Global existence of weak solutions for the Maxwell–Stefan system in the whole space
Stefanos Georgiadis · 2026
Read Global existence of weak solutions for the Maxwell–Stefan system in the whole space by Stefanos Georgiadis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-25536.txt.
Introduction
Mixtures diffuse strangely: each species in a gas or liquid is dragged by friction against every other, so fluxes couple across the whole mixture. Stefanos Georgiadis proves in 2026 that the Maxwell Stefan equations describing this transport always have global weak solutions on all of three dimensional space. Bounded container results existed, but open space breaks the standard tools because total mass can be infinite and the usual entropy is not finite either.
The fix is to measure entropy relative to a strictly positive constant equilibrium state instead. Georgiadis first solves the system on large balls with no flux through the boundary, harvesting estimates that do not depend on the ball's radius. Local compactness and a diagonal argument then pass to the whole space limit. The limiting solution obeys the equations in the distribution sense and keeps a global relative entropy inequality.
Catalog record
| CREATOR | Stefanos Georgiadis |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.25536
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Global existence of weak solutions for the Maxwell–Stefan system in the whole space Stefanos Georgiadis Abstract . We prove the global-in-time existence of weak solutions to the isothermal Maxwell–Stefan system on the whole space ℝ 3 \mathbb{R}^{3} . The main difficulty is that, unlike in bounded domains, the concentrations generally have infinite mass and the standard mixing entropy is not finite. We therefore work with the relative entropy with respect to a strictly positive constant equilibrium state. The proof proceeds by solving the problem on balls B R B_{R} with no-flux boundary conditions, using the bounded-domain entropy theory, and deriving estimates independent of R R . These estimates yield uniform control of the relative entropy, the gradients ∇ c i \nabla\sqrt{c_{i}} , and the fluxes J i J_{i} . Passing to the limit R → ∞ R\to\infty is achieved by local compactness and a diagonal argument. The resulting weak solution satisfies the Maxwell–Stefan system in the sense of distributions and obeys a global relative entropy inequality. 2020 Mathematics Subject Classification 35D30, 35Q35, 76N10, 76R50, 76T30. 1. Introduction The Maxwell–Stefan system is a fundamental model for multicomponent mass transport in fluid mixtures. It is derived from a balance between thermodynamic driving forces and interspecies friction, rather than from a direct proportionality between fluxes and concentration gradients as in Fickian diffusion. The dependence of each species flux on the state and motion of the other species leads to a strongly coupled cross-diffusion system, in which the evolution of one component is influenced by the concentration gradients of all components. This intrinsic coupling between species makes the Maxwell–Stefan approach particularly suitable for modeling transport processes in gases, liquids, porous media, membrane separations, and electrochemical devices. Consider the isothermal Maxwell–Stefan system (1.1) \partial_{t}c_{i}+\operatorname{div}J_{i}=0,\qquad i\in\{1,\dots,n+1\}, where c=(c_{1},\dots,c_{n+1}) is the vector of concentrations and J=(J_{1},\dots,J_{n+1}) the vector of mass fluxes, which are determined by solving uniquely the constrained linear system (1.2) \displaystyle-\sum_{j\not=i}^{n+1}\frac{c_{j}J_{i}-J_{j}c_{i}}{D_{ij}} \displaystyle=c_{i}\nabla\mu_{i},\quad i\in\{1,\dots,n+1\} (1.3) \displaystyle\sum_{i=1}^{n+1}J_{i} \displaystyle=0, where D_{ij} are positive symmetric interaction coefficients and \mu_{i} denotes the chemical potential of the i -th species, given by the variational derivative of the associated entropy (or free energy) functional H , namely \mu_{i}=\frac{\delta H}{\delta c_{i}}. The concentrations c_{i} satisfy the bounds 0\leq c_{i}\leq 1 , as well as the normalization constraint (1.4) \sum_{i=1}^{n+1}c_{i}=1. The mathematical analysis of the Maxwell–Stefan system has attracted considerable attention in recent years. The existence of local-in-time strong solutions has been established in [ 4 ] for the Cauchy problem and in [ 1 ] for bounded domains, while weak solutions were shown to exist globally in time in [ 8 ] on a bounded subset of \mathbb{R}^{3} . For results on the nonisothermal case we refer to [ 5 , 3 ] for weak solutions and [ 6 , 7 , 9 ] for strong solutions. The critical estimates for the existence theory of weak solutions come from the idea of associating an entropy functional to the system, which plays the role of a Lyapunov functional. In the case of a bounded domain \Omega , the entropy is the standard Boltzmann/mixing entropy, namely (1.5) H_{0}[c(t)]=\int_{\Omega}h_{0}(c)\,dx=\int_{\Omega}\sum_{i=1}^{n+1}c_{i}(\log{c_{i}}-1)\,dx. The objective of the present work is to show the existence of global-in-time weak solutions on the whole space \mathbb{R}^{3} . Contrary to the bounded domain case, one cannot use the same entropy because c_{i} do not have finite mass [ 2 ] . Instead, we fix a constant equilibrium state \bar{c}=(\bar{c}_{1},\dots,\bar{c}_{n+1}) , \bar{c}_{i}>0 , \sum_{i=1}^{n+1}\bar{c}_{i}=1 , and use the relative entropy (1.6) H[c|\bar{c}]=\int_{\mathbb{R}^{3}}h(c|\bar{c})\,dx=\int_{\mathbb{R}^{3}}\sum_{i=1}^{n+1}c_{i}\log\frac{c_{i}}{\bar{c}_{i}}\,dx. The use of the relative entropy does not change the underlying Maxwell–Stefan dynamics. Indeed, the relative entropy density differs from the standard mixing entropy density only by a constant and an affine term in the concentrations. More precisely, since h_{0}(c)=\sum_{i=1}^{n+1}c_{i}(\log c_{i}-1), we have h(c|\bar{c})=\sum_{i=1}^{n+1}c_{i}\log\frac{c_{i}}{\bar{c}_{i}}=h_{0}(c)+1-\sum_{i=1}^{n+1}c_{i}\log\bar{c}_{i}, where we used the constraint \sum_{i}c_{i}=1 . Hence the corresponding chemical potentials differ only by constants, and therefore their gradients coincide: \nabla\log\frac{c_{i}}{\bar{c}_{i}}=\nabla\log c_{i}. Thus the relative entropy is a natural Lyapunov functional for the whole-space problem, while leading to the same Maxwell–Stefan flux relations. The main difficulty in passing from bounded domains to the whole space is twofold. First, the concentrations are not expected to have finite mass on \mathbb{R}^{3} . In fact, if c_{i} approaches a strictly positive constant state \bar{c}_{i} at spatial infinity, then \int_{\mathbb{R}^{3}}c_{i}\,dx=\infty. Consequently, the standard entropy ( 1.5 ) is not the appropriate quantity on \mathbb{R}^{3} . Instead, a finite-entropy condition is imposed on the perturbation from equilibrium through the relative entropy ( 1.6 ). Second, the compact embedding H^{1}(\Omega)\hookrightarrow L^{2}(\Omega) , which is available on bounded domains, has no direct analogue on \mathbb{R}^{3} . Hence one cannot pass to the limit by using global compactness. The proof must instead rely on local compactness and a careful control of the behavior at infinity. The purpose of this paper is to prove the global-in-time existence of weak solutions to the isothermal Maxwell–Stefan system on \mathbb{R}^{3} for initial data with finite relative entropy with respect to a strictly positive constant equilibrium state. More precisely, assuming 0\leq c_{i}^{0}\leq 1,\qquad\sum_{i=1}^{n+1}c_{i}^{0}=1, and H[c^{0}|\bar{c}]<\infty, we prove that, for every T>0 , there exists a weak solution (c,J) on (0,T)\times\mathbb{R}^{3} satisfying 0\leq c_{i}\leq 1,\qquad\sum_{i=1}^{n+1}c_{i}=1,\qquad\sum_{i=1}^{n+1}J_{i}=0, as well as the estimates c_{i}-\bar{c}_{i}\in L^{\infty}(0,T;L^{2}(\mathbb{R}^{3})), \nabla\sqrt{c_{i}}\in L^{2}((0,T)\times\mathbb{R}^{3};\mathbb{R}^{3}), and J_{i}\in L^{2}((0,T)\times\mathbb{R}^{3};\mathbb{R}^{3}). Moreover, the solution satisfies the relative entropy inequality H[c(t)|\bar{c}]+C\int_{0}^{t}\int_{\mathbb{R}^{3}}\sum_{i=1}^{n+1}|\nabla\sqrt{c_{i}}|^{2}\,dx\,ds\leq H[c^{0}|\bar{c}] for almost every t>0 , where C>0 depends only on the Maxwell–Stefan diffusivities. The proof follows the entropy method of Jüngel and Stelzer [ 8 ] on bounded domains, but with several modifications required by the whole-space setting. We first solve the problem on balls B_{R}\subset\mathbb{R}^{3} with homogeneous no-flux boundary conditions. The bounded-domain theory gives weak solutions on each B_{R} . Since the no-flux condition conserves the mass of each species on B_{R} , the standard entropy inequality can be rewritten as a relative entropy inequality with respect to the fixed constant state \bar{c} . This yields H_{B_{R}}[c^{R}(t)|\bar{c}]+\int_{0}^{t}\mathcal{D}_{B_{R}}[c^{R},J^{R}]\,ds\leq H_{B_{R}}[c^{0,R}|\bar{c}]\leq H[c^{0}|\bar{c}], This uniform estimate is the key point of the argument. The entropy dissipation controls both the gradients of the square roots of the concentrations and the fluxes. More precisely, using the Maxwell–Stefan relations and the constraint \sum_{i}J_{i}=0 , one obtains bounds of the form \int_{0}^{T}\int_{B_{R}}\sum_{i=1}^{n+1}|\nabla\sqrt{c_{i}^{R}}|^{2}\,dx\,dt\leq CH[c^{0}|\bar{c}], and \int_{0}^{T}\int_{B_{R}}\sum_{i=1}^{n+1}|J_{i}^{R}|^{2}\,dx\,dt\leq CH[c^{0}|\bar{c}], where C is independent of R . These estimates are purely entropic and algebraic; in particular, no Poincaré inequality on B_{R} is used. This is essential, since the Poincaré constant would diverge as R\to\infty . The passage to the limit R\to\infty is obtained by local compactness. On every fixed ball B_{K} , the estimates imply c_{i}^{R}-\bar{c}_{i}\quad\text{bounded in}\quad L^{\infty}(0,T;L^{2}(B_{K})), c_{i}^{R}\quad\text{bounded in}\quad L^{2}(0,T;H^{1}(B_{K})), and \partial_{t}c_{i}^{R}=-\operatorname{div}J_{i}^{R}\quad\text{bounded in}\quad L^{2}(0,T;W^{-1,2}(B_{K})). The Aubin–Lions lemma then gives strong convergence of c_{i}^{R} in L^{2}((0,T)\times B_{K}) . A diagonal argument over K yields strong local convergence on \mathbb{R}^{3}\times(0,T) , which is sufficient to pass to the nonlinear Maxwell–Stefan relations. Finally, lower semicontinuity of the relative entropy and of the dissipation gives the entropy inequality on the whole space. We emphasize that the far-field condition is not imposed pointwise. The condition H[c^{0}|\bar{c}]<\infty is the natural finite-entropy replacement for finite mass in the present setting. It implies, in particular, that c_{i}^{0}-\bar{c}_{i}\in L^{2}(\mathbb{R}^{3}), and the entropy inequality propagates this information in time. Thus the solution remains close to the equilibrium state in the relative-entropy sense, although the individual