The upper bound on the ground state energy of the two-component charged Bose gas derived in \[Comm. Math. Phys. 266 (2006), 797–818] is extended to the more general case, where the mass of the positive bosons can be different from the mass of the negative bosons. The analysis in principle is quite similar, however the formerly scalar problem becomes two-dimensional (one dimension for the
+ and one for the
− case). This leads to a non-linear
2×2-matrix minimization problem. Its minimizer is calculated and results in a non-linear mass-dependent coefficient in the upper bound. These new results agree in the equal-mass case with the formerly known results. It is also discussed how the result here interpolates between the equal mass case and the case where one mass is infinite.
We consider a system of N bosons in a unitary box in the grand-canonical setting interacting through a potential with the scattering length scaling as N−1+κ, κ ∈ (0, 2/3). This regimes interpolate between the Gross–Pitaevskii regime (κ = 0) and the thermodynamic limit (κ = 2/3). In the work of Basti et al. [Forum Math., Sigma 9, E74 (2021)], as an intermediate step to prove an upper bound in agreement with the Lee–Huang–Yang formula in the thermodynamic limit, a second order upper bound on the ground state energy for κ < 5/9 was obtained. In this paper, thanks to a more careful analysis of the error terms, we extend the mentioned result to κ < 7/12.
When calculating the ground-state energy of a weakly interacting Bose gas with the help of the customary contact pseudopotential, one meets an artifical ultraviolet divergence which is caused by the incorrect treatment of the true interparticle interactions at small distances. We argue that this problem can be avoided by retaining the actual, momentum-dependent interaction matrix elements, and use this insight for computing both the ground-state energy and the depletions of a binary Bose gas mixture. Even when considering the experimentally relevant case of equal masses of both species, the resulting expressions are quite involved, and they are no straightforward generalizations of the known single-species formulas. On the other hand, we demonstrate in detail how these latter formulas are recovered from our two-species results in the limit of vanishing interspecies interaction.
Consider
N bosons in a finite box $\ensuremath{\Lambda}={[0,L]}^{3}\ensuremath{\subset}{\mathbb{R}}^{3}$ interacting via a two-body non-negative soft potential $V=\ensuremath{\lambda}\stackrel{\ifmmode \tilde{}\else \~{}\fi{}}{V}$ with $\stackrel{\ifmmode \tilde{}\else \~{}\fi{}}{V}$ fixed and $\ensuremath{\lambda}>0$ small. We will take the limit $L,N\ensuremath{\rightarrow}\ensuremath{\infty}$ by keeping the density $\ensuremath{\varrho}=N∕{L}^{3}$ fixed and small. We construct a variational state, which gives an upper bound on the ground-state energy per particle $\ensuremath{\epsilon}$, $\ensuremath{\epsilon}\ensuremath{\leqslant}4\ensuremath{\pi}\ensuremath{\varrho}a[1+(128∕15\sqrt{\ensuremath{\pi}}){(\ensuremath{\varrho}{a}^{3})}^{1∕2}{S}_{\ensuremath{\lambda}}]+O({\ensuremath{\varrho}}^{2}\ensuremath{\mid}\mathrm{ln}\phantom{\rule{0.2em}{0ex}}\ensuremath{\varrho}\ensuremath{\mid})$, as $\ensuremath{\varrho}\ensuremath{\rightarrow}0$, with a constant satisfying $1\ensuremath{\leqslant}{S}_{\ensuremath{\lambda}}\ensuremath{\leqslant}1+C\ensuremath{\lambda}$. Here
a is the scattering length of
V and thus depends on $\ensuremath{\lambda}$. In comparison, the prediction by Lee and Yang [Phys. Rev. 105, 1119 (1957)] and Lee, Huang, and Yang [Phys. Rev. 106, 1135 (1957)] asserts that ${S}_{\ensuremath{\lambda}}=1$ independent of $\ensuremath{\lambda}$.
A simple upper bound to the ground-state energy of a quantal system composed of
N equal particles obeying Boltzmann or Bose statistics and interacting through the interparticle potential
V(r) is established. This bound applies in the limit of large
N; different expressions obtain depending on whether the potential is or is not singularly attractive at the origin. In the former case, the bound reads - $C{N}^{\frac{(4\ensuremath{-}q)}{(2\ensuremath{-}q)}}$, -
q being the (negative) exponent characterizing the behavior of the pair potential at the origin through ${\mathrm{lim}}_{r\ensuremath{\rightarrow}0}[{r}^{q}V(r)]={V}_{0}$, $\ensuremath{-}\ensuremath{\infty}<{V}_{0}<0$ (of course,
q<2 to prevent two-body collapse); in the latter case, it reads $\ensuremath{-}B{N}^{2}$. Explicit expressions for the constants
C and
B in terms of the pair potential
V(r) are obtained; it is expected (and, in some cases, demonstrated) that the associated upper bound to the ground-state energy of the system approximates closely the actual value. It is also noted that, if for some non-negative value of
p the quantity $\ensuremath{\int}{0}^{\ensuremath{\infty}}\mathrm{dr}{r}^{2} \mathrm{exp}(\ensuremath{-}{p}^{2}{r}^{2})V(r)$ is negative, the
N-body system is unstable in the sense that, at large
N, its total energy is negative and increases in modulus at least as
N2 (so that the binding energy per particle increases in modulus at least as
N). Examples illustrating the nontrivial nature of this conclusion are displayed.
Now that the low temperature properties of quantum-mechanical many-body systems (bosons) at low density,
ρ, can be examined experimentally it is appropriate to revisit some of the formulas deduced by many authors 4-5 decades ago. For systems with repulsive (i.e. positive) interaction potentials the experimental low temperature state and the ground state are effectively synonymous -- and this fact is used in all modeling. In such cases, the leading term in the energy/particle is
2πℏ2aρ/m where
a is the scattering length of the two-body potential. Owing to the delicate and peculiar nature of bosonic correlations (such as the strange
N7/5 law for charged bosons), four decades of research failed to establish this plausible formula rigorously. The only previous lower bound for the energy was found by Dyson in 1957, but it was 14 times too small. The correct asymptotic formula has recently been obtained by us and this work will be presented. The reason behind the mathematical difficulties will be emphasized. A different formula, postulated as late as 1971 by Schick, holds in two-dimensions and this, too, will be shown to be correct. With the aid of the methodology developed to prove the lower bound for the homogeneous gas, two other problems have been successfully addressed. One is the proof by us that the Gross-Pitaevskii equation correctly describes the ground state in the `traps' actually used in the experiments. For this system it is also possible to prove complete Bose condensation, as we have shown. Another topic is a proof that Foldy's 1961 theory of a high density Bose gas of charged particles correctly describes its ground state energy.