RPA corrections
The following instructions assume familiarity with the main programs of the pCI package.
The Random Phase Approximation (RPA) accounts for core-polarization corrections to radial matrix elements of external-field operators. When an external perturbation (e.g., an electromagnetic field) acts on the atom, it distorts the core electron cloud in addition to acting directly on the valence electrons. RPA captures this effect self-consistently by solving a set of linear equations for the induced change in each core orbital.
The corrections are computed for radial matrix elements of the following operators:
Label |
Operator |
|---|---|
|
Hyperfine constant A |
|
Hyperfine constant B |
|
Electric dipole (length form) |
|
Electric dipole (velocity form) |
|
Electric quadrupole |
|
Electric octupole |
|
Magnetic dipole |
|
Magnetic quadrupole |
|
Magnetic octupole |
|
Electric dipole moment |
|
Parity non-conserving amplitude |
|
Anapole moment |
|
Magnetic quadrupole moment |
Programs
rpa
Solves the RPA equations and writes corrected radial integrals for each operator. It runs in two modes selected interactively:
Core mode (
1) — solves the RPA equations self-consistently for the core shells and writes core-core matrix elements.Valence mode (
2) — computes RPA corrections for valence shells. This is the mode used in the typical workflow.
Each run produces one RPA_N.INT file per active operator, along with a summary in RPA.RES.
Input files:
MBPT.INP— defines the shell structure, active operators, and RPA optionsHFD.DAT— orbitals fromhfdCONF.DAT— CI wave functions (valence mode only)
Output files:
RPA.RES— printed summary of RPA correctionsRPA_N.INT— corrected radial integrals for each operator (one file per operator)
rpa_dtm
Substitutes the zeroth-order radial integrals in DTM.INT with the RPA-corrected integrals from the RPA_N.INT files produced by rpa. Run this after pdtm has generated DTM.INT.
Input files:
DTM.INT— transition matrix integrals frompdtmRPA_N.INT— corrected integrals fromrpa
Output files:
RPA_DTM.RES— summary of substituted integralsDTM.INT— updated in place with RPA-corrected integrals
Workflow
Run
pdtmto compute non-RPA radial integrals and produceDTM.INT.Run
rpain valence mode (enter2when prompted) to compute RPA-corrected radial integrals.Run
rpa_dtmto substitute the integrals inDTM.INTwith the RPA-corrected ones.
MBPT.INP configuration
The rpa program reads shell and operator settings from the MBPT and RPA blocks in MBPT.INP:
MBPT
Nso= n # number of core shells (CI core)
Nsh= n # last shell included in the self-consistent RPA field
Nss= n # last virtual shell for the RPA sum
Nsv= n # first valence shell (usually Nso+1)
Nmax= n # last valence shell for effective radial integrals
Lmax= n # maximum angular momentum for valence integrals
n # max multipolarity (not used)
Kt2= n # accuracy for Coulomb radial integrals
Kbrt= n # Breit corrections (0 = off, 1 = on)
RPA
n # A_hf (1 = compute, 0 = skip)
n # B_hf
n # E1_L
n # EDM
n # PNC
n # E1_V
n # ANM
n # MQM
n # M1
n # E2
n # E3
n # M2
n # M3
n # Nhf: number of core shells in the SCF procedure
n # Kmg: frequency type (0 = static)
n.nnnnnn # Omega: perturbation frequency (ignored if Kmg=0)
n # Kex: include exchange (0 = skip, 1 = include)