arXiv subjects

Seiki Saito

Publications and source records attributed to Seiki Saito.

11 recordsLinked to original sources

arXiv2026

Two-state generator extraction: property currents and a two-layer arrow of time in pre- and post-selected quantum dynamics

Conditioning on both past and future assigns intermediate-time properties a causal observer does not; these time-symmetric assignments obey exact symmetry theorems and are measurable from trajectories. We use two-state generator extended dynamic mode decomposition (gEDMD): because weak values obey $dA_w/dt=i\langle[H,A]\rangle_w$ exactly, generator extraction, with an exact-derivative baseline, applies unchanged to them. First, a reflection involution on the pre-/post-selected ensemble splits every window-fitted friction uniquely as $γ_{fwd}=γ_A+γ_S$: $γ_A$, antisymmetric about the midpoint, carries the modes' boundary-condition physics; $γ_S$, symmetric, comes from the differencing scheme; both follow from the same data as $(γ_{fwd}\pmγ_{bwd})/2$. At a fixed inference resolution the arrow of time has two layers: the coherent-mode arrow reverses at the midpoint, the fluctuation-level one does not, $γ_S$ dominating $γ_A$ at every size and class. The difference is one of degree: $γ_S$ is 34 times larger there than at the mode layer, and with the exact derivative both layers reverse: immunity belongs to the inference, not the ensemble. Second, in a lattice interferometer conditioned only at its ports, the quantum Cheshire-cat structure emerges unimposed: particle and polarization obey separate continuity equations, and a local field in the polarization-carrying arm rotates that phase alone, at exactly twice the field strength, entering the generator as a rigid imaginary shift, while the particle's weak density stays invariant to machine precision. We verify the sample-level identity and the two layers from $2^8$ to $2^{20}$ dimensions: $|γ_A|/γ_S=0.09$ to $0.27$ across five classes; self-averaging makes it insensitive to class among those sharing a boundary modulation, removing the $2^{-N/2}$ overlap obstruction for $N$ qubits.

quant-ph↗

arXiv2026

Temporal Coarse-Graining as the Origin of Macroscopic Friction in Quantum Spin Chains via Data-Driven Liouvillian Extraction

Understanding how macroscopic irreversible hydrodynamics emerges from the reversible unitary dynamics of isolated quantum many-body systems remains a central challenge. Conventional approaches force spin density dynamics into purely diffusive models, obscuring the interplay of pressure, spin current, and local friction, and leave open the role of the observer's temporal resolution. We introduce a fully data-driven framework based on generalized Extended Dynamic Mode Decomposition (gEDMD) integrated with the Mori-Zwanzig projection. By expanding the observable dictionary to include spin currents, we extract the Navier-Stokes hydrodynamic coefficients from a chaotic XXZ spin chain across temporal coarse-graining scales. Our unconstrained extraction reveals a dichotomy: the mechanical elasticity ($c^2$) derives from the exact unitary dynamics, preserving microscopic reversibility, whereas the macroscopic friction ($γ$) and kinematic viscosity ($ν$) show zero net dissipation, oscillating rapidly around zero in the exact-derivative limit. Genuine macroscopic transport therefore requires finite temporal coarse-graining: at a finite observation timescale ($Δt_{cg} > 0$) the system passes through a crossover where these fluctuations average out, establishing an intermediate functional regime with strictly positive friction and viscosity. We further show that the coarse-grained generator is a predictive reduced model requiring only $O(N)$ macroscopic observables, and that the sign of the emergent friction is inherited from the causal, forward-in-time direction of the inference: time-symmetric estimators yield zero net dissipation at any coarse-graining scale. Macroscopic friction in isolated quantum systems is thus not an absolute property but an emergent phenomenon dictated by the temporal resolution and causal direction of the observer's description.

quant-ph↗

arXiv2026

Development of a 3D-CNN-based Prediction Model for Migration Barriers in Plasma-Wall Interactions

Understanding the long-term transport of hydrogen isotopes in plasma-facing materials, such as tungsten, is critical for the steady-state operation of magnetic confinement fusion reactors. However, dynamically updating the transition parameters for kinetic Monte Carlo (kMC) simulations as the atomic structure evolves under continuous plasma irradiation remains a severe computational bottleneck. Conventionally, calculating these migration barriers requires the iterative and computationally expensive Nudged Elastic Band (NEB) method. To overcome this limitation, this article presents a highly efficient surrogate model for predicting migration barriers using a three-dimensional Convolutional Neural Network (3D-CNN), establishing the final component necessary to realize on-the-fly molecular dynamics (MD) and kMC hybrid simulations. The proposed deep learning model takes a two-channel volumetric input, the local three-dimensional potential energy distribution and the voxelized spatial coordinates of the initial and final trapping sites, to directly output the migration barrier as a scalar value. Trained on a comprehensive dataset of tungsten-hydrogen configurations evaluated using the Embedded Atom Method (EAM) potential, the model demonstrated robust predictive accuracy, achieving a Mean Absolute Error (MAE) of 0.124 eV and a high coefficient of determination of 0.890. Furthermore, utilizing GPU acceleration, the inference time is reduced to approximately 2.7 milliseconds per barrier, achieving a speed-up ratio of over 23,000 compared to conventional NEB calculations. This extraordinary acceleration effectively resolves the computational barrier of transition rate evaluations, paving the way for large-scale, dynamic modeling of plasma-wall interactions.

physics.plasm-ph↗

arXiv2026

Deep Learning-Accelerated Dynamic Kinetic Monte Carlo Simulation for Hydrogen Transport in Tungsten

In magnetic confinement fusion reactors, hydrogen plasma irradiation causes material saturation and recycling, where hydrogen released from the tungsten wall significantly impacts the peripheral plasma. Kinetic Monte Carlo (kMC) simulations are essential for investigating the dynamic balance between incident and emitted fluxes at the atomic scale. However, standard kMC frameworks are inadequate for handling realistic material complexities, such as polycrystalline structures and dynamic evolution under irradiation, being computationally bottlenecked by continuous transition parameter updates. Conventionally, evaluating migration barriers in disordered systems (e.g., grain boundaries) relies on computationally prohibitive on-the-fly atomistic calculations like the Nudged Elastic Band (NEB) method. Here, we present a deep learning-accelerated Dynamic kMC framework that eliminates this reliance. Our approach integrates a three-stage deep learning pipeline: a pix2pix model for predicting local 3D potential energy distributions, a U-Net for extracting hydrogen trapping sites, and a 3D-CNN for directly evaluating migration barriers. To achieve macroscopic timescales, we implemented a hierarchical spatial index combined with a differential local-update algorithm operating in O(1) complexity. This architecture restricts recalculations to the immediate vicinity of moving atoms, accelerating updates. Demonstrated on a large-scale realistic polycrystalline tungsten model, the framework successfully reproduces preferential hydrogen trapping along grain boundaries, bridging the gap between atomic-scale accuracy and macroscopic timescales for full-scale plasma-wall interaction simulations.

cond-mat.mtrl-sci↗

arXiv2024

Reactive Molecular Dynamics Simulation on DNA Double Strand Breaks Induced by Hydrogen Elimination

We propose a scar model to reproduce how double-strand breaks occur in the telomeric DNA damaged by the effect of the $β$-decay of tritium to helium. In this scar model, the two hydrogens bonded to the 5$^\prime$ carbon connecting the pent saccharides and phosphate are removed. Molecular dynamics simulations using the reactive force field are carried out for 10 cases for the telomeric DNA consisting of 16 base pairs (32 nucleotides). It results in double-strand breaks (DSBs) being observed for structures with more than 24 scars. For 16 scar cases, only single-strand breaks (SSB) are observed. Moreover, in the case of $\left\{16, 0 \right\}$ and $\left\{ 0, 16 \right\},$ where only one of the strands had scars, SSB occurs only in the scarred strand. Secondly, in the $\left\{16, 8 \right\}$ and $\left\{ 8, 16 \right\}$ cases, DSBs occurs. Therefore, we conclude that the following conditions are necessary for DSBs:(i) Scars must occur on both the L and R strands. (ii) A large number of scars (24 or more) must occur in close proximity to each other.

cond-mat.soft↗

arXiv2024

Molecular dynamics simulation for coalescence of vacancies in tungsten crystal

We performed molecular dynamics simulations of coalescence of two vacancies in a tungsten (W) crystal to elucidate the effect of temperature and hydrogen atoms. Simulations were performed for two types of vacancy structures, $\mathrm{V}_9 + \mathrm{W}_1 + \mathrm{V}_9$ and $\mathrm{V}_{10} + \mathrm{W}_4 + \mathrm{V}_{10}$ ($\mathrm{V}_{n}$ means that a vacancy corresponds to the absence of $n$ W atoms, and $\mathrm{W}_{m}$ indicates that there are $m$ W atoms between two vacancies) in various cases of temperature and hydrogen atom concentration. Under the vacancy structure $\mathrm{V}_9 + \mathrm{W}_1 + \mathrm{V}_9$, we observed vacancy coalescence for all the cases of the temperature and the number of hydrogen atoms. Evaluating the potential energy required for removing one of the W atoms between two vacancies, we found that high temperature and existing hydrogen atoms in the vacancies facilitate vacancy coalescence, and that under the structure $\mathrm{V}_{10} + \mathrm{W}_4 + \mathrm{V}_{10}$, hydrogen atoms facilitate vacancy coalescence most strongly when the number is around 45 to 54 in each vacancy.

cond-mat.mtrl-sci↗

arXiv2012

Formation and Classification of Amorphous Carbon by Molecular Dynamics Simulation

By using molecular dynamics simulation, formation mechanisms of amorphous carbon in particular sp${}^3$ rich structure was researched. The problem that reactive empirical bond order potential cannot represent amorphous carbon properly was cleared in the transition process from graphite to diamond by high pressure and the deposition process of amorphous carbon thin films. Moreover, the new potential model which is based on electron distribution simplified as a point charge was developed by using downfolding method. As a result, the molecular dynamics simulation with the new potential could demonstrate the transition from graphite to diamond at the pressure of 15 GPa corresponding to experiment and the deposition of sp${}^3$ rich amorphous carbon.

cond-mat.mtrl-sci↗

arXiv2010

Molecular Dynamics Simulation of Chemical Vapor Deposition of Amorphous Carbon: Dependence on H/C Ratio of Source Gas

By molecular dynamics simulation, the chemical vapor deposition of amorphous carbon onto graphite and diamond surfaces was studied. In particular, we investigated the effect of source H/C ratio, which is the ratio of the number of hydrogen atoms to the number of carbon atoms in a source gas, on the deposition process. In the present simulation, the following two source gas conditions were tested: one was that the source gas was injected as isolated carbon and hydrogen atoms, and the other was that the source gas was injected as hydrocarbon molecules. Under the former condition, we found that as the source H/C ratio increases, the deposition rate of carbon atoms decreases exponentially. This exponential decrease in the deposition rate with increasing source H/C ratio agrees with experimental data. However, under the latter molecular source condition, the deposition rate did not decrease exponentially because of a chemical reaction peculiar to the type of hydrocarbon in the source gas.

cond-mat.mtrl-sci↗

arXiv2010

Hybrid Simulation between Molecular Dynamics and Binary Collision Approximation Codes for Hydrogen injection onto Carbon Materials

Molecular dynamics (MD) simulation with modified Brenner's reactive empirical bond order (REBO) potential is a powerful tool to investigate plasma wall interaction on divertor plates in a nuclear fusion device. However, MD simulation box's size is less than several nm for the performance of a computer. To extend the size of the MD simulation, we develop a hybrid simulation code between MD code using REBO potential and binary collision approximation (BCA) code. Using the BCA code instead of computing all particles with a high kinetic energy for every step in the MD simulation, considerable computation time is saved. By demonstrating a hydrogen atom injection on a graphite by the hybrid simulation code, it is found that the hybrid simulation code works efficiently in a large simulation box.

cond-mat.mtrl-sci↗

arXiv2010

Extension of the simulation code ACAT to treat real atomic positions

We have investigated plasma-surface interactions with molecular dynamics (MD) simulations. It, however, is high cost computation and is limited to simulations for materials of nanometer order. In order to overcome the limitation, a complementary model based on binary collision approximation (BCA) can be established. We employed a BCA-based simulation code ACAT and extended to handle any structure involving crystalline and amorphous. The extended code, named "ACaT", stores all positions of projectile and target atoms and velocities of recoil atoms, so it can be combined with the MD code. It also holds the potential to reproduce channeling phenomena. Thus it is expected to be useful for evaluation of channeling effects.

cond-mat.mtrl-sci↗

arXiv2009

Incident angle dependence of reactions between graphene and hydrogen atom by molecular dynamics simulation

Incident angle dependence of reactions between graphene and hydrogen atoms are obtained qualitatively by classical molecular dynamics simulation under the NVE condition with modified Brenner reactive empirical bond order (REBO) potential. Chemical reaction depends on two parameters, i.e., polar angle $θ$ and azimuthal angle $ϕ$ of the incident hydrogen. From the simulation results, it is found that the reaction rates strongly depend on polar angle $θ$. Reflection rate becomes larger with increasing $θ$, and the $θ$ dependence of adsorption rate is also found. The $θ$ dependence is caused by three dimensional structure of the small potential barrier which covers adsorption sites. $ϕ$ dependence of penetration rate is also found for large $θ$.

physics.plasm-ph↗