Answers are grounded in this paper and cite their sources.
We study global minimizers in Rᵈ of the free energy E_κ=W+κU, where W is a power-law interaction energy with mild repulsion and U is the porous-medium internal energy with exponent $m>1$. In a range of exponents that we call the strict simplex regime, results of Lim--McCann and Davies--Lim--McCann show that W is minimized, uniquely up to rigid motions, by equal point masses at the vertices of a regular unit d-simplex. Diffusion excludes such atomic minimizers for every $κ>0$. We prove that uniqueness nevertheless persists: for every fixed $m>1$ and all sufficiently small $κ>0$, the free energy has a unique global minimizer up to rigid motions. After a suitable rigid motion, this minimizer is invariant under every symmetry of the simplex, and its support splits into $d+1$ clusters near the vertices, each of mass exactly $1/(d+1)$; in particular, it is not radially symmetric when d≥2. The clusters have width of order κ1/(d(m-1)+2), which is also the exact order of the quadratic Wasserstein distance from the minimizer to the atomic simplex measure. Rescaled at this scale, the clusters converge in L¹ and L^∞ to explicit Barenblatt-type profiles supported on ellipsoids determined by the Hessian of the simplex potential. For d≥2 these ellipsoids are spheroids flattened along the directions of the vertices; when the repulsive exponent exceeds two, their aspect ratio equals (d+1)-1/2, independently of all other parameters. At the endpoint $(4,2)$ of this regime in dimensions d≥2, where the zero-diffusion minimizer is not unique, we show instead that vanishing diffusion selects the uniform spherical shell.
No takes yet. Share an insight, caveat, or question.
Lim et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: