Barra and Moro (Annals of Physics 359, 290 (2015)) proposed an exact finite-NN mean-field partition function for the van der Waals fluid: a one-dimensional integral over the system volume, ZN=∫b∞eN(xv+ta/v+log⁡(v−b)) dvZ_N=\int_b^\infty e^{N(xv+ta/v+\log(v-b))}\,dv, which reproduces the classical van der Waals isotherms away from criticality, encodes the Maxwell equal-area rule in the thermodynamic limit, and resolves the critical singularity through a Pearcey-integral scaling of the volume at order N−1/4N^{-1/4}. The construction rests on an unproven mean-field ansatz — that all configurations of equal convex-hull volume are equally weighted — with the interaction and entropy functions ϕ(v)\phi(v) and σ(v)\sigma(v) fixed by matching to the macroscopic equation of state rather than derived from a microscopic two-body potential. No resolution is known in the literature to date: for an explicitly specified microscopic Hamiltonian class, either prove the volume-measure reduction with error bounds uniform through the critical region, deriving ϕ\phi and σ\sigma from the potential and establishing or correcting the Pearcey scaling, or prove that the reduction fails at the critical scale.

Background

The van der Waals (vdW) equation of state, (P+a/v2)(v−b)=nRT(P+a/v^2)(v-b)=nRT, augments the ideal-gas law with a hard-core excluded volume bb and an attractive pressure correction a/v2a/v^2 (PP pressure, vv volume, TT temperature, nn the number of moles, RR the gas constant). Below a critical temperature its isotherms develop an unphysical loop — a region where the volume increases with pressure. The repair is the Maxwell equal-area rule: the loop is replaced by a horizontal tie line placed so that the two areas enclosed between the isotherm and the tie line are equal, expressing liquid–gas coexistence at a single pressure. The rule is a thermodynamic consistency condition, not an output of the equation of state, and its reconciliation with a finite-NN statistical-mechanical description — in which the partition function is analytic and no loop can occur — is a classical problem of mean-field theory.

Barra and Moro (2015) proposed an answer. They consider NN particles interacting through a two-body potential and coupled to an external pressure field through the volume of the configuration, defined as the minimum convex hull of the particle positions (the smallest convex set containing them). In the rescaled thermodynamic variables x=−P/(NKBT)x=-P/(NK_B T) and t=1/(NKBT)t=1/(NK_B T) (KBK_B is Boltzmann's constant; note x=−Ptx=-Pt), they postulate a mean-field measure: all configurations of equal volume are equally weighted, so that the configurational partition function collapses to a one-dimensional integral

ZN=∫b∞eN(xv+t2ϕ(v)+σ(v)) dv,Z_N = \int_b^{\infty} e^{N\left(xv+\frac{t}{2}\phi(v)+\sigma(v)\right)}\,dv,

where vv is the system volume with hard-core floor bb, ϕ(v)\phi(v) collects the interaction energy as a function of volume, and σ(v)\sigma(v) is interpreted as the configurational entropy (the logarithm of the number of configurations of prescribed volume). Requiring that the saddle-point evaluation of this integral — the Laplace method, i.e. the standard asymptotic technique for integrals with a large parameter NN, dominated by the minima of the exponent — reproduce the classical vdW equation away from the critical region fixes the functions uniquely as ϕ(v)=2a/v\phi(v)=2a/v and σ(v)=log⁡(v−b)\sigma(v)=\log(v-b). The resulting ZNZ_N satisfies a linear hyperbolic equation of Klein–Gordon type in the characteristic variables, ∂2Z/∂x ∂t=N2aZ\partial^2 Z/\partial x\,\partial t=N^2 a Z. At finite NN it is analytic; as N→∞N\to\infty (the thermodynamic limit) the isotherms behave like nonlinear waves that develop a gradient catastrophe at the vdW critical point (xc,tc,vc)=(−1/(8b), 27b/(8a), 3b)(x_c,t_c,v_c)=(-1/(8b),\,27b/(8a),\,3b) and steepen into shock discontinuities whose Rankine–Hugoniot condition (the standard jump condition relating the speed of a discontinuity to the jump of the flux across it) reproduces the Clausius–Clapeyron relation (the thermodynamic relation giving the slope of the coexistence curve from the latent heat), and whose equal-area resonance condition is exactly the Maxwell rule. Near the critical point the finite-NN volume obeys the scaling

v−vc=N−1/4 u ⁣(x−xc+a(t−tc)/vc2N−3/4, t−tcN−1/2),u=−2 ∂ξlog⁡Λ(ξ,τ),v-v_c = N^{-1/4}\,u\!\left(\frac{x-x_c+a(t-t_c)/v_c^2}{N^{-3/4}},\,\frac{t-t_c}{N^{-1/2}}\right),\qquad u=-2\,\partial_\xi\log\Lambda(\xi,\tau),

where Λ(ξ,τ)=∫−∞∞e−(z4−2τz2+4ξz)/8 dz\Lambda(\xi,\tau)=\int_{-\infty}^{\infty}e^{-(z^4-2\tau z^2+4\xi z)/8}\,dz is the Pearcey integral, a canonical cusp-catastrophe special function governing universal asymptotic behavior when two saddle points of an integral coalesce; this scaling analysis was developed in Moro's earlier work on the shock dynamics of phase diagrams.

The construction is, however, an ansatz. The reduction of the many-body interaction energy to a volume function ϕ(v)\phi(v) is supported only by a heuristic nearest-neighbour argument (particles assumed on average equidistant), and ϕ\phi, σ\sigma are fixed by matching to the macroscopic equation of state rather than computed from the microscopic two-body potential ψ\psi. The rigorous theory of the vdW liquid–gas transition works differently: van Kampen (1964) and Lebowitz and Penrose (1966) proved, for Kac-type potentials — weak, long-range attractions of the form γdψ(γr)\gamma^d\psi(\gamma r) whose range γ−1\gamma^{-1} is sent to infinity after the thermodynamic limit — that the exact free energy converges to the vdW–Maxwell construction; Kac, Uhlenbeck and Hemmer (1963) solved the one-dimensional version exactly in the same limit. These results concern the limiting free energy: they give no finite-NN partition-function representation, no finite-NN critical scaling, and no error estimates. In the Curie–Weiss model — the fully connected mean-field Ising spin model — the critical finite-size scaling (magnetization per spin of order N−1/4N^{-1/4} with a quartic-exponential limit law) is a rigorous theorem of Ellis and Newman (1978). The fluid analogue for the Barra–Moro volume-measure representation has no counterpart theorem, and that missing microscopic foundation is the open problem.

Problem Statement

Specify an explicit microscopic Hamiltonian class for the fluid — for instance, NN particles in dimension d≥1d\ge 1 interacting through a hard core of volume bb plus an attractive pair potential, taken either of Kac type (−γd2∑i≠jψ(γ∣ri−rj∣)-\frac{\gamma^d}{2}\sum_{i\ne j}\psi(\gamma|\mathbf r_i-\mathbf r_j|) with range γ−1\gamma^{-1}) or of fixed range — in the Barra–Moro isothermal-isobaric setting (fixed NN, pressure and temperature, volume fluctuating) where the external pressure couples to the convex-hull volume of the configuration, and settle the status of the volume-measure reduction for the chosen class by one of the following rigorous outcomes:

(1) Prove that the configurational partition function of the microscopic model admits the one-dimensional volume-measure representation ZN=∫b∞eN(xv+t2ϕ(v)+σ(v)) dvZ_N=\int_b^{\infty}e^{N(xv+\frac{t}{2}\phi(v)+\sigma(v))}\,dv exactly, or asymptotically as N→∞N\to\infty with explicit error bounds uniform in the thermodynamic variables (x,t)(x,t) including the critical region; within the same framework derive ϕ(v)\phi(v) and σ(v)\sigma(v) from the microscopic potential — recovering ϕ(v)=2a/v\phi(v)=2a/v and σ(v)=log⁡(v−b)\sigma(v)=\log(v-b) for the vdW case, or identifying the correct forms and the precise sense in which matching to the macroscopic equation of state fixes them; and establish or correct the finite-NN critical scaling v−vc=N−1/4u((x−xc+a(t−tc)/vc2)/N−3/4,(t−tc)/N−1/2)v-v_c=N^{-1/4}u\big((x-x_c+a(t-t_c)/v_c^2)/N^{-3/4},(t-t_c)/N^{-1/2}\big) with u=−2 ∂ξlog⁡Λu=-2\,\partial_\xi\log\Lambda and Λ\Lambda the Pearcey integral, for the microscopic model — including the order in which N→∞N\to\infty and any Kac range parameter γ−1→∞\gamma^{-1}\to\infty limits are taken.

(2) Or prove a rigorous negative result for the specified class: that the volume-measure reduction and/or the Pearcey critical scaling fails at the critical scale for that class of microscopic Hamiltonians, identifying the correct scaling or the scale at which the deviation appears.

The three requested results in outcome (1) — the reduction theorem, the derivation of ϕ\phi and σ\sigma, and the critical-region scaling — form one joint objective: to determine whether the Barra–Moro exact solution is a theorem about a microscopic model or a self-consistent macroscopic postulate. The microscopic class is the solver's to choose and must be fully specified; settling the outcome for that single class is a complete resolution, and a theorem covering every member of a broader family is not required. Outcome (1) requires all three positive results jointly for the chosen class; outcome (2) is a complete resolution as soon as any one of the three fails for that class — a rigorous failure of the reduction, or of the derivation of ϕ\phi and σ\sigma, or of the critical-region scaling at the critical scale is sufficient on its own.

References

  1. Adriano Barra, Antonio Moro, Exact solution of the van der Waals model in the critical region, Annals of Physics 359, 290–299 (2015); DOI 10.1016/j.aop.2015.04.032; arXiv:1412.1951; https://arxiv.org/abs/1412.1951
  2. Antonio Moro, Shock dynamics of phase diagrams, Annals of Physics 343, 49–60 (2014); DOI 10.1016/j.aop.2014.01.011; https://doi.org/10.1016/j.aop.2014.01.011
  3. N. G. van Kampen, Condensation of a Classical Gas with Long-Range Attraction, Physical Review 135, A362–A369 (1964); DOI 10.1103/PhysRev.135.A362; https://doi.org/10.1103/PhysRev.135.A362
  4. J. L. Lebowitz, O. Penrose, Rigorous Treatment of the Van Der Waals-Maxwell Theory of the Liquid-Vapor Transition, Journal of Mathematical Physics 7, 98–113 (1966); DOI 10.1063/1.1704821; https://doi.org/10.1063/1.1704821
  5. M. Kac, G. E. Uhlenbeck, P. C. Hemmer, On the van der Waals Theory of the Vapor-Liquid Equilibrium. I. Discussion of a One-Dimensional Model, Journal of Mathematical Physics 4, 216–228 (1963); DOI 10.1063/1.1703946; https://doi.org/10.1063/1.1703946
  6. Richard S. Ellis, Charles M. Newman, Limit theorems for sums of dependent random variables occurring in statistical mechanics, Zeitschrift für Wahrscheinlichkeitstheorie und Verwandte Gebiete 44, 117–139 (1978); DOI 10.1007/BF00533049; https://doi.org/10.1007/BF00533049
  7. Francesco Giglio, Giulio Landolfi, Antonio Moro, Integrable extended van der Waals model, Physica D: Nonlinear Phenomena 333, 293–300 (2016); DOI 10.1016/j.physd.2016.02.010; arXiv:1602.03975; https://arxiv.org/abs/1602.03975
  8. Paolo Lorenzoni, Antonio Moro, Exact analysis of phase transitions in mean-field Potts models, Physical Review E 100, 022103 (2019); DOI 10.1103/PhysRevE.100.022103; arXiv:1903.03384; https://arxiv.org/abs/1903.03384
  9. Francesco Giglio, Giulio Landolfi, Luigi Martina, Antonio Moro, Symmetries and criticality of generalised van der Waals models, Journal of Physics A: Mathematical and Theoretical 54, 405701 (2021); DOI 10.1088/1751-8121/ac2009; arXiv:2106.02456; https://arxiv.org/abs/2106.02456
  10. Xin An, Francesco Giglio, Giulio Landolfi, Phase transitions and finite-size effects in integrable virial statistical models, Physical Review E 113, L042103 (2026); DOI 10.1103/44vp-9qv9; arXiv:2503.15719; https://arxiv.org/abs/2503.15719
  11. Chung-Yang Wang, Yih-Yuh Chen, A mechanical approach to one-dimensional interacting gas, Chinese Journal of Physics 56, 2006–2021 (2018); DOI 10.1016/j.cjph.2018.09.016; arXiv:1809.10552; https://arxiv.org/abs/1809.10552