| name | method-dfpt |
| description | Use when a DFPT electron-phonon coupling and superconductivity calculation needs method-level route and tool selection — native PHonon/DFPT workflow for alpha2F(omega), lambda, omega_log, McMillan/Allen-Dynes Tc, q/k convergence. This skill owns the method understanding, method-vs-method routing (DFPT vs EPW), and the knob JUDGMENT. This is for QE PHonon workflows, not EPW. |
Method: DFPT (density-functional perturbation theory) electron-phonon coupling
Method-level skill for studying electron-phonon superconductivity via the native DFPT-EPC route. It owns the method understanding, method-vs-method routing, and the knob JUDGMENT. Literal keywords, values, templates, and run mechanics live in using-quantum-espresso.
Scope note. This skill is for QE's PHonon route (pw.x, ph.x, q2r.x,
matdyn.x, alpha2f.x). It is not an EPW workflow. Use it when the user
wants native QE electron-phonon coupling, lambda_tetra, alpha2F(omega),
lambda, omega_log, or McMillan/Allen-Dynes Tc. If the user asks for
Wannier interpolation, dense electron-phonon matrix interpolation, or EPW input
generation, switch to an EPW-specific workflow instead.
Operating principle
Branch-separated, evidence-first, convergence-qualified.
- Branch-separated - keep the normal IFC/phonon-dispersion branch separate
from the shifted-q EPC branch. Shared PHonon scratch is a common source of
false k-grid failures.
- Evidence-first - judge jobs from
run_steps.tsv, output tails, q-point
progress, and concrete QE error text, not from scheduler state alone.
- Convergence-qualified - report
Tc only with its q-grid, dense k-grid,
structure, imaginary-frequency status, and remaining convergence caveats.
The workflow spine
Use this sequence unless the active QE example or project convention gives a
validated reason to deviate:
pw.x scf
ph.x ph_ifc # unshifted IFC branch
q2r.x
matdyn.x
archive tmp/_ph0 from IFC branch if one outdir is reused
ph.x ph_dvscf_epc # shifted-q dVscf branch
ph.x elph_densek # lambda_tetra on dense electron k grid
alpha2f.x # serial post-processing in common QE 7.x builds
The per-stage input files, directory layout, and exact invocation mechanics live in using-quantum-espresso. -> using-quantum-espresso "The workflow spine".
Phase 0 - Preflight
Resolve the calculation context before writing or submitting anything:
If starting a new calculation, first walk through
using-quantum-espresso/references/dfpt/prepare-qe-todo.md and keep the completed copy or answers beside the
run tree.
- QE executables. Verify the QE version and paths for
pw.x, ph.x,
q2r.x, matdyn.x, and alpha2f.x. Confirm the MPI launcher expected by
the cluster or workstation.
- Structure. Use the relaxed structure that the project actually wants to
study. Check whether it is close to a phonon instability; EPC
Tc is not
meaningful if the relevant q grid has unresolved imaginary modes.
- Pseudopotentials and cutoffs. Keep the exchange-correlation functional
and pseudopotential family consistent across species. Copy or symlink UPF
files into
pseudo/; do not assume this skill ships them.
- Grid plan. Record the SCF k grid, EPC shifted q grid, dense electron k
grid, and at least one follow-up convergence comparison.
- Scratch policy. Inspect whether old
tmp/, _ph0/, .save/, .dyn*,
.elph*, .out, or .err files exist. Decide whether this is a fresh run,
a full rerun, or a post-processing continuation.
Phase 1 - Prepare inputs (knob JUDGMENT)
The literal input keywords, per-file SCF / IFC / EPC value patterns, and templates live in using-quantum-espresso. -> using-quantum-espresso "Phase 1 - Prepare inputs". Two of those bullets carry the method JUDGMENT rather than a literal value: the ASR-choice rationale (the IFC-branch bullet "asr / zasr selected according to the material and project convention.") and the dense-k restart rationale (the EPC-branch bullet "no recover = .true. in this dense-k lambda input;").
Convergence judgment
Do not call one parameter point final. At minimum compare:
- q-grid convergence at fixed dense k;
- dense-k convergence at fixed q grid;
- structural or lattice-constant sensitivity when the material is close to an
instability;
- direct QE values against integrated
alpha2F(omega) values.