Answers are grounded in this paper and cite their sources.
Based on a nonperturbative scheme to determine the self-energy Σ(k,iωₙ) with automatically satisfying the Ward identity and the total-momentum conservation law, a fully self-consistent calculation is done in the electron gas at various temperatures T to obtain G(k,iωₙ) the one-particle Green's function with fulfilling all known conservation laws, sum rules, and correct asymptotic behaviors; here, T is taken unprecedentedly low, namely, kBT/εF down to 10^-4 with εF the Fermi energy, and tiny mesh Δk as small as 10^-4kF is chosen near the Fermi surface in k space with kF the Fermi momentum. By analytically continuing G(k,iωₙ) to the retarded function GR(k,ω), we find a novel low-energy peak, in addition to the quasiparticle (QP) peak and one- and two-plasmon high-energy satellites, in the spectral function A(k,ω)[=-ImGR(k,ω)/π] for kBT10^-3εF in the simple-metal density region (2<rₛ<6 with rₛ the dimensionless density parameter). This new peak is attributed to the effect of excitonic attraction on Σ(k,iωₙ) arising from multiple excitations of tightly bound electron-hole pairs in the polarization function Π(q,iωq) for |q|≈2kF and |ωq|εF and thus it is dubbed ``excitron.'' Although this excitron peak height is only about a one-hundredth of that of QP, its excitation energy is about a half of that of QP for |k|≈kF, seemingly in contradiction to the Landau's hypothesis as to the one-to-one correspondence of low-energy excitations between a free Fermi gas and an interacting normal Fermi liquid. As for the QP properties, our results of both the effective mass m* and the renormalization factor z* are in good agreement with those provided by recent quantum Monte Carlo simulations and available experiments.
No takes yet. Share an insight, caveat, or question.
Y. Takada (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: