexagonal flakes to a continuous and uniform thin film, which is confirmed by the scanning electron microscope. Raman spectroscopy shows that the characteristic peaks of In2Se3 are located at 103, 180, and 195 cm–1, respectively, indicating that the CVD-grown In2Se3 is α-phase. Furthermore, energy dispersive spectrometer and X-ray photoelectron spectroscopy indicate that the elemental composition is close to the ideal stoichiometric ratio, confirming the successful synthesis of the α-In2Se3. Then, the as-prepared α-In2Se3 is transferred onto Si/SiO2 substrate for device fabrication. Atomic force microscope indicates that the film is uniform, with an approximate thickness of 63 nm. The fabricated two-terminal memristors exhibit analogous resistive switching behaviors. And such memristors are used to achieve synaptic functions of long-term potentiation/long-term depression. For artificial neural network simulations based on the synaptic memristors, the recognition accuracy for hand-written digit image exceeds 90%. This work provides a practical method for growing two-dimensional ferroelectric α-In2Se3 at low temperatures for applications in synaptic devices and neuromorphic computing.
- 1
- 2
- 3
- 4
- 5
- ...
- 14
- 15
