Affine relationships between steady currents
Faezeh Khodabandehlou, Christian Maes, Karel Netočný · 2024
Read Affine relationships between steady currents by Faezeh Khodabandehlou, Christian Maes, Karel Netočný (2024) 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-2412-05019.txt.
Introduction
Faezeh Khodabandehlou, Christian Maes, and Karel Netočný study affine relationships between steady currents in nonequilibrium systems. Their 2024 paper deals with systems modeled by a Markov jump process in a steady state.
Perturbing transition rates changes local currents, and their susceptibility is usually given by Green-Kubo type relations. A different question is how currents relate to each other directly. Past work using algebraic graph theory found a mutual linearity of currents over graph edges. The authors give a novel, shorter derivation. They express the current-current susceptibility as a difference in mean first-passage times. The relation extends to multiple currents, where it stays affine but not additive.
Catalog record
| CREATOR | Faezeh Khodabandehlou, Christian Maes, Karel Netočný |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-05019
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Affine relationships between steady currents Faezeh Khodabandehlou, Christian Maes, Karel Netočný Abstract Perturbing transition rates in a steady nonequilibrium system, e.g. modeled by a Markov jump process, causes a change in the local currents. Their susceptibility is usually expressed via Green-Kubo relations or their nonequilibrium extensions. However, we may also wish to directly express the mutual relation between currents. Such a nonperturbative interrelation was discovered by P.E. Harunari et al. in viv1 by applying algebraic graph theory showing the mutual linearity of currents over different edges in a graph. We give a novel and shorter derivation of that current relationship where we express the current-current susceptibility as a difference in mean first-passage times. It allows an extension to multiple currents, which remains affine but the relation is not additive. I Introduction The maintenance of currents or fluxes is a key-feature of steady nonequilibrium systems. It is of great importance there to understand the current susceptibility, i.e. , how it relates to (thermodynamic) forces or possible changes in architecture etc . Around equilibrium, the result gets traditionally summarized by means of linear response or transport coefficients such as mobilities or conductivities. That is part of standard linear response theory and goes under the name of Green-Kubo relations, green ; kubo . The transport coefficients usually disclose a more microscopic specification of the system’s susceptibility to driving. Around equilibrium, they directly reveal the Onsager(-Casimir) reciprocity as a consequence of time-reversal invariance, onsa . Around nonequilibrium conditions, the extended Green-Kubo relations become more involved as they need to incorporate corrections to the equilibrium version of the fluctuation-dissipation theorem; see e.g. resp ; gaspard . Recently, an alternative to characterizing the current response has appeared, at least in the context of continuous-time Markov jump processes, e.g. for describing unimolecular chemical reactions in open networks. Harunari, Dal Cengio, Lecomte and Polettini have proven that any two stationary currents are linearly related upon perturbations of a single edge’s transition rates, arbitrarily far from equilibrium, viv1 . They give explicit expressions for the current-current susceptibility in terms of the network geometry, viv2 . The present paper continues that work by giving an elementary derivation, extension and interpretation of their relation. We start by expressing how the stationary distribution gets modified when changing transition rates. It constitutes a slight generalization of the setup in viv1 , where we work with multigraphs and hence allow for different transition channels over an edge. In that way, we can compare the stationary distributions when we remove or add channels. From there we readily get the mutual linearity between steady currents in different multigraphs, which we generalize to an affine relation between multiple currents. Interesting and also new is that the current-current susceptibility gets a physical interpretation in terms of first-passage times. The connection between a response coefficient and first-passage times is original but natural as it expresses the causal structure between the directed motion of particles in different locations. In the Appendix, we add a graphical expression for the current-current susceptibility, starting from the Kirchhoff formula for the steady currents. Such graphical representations clarify the relevant graph properties for the transport of current, and hence make it possible to micro-engineer or control how currents propagate through the network. We have in mind applications like the regulation of traffic flow in complex networks to the control of biological pathways subject to boundary currents. However, such applications may require the extension of our results to hypergraphs, which are not being considered here. Plan of the paper: We start in the next section by giving the setup for comparing two Markov jump processes on a multigraph. To comprehend the response in steady currents, it is natural to first understand how a perturbation affects the stationary distribution; that is done in Section III . The modification in stationary distribution when a transition rate is changed, is expressed via differences in mean first-passage times in the original process, Eq.( III.1 ). Section IV gives the corollary, yielding the affine current-current relation. That gets generalized in Section V for multiple currents. The Example V.1 emphasizes that the current-current relationship is not additive. Appendix A expresses the current-current susceptibility via the Kirchhoff formula. II Setup The nonequilibrium system has a finite number of states x,y,z,\ldots\in\mathcal{V} , each for instance representing a configuration of spins, energies or molecules in the system under consideration. To understand the dynamics, we imagine a random walker hopping between those states. The hopping may proceed through a number of possible channels. That Markovian stochastic dynamics is then mathematically specified by transition rates over the various channels that connect any two states. More specifically, for the i th channel x\overset{i}{\leftrightarrow}y , let k_{i}(x,y) and k_{i}(y,x) denote the transition rates. In that way, we build a directed multigraph G(\mathcal{V},\mathcal{E}) where \mathcal{E} is the set of all directed edges e_{i}(x,y) between vertices x and y where k_{i}(x,y)>0 ; see Fig. 1 . (a) (b) Figure 1 : The system is modeled as a random walker on a multigraph (a), where each connection x\leftrightarrow y may have different channels (b). The above defines the Markov jump process with rates k(x,y)=\sum_{i}k_{i}(x,y) and with forward generator L^{\dagger} having matrix elements (II.1) \displaystyle[L^{\dagger}]_{xy} \displaystyle=k(y,x)\quad x\not=y,\qquad[L^{\dagger}]_{xx}=-\sum_{y}k(x,y) Assuming the irreducibility of the graph, the unique stationary distribution is denoted by \rho , normalized \sum_{x}\rho(x)=1 , with probabilities \rho(x)>0 , and verifies the stationary Master equation L^{\dagger}\rho=0 , for all x\in\mathcal{V} , (II.2) \displaystyle\sum_{y}[k(y,x)\rho(y)-k(x,y)\rho(x)]=0 The steady current over the edge e_{i}(x,y) is defined as (II.3) j_{i}(x,y)=k_{i}(x,y)\rho(x)-k_{i}(y,x)\rho(y) The sum j(x,y)=\sum_{i}j_{i}(x,y) enters the steady Master equation ( II.2 ) as (II.4) L^{\dagger}\rho(x)=-\sum_{y}j(x,y)=0 In what follows, the above notation refers to the situation after a change in transition rates. The original process is a Markov jump process on the same multigraph with forward generator L_{0}^{\dagger} and (total) rates k^{0}(x,y) . The change in those rates defines the forward generator L^{\dagger} , with (total transition) rates (II.5) k(x,y)=k^{0}(x,y)+\Delta(x,y) for x\rightarrow y . The \Delta(x,y) determine the change of transition rates. In some cases, we can use that change to really change the channel structure. The simplest to imagine is a single graph and to focus on one specific pair \{a,b\} of states: by taking k^{0}(a,b)=0=k^{0}(b,a),\quad k(a,b)=\Delta(a,b),k(b,a)=\Delta(b,a) we truly add an edge a\leftrightarrow b with transition rates \Delta(a,b),\Delta(b,a) . Similarly and more generally, we detail \Delta(x,y)=\sum_{i}\Delta_{i}(x,y) in ( II.5 ) by writing per channel (II.6) k_{i}(x,y)=k^{0}_{i}(x,y)+\Delta_{i}(x,y) where we are especially interested in the case where k_{i}^{0}(x,y)=0,\Delta_{i}(x,y)>0 (adding a channel for x\rightarrow y ) and in the case where \Delta_{i}(x,y)=-k_{i}^{0}(x,y) (removing a channel for x\rightarrow y ). In that sense, the change ( II.6 ) modifies the channel structure; it is like changing the set of edges \mathcal{E}_{0}\rightarrow\mathcal{E} over which there is a nonzero transition rate. The set of states \mathcal{V}_{0}=\mathcal{V} is never changing. At any rate, mathematically it is most convenient to not change the graph geometry at all and to think of comparing two (or even more than two) Markov processes on the same graph. We always assume that irreducibility is preserved. III Change in stationary distribution We wish to express the new stationary distribution \rho in terms of the original (unperturbed) \rho^{0} (where we always use sub- or superscript ‘ 0 ’ to indicate the process with