Abstract
We consider the zero-range process with long jumps and in contact with infinitely extended reservoirs in its non-equilibrium stationary state. We derive the hydrostatic limit and the Fick’s law, which are a consequence of a static relationship between the exclusion process and the zero-range process. We also obtain the large deviation principle for the empirical density, i.e. we compute the non-equilibrium free energy.
Similar content being viewed by others
Notes
A sequence of integrable random variables \((X_N)_{N}\) is said to converge in mean if \(\{{\mathbb {E}} [X_N)]\}_{N}\) converges.
We recall that a sequence \(\{\mu _n\}_n \in {{\mathcal {M}}}\) converges \(\star \)-weakly to \(\mu \in {{\mathcal {M}}}\) if, and only if, for all \(G\in C([0,1])\) we have that \(\int _0^1 G(x) d\mu _n (x)\) converges to \(\int _0^1 G(x) d\mu (x)\). This coincides with the notion of ‘weak convergence’ used in probability theory.
We assume for simplicity that \(\varepsilon N\) is an integer.
Theorem A.1 is established for continuous functions and we need to apply it to the non continuous function \(\iota ^{\varepsilon }_u (\cdot ) = (2\varepsilon )^{-1} {\mathbb 1}_{\vert \cdot - u \vert \le \varepsilon }\), \(u\in [0,1]\). This can be done by a standard approximation argument.
It means that for any \(h \in C_c^\infty ((0,1))\), \( \int _0^1 f(u) \vert \Delta \vert ^{\gamma /2} h (u) \, du \, +\, \int _0^1 f(u) q(u) \, du =0 \).
References
Bahadoran, C.: A quasi-potential for conservation laws with boundary conditions. arXiv Preprint arXiv:1010.3624 (2010)
Barré, J., Bernardin, C., Chétrite, R.: Density large deviations for multidimensional stochastic hyperbolic conservation laws. J. Stat. Phys. 170(3), 466–491 (2018)
Bellettini, G., Bertini, L., Mariani, M., Novaga, M.: \(\Gamma \)-entropy cost for scalar conservation laws. Arch. Ration. Mech. Anal. 195, 261–309 (2010)
Bernardin, C.: Stationary nonequilibrium properties for a heat conduction model. Phys. Rev. E 78, 021134 (2008)
Bernardin, C., Cardoso, P., Goncalves, P., Scotta, S.: Hydrodynamic limit for a boundary driven super-diffusive symmetric exclusion. arXiv Preprint arXiv:2007.01621 (2020)
Bernardin, C., Gonçalves, P., Jara, M.: 3/4-Fractional superdiffusion in a system of harmonic oscillators perturbed by a conservative noise. Arch. Ration. Mech. Anal. 220(2), 505–542 (2016)
Bernardin, C., Gonçalves, P., Jara, M., Sasada, M., Simon, M.: From normal diffusion to superdiffusion of energy in the evanescent flip noise limit. J. Stat. Phys. 159(6), 1327–1368 (2015)
Bernardin, C., Gonçalves, P., Jara, M., Simon, M.: Interpolation process between standard diffusion and fractional diffusion. Annales de l’Institut Henri Poincaré, Probabilités et Statistiques 54(3), 1731–1757 (2018)
Bernardin, C., Gonçalves, P., Jara, M., Simon, M.: Nonlinear perturbation of a noisy Hamiltonian lattice field model: universality persistence. Commun. Math. Phys. 361(2), 605–659 (2018)
Bernardin, C., Gonçalves, P., Jiménez-Oviedo, B.: A microscopic model for a one parameter class of fractional Laplacians with Dirichlet boundary conditions. Arch. Ration. Mech. Anal. 239(1), 1–48 (2020)
Bernardin, C., Gonçalves, P., Jiménez-Oviedo, B.: Slow to fast infinitely extended reservoirs for the symmetric exclusion process with long jumps. Markov Process. Relat. Fields 25, 217–274 (2019)
Bernardin, C., Gonçalves, P., Sethuraman, S.: Occupation times of long-range exclusion and connections to KPZ class exponents. Probab. Theory Relat. Fields 166(1), 365–428 (2016)
Bernardin, C., Jiménez-Oviedo, B.: Fractional Fick’s law for the boundary driven exclusion process with long jumps. ALEA 14(1), 473–501 (2017)
Bernardin, C., Kannan, V., Lebowitz, J.L., Lukkarinen, J.: Harmonic systems with bulk noises. J. Stat. Phys. 146(4), 800–831 (2012)
Bertini, L., De Sole, A., Gabrielli, D., Jona-Lasinio, G., Landim, C.: Macroscopic fluctuation theory. Rev. Mod. Phys. 87(2), 593 (2015)
Bodineau, T., Derrida, B.: Phase fluctuations in the ABC model. J. Stat. Phys. 145, 745–762 (2011)
Bogdan, K., Byczkowski, T.: Potential theory of Schrödinger operator based on fractional Laplacian. Probab. Math. Stat. 20(2), 293–335 (2000)
Cane, G.: Superdiffusion transition for a noisy harmonic chain subject to a magnetic field. arXiv Preprint arXiv:2201.03373 (2022)
Cardoso, P., Gonçalves, P., Jiménez-Oviedo, B.: Hydrodynamic behavior of long-range symmetric exclusion with a slow barrier: diffusive regime to appear in AIHP Sec. B. arXiv Preprint arXiv:2111.02868 (2021)
Chebloun, P., Grosskinsky, S.: Condensation in stochastic particle systems with stationary product measures. J. Stat. Phys. 154(1–2), 432–465 (2014)
Das, S.G., Dhar, A., Saito, K., Mendl, C.B., Spohn, H.: Numerical test of hydrodynamic fluctuation theory in the Fermi-Pasta-Ulam chain. Phys. Rev. E 90(1), 012124 (2014)
De Masi, A., Ferrari, P.A.: A remark on the hydrodynamics of the zero range process. J. Stat. Phys. 36, 81–87 (1984)
Derrida, B.: Non-equilibrium steady states: fluctuations and large deviations of the density and of the current. J. Stat. Mech. 2007, P07023 (2007)
Di Nezza, E., Palatucci, G., Valdinoci, E.: Hitchhiker’s guide to the fractional Sobolev spaces. Bull. Sci. Math. 136(5), 521–573 (2012)
Evans, M.R.: Phase transitions in one-dimensional nonequilibrium systems. Braz. J. Phys. 30(1), 42–57 (2000)
Evans, M.R., Hanney, T.: Nonequilibrium statistical mechanics of the zero-range process and related models. J. Phys. A 38(19), R195 (2005)
Gonçalves, P.: Hydrodynamics for symmetric exclusion in contact with reservoirs. In: Stochastic Dynamics Out of Equilibrium, Institut Henri Poincaré, Paris, France, 2017, Springer Proceedings in Mathematics and Statistics book series, pp. 137–205 (2019)
Gonçalves, P., Jara, M.: Density fluctuations for exclusion processes with long jumps. Probab. Theory Relat. Fields 170(1), 311–362 (2018)
Gonçalves, P., Scotta, S.: Diffusive to super-diffusive behavior in boundary driven exclusion. Markov Process. Relat. Fields 28, 149–178 (2022)
Guan, Q.-Y., Ma, Z.-M.: Reflected symmetric \(\alpha \)-stable processes and regional fractional Laplacian. Probab. Theory Relat. Fields 134(4), 649–694 (2006)
Hurtado, P.I., Garrido, P.L.: A violation of universality in anomalous Fourier’s law. Sci. Rep. 6(1), 1–10 (2016)
Jara, M.: Current and density fluctuations for interacting particle systems with anomalous diffusive behavior. eprint at arXiv:0901.0229 (2009)
Jara, M., Komorowski, T., Olla, S.: Superdiffusion of energy in a chain of harmonic oscillators with noise. Commun. Math. Phys. 339(2), 407–453 (2015)
Jensen, L.: The asymmetric exclusion process in one dimension. Ph.D. Dissertation, New York University, New York (2000)
Kipnis, C.: Central limit theorems for infinite series of queues and applications to simple exclusion. Ann. Probab. 14(2), 397–408 (1986)
Kipnis, C., Landim, C.: Scaling Limits of Interacting Particle Systems. Springer-Verlag, New York (1999)
Kundu, A., Bernardin, C., Saito, K., Kundu, A., Dhar, A.: Fractional equation description of an open anomalous heat conduction set-up. J. Stat. Mech.: Theory Exp. 2019(1), 013205 (2019)
Lepri, S., Politi, A.: Density profiles in open superdiffusive systems. Phys. Rev. E 83, 030107(R) (2011)
Levine, E., Mukamel, D., Schütz, G.M.: Zero-range process with open boundaries. J. Stat. Phys. 120(5–6), 759–778 (2005)
Liggett, T.M.: Interacting Particle Systems. Classics in Mathematics. Springer-Verlag, Berlin (2005)
Mariani, M.: Large deviations principles for stochastic scalar conservation laws. Probab. Theory Relat. Fields 147(3–4), 607–648 (2010)
Popkov, V., Schadschneider, A., Schmidt, J., Schütz, G.: Fibonacci family of dynamical universality classes. PNAS 11241, 12645–12650 (2015)
Popkov, V., Schmidt, J., Schütz, G.M.: Universality classes in two-component driven diffusive systems. J. Stat. Phys. 160(4), 835–860 (2015)
Saito, K., Sasada, M., Suda, H.: 5/6-Superdiffusion of energy for coupled charged harmonic oscillators in a magnetic field. Commun. Math. Phys. 372(1), 151–182 (2019)
Sethuraman, S.: On microscopic derivation of a fractional stochastic Burgers equation. Commun. Math. Phys. 341(2), 625–665 (2016)
Spitzer, F.: Interaction of Markov processes. Adv. Math. 5(2), 246–290 (1970)
Sethuraman, S., Shahar, D.: Hydrodynamic limits for long-range asymmetric interacting particle systems. Electron. J. Probab. 23, 1–54 (2018)
Spohn, H.: Large Scale Dynamics of Interacting Particles. Springer-Verlag, Berlin (1991)
Spohn, H.: Nonlinear fluctuating hydrodynamics for anharmonic chains. J. Stat. Phys. 154(5), 1191–1227 (2014)
Spohn, H., Stolz, G.: Nonlinear fluctuating hydrodynamics in one dimension: the case of two conserved fields. J. Stat. Phys. 160, 861–884 (2015)
Tsunoda, K.: Hydrostatic limit for exclusion process with slow boundary revisited. RIMS Kôkyûroku Bessatsu B79, 149–162 (2020)
Vanicat, M.: Exact solution to integrable open multi-species SSEP and macroscopic fluctuation theory. J. Stat. Phys. 166, 1129–1150 (2017)
Varadhan, S.R.S.: Large deviations for the asymmetric simple exclusion process. Stochastic analysis on large scale interacting systems. Adv. Stud. Pure Math. 39, 1–27 (2004)
Zeitouni, A.D.O., Dembo, O.: Large Deviations Techniques and Applications. Springer-Verlag, New York (1998)
Acknowledgements
The work of C.B. has been supported by the projects LSD ANR-15-CE40-0020-01 of the French National Research Agency (ANR). B.J.O. thanks Universidad Nacional de Costa Rica for sponsoring the participation in this article through the project 0497-18. P.G. and S.S. thank FCT/Portugal for financial support through CAMGSD, IST-ID, projects UIDB/04459/2020 and UIDP/04459/2020. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovative programme (Grant Agreement No. 715734).
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Abhishek Dhar.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A: Macroscopic Properties of NESS in Boundary Driven Long Range Exclusion
Appendix A: Macroscopic Properties of NESS in Boundary Driven Long Range Exclusion
In this section, we present some results for the stationary behaviour of the open boundary exclusion process with long jumps based on previous results from [5, 10, 13]. Recall (2.12) and note that we consider the process speeded up in the time scale \(\Theta (N)\) as in (2.11). Recall that \(\tilde{\alpha }, \tilde{\beta }\in (0,1)\).
1.1 A.1. Hydrostatic Limit
Let \({{\mathcal M}}^+\), be the space of positive measures on [0, 1] with total mass bounded by 1 and equipped with the weak \(\star \)-topology. For any \(\eta \in \Omega ^N\) the empirical measure \({ \pi }^N_{ex}:={ \pi }^N_{ex} (\eta ) \in {{\mathcal M}}^+\) is defined by
Let \(P^N\) be the probability measure on \({{\mathcal M}}^+\) obtained as the pushforward of \( \mu _{ss}^N\) by \(\pi ^N_{ex}\). We denote the action of \(\pi _{ex}^N \in {{\mathcal M}}^+\) on a continuous function \(G:[0,1]\rightarrow {{\mathbb {R}}}\) by
Theorem A.1
(Hydrostatic limit in mean) For any \(G \in C^0 ([0,1])\) we have
where
-
(a)
for \(\theta <0\) and \(\gamma \in (0,2)\), \( {\bar{\rho }} (u) = \dfrac{V_{0}(u)}{V_{1}(u)}. \)
-
(b)
for \(\theta =0\) and \(\gamma \in (0,2)-\{1\}\), \(\bar{\rho }(\cdot )\) is the unique weak solution of (2.23), with \(\hat{\kappa }=\kappa \).
-
(c)
for \(\theta \in (0,\gamma -1)\) and \(\gamma \in (1,2)\), \(\bar{\rho }(\cdot )\) is the unique weak solution of (2.24).
-
(d)
for \(\theta = \gamma -1\) and \(\gamma \in (1,2)\), \(\bar{\rho }(\cdot )\) is the weak solution of (2.25) with \(\hat{\kappa }=\kappa d\).
-
(e)
for \(\theta > 0\) and \(\gamma \in (0, 1]\) or for \(\theta > \gamma - 1\) and \(\gamma \in (1, 2)\), \(\bar{\rho }(\cdot )\) is the weak solution of (2.27) with \({\hat{M}} =\tfrac{\tilde{\alpha }+\tilde{\beta }}{2}\), i.e. \({\bar{\rho }}=\tfrac{\tilde{\alpha }+\tilde{\beta }}{2} \).
Proof
We start by noting that a simple computation based on the fact that the mass of the system is finite, shows that the sequence \(\{P^N\}_{N\ge 2}\) is tight and that all limit points are concentrated on measures \(\pi (du)\) which are absolutely continuous with respect to Lebesgue measure on [0, 1], i.e. \(\pi (du)=\rho (u)du\). Let us also introduce \(\{\bar{\pi }^N\}_{N\ge 2}= \left\{ E_{\mu _{ss}^N} \left[ \pi _{ex}^N\right] \right\} _{N \ge 2}\) which forms a sequentially compact sequence of \({{\mathcal {M}}}^+\) whose limit points \(\bar{\pi }(du)\) are absolutely continuous with respect to the Lebesgue measure on [0, 1], i.e. \(\bar{\pi }(du)=\bar{\rho }(u)du\). Let \(\bar{\pi }(du)=\bar{\rho }(u)du\) be a limit point of \(\{\bar{\pi }^N\}_{N\ge 2}\). Without loss of generality we can consider a subsequence for which \(\{P^N\}_{N\ge 2}\) is also converging to a limit point denoted by \(P^*\) (the corresponding expectation is denoted by \({E} ^*\)). To lighten notation, in the sequel, we assume that we are taking the limit according to this subsequence, even if it is not specified. Observe that \(\bar{\rho }= E^* [\rho ]\). Our goal is to show that \(\bar{\rho }\) is unique and given as in Theorem A.1.
If \(\gamma \in (0,2)\), the energy estimates of Sect. 3.3 of [10] show that for \(\theta \ge 0\)
and for \(\theta \le 0\)
From Jensen’s inequality, we have also
It remains now to check that \(\bar{\rho }\) satisfies the other conditions in the notions of stationary weak solutions.
Recall (2.11). Note that
where the action of \(\mathbb {L}_{N}\) on functions G is defined by
Taking the expectation with respect to \(\mu _{ss}^N\) on (A.2), we get, from stationarity, that
Recall (2.9). We define the functions \(r_{N}^{\pm }: [0,1]\rightarrow {{\mathbb {R}}}\) as the linear interpolation of \(r_N^- \left( \tfrac{x}{N}\right) \) and \( r_N^+ \left( \tfrac{x}{N}\right) \) for all \(x \in \Lambda _{N}\) with \(r_{N}^{\pm }(0) = r_{N}^{\pm }\left( \tfrac{1}{N}\right) \) and \(r_{N}^{\pm }(1) = r_{N}^{\pm }\left( \tfrac{N-1}{N}\right) \). By Lemma 3.3 in [13], for \(0<\gamma <2\) we have that
uniformly in \([a,1-a]\) for \(a\in (0,1)\) and from that lemma it also follows that
uniformly in \([a,1-a]\), for all functions G with compact support included in \([a,1-a]\).
Now, we split the analysis by taking into account the value of \(\theta \).
Case \(\theta < 0\): In this regime we take \(G\in C_c^\infty ((0,1))\) and \(\Theta (N) = N^{\theta +\gamma }\). Observe that (A.4) can be written as
From the weak convergence together with (A.5) and (A.6), last display converges, as \(N\rightarrow +\infty ,\) to
which implies that \(\bar{\rho }(u)=\frac{V_0(u)}{V_1(u)}\) for almost every u (see also Remark (2.14) in [10]).
Case \(\theta = 0\): In this regime we take \(G\in C_c^\infty ((0,1))\) and \(\Theta (N) = N^{\gamma }\). From weak convergence together with (A.5) and (A.6) we get that
Hence we have that \(F_{RD} (\bar{\rho }, G)=0\) for any \(G\in C_c^\infty ((0,1))\) (with \({\hat{\kappa }}=\kappa )\).
Case \(\theta \in (0,\gamma -1)\) and \(\gamma \in (1,2)\): In this regime we take \(G\in C_c^\infty ((0,1))\) and \(\Theta (N) = N^{\gamma }\). From the weak convergence, the rightmost term in the first line of (A.4) converges, as \(N\rightarrow +\infty \), to
Moreover, the term on the second line of (A.2) can be bounded from above by a constant times
plus lower order terms in N which vanish as \(N\rightarrow +\infty \). Hence we have that \(F_{Dir} (\bar{\rho }, G)=0\) for any \(G\in C_c^\infty ((0,1))\).
Case \(\theta = \gamma -1\) and \(\gamma \in (1,2)\): In this regime we take \(G\in C^\infty ([0,1])\) and \(\Theta (N) = N^{\gamma }\). We start by noting that from Lemma 5.1 of [5], we have that
for functions \(G\in C^\infty ([0,1])\) and \(\gamma \in (0,2).\) Moreover, the first term on the second line of (A.4) can be rewritten as
plus analogous terms with respect to the right boundary. The expectation of the second term above converges to zero as \(N\rightarrow +\infty \) from Lemma 5.9 of [5] (see Remark 5.10 there). We note that, to apply Lemma 5.9 of [5], which is written with the time integral and in the \(L^1\) sense, in last argument, we need to use the fact that \(\mu _{ss}^N\) is a stationary measure, to introduce the time integral in the rightmost term of (A.8) and then use the aforementioned lemma. Now, for the first term, we can perform a Taylor expansion on G obtaining the following expression,
plus lower order terms in N. Observe that
which goes to zero as N goes to infinity. Finally, the remaining term can be treated by using the weak convergence, the fact that the limiting measure \(\bar{\pi }(du)\) is absolutely continuous with respect to the Lebesgue measure with density \(\bar{\rho }\) and also that
Hence we get that \(F_{Rob} (\bar{\rho }, G)=0\) (with \(\hat{\kappa }=\kappa d\)) for any \(G\in C^\infty ([0,1])\).
Case \(\theta >\gamma -1\) and \(\gamma \in (1,2)\): The proof in this regime is completely analogous to the previous case. Nevertheless, we could also obtain the result as a consequence of the hydrostatic limit, with a convergence in probability, by following the same strategy described in [51]. Since we do not need this stronger convergence to attain our results, we did not pursue this issue here.
Case \(\theta > 0\) and \(\gamma \in (0,1)\): In this case the analysis of the first term of (A.4) is given by an approximation argument of the operator \(\mathbb {L}\) given in Lemma 5.1 of [5]. Since G(x/N) and \( \eta (x) \) are bounded we know that the second term of (A.4) is bounded by a constant times
Since \(\gamma \in (0,1)\), we lose the convergence of the partial sum above. However, it is not difficult to see that
which vanishes as N goes to \(\infty \). Hence \(F_{Neu} (\bar{\rho }, G)=0\) for any \(G\in C^\infty ([0,1])\).
In the two last cases, we have also to show that \(\int _0^1 {\bar{\rho }} (u) du =\tfrac{\tilde{\alpha }+ \tilde{\beta }}{2}\). In fact, since in these cases, \(F_{Neu} (\bar{\rho }, G)=0\) for any \(G\in C^\infty ([0,1])\), we can conclude that \(\bar{\rho }\) is equal to a constant M by using the same argument as in Lemma 2.14 when showing that \(\Vert {\bar{\rho }}\Vert _{\gamma /2}=0\).
Recall the definition of \({\bar{\pi }}^N\) given at the beginning of the proof. We consider a sequence \((\chi ^\varepsilon )_{0<\varepsilon < 1/4}\) of smooth functions with values in [0, 1], symmetric with respect to 1/2, equal to 0 on \([0,\varepsilon /2]\) and to 1 on \([\varepsilon ,1/2]\). Observe that \((\chi ^\varepsilon )_\varepsilon \) converges in \(L^1\) as \(\varepsilon \) goes to 0 to the constant function equal to 1. Taking \(G=\chi ^\varepsilon \) in (A.4) we get
By (A.6) and (A.5), we can then replace in the previous expression \(N^\gamma {\mathbb {L}}_N\) by \({\mathbb {L}}\) and \(N^\gamma r_N^\pm \) by \(r^\pm \). Recall that \({\bar{\pi }}^N (du)\) converges weakly to \(\bar{\rho }(u) du=M du\). Since \(\theta >0\) we get
In the case \(\gamma \in (0,1)\) the functions \(V_0\) and \(V_1\) are integrable and we get by sending \(\varepsilon \) to 0 that
In the case \(\gamma \in (1,2)\) the integrals are diverging but
therefore we get again that
Conclusion: All limit points of the sequence \(\{\bar{\pi }^N (du) \}_{N\ge 2}\) are in the form \(\bar{\rho }(u) du\) where \(\bar{\rho }\) is a weak solution of the hydrostatic equation. By uniqueness of weak solutions for these equations, \(\bar{\rho }\) is unique and therefore the sequence is converging, without extracting a subsequence, to this unique weak solution. \(\square \)
Lemma A.2
The profiles \(\bar{\rho }\) in Theorem A.1 are continuous in (0, 1).
Proof
In the case \(\theta <0\) the claim follows easily since the profile is explicit. We consider now \(\theta \ge 0\). If \(\gamma \in (1,2)\), by definition of a weak solution, we know that \(\bar{\rho }\) is bounded and belongs to \({\mathcal {H}}^{\gamma /2}\) and from Theorem 8.2 of [24], we conclude that \(\bar{\rho }\) is \(\tfrac{\gamma -1}{2}\)-Hölder in [0, 1], therefore continuous in (0, 1). If \(\gamma \in (0,1)\) and \(\theta >0\), the profile is constant and therefore continuous. The only missing case is \(\gamma \in (0,1)\) and \(\theta =0\). We have hence to prove that the stationary solution \(\rho \) of the regional fractional reaction–diffusion equation (2.23) is continuous in (0, 1) when \(\gamma \in (0,1)\). It is known that if \(\gamma \in (0,1]\), the condition \(f \in {{\mathcal {H}}}^{\gamma /2}\) does not guarantee, contrarily to the case \(\gamma \in (1,2)\), that f is continuous. Therefore the continuity property of \(\rho \) can only result from the fact that \(\rho \) satisfies the weak formulation of (2.23). This property is a consequence of potential theory for (fractional) Shrödinger theory developed in [17], more exactly of Proposition 6.1 of that article that we restate in our particular context. Before doing so, we introduce a few notations.
The fractional Laplacian \(\vert \Delta \vert ^{\gamma /2}\) on \({\mathbb {R}}\) is the operator acting on functions \(f:{\mathbb {R}} \rightarrow {{\mathbb {R}}}\) such that
as
for any \(u \in {\mathbb {R}}\) if the limit exists (which is for example the case for smooth compactly supported functions). It can be extended into the weak fractional Laplacian (that we denote abusively by the same notation) by duality: For any f satisfying (A.11), \(\vert \Delta \vert ^{\gamma /2} f\) is the distribution (or generalized function) on \({\mathbb {R}}\) satisfying the identity \(\langle \vert \Delta \vert ^{\gamma /2} f, h \rangle = \langle f, \vert \Delta \vert ^{\gamma /2} h \rangle \), for any compactly supported function h on \({\mathbb {R}}\).
Property 6.1 of [17] claims that if \(f:{\mathbb {R}} \rightarrow {\mathbb {R}}\) is a solution, in the distributional senseFootnote 5, on (0, 1) of the equation
then f is continuous on (0, 1) as soon as the function \(q:(0,1) \rightarrow {\mathbb {R}}\) belongs to the (local) Kato class of exponent \(\gamma \), i.e.
for any \(\varepsilon \in (0,1/2)\). It is not difficult to see that if \(\rho \) is the weak solution of (2.23) and is extended by 0 outside of (0, 1), then \(\rho \) satisfies, in the distributional sense, (A.13) on (0, 1) for q given as a linear combination of \(r^-\) and \(r^+\) (defined by (2.22)). Hence to conclude the proof of the lemma, it is sufficient to prove that \(r^\pm \) belong to the (local) Kato class of exponent \(\gamma \). This exercise is left to the interested reader. \(\square \)
1.2 A.2. Fractional Fick’s Law of the Boundary Driven Exclusion
By adapting the strategy of [13] we can obtain the “fractional Fick’s Law” which is given in the next theorem. The expression of the current is given by
Theorem A.3
(Fractional Fick’s law) Let \(\bar{\rho }(\cdot )\) be the hydrostatic profile of the boundary driven exclusion given in Theorem A.1. For \(u\in (0,1)\) the following fractional Fick’s law holds, apart from the case \(\theta =0\) and \(\gamma =1\) :
where
the function \(h_\theta :(0,1)\rightarrow \mathbb {R}\) is given by
and
This implies that
-
(a)
for \(\theta < 0\),
$$\begin{aligned} \begin{aligned} \lim _{N \rightarrow \infty }\frac{1}{N^{1-\theta -\gamma }}{{\mathbb {E}}}_{ \mu _{ss}^N}\left[ {W_{[uN]}^{ex}}\right]&= \kappa \int _{u}^{1}(\tilde{\alpha }-\bar{\rho }(v))r^{-}(v)dv- \kappa \int _{0}^{u}(\tilde{\beta }-\bar{\rho }(v))r^{+}(v)dv\\&=\kappa c_\gamma \gamma ^{-1} \int _0^1\dfrac{\tilde{\alpha }-\tilde{\beta }}{v^\gamma +(1-v)^\gamma }dv; \end{aligned} \end{aligned}$$(A.19) -
(b)
for \(\theta = 0\),
$$\begin{aligned} \begin{aligned} \lim _{N \rightarrow \infty } \frac{1}{N^{1-\gamma }} {{\mathbb {E}}}_{\mu _{ss}^N}\left[ {W_{[uN]}^{ex}}\right] =c_{\gamma } \int _{0}^u \; \int _{u}^{1} \, \frac{{ \bar{\rho }} (v) -{ \bar{\rho }} (w)}{(w-v)^{1+\gamma }}\, dw dv&+ \kappa \int _{u}^{1}(\tilde{\alpha }-\bar{\rho }(v))r^{-}(v)dv\\&- \kappa \int _{0}^{u}(\tilde{\beta }-\bar{\rho }(v))r^{+}(v)dv; \end{aligned} \end{aligned}$$(A.20) -
(c)
for \(\theta >0\),
$$\begin{aligned} \begin{aligned} \lim _{N \rightarrow \infty } \frac{1}{N^{1-\gamma }}{{\mathbb {E}}}_{\mu _{ss}^N}\left[ {W_{[uN]}^{ex}}\right]&=c_{\gamma } \int _{0}^u \; \int _{u}^{1} \, \frac{{ \bar{\rho }} (v) -{ \bar{\rho }} (w)}{(w-v)^{1+\gamma }}\, dw dv. \end{aligned} \end{aligned}$$(A.21)
Proof
Since the measure \(\mu _{ss}^N\) is stationary and \({\mathcal {L}}_N \eta _ x = W_x^{ex}(\eta )-W_{x+1}^{ex}(\eta )\) for all \(x\in \Lambda _N\) and for all \(\eta \), it follows that \({E}_{\mu _{ss}^N}[ W_x^{ex}] = {E}_{\mu _{ss}^N}[ W_1^{ex}],\) for all \(x\in \Lambda _N\). Then we can write
We define the linear interpolation functions \( {\tilde{r}}_{N}^{\pm } :[0,1]\rightarrow {{\mathbb {R}}}\), such that for all \(z\in \Lambda _N\) we have that
Using (A.22) it is not difficult to see that
where
and
From (A.23), we get the following convergence
which holds uniformly for \(u \in (a,1-a)\) (with \(0<a<1\) fixed), and the function \(h_\theta :(0,1)\rightarrow \mathbb {R}\) is given in (A.17). We note that \(h_\theta \) is singular at \(u=0\) and \(u=1\) if \(\gamma \in (1,2)\), but it is integrable in [0, 1] for \(\gamma \in (0,2)\). Moreover, it is easy to see that, for \(u\in (0,1)\) and for \(\gamma \ne 1\) we have that
Regarding the last term of (A.24), a simple computation shows that
where \( C(\tilde{\alpha },\tilde{\beta },\theta )\) was defined in (A.18). Now, note that from (A.25) and the fact that \(| \eta (x)|\le 1 \), we get that
where \(I(a) = [a,1-a]\) and \(NI(a)= [Na,N(1-a)]\cap \mathbb {N}.\) From (A.26) and the fact that \(| \eta (x)|\le 1 \) we get that
Moreover, from Theorem A.1 and, for \(h_\theta ^a\) a continuous extension of the function \(h_\theta \) restricted to I(a), we get that
Now, from (A.24) and (A.28) sending first \(N\rightarrow \infty \) and then \(a\rightarrow 0\) we obtain (A.15).
The other expressions of the limiting current given at the end of the theorem are achieved by using properties of the integrals and the fact that the limit does not depend on the variable u. To check them properly, let us consider for instance \(\theta >0\) and \(\gamma \ne 1\). Since the limit does not depend on u, we have that
Using Fubini’s theorem twice, last display equals to
Finally, a simple computation, based again on Fubini’s theorem, shows that last display is equal to
This ends the case \(\theta >0\). The cases \(\theta \le 0\) can be obtained by performing similar computations to the ones above, plus the fact that
Finally, we note that the second equality in item a) is obtained by algebraic manipulations using the fact that \(\bar{\rho }(u) = \dfrac{ V_0(u)}{V_1(u)}\). \(\square \)
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Bernardin, C., Gonçalves, P., Jiménez-Oviedo, B. et al. Non-equilibrium Stationary Properties of the Boundary Driven Zero-Range Process with Long Jumps. J Stat Phys 189, 32 (2022). https://doi.org/10.1007/s10955-022-02987-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10955-022-02987-3