One-dimensional polymers in random environments: stretching vs. folding

Abstract: In this article we study a \emph{non-directed polymer model} on $\mathbb Z$, that is a one-dimensional simple random walk placed in a random environment. More precisely, the law of the random walk is modified by the exponential of the sum of "rewards" (or penalities) $\beta \omega_x -h$ sitting on the range of the random walk, where $(\omega_x)_{x\in \mathbb Z}$ are i.i.d.\ random variables (the disorder), and where $\beta\geq 0$ (disorder strength) and $h\in \mathbb{R}$ (external field) are two parameters. When $\beta=0,h>0$, this corresponds to a random walk penalized by its range; when $\beta>0, h=0$, this corresponds to the "standard" polymer model in random environment, except that it is non-directed. In this work, we allow the parameters $\beta,h$ to vary according to the length of the random walk, and we study in detail the competition between the \emph{stretching effect} of the disorder, the \emph{folding effect} of the external field (if $h\ge 0$), and the \emph{entropy cost} of atypical trajectories. We prove a complete description of the (rich) phase diagram. For instance, in the case $\beta>0, h=0$ of the non-directed polymer, if $\omega_x$ ha a finite second moment, we find a transversal fluctuation exponent $\xi=2/3$, and we identify the limiting distribution of the rescaled log-partition function.