Coulomb Screening, Dielectric Response, and RPA Validity
Screening is the self-consistent rearrangement of charge induced by a perturbing potential. The random-phase approximation (RPA) is the geometric resummation of polarization bubbles that produces this rearrangement. It is exact for a specified noninteracting polarization inside that diagram class, but its accuracy for an interacting material depends on density, dimensionality, degeneracy, and the observable. Its controlled high-density correlation-energy use is demonstrated in Gell-Mann and Brueckner 1957, pp. 364–367.
Required background. Polarization, the Lindhard Function, and Particle–Hole Continua fixes the sign and normalization of .
Helpful background. Sources, Linear Response, and Kubo Formulae supplies the response and analytic-continuation framework.
Self-consistent potential and dielectric function
Section titled “Self-consistent potential and dielectric function”Let couple to number density and let be the repulsive interaction. In the convention , the induced Hartree potential is . Therefore
The RPA dielectric function, density response, and screened interaction are consequently
Diagrammatically, . This identity also fixes every sign: in the static limit , so repulsion makes the denominator larger than one and reduces the long-range potential.
Thomas–Fermi screening in three dimensions
Section titled “Thomas–Fermi screening in three dimensions”For Coulomb interactions in a background dielectric medium,
At and , . Hence
and
Fourier transformation gives a Yukawa potential . The approximation captures the long-wavelength screening length; keeping the full Lindhard function also produces the nonanalyticity and Friedel tail, which a constant misses.
What RPA sums and what it omits
Section titled “What RPA sums and what it omits”RPA sums direct ring diagrams built from bare or otherwise declared propagators and bare density vertices. It omits exchange diagrams, vertex corrections, and local-field factors. If dressed propagators are inserted without changing the vertex, conservation identities can be lost.
Controlled regimes include the high-density electron gas for leading correlation-energy logarithms, large flavor degeneracy with an appropriate interaction scaling, and long wavelengths where the Coulomb singularity enhances direct density fluctuations. These are regime statements, not a uniform error bound for every , , or observable. Short-range correlations, low density, and proximity to an instability generally require corrections.
For a neutral system, the Coulomb component is removed or cancelled by the uniform positive background. Failing to state that prescription creates a spurious divergent Hartree energy.
Analytic structure and measured loss
Section titled “Analytic structure and measured loss”Retarded screening uses analytic for . The energy-loss function
can contain both a collective pole and particle–hole background. A peak in is not by itself proof of an isolated pole: one must continue the inverse response and distinguish a zero of from a maximum created by a continuum edge.
Common pitfalls
Section titled “Common pitfalls”Using with a negative static polarization. That convention would antiscreen. Derive the denominator from the declared coupling rather than memorizing a sign.
Calling RPA generically controlled. Its leading terms are controlled only in specified limits. A small interaction parameter for one quantity need not control spectra near a continuum threshold.
Mixing background and vacuum Coulomb conventions. State whether and the neutralizing background have already been included in .
Exercises
Section titled “Exercises”Sum the bubble series
Section titled “Sum the bubble series”Starting from , derive the screened interaction for commuting scalar functions.
Solution
Factor out and recognize a geometric series: when the series converges. The retarded expression elsewhere follows by analytic continuation of this identity.
Check the screening sign
Section titled “Check the screening sign”A positive static test potential is applied to an electron gas. Explain why at small momentum.
Solution
The perturbation depletes number density, so . Thus , and the total potential is smaller than the bare potential. The induced charge opposes the test charge.
Continue
Section titled “Continue”Plasmons and Collective Charge Modes finds zeros of the dielectric function. Hedin Equations, Screened Interactions, and GW promotes to a self-consistent many-body object. Baym–Kadanoff Conservation and Validity separates conservation from quantitative accuracy.
References
Section titled “References”- Gell-Mann, Murray, and Keith A. Brueckner. “Correlation Energy of an Electron Gas at High Density.” Physical Review 106 (1957): 364–368. DOI.
Further reading
Section titled “Further reading”- Bohm, David, and David Pines. “A Collective Description of Electron Interactions: III. Coulomb Interactions in a Degenerate Electron Gas.” Physical Review 92 (1953): 609–625. DOI.
- Giuliani, Gabriele F., and Giovanni Vignale. Quantum Theory of the Electron Liquid. Cambridge: Cambridge University Press, 2005. DOI.
- Pines, David, and Philippe Nozières. The Theory of Quantum Liquids, Volume I: Normal Fermi Liquids. Boca Raton, FL: CRC Press, 2018; originally published 1966. Publisher record.