1:
2: \documentclass[prl,twocolumn,preprintnumbers]{revtex4}
3: \usepackage{graphicx,psfrag,dcolumn,bm,amsmath,amssymb}
4: \usepackage[dvips]{color}
5:
6:
7: \renewcommand{\vec}[1]{\boldsymbol{#1}}
8: \newcommand{\avg}[1]{\left\langle #1\right\rangle}
9: \newcommand{\mint}[4]{\int_{#2}^{#3}\!\!#1\,#4}
10: \newcommand{\fex}{\mathfrak f}
11: \newcommand{\Or}[1]{{\cal O}(#1)}
12:
13: \newcommand{\crm}[1]{\color{red}{#1 }\color{black}}
14: \newcommand{\rem}[1]{\textbf{#1}}
15:
16: \newcounter{fnt}
17:
18:
19: \begin{document}
20:
21:
22:
23: \title{Propagation and Relaxation of Tension in Stiff Polymers}
24:
25: \author{Oskar Hallatschek$^1$} \author{Erwin Frey$^{1,2}$}
26: \author{Klaus Kroy$^{1,3}$} \affiliation{$^1$Hahn--Meitner Institut,
27: Glienicker Stra\ss e 100, 14109 Berlin, Germany \\ $^2$Arnold
28: Sommerfeld Center, Department of Physics,
29: Ludwigs-Maximilians-Universit\"at M\"unchen, Theresienstra\"se 37,
30: 800333 M\"unchen, Germany \\ $^3$Institut f\"ur Theoretische Physik,
31: Universit\"at Leipzig, Augustusplatz 10/11, 04109 Leipzig, Germany}
32:
33:
34: \date{\today}
35:
36:
37: \begin{abstract}
38: We present a unified theory for the longitudinal dynamic response of
39: a stiff polymer in solution to various external perturbations
40: (mechanical excitations, hydrodynamic flows, electrical fields,
41: temperature quenches \dots) that can be represented as sudden
42: changes of ambient/boundary conditions. The theory relies on a
43: comprehensive analysis of the non-equilibrium propagation and
44: relaxation of backbone stresses in a wormlike chain. We recover and
45: substantially extend previous results based on heuristic arguments.
46: New experimental implications are pointed out. (LMU-ASC 15/05)
47: \end{abstract}
48:
49:
50:
51: \maketitle
52:
53: Despite considerable practical and interdisciplinary interest, it is
54: theoretically not yet fully understood how polymers respond to
55: external fields \cite{perkins-smith-chu:97,bohbot-raviv-etal:2004}.
56: Consider, e.g., the simple problem of an inextensible semiflexible
57: polymer suddenly stretched along its end-to-end vector by an external
58: force $\fex$ (\emph{Pulling}). It was recognized before
59: \cite{seifert-wintz-nelson:96} that tension propagation (from the ends
60: into the bulk) is the key to understanding its dynamics: in response
61: to the spreading tension, the polymer stretches within a growing
62: boundary layer of length $\ell_\|$, as depicted in Fig.~\ref{fig:def}.
63: Depending on the setup, different tension propagation laws
64: $\ell_\|(t)$ have been
65: predicted~\cite{seifert-wintz-nelson:96,everaers-etal:99,%
66: brochard-buguin-de_gennes:99,shankar-pasquali-morse:2002}. In
67: particular, we contrast the above \emph{Pulling}-scenario with the
68: (inverse) \emph{Release}-scenario, where a constant pre-stretching
69: force $\fex$ is suddenly removed. While for small $\fex$, one expects
70: $\ell_\|(t)\propto t^{1/8}$ in both cases \cite{everaers-etal:99}, the
71: predictions for strong force are markedly different:
72: $\ell_\|(t)\propto (\fex\, t)^{1/4}$ \cite{seifert-wintz-nelson:96}
73: for \emph{Pulling}, and $\ell_\|(t)\propto \fex^{3/4} t^{1/2}$
74: \cite{brochard-buguin-de_gennes:99} for \emph{Release}. However, the
75: precise meaning of ``strong'' and ``weak'', and the validity of the
76: diverse assumptions in
77: Refs.~\cite{seifert-wintz-nelson:96,brochard-buguin-de_gennes:99} are
78: not obvious. Here we develop from first principles a theory of stress
79: propagation and relaxation that allows us to unify and systematically
80: extend these previous results, and to derive (often analytically) the
81: longitudinal nonlinear response to various external perturbations.
82: After introducing the standard model of a semiflexible polymer, we
83: extend a heuristic argument of Ref.~\cite{everaers-etal:99} for
84: \emph{Pulling}. This elucidates the crossover from ``weak-'' to
85: ``strong-force'' behavior and reveals the crucial length-scale
86: separation underlying our subsequent systematic analysis.
87:
88:
89:
90: \begin{figure}[t]
91: \includegraphics[width=\columnwidth]{driven-chain-2bein-psfraged.eps}
92: \caption{\emph{Pulling} (schematic): In response to an external force
93: $\fex$, the thermally undulated contour $\vec r(s)=(\vec
94: r_{\!\perp},s-r_\|)^T$ is straightened within boundary layers of
95: growing width $\ell_\|(t)$.}
96: \label{fig:def}
97: \end{figure}
98:
99:
100:
101: The wormlike chain model represents the polymer by a smooth
102: inextensible contour $\vec r(s,t)$ subject to the energy
103: \begin{equation}\label{eq:h}
104: {\cal H}=\frac\kappa2 \mint{ds}{0}{L}{\vec r''^2} +
105: \frac12\mint{ds}{0}{L}{f\vec r'^2}\;.
106: \end{equation}
107: The scalar force $f(s,t)$ (backbone ``tension'') is a Lagrange
108: multiplier function introduced \cite{goldstein-langer:95} to enforce
109: the inextensibility constraint $\vec r'^2=1$ for the tangent vector
110: $\vec r'\equiv \partial \vec r/\partial s$. We require a bending
111: stiffness $\kappa$ such that the persistence length
112: $\ell_p=\kappa/k_BT$ is much larger than the contour length $L$, which
113: entails relative mean square transverse displacements of order
114: $\epsilon\equiv L/\ell_p\ll1$. The elastic forces derived from $\cal
115: H$ have to balance thermal forces $\vec \xi$ (Gaussian white noise)
116: and Stokes friction, which (in the ``free-draining'' approximation)
117: enters through two local friction coefficients per unit length
118: $\zeta_\perp$, $\zeta_\|\approx \zeta_\perp/2$ for motion
119: perpendicular and parallel to $\vec r'$, respectively: $[\zeta_\| \vec
120: r'\vec r'+ \zeta_\perp(1- \vec r'\vec r')]\cdot\partial_t\vec
121: r=-\delta {\cal H}/\delta\vec r+ \vec \xi$. For the following, we
122: choose convenient units such that $\kappa \equiv \zeta_\perp \equiv
123: 1$. Then, all dimensional quantities represent powers of length
124: (e.g.~$k_B T=\ell_p^{-1}$) and $\zeta_\|\equiv\zeta\approx1/2$.
125:
126: We now turn to a heuristic discussion of \emph{Pulling} to leading
127: order in $\epsilon$. In the parameterization introduced in
128: Fig.~\ref{fig:def}, the exact equations of motion reduce to an
129: equation for the transverse displacements $\vec r_\perp$ alone,
130: \begin{equation}\label{eq:eom1}
131: \partial_t \vec r_{\!\perp} = -\vec r_{\!\perp}''''+f \vec r_{\!\perp}'' +
132: \vec\xi_{\!\perp} \;,
133: \end{equation}
134: with a spatially uniform tension $f=\fex\, \Theta(t)$ fixed by the
135: driving force at the boundaries; $\langle\vec\xi_{\!\perp}(s,t) \cdot
136: \vec\xi_{\!\perp}(s',t') \rangle=4 \ell_p^{-1} \delta(s-s')
137: \delta(t-t')$. The (higher order) longitudinal displacements $r_\|$
138: are slaved by the arclength constraint $r_\|'=\vec r_{\!\perp}'^2/2$.
139:
140: From a simple scaling analysis of Eq.~(\ref{eq:eom1}), $\avg{\vec
141: r_{\!\perp}}\!/t\approx \avg{\vec r_{\!\perp}}(\ell_{\!\perp}^{-4}+
142: \fex\, \ell_{\!\perp}^{-2})$, we deduce the characteristic dynamic
143: wavelength $\ell_{\!\perp}(t)$ corresponding to the (lowest) mode
144: equilibrated at time $t$. For instance,
145: $\ell_{\!\perp}(t_L^\perp)\equiv L$ defines the longest relaxation
146: time. Due to the competition between bending forces ($\propto
147: r_\perp\ell_{\!\perp}^{-4}$) and tension ($\propto r_\perp \fex\,
148: \ell_{\!\perp}^{-2}$), the growth of $\ell_{\!\perp}$ exhibits a
149: dynamic crossover from free relaxation ($\ell_{\!\perp}\propto
150: t^{1/4}$) to relaxation under tension ($\ell_{\!\perp}\propto
151: \sqrt{\fex\, t}$) at a characteristic time $t_\fex\equiv \fex^{-2}$
152: (Tab.~\ref{tab:pulling-growth-laws}/left).
153:
154: By the above interpretation of $\ell_{\!\perp}$, the longitudinal
155: elongation of a subsection of arclength $\ell_{\!\perp}$ under a given
156: tension $\fex$ can be estimated by equilibrium theory. One has to
157: distinguish weak and strong tension relative to the internal
158: characteristic force scale $\ell_{\!\perp}^{-2}$ of the subsection,
159: which corresponds to the Euler buckling force of the subsection. For
160: \emph{weak} tension $\fex\ll \ell_{\!\perp}^{-2}$, the elongation
161: $\fex\, \ell_{\!\perp}^4 \ell_p^{-1}$ follows from linear response
162: \cite{mackintosh-kaes-janmey:95}. For \emph{strong} tension $\fex\gg
163: \ell_{\!\perp}^{-2}$ the subsection is virtually straight, so that the
164: elongation is equal to its equilibrium thermal contraction
165: $\ell_{\!\perp}^2\ell_p^{-1}$ caused by the bending undulations. Since
166: the whole polymer is subject to the same constant tension $\fex$, it
167: can be divided (at any time $t$) into $L/\ell_{\!\perp}(t)$
168: independent equilibrated subsections of length $\ell_{\!\perp}(t)$.
169: The total change $\Delta R(t)\equiv \lvert\avg{R(t)-R(0)}\rvert$ of
170: the end-to-end distance $R$ thus crosses over from $\Delta R\propto
171: L\fex\,\ell_{\!\perp}^{3} \ell_p^{-1}\propto t^{3/4}$ \cite{granek:97}
172: for $t\ll t_\fex$ to $\Delta R\propto L\ell_{\!\perp}
173: \ell_p^{-1}\propto t^{1/2}$ for $t\gg t_\fex$.
174:
175: \begin{table}[t]
176: \caption{The transverse equilibration length
177: $\ell_{\!\perp}(t)$ and the tension propagation length
178: $\ell_\|(t)$ both exhibit a crossover at $t_\fex\equiv\fex^{-2}$
179: (here, for the \emph{Pulling} problem with $\fex\gg L^{-2}$, $t\ll
180: t_L^\perp$).}
181: \label{tab:pulling-growth-laws}
182: \begin{ruledtabular}
183: \begin{tabular}{r|cc|c} &
184: $\ell_{\!\perp}(t)$ & & $\ell_\|(t)$ \qquad \quad \mbox{}\\ \hline
185: $t\ll t_\fex$ & $t^{1/4}$ & & $t^{1/8} (\ell_p/ \zeta )^{1/2}$
186: \hfill \cite{everaers-etal:99}\\
187: $t\gg t_\fex$ & $t^{1/2}\fex^{1/2}$ & & $t^{1/4}\fex^{1/4}
188: (\ell_p/\zeta)^{1/2}$ \quad \cite{seifert-wintz-nelson:96}
189: \end{tabular}
190: \end{ruledtabular}
191: \end{table}
192:
193:
194: These results comprise the predictions of ordinary perturbation theory
195: (OPT) to leading order. As evident from Eq.~(\ref{eq:eom1}), it neglects
196: longitudinal friction forces, which are of higher order in $\epsilon$.
197: However, the resulting $\Delta R$ obtained above implies a total
198: longitudinal friction $\zeta L \partial_t \Delta R$ crossing over from
199: $\zeta L^2 \fex\, t^{-1/4} \ell_p^{-1}$ to $\zeta
200: L^2(\fex/t)^{1/2}\ell_p^{-1}$ at $t_\fex$. Both expressions diverge
201: \cite{morse:98,everaers-etal:99} for $t\to 0$ indicating the breakdown
202: of OPT at short times. More precisely, for
203: times shorter than a certain $t_\star$~\footnote{$t_\star=(\zeta
204: L^2\!/\ell_p)^4$ if $\fex \ll \ell_p^2/L^4$ and $t_\star=(\zeta
205: L^2\!/\ell_p)^2 \fex^{-1}$ if $\fex\gg\ell_p^2/L^4$. See
206: Fig.~\ref{fig:tf}.} the longitudinal friction would exceed the
207: driving force $\fex$. This was recognized as a consequence of the
208: omission of tension propagation in Eq.~(\ref{eq:eom1}): it was argued
209: \cite{everaers-etal:99} that actually only segments up to a distance
210: $\ell_\|(t)$~\footnote{Identifying $\ell_{\perp,\|}$ as correlation
211: length of $\partial_t r_{\perp,\|}$, respectively, justifies the
212: notation.} from the ends are set into longitudinal motion causing
213: longitudinal friction. The proper expression for the total
214: longitudinal friction thus follows from the above upon substituting
215: $L$ by $\ell_\|$. The balance of longitudinal friction and external
216: force can now be satisfied by choosing the size $\ell_\|(t)$ of the
217: boundary layer according to Tab.~\ref{tab:pulling-growth-laws}/right.
218: Hence, the putative ``weak- and strong- force'' cases
219: $\ell_\|\propto t^{1/8}$ \cite{everaers-etal:99} and $\ell_\|\propto
220: t^{1/4}$ \cite{seifert-wintz-nelson:96} are identified as asymptotes of
221: a ``short-long time'' crossover (still) signaling the change from
222: ``free'' to ``forced'' relaxation at $t=t_\fex$.
223:
224:
225: In summary, the foregoing discussion reveals: (\emph{i}) the breakdown
226: of OPT at times $t<t_\star$, where
227: (\emph{ii}) the neglected longitudinal friction actually limits the
228: relaxation to boundary layers of size $\ell_\|$; (\emph{iii}) the
229: crossover from free to forced relaxation at $t=t_\fex$; (\emph{iv})
230: the \emph{scale separation} $\ell_{\!\perp}/\ell_\| \propto
231: \epsilon^{1/2}\ll 1$ (Tab.~\ref{tab:pulling-growth-laws}).
232:
233: The latter lends itself as starting point for a \emph{multiple-scale}
234: calculus to separate the physics on different dynamic scales and
235: obtain an improved (``multiple-scale'') perturbation theory (MSPT)
236: that is regular in the limit $t\to0$ while $\epsilon\ll1$ is fixed.
237: The procedure is similar to that for athermal dynamics
238: \cite{hallatschek-frey-kroy:04} and will be documented in detail
239: elsewhere \cite{hallatschek-etal:tbp}. The basic idea is to regard
240: functions $g(s)$ as functions $g(s,\bar s\epsilon^{1/2})$ of rapidly
241: and slowly varying arclength parameters $s$ and $\bar s
242: \epsilon^{1/2}$ that are treated as independent variables. Closed
243: equations for the dynamics on the scale $\bar s \epsilon^{1/2}$ are
244: obtained upon averaging $\bar g(\bar s)\equiv
245: \mint{ds}{l}{\phantom{l}}{g(s,\bar s \epsilon^{1/2})/l}$ over the
246: fluctuations on the microscale ($\ell_{\!\perp}\ll l \ll \ell_\|$).
247: To leading order, we get \cite{hallatschek-etal:tbp} $f= \bar f(\bar
248: s,t)$,
249: \begin{equation}\label{eq:eom2}
250: \partial_t \vec r_{\!\perp} = -\vec r_{\!\perp}''''+ \bar f\vec
251: r_{\!\perp}'' + \vec \xi_{\!\perp} \;, \quad \text{and} \quad
252: \partial_{\bar s}^2\bar f = - \zeta \overline {\partial_t r_\|'} \;.
253: \end{equation}
254: This provides the sought-after rigorous local generalization of the
255: above, heuristically renormalized force balance. The arclength
256: average extending over many ($l/\ell_{\!\perp}\gg1$) uncorrelated
257: sections of length $\ell_{\!\perp}$ subject to the same tension $\bar
258: f(\bar s)$ can be interpreted as a coarse-graining that
259: \emph{effectively generates a local ensemble average}:
260: $\overline{\partial_t r_\|'} \sim \langle \partial_t r_\|'\rangle$ for
261: $\epsilon\to0$. Only the ``systematic'' $\Or1-$variations of the
262: tension are retained, while its $\Or\epsilon-$noise is leveled out, so
263: that the longitudinal Eq.~(\ref{eq:eom2}) represents
264: \emph{deterministic} dynamics: local longitudinal motion is driven by
265: tension gradients (like in a thread pulled through a viscous medium).
266:
267:
268: Integrating the longitudinal Eq.~(\ref{eq:eom2}) over time expresses
269: the change of the thermal contraction $r_\|$ in terms of the
270: time-integrated tension $F \equiv \mint{d t'}{0}{t}{f}$, namely
271: $\bigl\langle r_\|'(\bar s,t)-r_\|'(\bar s,
272: 0)\bigr\rangle=-\partial_{\bar s}^2\bar F/\zeta$. Since $2r_\|'= \vec
273: r_{\!\perp}'^2$ from the arc\-length constraint, we integrate the
274: transverse Eq.~(\ref{eq:eom2}) for $\langle\vec
275: r_{\!\perp}'^2\rangle(\bar s,t)=
276: {\bigl\langle\bigl[\frac1L\sum_q\mint{dt'}{-\infty}{t}{q
277: \chi_{\!\perp}(q;t,t')\vec\xi_{\!\perp}(q,t')}\bigr]^2\bigr\rangle}$.
278: Here
279: \begin{equation}
280: \label{eq:tszbility}
281: \chi_{\!\perp}(q;t,t')\equiv e^{ - q^4(t-t')-q^2[\bar
282: F(\bar s, t)-\bar F(\bar s,t')]}
283: \end{equation}
284: should be recognized as the susceptibility for the response of the
285: Fourier modes of $\vec r_{\! \perp}$ to transverse forces. Note that
286: the $\bar s-$dependence of $\langle\vec r_{\!\perp}'^2\rangle$ is
287: purely adiabatic, as it is parametrically inherited from $\bar F(\bar
288: s,t)$. Altogether, Eq.~(\ref{eq:eom2}) is condensed into a single
289: equation for $\bar F$~\footnote{The right-hand side of this equation
290: has recently independently been derived in a related context
291: \cite{bohbot-raviv-etal:2004}.},
292: \begin{equation}\label{eq:master}
293: \frac{\partial_{\bar s}^2\bar
294: F}\zeta =\mint{\frac{dq}{\pi}}{0}{\infty}{\left[
295: \frac{1-\chi_{\!\perp}^2(q;t,0)}{\ell_p^-(q^2+f^-)} -\frac{2
296: q^2}{\ell_p}\mint{dt'}{0}{t}\chi^2_\perp(q;t,t')\right]}\;.
297: \end{equation}
298: Indices ``$-$'' referring to $t<0$ were introduced to allow the system
299: to be prepared in equilibrium with ambient/boundary conditions
300: different from those prescribed for $t\geq0$. Taking $\ell_p$ (for
301: $t\geq0$) to $\infty$, only the first term in the integrand remains,
302: which thus accounts for the deterministic relaxation of the initial
303: thermal contraction (set by $\ell_p^-$, $f^-$). In this limit, the
304: ``zero-temperature'' buckling dynamics analyzed in
305: Ref.~\cite{hallatschek-frey-kroy:04} is recovered. For finite
306: $\ell_p$ the second term in the brackets represents the action of the
307: thermal forces for $t>0$, which aim to establish the equilibrium
308: contraction.
309:
310:
311:
312: To further unravel the physical implications of Eq.~(\ref{eq:master}),
313: we begin with the tension propagation regime $\ell_\|\ll L$, where $L$
314: is irrelevant. Problems like \emph{Pulling} and \emph{Release} still
315: depend on four length scales ($\ell_p^-=\ell_p, \fex^{-1/2},s ,
316: t^{1/4}$). Yet, Eq.~(\ref{eq:master}) is \emph{solved exactly}
317: by the two-variable scaling ansatz
318: \begin{equation}\label{eq:xover}
319: f(\bar s,t)= \fex \; \Phi(\bar s/s_\fex, t/t_\fex) \;.
320: \end{equation}
321: With $t_\fex\equiv \fex^{-2}$ and $s_\fex\equiv
322: (\ell_p/\zeta)^{1/2}t_\fex^{1/8}$ Eq.~(\ref{eq:xover}) can be shown to
323: eliminate the parameter dependencies in Eq.~(\ref{eq:master}) and in
324: its boundary conditions. The scaling function $\Phi$ describes how
325: sudden changes of the tension at the ends spread into the bulk of the
326: polymer. In the limits $t \ll t_\fex$ and $t\gg t_\fex$
327: Eq.~(\ref{eq:xover}) reduces to the simple (one-variable) scaling
328: laws $\Phi\sim (t/t_\fex)^\alpha \phi (\bar s/\ell_\|)$ with
329: the tension propagation length $\ell_\|\equiv s_\fex\,(t/t_\fex)^z$.
330: Note that this crossover scenario, as well as the expressions for
331: $t_\fex$, $s_\fex\approx\ell_\|(t_\fex)$, and $\ell_\|$ are consistent
332: with our above heuristic observations for \emph{Pulling}
333: (Tab.~\ref{tab:pulling-growth-laws}). In fact, this structure
334: generally emerges for all problems involving a single external force
335: scale. It is implicitly understood that $ \phi$, $\alpha$, and
336: $z$ will generally not only depend on the kind of external
337: perturbation applied, but will also be different in both scaling
338: limits $t/t_\fex\to 0,\infty$. In the following, these limits are
339: analyzed in more detail.
340:
341: For $t\ll t_\fex$ Eq.~(\ref{eq:master}) may be linearized in $f$ and
342: the scaling function $ \phi$ can be obtained analytically
343: \cite{hallatschek-etal:tbp}. In contrast to $ \phi$, the
344: corresponding exponent $z=1/8$ does not depend on the boundary
345: conditions. It already follows from requiring $\phi$ to become
346: $\fex-$independent, as in linear response. The short-time dynamics
347: for strong external force is thus closely related to the linear
348: response. As established by our heuristic discussion of
349: \emph{Pulling}, this is due to the relaxation of subsections with
350: Euler forces $\ell_{\!\perp}^{-2}$ much larger than the external
351: force. Note, however, that the limit $\fex\to0$ is problematic, as it
352: does not interchange with $\epsilon\to0$~\footnote{Our
353: identification of arclength averages with (local) ensemble
354: averages after Eq.~(\ref{eq:eom2}) breaks down for
355: $\fex<(\zeta/\ell_p)^{1/4}t^{-7/16}$, where fluctuations in the
356: tension become comparable to its average
357: \cite{hallatschek-etal:tbp}.}. In fact, extending
358: Eq.~(\ref{eq:master}) to linear response amounts to an uncontrolled
359: factorization approximation $\avg{f r_{\!\perp}^2}\to
360: \avg{f}\avg{r_{\!\perp}^2}$.
361:
362: For $t\gg t_\fex$ the dynamics becomes nonlinear in the external force
363: and starts to depend on the force protocol. Previously predicted
364: power laws can be recovered from Eq.~(\ref{eq:master}) by employing
365: different approximations to its right hand side. In the
366: \emph{taut-string approximation} of
367: Ref.~\cite{seifert-wintz-nelson:96} one neglects for $t>0$ bending and
368: thermal forces against the tension, i.e.\ one drops the $q^4-$term in
369: the expression Eq.~(\ref{eq:tszbility}) for $\chi_\perp$ and sets
370: $\ell_p\to\infty$. The complementary \emph{quasi-static
371: approximation} of Ref.~\cite{brochard-buguin-de_gennes:99} amounts
372: to the omission of memory effects, i.e.\ to the assumption of
373: instantaneous equilibration of tension and stored length,
374: $F(t)-F(t')\to f(t)(t-t')$. A careful analysis of
375: Eq.~(\ref{eq:master}) \cite{hallatschek-etal:tbp} shows that either of
376: these approximations becomes rigorous in the intermediate asymptotic
377: regime defined by $t\gg t_\fex$, $\ell_\|\ll L$. The ``pure''
378: \footnote{We postpone ``mixed'' scenarios (e.g.\ strong stretching of
379: a pre-stretched polymer) that involve more than one external force
380: scale. Contrary to previous expectations
381: \cite{seifert-wintz-nelson:96,brochard-buguin-de_gennes:99}, they
382: lead to multiple crossovers even for $t\gg t_\fex$.} scenarios are
383: summarized in Fig.~\ref{fig:lf} and for the cases \emph{Pulling} and
384: \emph{Release} also in Fig.~\ref{fig:tf} and Tab.~\ref{tab:tf}.
385:
386: \setcounter{fnt}{\value{footnote}}
387:
388: In Fig.~\ref{fig:lf} we have moreover displayed results for sudden
389: changes in persistence length from $\ell_p^-$ to $\ell_p\neq\ell_p^-$
390: (``$\ell_p-$\emph{Quench}'') and \emph{Electrophoretic Pulling}, which
391: have not been discussed before. The second is a variant of the
392: \emph{Pulling} problem, where the external force is applied along the
393: whole contour of an end-grafted polymer as is the case for a
394: hydrodynamic flow or an electric field. The $\ell_p$-\emph{Quench}
395: is exceptional in that there is \emph{no external force scale}
396: involved, so that Eq.~(\ref{eq:master}) can be solved by a simple
397: one-variable scaling ansatz $f(\bar s,t)=t^{-1/2}\varphi (\bar
398: s/\ell_\|)$ with $\ell_\| \approx (\ell_p/\zeta)^{1/2}t^{1/8}$.
399: Neither the \emph{taut-string approximation} nor the
400: \emph{quasi-static approximation} can be applied, and in contrast to
401: the other cases the scaling function has to be evaluated numerically.
402:
403: \begin{figure}[t]
404: \includegraphics[width=\columnwidth]{growth-laws-psfraged.eps}
405: \caption{Double-logarithmic sketch of the tension propagation laws
406: $\ell_\|(t)\propto t^z$. At $t_\fex=\fex^{-2}$ they cross over
407: from a universal short-time regime to (problem-specific)
408: tension-dominated intermediate asymptotics, except for weak
409: forces, $\fex < \ell_p^2/L^4$, and for
410: $\ell_p-$\emph{Quenches}. The propagation ends when
411: $\ell_\|(t)\approx L$.} \label{fig:lf}
412: \end{figure}
413:
414:
415: Eventually, at a time $t_L^\|$, the tension will have propagated
416: through the whole polymer, i.e.\ $\ell_\|(t_L^\|)\approx L$.
417: Subsequently, the characteristic longitudinal scale is the contour
418: length $L$. One would expect that regular perturbation theory would
419: then become valid. Surprisingly, for $t_L^\|\gg t_\fex$ the
420: \emph{Release} scenario provides an exception. The contraction
421: dynamics exhibits an intermediate regime of \emph{homogeneous tension
422: relaxation} (white in Fig.~\ref{fig:tf}). Its asymptotic power-law
423: form is revealed by the same \emph{quasi-static approximation}
424: applicable during the preceding tension propagation, but with the
425: separation ansatz $f(\bar s,t)\sim h(\bar s)(\zeta L^2\!/\ell_p t
426: )^{2/3}$ instead of the single-variable scaling ansatz. It solves
427: Eq.~(\ref{eq:master}) analytically with a roughly parabolic stationary
428: tension profile $h(\bar s)$. The homogeneous tension relaxation
429: dominates the short-time relaxation for $t\ll t_\star=(\zeta
430: L^2/\ell_p)^4$ if $\fex\to\infty$ (i.e.~$t^\|_L \to0$).
431:
432:
433: \begin{figure}[t]
434: \includegraphics[width=\columnwidth]{lelo-vs-pulling-psfraged.eps}
435: \caption{Characteristic times (logarithmic scale) for \emph{Pulling} and
436: \emph{Release} against the applied external force $\fex$ (linear
437: scale). The time $t_\star$ (stars) separates regions where ordinary
438: perturbation theory (OPT) applies (dark shaded) from regions (light
439: shaded) of linear (hatched) and nonlinear tension propagation and
440: from homogeneous tension relaxation (white). Whereas longitudinal
441: friction is negligible for $t>t_\star$, it limits the dynamics for
442: $t<t_\star$.}\label{fig:tf}
443: \end{figure}
444:
445: To make contact with experiments, it is desirable to derive the
446: consequences for pertinent observables such as the end-to-end
447: distance. Integrating the longitudinal Eq.~(\ref{eq:eom2}) over $\bar
448: s$ and $t$ shows that the growth laws $\Delta \bar R(t)=
449: 2\zeta^{-1}\left|\bar F'(\bar s=0,t)\right|$ are directly related to
450: the tension profiles discussed above. Tab.~\ref{tab:tf} summarizes our
451: results for the intermediate asymptotic regimes. Note that $\Delta
452: \bar R$ is a coarse-grained quantity that does not resolve the
453: ``microscopic'' details below the coarse-graining scale $l$. Near the
454: polymer ends these give relevant contributions obliterating the
455: predicted $t^{7/8}$ in experiments
456: \cite{everaers-etal:99,legoff-etal:2002}. During homogeneous tension
457: relaxation $\Delta \bar R\propto t^{1/3}$, which we expect to hold for
458: strongly stretched polymers even if $L\gg\ell_p$ (e.g.\ DNA), at
459: variance with earlier predictions \cite{brochard-buguin-de_gennes:99}.
460: The exponent $1/3$ coincides with that obtained by adiabatic
461: application of the stationary force-extension relation
462: \cite{bustamante-marko-siggia-smith:94} to a ``frictionless''
463: \cite{bohbot-raviv-etal:2004} polymer with attached beads at its ends.
464: Finally, $\Delta \bar R\approx (\zeta \ell_p)^{-1/2}t^{3/8}$ in
465: $\ell_p-$\emph{Quenches} for $t\ll t_L^\|$. Interestingly, the tension
466: propagation/relaxation itself can in some situations be directly
467: monitored experimentally. In $\ell_p$-\emph{Quenches}, the
468: (longitudinal) radius of gyration mirrors the characteristic bulk
469: relaxation $f\propto t^{-1/2}$ of the tension
470: \cite{hallatschek-frey-kroy:04}. In \emph{Electrophoretic-Pulling},
471: where $\Delta \bar R\propto t$ for $t\ll t_L^\|$, the force on the
472: grafted end obeys $\fex\propto\ell_\|$.
473:
474:
475: In conclusion, we have developed a unified theory of non-equilibrium
476: tension dynamics in stiff polymers based on the scale separation
477: between the two dynamic correlation lengths $\ell_\perp$ and
478: $\ell_\|$. The recovered known results and our new predictions are
479: summarized in Figs.~\ref{fig:lf},~\ref{fig:tf} and Tab.~\ref{tab:tf}.
480: Various dynamic regimes should be well realizable for certain
481: biopolymers and it is an intriguing question, whether the tension
482: propagation laws $\ell_\|(t)$ govern mechanical signal transduction
483: through the cytoskeleton
484: \cite{shankar-pasquali-morse:2002,gardel-etal:2003}. Inclusion of
485: hydrodynamic interactions merely produce logarithmic corrections but
486: would give rise to more interesting effects for membranes. Other
487: natural generalizations including the transverse nonlinear response
488: and more complex force protocols (e.g.~\footnotemark[\value{fnt}]) are
489: currently also under investigation.
490:
491: \begin{table}[t]
492: \caption{Growth laws for the end-to-end distance $\Delta \bar
493: R(t)$ in the intermediate asymptotic regimes marked in
494: Fig.~\ref{fig:tf}. OPT and MSPT refer to ``ordinary perturbation
495: theory'' and ``multiple-scale perturbation theory'',
496: respectively.}\label{tab:tf}
497: \begin{ruledtabular}
498: \begin{tabular}{l|c|c} & {\emph{Release}} & {\emph{Pulling}} \\
499: \hline linear MSPT &
500: \multicolumn{2}{c}{ $2^{{\frac{5}8}^{\phantom 2}}\!\!\!\Gamma(\frac{15}8)^{-1}
501: (\zeta \ell_p)^{-\frac{1}2}\,\fex\, t^{\frac78}$} \quad\mbox{} \\ linear OPT
502: \cite{granek:97} & \multicolumn{2}{c}{$2^{-\frac34}
503: \Gamma(\frac74)^{-1}(L/\ell_p)\,\fex\, t^{\frac34} $} \\ nonlin.~MSPT
504: & $ 3.503\, (\zeta
505: \ell_p)^{-\frac12}\,\fex^{\frac14}t^{\frac12}$ & $
506: \left(\frac{512}{9\pi}\right)^{\!\frac14} (\zeta
507: \ell_p)^{-\frac12} (\fex \, t)^{\frac34}$\\ hom.~relaxation & ${\displaystyle
508: 2\!\cdot\!54^{\frac23} (L/\zeta \ell_p^2)^{\frac13} t^{\frac13}}$ & --- \\
509: nonlin.~OPT & ${ 2^{\frac34}\Gamma(\frac14)^{-1} (L/\ell_p)\,
510: t^{\frac14}} $ & ${(2/\pi)^{\frac12}(L/\ell_p)
511: (\fex\, t)^{\frac12}}$
512: \end{tabular}
513: \end{ruledtabular}
514: \end{table}
515:
516: %\bibliographystyle{apsrev}
517: %\bibliography{bibis/semiflexibleA04.bib,bibis/journals,bibis/mysf,bibis/mysfnet,bibis/unpub,bibis/sf}
518:
519: %\end{document}
520:
521: \begin{thebibliography}{16}
522: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
523: \expandafter\ifx\csname bibnamefont\endcsname\relax
524: \def\bibnamefont#1{#1}\fi
525: \expandafter\ifx\csname bibfnamefont\endcsname\relax
526: \def\bibfnamefont#1{#1}\fi
527: \expandafter\ifx\csname citenamefont\endcsname\relax
528: \def\citenamefont#1{#1}\fi
529: \expandafter\ifx\csname url\endcsname\relax
530: \def\url#1{\texttt{#1}}\fi
531: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
532: \providecommand{\bibinfo}[2]{#2}
533: \providecommand{\eprint}[2][]{\url{#2}}
534:
535: \bibitem[{\citenamefont{Perkins et~al.}(1997)\citenamefont{Perkins,
536: Smith, and Chu}}]{perkins-smith-chu:97}
537: \bibinfo{author}{\bibfnamefont{T.~T.} \bibnamefont{Perkins}},
538: \bibinfo{author}{\bibfnamefont{D.~E.} \bibnamefont{Smith}},
539: \bibnamefont{and}
540: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Chu}},
541: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{276}},
542: \bibinfo{pages}{2016}
543: (\bibinfo{year}{1997});
544: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Maier}},
545: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Seifert}}, \bibnamefont{and}
546: \bibinfo{author}{\bibfnamefont{J.~O.} \bibnamefont{R{\"a}dler}},
547: \bibinfo{journal}{Europhys.~Lett.} \textbf{\bibinfo{volume}{60}},
548: \bibinfo{pages}{622} (\bibinfo{year}{2002});
549: \bibinfo{author}{\bibfnamefont{N.-K.} \bibnamefont{Lee}} \bibnamefont{and}
550: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Thirumalai}},
551: \bibinfo{journal}{Biophys. J.} \textbf{\bibinfo{volume}{86}},
552: \bibinfo{pages}{2641} (\bibinfo{year}{2004}).
553:
554:
555:
556: \bibitem[{\citenamefont{Bohbot-Raviv et~al.}(2004)\citenamefont{Bohbot-Raviv,
557: Zhao, Feingold, Wiggins, and Granek}}]{bohbot-raviv-etal:2004}
558: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Bohbot-Raviv}},
559: \bibinfo{author}{\bibfnamefont{W.~Z.} \bibnamefont{Zhao}},
560: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Feingold}},
561: \bibinfo{author}{\bibfnamefont{C.~H.} \bibnamefont{Wiggins}},
562: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Granek}},
563: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{92}},
564: \bibinfo{pages}{098101} (\bibinfo{year}{2004}).
565:
566: \bibitem[{\citenamefont{Seifert et~al.}(1996)\citenamefont{Seifert, Wintz, and
567: Nelson}}]{seifert-wintz-nelson:96}
568: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Seifert}},
569: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wintz}}, \bibnamefont{and}
570: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Nelson}},
571: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{77}},
572: \bibinfo{pages}{5389} (\bibinfo{year}{1996}).
573:
574:
575: \bibitem[{\citenamefont{Everaers et~al.}(1999)\citenamefont{Everaers,
576: J{\"u}licher, Ajdari, and Maggs}}]{everaers-etal:99}
577: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Everaers et al.}},
578: %\bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{J{\"u}licher}},
579: %\bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ajdari}},
580: %\bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~C.}
581: %\bibnamefont{Maggs}} et~al.,
582: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{82}},
583: \bibinfo{pages}{3717} (\bibinfo{year}{1999}).
584:
585: \bibitem[{\citenamefont{Brochard-Wyart
586: et~al.}(1999)\citenamefont{Brochard-Wyart, Buguin, and
587: de~Gennes}}]{brochard-buguin-de_gennes:99}
588: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Brochard-Wyart}},
589: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Buguin}},
590: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.~G.}
591: \bibnamefont{de~Gennes}}, \bibinfo{journal}{Europhys.~Lett.}
592: \textbf{\bibinfo{volume}{47}}, \bibinfo{pages}{171}
593: (\bibinfo{year}{1999}).
594:
595:
596: \bibitem[{\citenamefont{Shankar et~al.}(2002)\citenamefont{Shankar, Pasquali,
597: and Morse}}]{shankar-pasquali-morse:2002}
598: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Shankar}},
599: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Pasquali}}, \bibnamefont{and}
600: \bibinfo{author}{\bibfnamefont{D.~C.} \bibnamefont{Morse}},
601: \bibinfo{journal}{J.~Rheol.} \textbf{\bibinfo{volume}{46}},
602: \bibinfo{pages}{1111} (\bibinfo{year}{2002}).
603:
604: \bibitem[{\citenamefont{Goldstein and Langer}(1995)}]{goldstein-langer:95}
605: \bibinfo{author}{\bibfnamefont{R.~E.} \bibnamefont{Goldstein}}
606: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~A.}
607: \bibnamefont{Langer}}, \bibinfo{journal}{Phys.~Rev.~Lett.}
608: \textbf{\bibinfo{volume}{75}}, \bibinfo{pages}{1094} (\bibinfo{year}{1995}).
609:
610: \bibitem[{\citenamefont{MacKintosh et~al.}(1995)\citenamefont{MacKintosh,
611: K{\"a}s, and Janmey}}]{mackintosh-kaes-janmey:95}
612: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{MacKintosh}},
613: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{K{\"a}s}}, \bibnamefont{and}
614: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Janmey}},
615: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{75}},
616: \bibinfo{pages}{4425} (\bibinfo{year}{1995}).
617:
618:
619: \bibitem[{\citenamefont{Granek}(1997)}]{granek:97}
620: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Granek}},
621: \bibinfo{journal}{J.~Phys.~II~France} \textbf{\bibinfo{volume}{7}},
622: \bibinfo{pages}{1761} (\bibinfo{year}{1997});
623: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Gittes}} \bibnamefont{and}
624: \bibinfo{author}{\bibfnamefont{F.~C.} \bibnamefont{MacKintosh}},
625: \bibinfo{journal}{Phys.~Rev.~E} \textbf{\bibinfo{volume}{58}},
626: \bibinfo{pages}{R1241} (\bibinfo{year}{1998}).
627:
628: \bibitem[{\citenamefont{Morse}(1998)\citenamefont{Morse}}]{morse:98}
629: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Morse}},
630: \bibinfo{journal}{Macromol.} \textbf{\bibinfo{volume}{31}},
631: \bibinfo{pages}{7044} (\bibinfo{year}{1998}).
632:
633:
634:
635:
636: \bibitem[{\citenamefont{Hallatschek et~al.}(2004)\citenamefont{Hallatschek,
637: Frey, and Kroy}}]{hallatschek-frey-kroy:04}
638: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Hallatschek}},
639: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Frey}}, \bibnamefont{and}
640: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kroy}},
641: \bibinfo{journal}{Phys.~Rev.~E} \textbf{\bibinfo{volume}{70}},
642: \bibinfo{pages}{031802} (\bibinfo{year}{2004}).
643:
644: %\bibitem[{\citenamefont{Hallatschek et~al.}()\citenamefont{Hallatschek, Frey,
645: % and Kroy}}]{hallatschek-frey-kroy:04}
646: %\bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Hallatschek}},
647: % \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Frey}}, \bibnamefont{and}
648: % \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kroy}},
649: % \bibinfo{note}{cond-mat/0402082}.
650:
651: \bibitem[{\citenamefont{Hallatschek~\emph{et al.}}()}]{hallatschek-etal:tbp}
652: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Hallatschek et al.}},
653: \bibinfo{note}{in preparation}.
654:
655: \bibitem[{\citenamefont{Le~Goff et~al.}(2002)\citenamefont{Le~Goff,
656: Hallatschek, Frey, and Amblard}}]{legoff-etal:2002}
657: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Le~Goff et al.}},
658: % \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Hallatschek}},
659: % \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Frey}}, \bibnamefont{and}
660: % \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Amblard}},
661: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{89}},
662: \bibinfo{pages}{258101} (\bibinfo{year}{2002}).
663:
664: \bibitem[{\citenamefont{Bustamante et~al.}(1994)\citenamefont{Bustamante,
665: Marko, Siggia, and Smith}}]{bustamante-marko-siggia-smith:94}
666: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Bustamante et al.}},
667: % \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Marko}},
668: % \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Siggia}}, \bibnamefont{and}
669: % \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Smith}},
670: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{265}},
671: \bibinfo{pages}{1599} (\bibinfo{year}{1994}).
672:
673: \bibitem[{\citenamefont{Gardel et~al.}(2003)\citenamefont{Gardel, Valentine,
674: Crocker, Bausch, and Weitz}}]{gardel-etal:2003}
675: \bibinfo{author}{\bibfnamefont{M.~L.} \bibnamefont{Gardel et al.}},
676: % \bibinfo{author}{\bibfnamefont{M.~T.} \bibnamefont{Valentine}},
677: % \bibinfo{author}{\bibfnamefont{J.~C.} \bibnamefont{Crocker}},
678: % \bibinfo{author}{\bibfnamefont{A.~R.} \bibnamefont{Bausch}},
679: % \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.~A.}
680: % \bibnamefont{Weitz}},
681: \bibinfo{journal}{Phys.~Rev.~Lett.} \textbf{\bibinfo{volume}{91}},
682: \bibinfo{pages}{158302} (\bibinfo{year}{2003}).
683:
684: \end{thebibliography}
685:
686: \end{document}
687: