Interaction-Range-Controlled Crossover from First-Order to Continuous Melting in Two-Dimensional Diplon Crystals — Code and Data

This deposit accompanies the article: P. S. Branicio, Interaction-Range-Controlled Crossover from First-Order to Continuous Melting in Two-Dimensional Diplon Crystals, APS Open Science (2026). https://doi.org/10.1103/r36h-bzv9 It contains the simulation/analysis/plotting code and the numerical data required to reproduce all four main-text figures, the four supplemental figures, and the tables of Supplemental Material Secs. S6–S9. The full bulk-trajectory data (>150 GB) is not deposited here; the ~850 MB included is sufficient to regenerate every figure and to verify every numerical claim in the article. Version 1.1.0 (19 August 2026) adds the material produced during peer review: the six analysis scripts behind Supplemental Secs. S6–S9 (melting-temperature criteria, equilibration/autocorrelation tests, random-seed replicas, Frank constant, dislocation core energy and latent-heat proxy), the three independent random-seed runs at d = 100 Å, N = 10,000, the corresponding analysis outputs, the --seed option of the production driver, and the η6 block-bootstrap and χ6(T) result files read by the figure scripts. The three large data tarballs (data-N2500.tar.gz, data-N10000.tar.gz, data-N40000.tar.gz) are unchanged from version 1.0.0. See README.md for the layout, reproduction instructions and version history. Abstract: Molecular dynamics simulations of two-dimensional diplon crystals (electrons on helium films) reveal a melting crossover from first-order to continuous as film thickness d grows relative to lattice constant a. At small d/a, hysteresis persists at N = 40,000 alongside a strong Binder-cumulant anomaly. At large d/a, hysteresis vanishes and orientational correlations decay algebraically with η6 = 0.31 ± 0.04, near the KTHNY bound 1/4, consistent with continuous melting within the investigated system sizes and numerical resolution; disclination unbinding limits any hexatic phase to ΔT/Tm < 0.5%. The diplon system thus tunes 2D melting, relevant to electrons-on-helium quantum computing. Included: Complete simulation code (MD engine, Ewald summation, GPU kernels, MPI) KTHNY analysis pipeline (g₆, g_G, defect classification, AIC fitting, bootstrap error bars) and the Secs. S6–S9 analysis scripts Energy time series for every (d, N, T) point with both heating and cooling protocols, in .npy format (2,000 samples per temperature) Per-temperature configuration snapshots (one final snapshot per T) used for Voronoi and g₆(r) calculations in Figs 1, 2 Pre-computed KTHNY observables (η₆, ξ₆, |ψ₆|, six-fold fraction, free dislocation and disclination counts) in .npz format Three independent random-seed replicas at d = 100 Å, N = 10,000 (Sec. S7) The .npz outputs of the Secs. S6–S9 analyses, the η₆ block-bootstrap results (Fig. 3(c) error bars), the |ψ₆| block statistics and the χ₆(T) cache of Fig. 2(d) Not included (available from the author on request): Full production trajectories (traj_T*/ with 400 snapshots per T) — ~12 GB per (d, N) combination, ~150 GB total Equilibration-monitor .cfg files (cfg_monitor/) — internal-use diagnostic only