Laboratoire Matière et Systèmes Complexes

Fluctuation theorem for autonomous measurement and feedback

Naoto Shiraishi

Abstract:

When we measure something and perform operations on it depending on the measurement outcomes (i.e. feedback processes), we use “information”. Such information processes are ubiquitous in biochemical systems such as molecular motors[1] and bacterial chemotaxis[2]. However, if we analyze a system with information processes, we cannot use the usual framework of thermodynamics naively, which is known as the Maxwell’s demon problem[3]. For the case that the measurement and the feedback processes are separated externally, it is established that by taking the mutual information into account we obtain thermodynamic relations[4,5]. In contrast, for the non-separated case, which includes autonomous information processes and almost all biochemical systems, whether the mutual information also plays a crucial role has been debated.

We have recently discovered a reasonable way to define a partial entropy production with individual transition paths, which is decomposition of total entropy production and satisfies the fluctuation theorem [6]. By applying this framework, we found that mutual information still plays a crucial role in non-separated measurement and feedback processes. In addition, our framework works in not only information processes but also general biochemical reactions. In this seminar, we will discuss these results.

References:

[1] K. Sekimoto, Compt. Rend. Phys. 8, 650 (2007).

[2] G. Lan, et. al., Nature Phys. 8, 422 (2012).

[3] H. S. Leff and A. F. Rex, eds., Maxwell’s Demon: Entropy, Information, Computing (Princeton University Press, New Jersey, 1990).

[4] T. Sagawa and M. Ueda, Phys. Rev. Lett. 109, 180602 (2012).

[5] T. Sagawa and M. Ueda, New Jour. Phys. 15, 125012 (2013).

[6] N. Shiraishi and T. Sagawa, arXiv: 1403.4018 (2014).

Vendredi 7 Novembre 2014 à 12:30 en salle 646A