Abstract
By using the quantum maximum entropy principle we formally derive, from a underlying kinetic description, isothermal (hydrodynamic and diffusive) quantum fluid equations for particles with Fermi-Dirac and Bose-Einstein statistics. A semiclassical expansion of the quantum fluid equations, up to \(\mathcal{O}(\hbar^{2})\)-terms, leads to classical fluid equations with statistics-dependent quantum corrections, including a modified Bohm potential. The Maxwell-Boltzmann limit and the zero temperature limit are eventually discussed.
Similar content being viewed by others
Notes
Reference [11] is a partial exception, because the fully-quantum model is deduced for a generic entropy. However, the semiclassical expansion assumes Boltzmann entropy.
Note that we have not to rescale the Wigner functions w and \(w_{w_{\mathrm{eq}}}\) because we are already using dimensionless Wigner functions (see Remark 2.1).
A rigorous proof of existence and uniqueness of the constrained minimization problem has been recently obtained by Méhats and Pinaud [26] for the one-dimensional case with periodic boundary conditions.
The proof given in Ref. [11], however, is still more general because the density-operator formalism covers the case of a system confined in a domain Ω∈ℝd, while the Wigner formalism is valid only in the whole-space case.
In the Bose-Einstein case, these models are only suited to describe the non-condensate phase; we shall discuss this point later on, in the semiclassical framework (see Proposition 3.2 and Remark 3.2).
Actually, in Ref. [11] the factor 1/T seems to be missing.
References
Arnold, A.: Self-consistent relaxation-time models in quantum mechanics. Commun. Partial Differ. Equ. 21(3–4), 473–506 (1996)
Barletti, L., Frosali, G.: Diffusive limit of the two-band k⋅p model for semiconductors. J. Stat. Phys. 139(2), 280–306 (2010)
Barletti, L., Méhats, F.: Quantum drift-diffusion modeling of spin transport in nanostructures. J. Math. Phys. 51(5), 053304 (2010), 20 pp.
Bohm, D.: A suggested interpretation of the quantum theory in terms of “hidden variables” I. Phys. Rev. 85, 166–179 (1952)
Bohm, D.: A suggested interpretation of the quantum theory in terms of “hidden variables” II. Phys. Rev. 85, 180–193 (1952)
Bourgade, J.P., Degond, P., Méhats, F., Ringhofer, C.: On quantum extensions to classical spherical harmonics expansion/Fokker-Planck models. J. Math. Phys. 47(4), 043302 (2006), 26 pp.
Brull, S., Méhats, F.: Derivation of viscous correction terms for the isothermal quantum Euler model. Z. Angew. Math. Mech. 90(3), 219–230 (2010)
Dalfovo, F., Giorgini, S., Pitaevskii, L.P., Stringari, S.: Theory of Bose-Einstein condensation in trapped gases. Rev. Mod. Phys. 71(3), 463–512 (1999)
Degond, P., Ringhofer, C.: Quantum moment hydrodynamics and the entropy principle. J. Stat. Phys. 112(3–4), 587–628 (2003)
Degond, P., Méhats, F., Ringhofer, C.: Quantum energy-transport and drift-diffusion models. J. Stat. Phys. 118(3–4), 625–667 (2005)
Degond, P., Gallego, S., Méhats, F.: Isothermal quantum hydrodynamics: derivation, asymptotic analysis, and simulation. Multiscale Model. Simul. 6(1), 246–272 (2007)
Dürr, D., Teufel, S.: Bohmian Mechanics. The Physics and Mathematics of Quantum Theory. Springer, Berlin (2009)
Ferry, D.K., Zhou, J.-R.: Form of the quantum potential for use in hydrodynamic equations for semiconductor device modeling. Phys. Rev. B 48(11), 7944–7950 (1993)
Folland, G.B.: Harmonic Analysis in Phase Space. Princeton University Press, Princeton (1989)
Gallego, S., Méhats, F.: Numerical approximation of a quantum drift-diffusion model. C. R. Math. Acad. Sci. Paris 339(7), 519–524 (2004)
Gallego, S., Méhats, F.: Entropic discretization of a quantum drift-diffusion model. SIAM J. Numer. Anal. 43(5), 1828–1849 (2005)
Gasser, I., Markowich, P.A.: Quantum hydrodynamics, Wigner transforms and the classical limit. Asymptot. Anal. 14(2), 97–116 (1997)
Jüngel, A.: Transport Equations for Semiconductors. Springer, Berlin (2009)
Jüngel, A.: Dissipative quantum fluid models. Riv. Mat. Univ. Parma 3(2), 217–290 (2012)
Jüngel, A., Matthes, D.: A derivation of the isothermal quantum hydrodynamic equations using entropy minimization. Z. Angew. Math. Mech. 85(11), 806–814 (2005)
Jüngel, A., Matthes, D., Milišić, J.P.: Derivation of new quantum hydrodynamic equations using entropy minimization. SIAM J. Appl. Math. 67(1), 46–68 (2006)
Jüngel, A., Krause, S., Pietra, P.: Diffusive semiconductor moment equations using Fermi-Dirac statistics. Z. Angew. Math. Phys. 62(4), 623–639 (2011)
Levermore, C.D.: Moment closure hierarchies for kinetic theories. J. Stat. Phys. 83(5–6), 1021–1065 (1996)
Lewin, L.: Polylogarithms and Associated Functions. North Holland, New York (1981)
Madelung, E.: Quantentheorie in hydrodynamischer Form. Z. Phys. 40, 322–326 (1926)
Méhats, F., Pinaud, O.: An inverse problem in quantum statistical physics. J. Stat. Phys. 140, 565–602 (2010)
Trovato, M., Reggiani, L.: Quantum maximum entropy principle for a system of identical particles. Phys. Rev. E 81, 021119 (2010), 11 pp.
Trovato, M., Reggiani, L.: Quantum maximum-entropy principle for closed quantum hydrodynamic transport within a Wigner function formalism. Phys. Rev. E 84, 061147 (2011), 29 pp.
Von Neumann, J.: Mathematical Foundations of Quantum Mechanics. Princeton University Press, Princeton (1955)
Wigner, E.: On the quantum correction for thermodynamic equilibrium. Phys. Rev. 40, 749–759 (1932)
Wood, D.C.: The computation of polylogarithms. University of Kent computing laboratory, Technical report 15/92 (1992)
Zachos, C.K., Fairlie, D.B., Curtright, T.L. (eds.): Quantum Mechanics in Phase Space. An Overview with Selected Papers. World Scientific Series in 20th Century Physics, vol. 34. World Scientific, Hackensack (2005)
Zamponi, N.: Some fluid-dynamic models for quantum electron transport in graphene via entropy minimization. Kinet. Relat. Models 5(1), 203–221 (2012)
Acknowledgements
This work was partially supported by the Italian Ministry of University, PRIN “Mathematical problems of kinetic theories and applications”, prot. 2009NAPTJF_003.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Moments of Fermi and Bose Distributions and Related Integrals
We recall that the polylogarithm of order s, with s∈ℝ, is defined in the complex unit disc by the power series
and can be analytically continued to a larger domain (depending on s). To our purposes it will be enough to know that \(\operatorname {\mathrm {Li}}_{s}(z)\) is always well defined, real-valued and regular for z∈(−∞,1), and
where ζ is the Riemann zeta function. The polylogarithms are strictly connected with the moments of FD and BE distributions.
Definition A.1
For λez>−1, \(\lambda \not= 0\), and s∈ℝ we define
where \(\operatorname {\mathrm {Li}}_{s}\) is the polylogarithm of order s. From known identities [24] we have that, for s>0, the above definition is equivalent to
(known as Fermi integral).
Note that ϕ s (z) is defined for all z∈ℝ if λ>0, and for z<−log|λ| if λ<0. In particular, in the FD case, λ=1, ϕ s (z) is defined on the whole real line while in the BE case, λ=−1, only for z<0.
The following properties of the functions ϕ s can be easily deduced from the properties of polylogarithms (see e.g. Refs. [24, 31]):
Here, “f(x)∼g(x) for x→y” means lim x→y f(x)/g(x)=1.
Starting from the identity (A.3), we shall now compute explicit expressions, in terms of the functions ϕ s , of all the kinds of integrals that have been encountered throughout this paper.
Lemma A.1
For λez>−1, k=1,2,3,… and s>0, let us consider the integrals
Then, \(I_{1}^{s}\) is given by Eq. (A.3) and higher values of k are recursively obtained by
In particular (omitting the argument z),
Proof
The recursive formula (A.6) follows immediately from a formal derivation of \(I_{k}^{s}(z)\) with respect to z; Eq. (A.7) follows from (A.3) and (A.6) by using the property (A.4e). □
Equation (A.7) suggests that we can look for an expression for \(I_{k}^{s}\) of this kind:
where \(c^{k}_{j}\) are numerical coefficients (independent on s) to be determined. Since
(as it is apparent from Lemma A.1), from the recursive relation (A.6) we can write the equivalent relation
Inserting (A.8) into (A.10) yields
Equating the coefficients of ϕ s (j=0) we obtain \(c^{k+1}_{0} = c^{k}_{0}\), and then (since \(c^{1}_{0} = 1\), as follows form (A.8) with k=1)
equating the coefficients of ϕ s−k (j=k) we obtain \(c^{k+1}_{k} = - \frac{1}{k} c^{k}_{k-1}\), and then
finally, equating the coefficients of ϕ s−j , with 1≤j≤k−1, we obtain
By using the recursive relations (A.11a)–(A.11c) one can easily generate all the coefficients of the expansion (A.8).
Proposition A.1
Let \(I_{k}^{s}(z)\) be given as in the previous lemma. Then,
where, as usual, \(n_{d} := (2\pi T)^{\frac{d}{2}}\).
Proof
By using polar coordinates it is easily shown that
This formula immediately yields Eq. (A.12a) and, by obvious symmetry considerations, Eq. (A.12b). The derivation of Eq. (A.12c) requires more explanations. We first show that
(which is of course independent on i). We proceed by double induction on d. The cases d=1,2 can be easily verified by direct computations. Then, assuming (A.14) to be valid for d, for d+2 we can write
From Eq. (A.9) we obtain therefore
which proves (A.14) by induction. On the other hand, Eq. (A.13) yields
and then, using \({\vert {p} \vert }^{4} = ( \sum_{i} p_{i}^{2} )^{2} = \sum_{i} p_{i}^{4} + \sum_{i\not=j} p_{i}^{2}p_{j}^{2}\) (and symmetry considerations), we obtain
for \(i \not= j\). The two cases, (A.14) and (A.15), are summarized by (A.12c). □
Appendix B: Postponed Proofs
2.1 B.1 Proof of Proposition 2.1
According to the short notations adopted from Sect. 3.1 on, let us denote \(\mathcal {G}_{A,B}\) by \(\mathcal {G}\) and \(\frac{1}{T}h_{A,B}\) by h (definition (3.1)). Recalling, moreover, the formalism introduced in Sect. 2.1, we can write
where we used the fact that \(\operatorname {Op}_{\epsilon }(\mathcal {G})\) is, by definition (2.31), a function of \(\operatorname {Op}_{\epsilon }(h )\) and then \(\{h,\mathcal {G}\}_{\#}= 0\), because it is the inverse Weyl quantization of a vanishing commutator. Since, from a direct computation,
then from the previous identity we have
(where Eq. (2.20) was used), which yields Eq. (2.32). □
2.2 B.2 Proof of Lemma 3.1
Lemma 3.1 follows from elementary manipulations of formal Taylor expansions. In order to shorten the notations, let us introduce the (d+1)-dimensional vectors
The constraint system (3.6), in these notations, reads as follows:
Note that f has a double dependence on ϵ: one is direct (which leads to the expansion (3.2), i.e. to the terms f (k)), and the other is through the Lagrange multipliers, which are expanded as μ=μ (0)+ϵμ (1)+ϵ 2 μ (2)+⋯. Then, we regard f as f(ϵ,μ(ϵ)), whose Taylor expansion at ϵ=0 can be written in this way:
(where we took into account that f (1)=0, from (3.4b), and, for the sake brevity, the third-order term was not shown). Since the moments m do not depend on ϵ, the constraint equation (B.1) is expanded as follows:
The first equation is Eq. (3.8a), the second one implies μ (1)=0 (i.e. A (1)=B (1)=0), the third one is Eq. (3.8b) and, finally, the fourth one (which has been omitted for brevity) implies μ (3)=0 (i.e. A (3)=B (3)=0). □
2.3 B.3 Proof of Lemma 3.2
By using Eq. (3.4c) with n=1 we obtain
The first term contains \(\operatorname {\mathcal {E}\!\mathit {xp}}_{2}(h)\), whose explicit expression can be taken from Eq. (5.14) of Ref. [10] and reads as follows:
where we introduced the following notations:
(we recall that \(h = \frac{1}{T} h_{A,B}\) is given by (3.1)). The other term, \(\mathcal {G}_{0}\,\#_{2}\,\mathrm {e}^{h}\), is a second-order Moyal product, given by (2.5). Using the notations just introduced, we obtain:
where, for k≥0, we define
Putting together the two terms we obtain the following expression for \(\mathcal {G}_{2}\):
Note that in the Maxwell-Boltzmann case, λ=0, we have F k =e−h for all k≥0 and, therefore,
Then, in this case, Eq. (B.5) reduces (as it should) to \(\operatorname {\mathcal {E}\!\mathit {xp}}_{2}(-h)\), which is given by (B.3) with the suitable changes of sign.
Now, defining q=p−B, taking the moments \({\langle \mathcal {G}_{2} \rangle }\) and \({\langle q_{i} \mathcal {G}_{2} \rangle }\) from Eq. (B.5), and taking account of vanishing integrals of odd functions, we get
and
The moments of functions F k can be reduced to integrals of type \(I_{k}^{s}\) (see Lemma A.1) by using
Recalling that
from (B.8) and (A.12a)–(A.12c) we obtain
Then, by using \({\langle p_{i}\mathcal {G}_{2} \rangle } = B_{i}{\langle \mathcal {G}_{2} \rangle } + {\langle q_{i}\mathcal {G}_{2} \rangle }\) and the identity
(from property (A.4e)), Eqs. (3.12a)–(3.12b) are easily obtained from the expressions (B.6) and (B.7). □
Rights and permissions
About this article
Cite this article
Barletti, L., Cintolesi, C. Derivation of Isothermal Quantum Fluid Equations with Fermi-Dirac and Bose-Einstein Statistics. J Stat Phys 148, 353–386 (2012). https://doi.org/10.1007/s10955-012-0535-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10955-012-0535-5