1: %% ****** Start of file template.aps ****** %
2: %%
3: %%
4: %% This file is part of the APS files in the REVTeX 4 distribution.
5: %% Version 4.0 of REVTeX, August 2001
6: %%
7: %%
8: %% Copyright (c) 2001 The American Physical Society.
9: %%
10: %% See the REVTeX 4 README file for restrictions and more information.
11: %%
12: %
13: % This is a template for producing manuscripts for use with REVTEX 4.0
14: % Copy this file to another name and then work on that file.
15: % That way, you always have this original template file to use.
16: %
17: % Group addresses by affiliation; use superscriqptaddress for long
18: % author lists, or if there are many overlapping affiliations.
19: % For Phys. Rev. appearance, change preprint to twocolumn.
20: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal
21: % Add 'draft' option to mark overfull boxes with black boxes
22: % Add 'showpacs' option to make PACS codes appear
23: % Add 'showkeys' option to make keywords appear
24: %\documentclass[aps,pre,preprint,groupedaddress,showpacs]{revtex4}
25: %%%\documentclass[aps,pre,groupedaddress,showpacs]{revtex4}
26: %\documentclass[aps,prl,preprint,superscriptaddress]{revtex4}
27: \documentclass[aps,pre,twocolumn,groupedaddress,showpacs]{revtex4}
28: \usepackage{amsfonts}
29: \usepackage{graphicx}
30: % You should use BibTeX and apsrev.bst for references
31: % Choosing a journal automatically selects the correct APS
32: % BibTeX style file (bst file), so only uncomment the line
33: % below if necessary.
34: \bibliographystyle{apsrev}
35:
36: \begin{document}
37: \newcommand{\ud}{{\mathrm d}}
38: \newcommand{\sech}{\mathrm{sech}}
39: % Use the \preprint command to place your local institutional report
40: % number in the upper righthand corner of the title page in preprint mode.
41: % Multiple \preprint commands are allowed.
42: % Use the 'preprintnumbers' class option to override journal defaults
43: % to display numbers if necessary
44: %\preprint{}
45:
46: %Title of paper
47: \title{Flux reversal in a simple random walk model on a fluctuating
48: symmetric lattice}
49:
50: % repeat the \author .. \affiliation etc. as needed
51: % \email, \thanks, \homepage, \altaffiliation all apply to the current
52: % author. Explanatory text should go in the []'s, actual e-mail
53: % address or url should go in the {}'s for \email and \homepage.
54: % Please use the appropriate macro foreach each type of information
55:
56: % \affiliation command applies to all authors since the last
57: % \affiliation command. The \affiliation command should follow the
58: % other information
59: % \affiliation can be followed by \email, \homepage, \thanks as well.
60: \author{Jes\'us Casado-Pascual}
61: %\email[]{Your e-mail address}
62: %\homepage[]{Your web page}
63: %\thanks{}
64: %\altaffiliation{}
65: \affiliation{F\'{\i}sica Te\'orica,
66: Universidad de Sevilla, Apartado de Correos 1065, Sevilla 41080, Spain}
67: %Collaboration name if desired (requires use of superscriptaddress
68: %option in \documentclass). \noaffiliation is required (may also be
69: %used with the \author command).
70: %\collaboration can be followed by \email, \homepage, \thanks as well.
71: %\collaboration{}
72: %\noaffiliation
73:
74: \date{\today}
75:
76: \begin{abstract}
77: A rather simple random walk model on a one-dimensional lattice is put
78: forward. The lattice as a whole switches randomly between two possible
79: states which are spatially symmetric. Both lattice states are
80: identical, but translated by one site with respect to each other, and
81: consist of infinite arrays of absorbing sites separated by two
82: non-absorbing sites. Exact explicit expressions for the long-time
83: velocity and the effective diffusion coefficient are obtained and
84: discussed. In particular, it is shown that the direction of the steady
85: motion can be reversed by conveniently varying the values of either
86: the mean residence times in the lattice states or the transition rates
87: to the absorbing and non-absorbing sites.
88:
89: \end{abstract}
90:
91:
92:
93: % insert suggested PACS numbers in braces on next line
94: \pacs{05.40.Fb, 05.60.Cd, 87.16.Nn}
95: % insert suggested keywords - APS authors don't need to do this
96: %\keywords{}
97:
98: %\maketitle must follow title, authors, abstract, \pacs, and \keywords
99: \maketitle
100: \section{Introduction}
101: The study of the transport of Brownian particles along periodic
102: structures has attracted considerable and ever growing interest in the
103: last years due to its relevance in a wide variety of problems in
104: physics, chemistry, biology, and nanotechnology \cite{Reimann,
105: Hanggi1}. As required by the second law of thermodynamics, Brownian
106: motion cannot induce a steady directed motion in a system at
107: equilibrium (see, for instance, Ref.~\cite{Reimann} for a detailed
108: proof). Nevertheless, a Brownian particle can display a directed
109: motion in a periodic potential which switches randomly or periodically
110: between two or more states
111: \cite{Bier,Prost,Chauwin,Rousselet,Faucheux}. For this to be possible,
112: certain spatiotemporal symmetry and supersymmetry conditions must be
113: broken \cite{Reimann1}. An important phenomenon
114: often observed in these systems is the reversal of the current direction
115: upon variation of a system parameter \cite{Doering,Elston}. Though in many cases the spatiotemporal symmetry breaking is achieved by using a spatially asymmetric periodic
116: potential, in Ref.~\cite{Chen} it was shown that it can also be
117: obtained by using a spatially symmetric periodic potential which
118: fluctuates between three states, one of them being of free diffusion.
119: As shown in Ref.~\cite{Kanada}, if the symmetry of glide
120: reflection is broken, a directed transport can also be induced in a spatially
121: symmetric potential which switches randomly between two states. Besides the purely theoretical interest, the study of this kind of
122: models is relevant because it may provide valuable information about
123: the workings of the so-called molecular motors
124: \cite{Reimann,Howard}. Molecular motors are able to travel along
125: polymer filaments by utilizing chemical energy from ATP hydrolysis,
126: and are responsible for the transport of different substances inside
127: the eukaryotic cells.
128:
129: Recently, it has been demonstrated experimentally that the Brownian
130: motion of a colloidal sphere can be rectified using a periodic
131: potential that alternates between three \cite{Grier1} or two
132: \cite{Grier2} states which differ only by a discrete translation. Each
133: state is spatially symmetric and consists of a large array of optical
134: traps, created with dynamic holographic optical tweezers, each of which can localize a colloidal sphere. The pattern of discrete optical traps can be modeled as a periodic array of Gaussian wells of width $\sigma$ and depth $V_0$, uniformly separated by a distance $L$ substantially larger than $\sigma$.
135: In the three-state case considered in Ref.~\cite{Grier1}, the initial array of optical traps is extinguished after time $T$ and replaced immediately with a second array, which is displaced from the first by $L/3$. This process is repeated after time $T$ with an additional displacement of $L/3$. A cycle is completed when, after time $T$, the array of traps is displaced once again by $L/3$ and returns to its initial state. In the two-state case reported in Ref.~\cite{Grier2}, the first step is the same as in \cite{Grier1}, i.e., the initial array of optical traps is extinguished after time $T_1$ and replaced immediately with a second array, which is displaced from the first by $L/3$. This second array is extinguished after time $T_2$ and replaced again by the first, thereby completing one cycle. One main
136: result of the three-state case is that the direction of the motion
137: depends sensitively on the duration of the states, $T$, relative to the time
138: required for the particle to diffuse the intertrap separation
139: \cite{Grier1}. By contrast, in the two-state case this direction
140: depends on the relative durations of the two states $T_2/T_1$
141: \cite{Grier2}. Specifically, the flux is directed from each optical trap in the longer-duration state toward the
142: nearest optical trap in the short-lived state, and vanishes for $T_1=T_2$. Thus, in both cases, by varying the value of the
143: corresponding parameter the flux can be reversed.
144:
145: To gain a deeper
146: insight into the behavior of these systems, it would be desirable to
147: develop simplified models, amenable to analytical treatment, which
148: capture the key features of the original complex systems. The main focus of this paper
149: is to present such a simplified model for the experiment described in Ref.~\cite{Grier2}.
150: The paper is organized as follows. In the following section, we introduce the model and provide definitions of the quantities of interest, namely, the long-time velocity and the effective diffusion coefficient. In Sec.~\ref{sect3}, we obtain explicit expressions for these two quantities. The main results arising from these expressions are discussed in Sec.~\ref{sect4}. Finally, we present conclusions for the main findings of our work.
151:
152: \section{Description of the model}
153:
154: In our simplified model the motion of the colloidal sphere in the periodic potential is
155: modelled by a random walk on a one-dimensional lattice with its sites
156: located at $x_n=n h$, with $n\in \mathbb{Z}$, where it has been
157: assumed that the distance between two consecutive sites
158: $x_{n+1}-x_n=h$ is independent of $n$. The lattice can be in two
159: possible states that will be indicated by the label $\alpha=\pm 1$. In
160: contrast with the approach described in Ref.~\cite{Grier2} in which
161: the potential switches periodically between its two states, here we
162: will assume that the lattice fluctuates randomly between its two
163: possible states. The evolution of the joint probabilities
164: $p_n^{(\alpha)}(t)$ that the particle is at site $x_n$ and the lattice
165: in state $\alpha$ at time $t$ are governed by the master equation
166: \begin{eqnarray}
167: \label{equ2}
168: \dot{p}_n^{(\alpha)}(t)&=&-\left[
169: L_{n}^{(\alpha)}+R_{n}^{(\alpha)}\right]
170: p_n^{(\alpha)}(t)+R_{n-1}^{(\alpha)} p_{n-1}^{(\alpha)} (t)\nonumber
171: \\&&+L_{n+1}^{(\alpha)} p_{n+1}^{(\alpha)}(t)+\gamma^{(-\alpha)}
172: p_n^{(-\alpha)}(t)-\gamma^{(\alpha)}
173: p_n^{(\alpha)}(t),\nonumber\\
174: &&
175: \end{eqnarray}
176: where $\gamma^{(\alpha)}$ is the transition rate from the lattice
177: state $\alpha$ to $-\alpha$, and $L_{n}^{(\alpha)}$ and $R_{n}^{(\alpha)}$
178: are the transition rates from site $x_n$ to site $x_{n-1}$ and $x_{n+1}$,
179: respectively, assuming that the lattice is in state $\alpha$. The
180: initial condition is irrelevant since we are interested in the
181: long-time behavior of the system.
182:
183: Retaining the main features of the model described in
184: Ref.~\cite{Grier2}, we will assume that both states of the lattice are
185: spatially symmetric (i.e., there exists a $n'\in \mathbb{Z}$ such that
186: $L_{n'+n}^{(\alpha)}=R_{n'-n}^{(\alpha)}$ for all $n\in \mathbb{Z}$)
187: and periodic with period $N$ (i.e.,
188: $L_{n+N}^{(\alpha)}=L_{n}^{(\alpha)}$ and
189: $R_{n+N}^{(\alpha)}=R_{n}^{(\alpha)}$ for all $n\in \mathbb{Z}$). This
190: is an important distinction between our model and other similar models
191: previously reported in the literature
192: \cite{Freund,Kolomeisky,Ambaye,Makhnovskii}, in which at least one of
193: the lattice states is spatially asymmetric. For simplicity, we will
194: restrict our study to the case $N=3$, which is the smallest value of
195: the period for which a non-zero steady directed motion is possible
196: under the above conditions of spatial symmetry and periodicity
197: \cite{prove}. The lattice state $\alpha=+1$ is obtained from the
198: lattice state $\alpha=-1$ by translating it to the right one site, so
199: that, $L^{(+1)}_{n}=L^{(-1)}_{n-1}$ and $R^{(+1)}_{n}=R^{(-1)}_{n-1}$.
200: When the lattice is in state $\alpha=-1$, there are absorbing sites at
201: the locations $x_{3 j+1}$, for all $j\in \mathbb{Z}$, and
202: consequently, $L^{(-1)}_{3 j+1}=R^{(-1)}_{3
203: j+1}=L^{(+1)}_{3j+2}=R^{(+1)}_{3 j+2}=0$. These absorbing sites play
204: the role of the optical traps in our simplified model. The transition
205: rates to an absorbing and a non-absorbing site will be denoted by $g$
206: and $k$, respectively, so that, $L^{(-1)}_{3 j+2}=R^{(-1)}_{3
207: j}=L^{(+1)}_{3 j}=R^{(+1)}_{3 j+1}=g$ and $R^{(-1)}_{3
208: j+2}=L^{(-1)}_{3 j}=R^{(+1)}_{3j}=L^{(+1)}_{3j+1}=k$. A sketch of the
209: model just described is shown in Fig.~\ref{Fig1}.
210:
211: \begin{figure}
212: \includegraphics[width=8cm]{Fig1.eps}
213: \caption{\label{Fig1} Sketch of the fluctuating lattice. The absorbing
214: sites are depicted by crosses.}
215: \end{figure}
216:
217: It is worthwhile to point out that the model just described is
218: formally equivalent to another random walk model which mimics the motion of
219: some molecular motors. In this model the walker switches
220: randomly between two possible states, indicated by the label
221: $\alpha=\pm 1$, $\gamma^{(\alpha)}$ being the transition rate from
222: the state $\alpha$ to $-\alpha$ (see Fig.~\ref{Fig2}). These states are
223: characterized by the position of two legs, each with a head-shaped object
224: attached at one of their ends. The walker ``walks'' on a rod which possesses an
225: infinite array of binding sites, depicted by grey squares in Fig.~\ref{Fig2}.
226: These binding sites are separated from each other by a distance $3 h$, and
227: translated by $h/2$ with respect to the sites of the lattice. Whenever any
228: of the walker's ``heads'' comes in contact with a binding site, the walker stays
229: trapped at that site until a fluctuation of its state occurs. For each
230: state of the walker, the sites of the lattice with this property
231: (absorbing sites) are depicted by crosses in Fig.~\ref{Fig2}. When
232: none of the walker's ``heads'' are in contact with a binding site, the
233: walker can hop to its nearest-neighbor sites. Then, the transition
234: rate from a site to its neighbors is $g$, if after hopping any of the
235: ``heads'' comes in contact with a binding site, or $k$, otherwise.
236: This interpretation of the model differs from
237: the one considered at the end of Ref.~\cite{Grier2}, in
238: three important aspects. First, ours is discrete and exactly
239: solvable. Second, in ours the walker fluctuates randomly between its two possible states, whereas in theirs it switches periodically between them.
240: In this sense, our model may be more effective
241: for describing real molecular motors, since in this class of molecules
242: the directed motion arises as a consequence of a competition between
243: two stochastic time scales (the ATP binding rate and the ATP
244: hydrolysis rate). Third, in \cite{Grier2} the direction of the steady
245: motion depends only on the relative durations of the two states,
246: whereas here, as we will see later on, it also depends on the sign of $g-k$.
247:
248:
249: \begin{figure}
250: \includegraphics[width=8cm]{Fig2.eps}
251: \caption{\label{Fig2} An alternative interpretation of the model sketched
252: in Fig.~\ref{Fig1}. The binding and absorbing
253: sites are depicted by grey squares and crosses, respectively.}
254: \end{figure}
255:
256: We are interested in the study of the long-time velocity or flux $v$
257: and the effective diffusion coefficient $D_{\mathrm{eff}}$, defined as
258: \begin{equation}
259: \label{vdef}
260: v=\lim_{t\rightarrow \infty} \frac{d}{d t}\langle x(t) \rangle
261: \end{equation}
262: and
263: \begin{equation}
264: \label{Ddef}
265: D_{\mathrm{eff}}=\frac{1}{2}\lim_{t\rightarrow \infty} \frac{d}{d
266: t}\left[\langle x^2(t) \rangle -\langle x(t) \rangle^2 \right],
267: \end{equation}
268: where the $l$th moment of the position is given by $\langle x^l(t)
269: \rangle=h^l \sum_{\alpha=\pm 1} \sum_{n=-\infty}^{+\infty} n^l
270: \,p_n^{(\alpha)}(t)$.
271:
272:
273: There are certain properties of $v$ and $D_{\mathrm{eff}}$ that can be
274: predicted by simple considerations. For instance, in the absence of
275: fluctuations between the two lattice states, both the long-time
276: velocity and the effective diffusion coefficient will be zero, as the
277: particle ends up being trapped by an absorbing site. Even if the sites
278: depicted by crosses in Fig.~\ref{Fig1} were not absorbing and,
279: therefore, there were a non-zero transition rate from them (the same
280: to the left and to the right), the long-time velocity would be also
281: zero in the absence of fluctuations between the two lattice
282: states. This is so because, in that case, both lattice states would be
283: invariant under reflections about any of the sites with
284: crosses. Another interesting property is obtained if one takes into
285: account that the fluctuating periodic lattice sketched in
286: Fig.~\ref{Fig1} is invariant under the transformation
287: $\{n,\alpha\}\Rightarrow \{-n,-\alpha\}$. Consequently, the long-time
288: velocity and the effective diffusion coefficient fulfill the symmetry
289: relations
290: $v\big[\gamma^{(+1)},\gamma^{(-1)}\big]=-v\left[\gamma^{(-1)},
291: \gamma^{(+1)}\right]$ and
292: $D_{\mathrm{eff}}\left[\gamma^{(+1)},\gamma^{(-1)}\right]=D_{\mathrm{eff}}
293: \left[\gamma^{(-1)}, \gamma^{(+1)}\right]$, where we have written
294: explicitly the dependence of $v$ and $D_{\mathrm{eff}}$ on
295: $\gamma^{(+1)}$ and $\gamma^{(-1)}$. Thus, if
296: $\gamma^{(+1)}=\gamma^{(-1)}$, the long-time velocity will be zero.
297:
298: \section{Explicit expressions for the long-time
299: velocity and the effective diffusion coefficient}
300:
301: \label{sect3}
302:
303: In order to go further in the analysis $v$ and $D_{\mathrm{eff}}$, it
304: is necessary to obtain explicit expressions for them. To do that, we
305: will extend the approach developed in Ref.~\cite{Derrida} to the case
306: of a fluctuating lattice. The procedure described here is quite
307: general and can be applied to any fluctuating periodic lattice with an
308: arbitrary period $N$. For this reason, in what follows it will be
309: described in general terms. An alternative approach based on the
310: gradient expansion of the probability distribution corresponding to a
311: coarse-grained description of the random walk model can be found in
312: Ref.~\cite{Freund}. First, let us introduce the quantities
313: \begin{equation}
314: \label{qdef}
315: q_n^{(\alpha)}(t)=\sum_{l=-\infty}^{\infty}p_{n+l N}^{(\alpha)}(t)
316: \end{equation}
317: and
318: \begin{equation}
319: \label{gdef}
320: s_n^{(\alpha)}(t)=h\sum_{l=-\infty}^{\infty}(n+l N)p_{n+l
321: N}^{(\alpha)}(t)-\langle x(t) \rangle q_n^{(\alpha)}(t),
322: \end{equation}
323: which are periodic in $n$ with period $N$, i.e.,
324: $q_{n+N}^{(\alpha)}(t)=q_{n}^{(\alpha)}(t)$ and
325: $s_{n+N}^{(\alpha)}(t)=s_n^{(\alpha)}(t)$ for all $n\in
326: \mathbb{Z}$. Substituting the expressions for the first and second
327: moments of the position into Eqs.~(\ref{vdef}) and (\ref{Ddef}), and
328: replacing $\dot{p}_n^{(\alpha)}(t)$ by the right-hand side of
329: Eq.~(\ref{equ2}), it is not difficult to prove that the long-time
330: velocity and the effective diffusion coefficient can be expressed as
331: \begin{equation}
332: \label{vtdef}
333: v=h \sum_{\alpha=\pm 1} \sum_{n=1}^{N}
334: \left[R_n^{(\alpha)}-L_n^{(\alpha)}\right]Q_n^{(\alpha)}
335: \end{equation}
336: and
337: \begin{eqnarray}
338: \label{dtdef}
339: D_{\mathrm{eff}}&=&\frac{h^2}{2} \sum_{\alpha=\pm 1} \sum_{n=1}^{N}
340: \left[R_n^{(\alpha)}+L_n^{(\alpha)}\right]Q_n^{(\alpha)}\nonumber \\
341: &&+ h \sum_{\alpha=\pm 1} \sum_{n=1}^{N}
342: \left[R_n^{(\alpha)}-L_n^{(\alpha)}\right]S_n^{(\alpha)},
343: \end{eqnarray}
344: where $Q_n^{(\alpha)}=\lim_{t\rightarrow \infty}q_n^{(\alpha)}(t)$ and
345: $S_n^{(\alpha)}=\lim_{t\rightarrow \infty}s_n^{(\alpha)}(t)$. Thus,
346: for the evaluation of $v$ and $D_{\mathrm{eff}}$ it suffices to calculate
347: the values of the $4N$ numbers $Q_n^{(\alpha)}$ and $S_n^{(\alpha)}$,
348: with $n=1,\dots,N$ and $\alpha=\pm 1$.
349:
350: Differentiating Eqs.~(\ref{qdef}) and (\ref{gdef}) with respect to
351: time, making use of Eq.~(\ref{equ2}), as well as of the periodicity of
352: $L_n^{(\alpha)}$ and $R_n^{(\alpha)}$, and taking into account that
353: $\lim_{t\rightarrow \infty}{\dot q}_n^{(\alpha)}(t)=\lim_{t\rightarrow
354: \infty}{\dot s}_n^{(\alpha)}(t)=0$, it is easy to see that the
355: long-time limits $Q_n^{(\alpha)}$ and $S_n^{(\alpha)}$ fulfill the set
356: of $4N$ algebraic equations
357: \begin{eqnarray}
358: \label{calcQ}
359: -\left[L_n^{(\alpha)}+R_n^{(\alpha)}\right]Q_n^{(\alpha)}\;+\;
360: R_{n-1}^{(\alpha)}Q_{n-1}^{(\alpha)}&+&L_{n+1}^{(\alpha)}
361: Q_{n+1}^{(\alpha)}\nonumber \\ +\;\gamma^{(-\alpha)}Q_n^{(-\alpha)}\;-\;
362: \gamma^{(\alpha)}Q_n^{(\alpha)}&=&0,\\
363: \label{calcS}
364: -\left[L_n^{(\alpha)}+R_n^{(\alpha)}\right]S_n^{(\alpha)}\;+\;
365: R_{n-1}^{(\alpha)}S_{n-1}^{(\alpha)}&+&L_{n+1}^{(\alpha)}
366: S_{n+1}^{(\alpha)}\nonumber \\ +\;\gamma^{(-\alpha)}S_n^{(-\alpha)}\;-\;
367: \gamma^{(\alpha)}S_n^{(\alpha)}&=&F_n^{(\alpha)},
368: \end{eqnarray}
369: for $n=1,\dots,N$ and $\alpha=\pm 1$, where
370: \begin{equation}
371: F_n^{(\alpha)}=h\left[L_{n+1}^{(\alpha)}Q_{n+1}^{(\alpha)}
372: -R_{n-1}^{(\alpha)}Q_{n-1}^{(\alpha)}\right]+v\,Q_{n}^{(\alpha)},
373: \end{equation}
374: and where the periodic boundary conditions
375: $Q_0^{(\alpha)}=Q_N^{(\alpha)}$, $Q_{N+1}^{(\alpha)}=Q_1^{(\alpha)}$,
376: $S_0^{(\alpha)}=S_N^{(\alpha)}$, and
377: $S_{N+1}^{(\alpha)}=S_1^{(\alpha)}$ are assumed.
378:
379: Summing up the $2 N$ equations contained either in the set of
380: equations (\ref{calcQ}) or (\ref{calcS}), it is easy to realize that
381: in each set there are only $2N-1$ independent equations (excluding
382: some pathological cases where the number can be even lower). Thus, to
383: univocally determine the $4 N$ quantities $Q_n^{(\alpha)}$ and
384: $S_n^{(\alpha)}$, it is also necessary to add the two additional
385: equations $\sum_{\alpha=\pm 1} \sum_{n=1}^{N} Q_n^{(\alpha)}=1$ and
386: $\sum_{\alpha=\pm 1} \sum_{n=1}^{N} S_n^{(\alpha)}=0$, which can be
387: straightforwardly proved from the definitions of $Q_n^{(\alpha)}$ and
388: $S_n^{(\alpha)}$.
389:
390: To apply the above procedure to the model sketched in Fig.~\ref{Fig1},
391: first one has to solve this set of equations, with $N=3$ and the
392: corresponding values for the transition rates $L_n^{(\alpha)}$ and
393: $R_n^{(\alpha)}$. Replacing the results obtained for $Q_n^{(\alpha)}$
394: and $S_n^{(\alpha)}$ in Eqs.~(\ref{vtdef}) and (\ref{dtdef}), one
395: obtains after some lengthy simplifications that
396: \begin{equation}
397: \label{velocity}
398: v=3 h \Delta \rho \left(1-\Delta \rho^2\right)(g-k)g k \Gamma^2 A^{-1}
399: \end{equation}
400: and
401: \begin{eqnarray}
402: \label{effdiff}
403: D_{\mathrm{eff}}&=&\frac{9}{2} h^2 \left(1-\Delta \rho^2\right)g k
404: (g+k)(2g+\Gamma)\Gamma A^{-1}\nonumber\\ &&-v^2 B \Gamma^{-1}A^{-1},
405: \end{eqnarray}
406: where
407: \begin{eqnarray}
408: \label{Adef}
409: A&=&4 g^2 (g+2k)^2+8 g (g+k)(g+2k)\Gamma\nonumber \\
410: &&+\left[4(g+k)^2-(1+3 \Delta \rho^2) k^2 \right]\Gamma^2,
411: \end{eqnarray}
412: \begin{eqnarray}
413: B&=&4 g^2 (g+2k)^2+24g (g+k)(g+2k)\Gamma\nonumber
414: \\&&+\left[28(g+k)^2-3(3+ \Delta \rho^2) k^2 \right]\Gamma^2+8(g+k)
415: \Gamma^3,\nonumber\\&&
416: \end{eqnarray}
417: $\Gamma=\gamma^{(+1)}+\gamma^{(-1)}$, and $\Delta
418: \rho=\rho_{\mathrm{eq}}^{(-1)}-\rho_{\mathrm{eq}}^{(+1)}$,
419: $\rho_{\mathrm{eq}}^{(\alpha)}$ being the equilibrium population of
420: the lattice state $\alpha$, i.e.,
421: $\rho_{\mathrm{eq}}^{(\alpha)}=\gamma^{(-\alpha)}/\Gamma$. This
422: quantity can alternatively be expressed as $\Delta
423: \rho=\left[T^{(-1)}-T^{(+1)}\right]/\left[T^{(-1)}+T^{(+1)} \right]$,
424: where $T^{(\alpha)}=1/\gamma^{(\alpha)}$ is the mean residence time of
425: the lattice in the state $\alpha$.
426:
427: \section{Results}
428:
429: \label{sect4}
430:
431: An important result that follows from Eq.~(\ref{velocity}) is that the
432: direction of the steady motion not only depends on the sign of $\Delta
433: \rho$, as would be expected from the previous symmetry considerations,
434: but also on the sign of the difference between the transition rates to
435: an absorbing and a non-absorbing site, $g-k$. Thus, for instance, if
436: $T^{(-1)}>T^{(+1)}$, the steady directed motion is from left to right
437: for $g>k$, whereas it is from right to left for $g<k$. Obviously, if
438: $T^{(-1)}<T^{(+1)}$, the behavior is just the opposite. Finally, no
439: steady directed motion is possible if either $T^{(-1)}=T^{(+1)}$ or
440: $g=k$. This subtle dependence of the direction of the steady motion
441: on the values of $g$ and $k$ cannot be inferred {\em a priori} from
442: symmetry considerations, and is one of the main results of this work.
443: It is important to point out that this model predicts correctly the
444: direction of the steady motion of the experiment described in
445: Ref.~\cite{Grier2}. Indeed, in that case, the values of $g$ and $k$
446: can be estimated as the inverses of mean times required for the
447: particle to diffuse a length $L/3-\sigma/2$ and $L/3$, respectively,
448: where $L$ is the intertrap distance and $\sigma$ is the width of an
449: optical trap. Therefore, $g=2D/(L/3-\sigma/2)^2$ and $k=2D/(L/3)^2$,
450: $D$ being the diffusion coefficient of the colloidal sphere in the
451: fluid. Since $g>k$, the motion is from left to right if
452: $T^{(-1)}>T^{(+1)}$ and from right to left if $T^{(-1)}<T^{(+1)}$,
453: i.e., from each absorbing site in the longer-duration state toward the
454: nearest absorbing site in the short-lived state. The above expressions for $g$ and $k$ also show that the finite width of the optical traps plays an important role in the appearance of the steady directed motion.
455:
456:
457: Some limiting values of the expressions (\ref{velocity}) and
458: (\ref{effdiff}) can be easily interpreted. For instance, as suggested
459: by the previous simple considerations, $v$ and $D_{\mathrm{eff}}$ tend
460: to zero both as $\Gamma \rightarrow 0$ (absence of fluctuations
461: between the two lattice states) and as $\Delta \rho \rightarrow \pm 1$
462: (absence of fluctuations between the two lattice states in the
463: long-time limit). As $g\rightarrow 0$, $v$ and $D_{\mathrm{eff}}$ also
464: go to zero because, in this limit, the particle stays trapped between
465: two consecutive absorbing sites, since it is not allowed to jump to
466: them. The expressions for $v$ and $D_{\mathrm{eff}}$ also tend to
467: zero both as $k\rightarrow 0$ and as $g\rightarrow \infty$. This is
468: due to the fact that, in these limits, the motion of the particle
469: consists of a series of forward and backward jumps from a
470: non-absorbing site to its nearest absorbing site, which arise as a
471: consequence of the fluctuations in the lattice state. Finally, the
472: limiting values $\lim_{\Gamma \rightarrow \infty} v=v_{\infty}$ and
473: $\lim_{\Gamma \rightarrow \infty}
474: D_{\mathrm{eff}}=D_{\mathrm{eff},\infty}$, which can be easily
475: calculated from Eqs.~(\ref{velocity}) and (\ref{effdiff}), also have a
476: simple interpretation. Using the results in Ref.~\cite{Derrida}, it is
477: not difficult to check that $v_{\infty}$ and $D_{\mathrm{eff},\infty}$
478: are the long-time velocity and the effective diffusion coefficient
479: corresponding to a random walk with the average transition rates
480: $L_n=\rho_{\mathrm{eq}}^{(+1)}L_n^{(+1)}+\rho_{\mathrm{eq}}^{(-1)}L_n^{(-1)}$
481: and
482: $R_n=\rho_{\mathrm{eq}}^{(+1)}R_n^{(+1)}+\rho_{\mathrm{eq}}^{(-1)}R_n^{(-1)}$.
483: In contrast with what happens in continuum models, the limiting value
484: $v_{\infty}$ does not vanish in general. This difference between
485: discrete and continuum models has been previously pointed out in the
486: literature (see, for example, Ref.~\cite{Ambaye}). From the
487: expressions for $v_{\infty}$ and $D_{\mathrm{eff},\infty}$, it can be
488: easily proved that $\lim_{g\rightarrow \infty} v_{\infty}=3 h \Delta
489: \rho(1-\Delta \rho^2) k/4$ and $\lim_{g\rightarrow
490: \infty}D_{\mathrm{eff},\infty} =9 h^2 (1-\Delta \rho^2) k/8$ which, in
491: general, are not equal to zero. This reflects the fact that the limits
492: $\Gamma \rightarrow \infty$ and $g \rightarrow \infty$ of
493: Eqs.~(\ref{velocity}) and (\ref{effdiff}) do not commute.
494:
495: In order to simplify the graphical presentation of results, it is
496: convenient to introduce dimensionless quantities by defining
497: characteristic length and time scales. The characteristic length scale
498: $L$ is chosen to be the distance between two consecutive absorbing
499: sites, i.e., $L=3 h$. The characteristic time scale $\tau$ is chosen
500: to be the mean time required for the particle to diffuse the length
501: $L$ in the absence of absorbing sites. Taking into account that in
502: the absence of absorbing sites the effective diffusion coefficient is
503: $D_{\mathrm{eff,NA}}=h^2 k$, it follows that $\tau=L^2/(2
504: D_{\mathrm{eff,NA}})=9/(2 k)$. Then, the number of independent
505: dimensionless parameters is reduced to only three (the values of
506: $\tau\Gamma$, $\Delta \rho$, and $\tau g $), since the other two take
507: the fixed values $L^{-1}h=1/3$ and $\tau k=9/2$.
508:
509: \begin{figure}
510: \includegraphics[width=8cm]{Fig3.eps}
511: \caption{\label{Fig3} Dependence of the dimensionless long-time
512: velocity $L^{-1}\tau v$ (top panels) and effective diffusion
513: coefficient $L^{-2}\tau D_{\mathrm{eff}}$ (bottom panels) on $\Delta
514: \rho$ (left panels) and $\log_{10}(\tau g)$ (right panels). The
515: dimensionless parameter values are: $L^{-1}h=1/3$, $\tau k=9/2$, $\tau
516: g=9$ (left panels), $\Delta \rho=0.6$ (right panels), $\tau \Gamma=0$
517: (solid lines), $\tau\Gamma =10$ (dotted lines), and $\tau\Gamma =100$
518: (long-dashed lines). The limiting values $ L^{-1}\tau v_{\infty}$ and
519: $L^{-2}\tau D_{\mathrm{eff,\infty}}$ are plotted with dash-dotted
520: lines.}
521: \end{figure}
522: In Fig.~\ref{Fig3}, we depict the dependence of the dimensionless
523: long-time velocity $L^{-1}\tau v$ (top panels) and effective diffusion
524: coefficient $L^{-2}\tau D_{\mathrm{eff}}$ (bottom panels) on $\Delta
525: \rho$ (left panels) and $\log_{10}(\tau g)$ (right panels). We have
526: taken $\tau g=9$ (left panels), $\Delta \rho=0.6$ (right panels), and
527: three values of $\tau \Gamma$, namely, $\tau \Gamma=0$ (solid lines),
528: $10$ (dotted lines), and $100$ (long-dashed lines). With dash-dotted
529: lines we have plotted the limiting values $ L^{-1}\tau v_{\infty}$ and
530: $L^{-2}\tau D_{\mathrm{eff,\infty}}$. Notice that both the absolute
531: value of $L^{-1}\tau v$ and $L^{-2}\tau D_{\mathrm{eff}}$ are
532: increasing functions of $\tau \Gamma$, except at those points at which
533: these functions vanish. This behavior is confirmed by a detailed study
534: of the signs of the derivatives $\partial v/\partial \Gamma$ and
535: $\partial D_{\mathrm{eff}}/\partial \Gamma$. The reversal of the
536: steady directed motion is clearly illustrated in panels (a) and
537: (b). These panels also reveal that, for fixed values of the remaining
538: parameters, there exist two optimal values of $\Delta \rho$ (opposite
539: to each other) and of $\tau g$ which maximize the absolute value of
540: the long-time velocity in each direction. In panel (b), one can also
541: see that the location of the second maximum is shifted towards larger
542: values of $\tau g$ as $\tau \Gamma$ is increased, and that this
543: maximum completely disappears in the limit $\Gamma \rightarrow
544: \infty$. Panel (c) shows that the dimensionless diffusion coefficient
545: as a function of $\Delta \rho$ is maximized for $\Delta \rho=0$, i.e.,
546: when the mean residence times in each lattice state are
547: equal. Analogously, there is also an optimal value of $\tau g$ which
548: maximizes the dimensionless diffusion coefficient as a function of
549: $\tau g$ [see panel (d)]. The location of this last maximum exhibits a
550: similar behavior with $\tau \Gamma$ as the second maximum in panel
551: (b).
552:
553: \section{Conclusions}
554:
555: \label{sect5}
556:
557: In summary, we have proposed a simple random walk model which allows
558: the analytical study of the flux reversal in a spatially symmetric
559: fluctuating lattice. We have obtained explicit analytical expressions for the long-time velocity and the effective diffusion coefficient and made a study of them.
560: In spite of its simplicity, by appropriately choosing the transition rates to an absorbing and a non-absorbing site,
561: our model is able to reproduce some important features of the experiment described in
562: Ref.~\cite{Grier2}. In particular, it provides the correct direction of the induced
563: flux and shows that this direction is closely related to the finite width of the optical traps. Furthermore, our model predicts the behavior of the effective diffusion coefficient that would be observed in the experiment reported in Ref.~\cite{Grier2}. Finally, it predicts that similar results to those contained
564: in \cite{Grier2} would be observed if the array of optical traps
565: switched randomly between its two possible configurations instead of periodically.
566:
567: We have also presented an alternative interpretation of our model which
568: mimics the motion of some molecular motors. In contrast to the one considered at the end of Ref.~\cite{Grier2}, we have proved that, in ours, the direction of the steady
569: motion depends not only on the relative durations of the two states,
570: but also on the sign of the difference between the transition rates to an absorbing and a non-absorbing site.
571:
572: \begin{acknowledgments}
573: The author acknowledge the support of the Ministerio de Educaci\'on y
574: Ciencia of Spain (FIS2005-02884) and the Junta de Andaluc\'{\i}a.
575: \end{acknowledgments}
576:
577: \begin{thebibliography}{100}
578: \bibitem{Reimann} P. Reimann, Phys. Rep. {\bf 361}, 57 (2002).
579: \bibitem{Hanggi1} R. D. Astumian and P. H\"anggi, Phys. Today {\bf
580: 55}, 33 (2002).
581: \bibitem{Bier} M. Bier and R. D. Astumian, Phys. Rev. Lett. {\bf 76},
582: 4277 (1996).
583: \bibitem{Prost} J. Prost, J. F. Chauwin, L. Peliti, and A. Ajdari,
584: Phys. Rev. Lett. {\bf 72}, 2652 (1994).
585: \bibitem{Chauwin} J. F. Chauwin, A. Ajdari, and J. Prost,
586: Europhys. Lett. {\bf 27}, 421 (1994); {\bf 32}, 373 (1995).
587: \bibitem{Rousselet} J. Rousselet, L. Salome, A. Ajdari, and J. Prost,
588: Nature (London) {\bf 370}, 446 (1994).
589: \bibitem{Faucheux} L. P. Faucheux, L. S. Bourdieu, P. D. Kaplan, and
590: A. J. Libchaber, Phys. Rev. Lett. {\bf 74}, 1504 (1995).
591: \bibitem{Reimann1} P. Reimann, Phys. Rev. Lett. {\bf 86}, 4992 (2001).
592: \bibitem{Doering} C. R. Doering, W. Horsthemke, and J. Riordan, Phys. Rev. Lett. {\bf 72}, 2984 (1994).
593: \bibitem{Elston} T. C. Elston and C. R. Doering, J. Stat. Phys. {\bf 83}, 359 (1996).
594: \bibitem{Chen} Y.-D. Chen, Phys. Rev. Lett. {\bf 79}, 3117 (1997).
595: \bibitem{Kanada} R. Kanada and K. Sasaki, J. Phys. Soc. Jpn. {\bf 68}, 3759 (1999).
596: \bibitem{Howard} J. Howard, {\it Mechanical of Motor Proteins and the
597: Cytoskeleton} (Sinauer Associates, Sunderland, 2001).
598: \bibitem{Grier1} S.-H. Lee, K. Ladavac, M. Polin, and D. G. Grier,
599: Phys. Rev. Lett. {\bf 94}, 110601 (2005).
600: \bibitem{Grier2} S.-H. Lee and D. G. Grier, Phys. Rev. E {\bf 71},
601: 060102(R) (2005).
602: \bibitem{Kolomeisky} A. B. Kolomeisky and B. Widom,
603: J. Stat. Phys. {\bf 93}, 633 (1998).
604: \bibitem{Freund} J. A. Freund and L. Schimansky-Geier, Phys. Rev. E
605: {\bf 60}, 1304 (1999).
606: \bibitem{Ambaye} H. Ambaye and K. W. Kehr. Physica A {\bf 267}, 111
607: (1999).
608: \bibitem{Makhnovskii} Y. A. Makhnovskii, V. M. Rozenbaum, D.-Y. Yang,
609: S. H. Lin, and T. Y. Tsong, Phys. Rev. E {\bf 69}, 021102 (2004).
610: \bibitem{prove} Indeed, if $N=2$, from the periodicity and the
611: symmetry property it follows that $R_1^{(\alpha)}=L_1^{(\alpha)}$ and
612: $R_2^{(\alpha)}=L_2^{(\alpha)}$. Consequently, from Eq.~(\ref{vtdef})
613: one has that $v=0$. Analogously, if $N=1$, $v=0$ trivially.
614: \bibitem{Derrida} B. Derrida, J. Stat. Phys. {\bf 31}, 433 (1983).
615:
616:
617:
618: \end{thebibliography}
619:
620:
621: \end{document}
622:
623:
624:
625:
626:
627:
628: