| name | paper_tutorial_plasmon_heating |
| description | Dual-purpose tutorial to reproduce manuscript Figure 5b, 5d, and 5e plasmonic flux and HCN heating maps (EM-only, TLS, and ASE/BOMD), and to transfer the workflow to related plasmon-molecule systems by adjusting geometry, molecular model, and run procedures. |
Paper Tutorial: Plasmonic Flux and HCN Heating (Figure 5b, 5d, 5e)
Use this skill to reproduce the scoped Pt/Si plasmonic manuscript results for:
- Figure 5b: EM-only flux/absorption spectrum.
- Figure 5d: MaxwellLink + TLS heating map (
nmol=256).
- Figure 5e: MaxwellLink + ASE/Psi4 BOMD heating map (
nmol=256) plus TLS-vs-BOMD consistency.
Core Simulation Strategy
- Run all simulations and postprocessing in one dated runtime folder under
projects/YYYY-MM-DD-<scope>/.
- Recommended scope slug:
plasmon-heating.
- Copy
assets/implementation_2025/ into the runtime folder, then run three branches with fixed paper anchors:
aa=2.79, rr=1.11, nmol=256, y-polarized excitation, coupled propagation t=60.
- EM-only branch (
meep_plasmon_empty) for Figure 5b.
- TLS branch (
meep_plasmon_HCN_excitation_tls_strong) for Figure 5d.
- BOMD branch (
meep_plasmon_HCN_excitation_bomd_strong) for Figure 5e.
- Use trend-level agreement targets (peak positions and anisotropy structure), not bitwise trajectory identity.
- Figure 5f (RT-Ehrenfest) is intentionally out of scope.
Minimal Execution Recipes
Run commands from repository root.
- Create runtime workspace and stage assets:
RUN_DATE="${RUN_DATE:-$(date +%F)}"
RUN_SCOPE="${RUN_SCOPE:-plasmon-heating}"
RUN_DIR="projects/${RUN_DATE}-${RUN_SCOPE}"
mkdir -p "$RUN_DIR"
cp -R skills/paper_tutorial_plasmon_heating/assets/implementation_2025 "$RUN_DIR/"
- Figure 5b (EM-only spectrum):
cd "$RUN_DIR/implementation_2025/meep_plasmon_empty"
mkdir -p vac no_mol_with_dielectric
cp template/* vac/
cp template/* no_mol_with_dielectric/
cd vac
mpirun -np 128 python -u emitter.py -empty -aa 2.79 > flux0_a2.79.out
grep flux1: flux0_a2.79.out | cut -d , -f2- > flux0_a2.79.dat
cd ../no_mol_with_dielectric
mpirun -np 128 python -u emitter.py -dielectric -aa 2.79 -rr 1.11 -nmol 1 > flux_a2.79_r1.11.out
grep flux1: flux_a2.79_r1.11.out | cut -d , -f2- > flux_a2.79_r1.11.dat
- Figure 5d (TLS coupled map via Slurm templates):
cd "$RUN_DIR/implementation_2025/meep_plasmon_HCN_excitation_tls_strong"
mkdir -p nmol_256_with_dielectric
cp -R template/* nmol_256_with_dielectric/
cd nmol_256_with_dielectric
sbatch submit_all.sh
- Figure 5e (BOMD coupled map via Slurm templates):
cd "$RUN_DIR/implementation_2025/meep_plasmon_HCN_excitation_bomd_strong"
mkdir -p nmol_256_with_dielectric
cp -R template/* nmol_256_with_dielectric/
cd nmol_256_with_dielectric
sbatch submit_all.sh
- Build scoped figures and enforce acceptance checks:
cd "$RUN_DIR/implementation_2025/plotting"
python fig5bde_postprocess.py \
--mode all \
--base-dir .. \
--nmol 256 \
--figure-out fig5bde_scope.pdf \
--report-out fig5bde_metrics.json \
--strict
- Runtime guardrails:
- Do not run simulations inside
skills/paper_tutorial_plasmon_heating/.
- Do not write runtime output under
skills/.
Figure Routing
fig_001 (Figure 5b): EM-only Pt/Si plasmonic baseline from vacuum-normalized flux, reproducing a strong absorption peak near 3492 cm^-1 and high absorption around HCN stretch; playbook: skills/paper_tutorial_plasmon_heating/playbooks/fig_001_plasmon_flux_spectrum.md.
fig_002 (Figure 5d): MaxwellLink+TLS 16x16 molecular-lattice heating map with y-gap-dominant anisotropy and y-edge hotspot structure under the same EM geometry; playbook: skills/paper_tutorial_plasmon_heating/playbooks/fig_002_tls_heating_map.md.
fig_003 (Figure 5e): MaxwellLink+ASE/Psi4 BOMD heating map with stronger y-gap anisotropy and BOMD mean gain exceeding TLS under matched settings; playbook: skills/paper_tutorial_plasmon_heating/playbooks/fig_003_bomd_heating_map.md.
Beyond Manuscript Exploration
- Sweep
rr around 1.11 (for example 1.00-1.20) at fixed aa=2.79 to quantify peak shifts.
- Sweep perfect-square molecular grids (
nmol=64, 100, 144, 256) while preserving lattice placement.
- Sweep TLS dipole strength
mu12 around 0.15 and track anisotropy ratio sensitivity.
- Benchmark BOMD settings (
basis, memory, num_threads) only after baseline reproduction passes strict checks.
- Keep EM geometry and source window fixed when comparing TLS vs BOMD to preserve interpretability.