24.12.2012 Views

5 Hirsch-Fye quantum Monte Carlo method for ... - komet 337

5 Hirsch-Fye quantum Monte Carlo method for ... - komet 337

5 Hirsch-Fye quantum Monte Carlo method for ... - komet 337

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Hirsch</strong>-<strong>Fye</strong> QMC 5.23<br />

Note the convergence factoreiωnη which is essential in order to get the correct result. Obviously,<br />

such a term is difficult to handle numerically; in practice, η may be replaced, e.g., by the time<br />

discretization parameter ∆τ, which also determines the cutoff frequency in the infinite sum.<br />

This approximation can be avoided by evaluating the noninteracting part separately. For the<br />

kinetic energy, this implies<br />

Ekin = lim<br />

η→0 +2T<br />

∞�<br />

e iωnη<br />

� ∞<br />

1<br />

dǫǫρ(ǫ)<br />

(43)<br />

iωn −ǫ+µ−Σ(iωn)<br />

� ∞<br />

= 2<br />

−∞<br />

� ∞<br />

≈ 2<br />

−∞<br />

n=−∞<br />

−∞<br />

ǫρ(ǫ)<br />

dǫ<br />

eβ(ǫ−µ) +1 +2T<br />

ǫρ(ǫ)<br />

dǫ<br />

eβ(ǫ−µ) +1 +2T<br />

∞�<br />

n=−∞<br />

L/2 �<br />

� ∞<br />

n=−L/2+1<br />

dǫ ǫρ(ǫ)<br />

−∞<br />

� Gǫ(iωn)−G 0 ǫ<br />

(iωn) �<br />

(44)<br />

� ∞<br />

dǫ ǫρ(ǫ) � Gǫ(iωn)−G 0 ǫ(iωn) � , (45)<br />

where we have assumed the paramagnetic case. Here, the interacting and noninteracting “momentumdependent”<br />

Green functions read<br />

Gǫ(iωn) =<br />

−∞<br />

1<br />

iωn −ǫ+µ−Σ(iωn) ; G0 ǫ (iωn) =<br />

1<br />

. (46)<br />

iωn −ǫ+µ<br />

In (45), the terms in the Matsubara sum fall off at least as1/ω 2 , which makes it well-defined also<br />

without convergence factor. At the same time, the truncation error is reduced significantly. The<br />

complementary ingredient to the energy is the double occupancyD withE = Ekin+UD. In the<br />

context of QMC calculations, this observable is best calculated directly from Wick’s theorem<br />

(i.e. as 〈ni↑ni↓〉 or the corresponding expression in Grassmann variables) when sampling over<br />

the auxiliary field. The overall behavior of D and E in the Hubbard model can be read off (<strong>for</strong><br />

T = 0.1) from the middle and lower parts of Fig. 13, respectively. For small U, the kinetic<br />

energy increases quadratically while D and, consequently, Epot and E increase linearly. The<br />

potential energy reaches a maximum below U = 3. A region of strong curvature of D, Epot,<br />

and Ekin near U = 4.6 gives a rough indication of the metal-insulator crossover. The total<br />

energyE, however, hardly shows any anomalies at this scale. Note also that the solutions <strong>for</strong>D<br />

and E are quite close to the results of plain zero-temperature second-order perturbation theory.<br />

The agreement actually becomes better <strong>for</strong> low-temperature QMC data, extrapolated to T → 0<br />

(not shown). The offset of the curves <strong>for</strong> E gives (<strong>for</strong> not too large U) an indication of the<br />

specific heatcV = dE/dT which is linear within the Fermi-liquid phase and is, in general, best<br />

evaluated by fitting the temperature dependence ofE. 18<br />

Susceptibilities<br />

The direct evaluation of the compressibility (as of most other susceptibilities) is numerically<br />

costly since it requires the QMC computation of 2-particle vertex functions. The <strong>for</strong>malism is<br />

omitted here; it can be found, e.g., in [1, 46].<br />

18 In Fermi liquid phases, the linear coefficient γ of the specific heat may also be obtained via the quasiparticle<br />

weightZ (extrapolated toT = 0).

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!