Premium
Exchange–correlation and temperature effects on plasmons in strongly correlated two‐dimensional electron systems: Finite‐temperature local‐field‐correction theory combined with angle‐resolved Raman spectroscopy
Author(s) -
Inaoka Takeshi,
Sugiyama Yosuke,
Sato Koji
Publication year - 2014
Publication title -
physica status solidi (b)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.51
H-Index - 109
eISSN - 1521-3951
pISSN - 0370-1972
DOI - 10.1002/pssb.201350147
Subject(s) - condensed matter physics , plasmon , physics , electron , raman spectroscopy , fermi gas , spectroscopy , local density approximation , electronic correlation , range (aeronautics) , fermi level , inverse , quantum mechanics , electronic structure , materials science , mathematics , geometry , composite material
Using the two types of the finite‐temperature ( T ) local‐field corrections, namely, the Singwi–Tosi–Land–Sjölander (STLS) and Hubbard (H) approximations, we investigate the exchange–correlation (XC) and T effects on plasmons (PLs) in strongly‐correlated two‐dimensional electron systems with ultralow density. In a density‐parameter range of 9.1 ≤ r s * ≤ 21.8 and in a Fermi‐ T normalized T range of 0.5 ≤ T / T F ≤ 8.4, we analyse those PL dispersions, which were observed extensively from single quantum wells by angle‐resolved Raman spectroscopy, but which have not been analysed theoretically so far. Even in our strongly correlated electron systems, the size of the XC holes can be well normalized by the inverse Fermi wave numberk F − 1. Close comparison with the experiment shows that the STLS approximation can be expected to give a quantitative description of the PLs in a smaller wave‐number ( q ) range of q ≲ k F , but that it overestimates the XC effect in a larger q range of q ≳ k F . The H approximation is insufficient to explain the remarkable XC effects. The conspicuous T dependence in the PL dispersion can be ascribed largely to the T dependence of the constituent electronic transitions of the PLs through the Fermi–Dirac distribution function.