---
title: Closed-String Polymer Model
url: https://www.emergentmind.com/topics/closed-string-polymer-model
type: topic
---

# Closed-String Polymer Model

Searching arXiv for the named topic and closely related closed-polymer/string models.
arxiv_search.query:
{"search_query":"all:\"Conservation of polymer winding states: a combinatoric approach\" OR ti:\"Conservation of polymer winding states\"","start":0,"max_results":5}
arxiv_search.query:
{"search_query":"all:\"Toy models of multibranched polymers: opened vs. circular structures\" OR ti:\"Toy models of multibranched polymers\"","start":0,"max_results":5}
arxiv_search.query:
{"search_query":"all:\"The closed piecewise uniform string revisited\" OR ti:\"closed piecewise uniform string revisited\"","start":0,"max_results":5}
A closed-string polymer model is a description of a polymer as a **closed one-dimensional object** whose admissible configurations are constrained by closure and, in some formulations, by a conserved topological invariant such as winding number. In the most explicit statistical-mechanical construction, the polymer is a **closed loop wound around an infinitely long rod**, and the model is built from half-space sub-arcs, topology-preserving augmentation rules, and a constrained partition function over valid arc sequences [1509.03528]. Closely related closed-loop frameworks include Gaussian polymers with **closed-ring backbones** and **closed-loop side branches**, semiflexible closed loops bound to a cylinder, and a **closed piecewise uniform string** used as an exactly solvable analogue of a periodic closed one-dimensional medium [2201.09053], [1503.01023], [2012.14301].

## 1. Definition and scope

In the winding-state formulation, the physical object is a **closed polymer loop** wound around an **infinitely long rod** placed along the \(y\)-axis in \(\mathbb R^3\). The polymer is projected into the plane perpendicular to the rod, and the central topological quantity is the **winding number** \(w\), the number of times the loop winds around the rod before closing. Once the polymer has been closed, and assuming no strand crossing or passage through the rod, \(w\) is a topological invariant [1509.03528].

The invariant is explicitly related to the planar winding-angle integral
\[
\oint \frac{x \dot y - y \dot x}{x^2+y^2}\,ds,
\]
with \(s\) the arc length. This formulation turns the model into a constrained statistical problem: the partition function is not taken over all closed curves, but only over those deformations that preserve the winding state [1509.03528].

Within the broader literature, the same emphasis on closure appears in several adjacent models. Gaussian multibranched polymers can be organized around either an **open backbone** or a **closed polymer ring**, with exact size ratios showing that loop closure compactifies the structure [2201.09053]. Semiflexible polymers constrained to a cylinder admit **closed completely bound states** labeled by a winding number and an oscillation number [1503.01023]. The **closed piecewise uniform string** provides a distinct but related relativistic analogue of a periodic closed one-dimensional object, explicitly noted as resembling a polymer ring or other closed periodic medium [2012.14301].

## 2. Winding-state construction around a rod

The rod-winding model decomposes the closed loop into **sub-arcs** obtained by cutting the loop at each piercing of the \(xy\)-plane. Each arc lies entirely in one half-space, either \(z>0\) or \(z<0\), so a polymer configuration becomes a sequence of arc weights [1509.03528].

For the simplest conserved state \(w=1\), the minimal loop consists of two arcs: one crossing from the \(x>0\) side to the \(x<0\) side and one making the reverse crossing. These are denoted \(T_{+-}\) and \(T_{-+}\). In condensed notation,
\[
T_{+-},\;T_{-+}\rightarrow T_c,\qquad T_{++},\;T_{--}\rightarrow T_s,
\]
where \(T_c\) is a **crossing arc** and \(T_s\) is a **same-side arc** [1509.03528].

The closed-loop condition is encoded locally: a product \(T_{\alpha\beta}T_{\gamma\delta}\) is valid only if \(\beta=\gamma\), so consecutive arc factors connect into a continuous strand. For higher winding number \(w>1\), the simplest loop is the repeated sequence
\[
(T_cT_c)^w.
\]
This is the basic \(w\)-fold configuration from which more elaborate but topologically equivalent strings are generated [1509.03528].

The same logic reappears in other closed-polymer settings, although with different geometric constraints. For a semiflexible polymer on a cylinder, closure requires \(\Delta\varphi=2\pi p\), where \(p\) is the winding number around the cylinder and is a topological invariant; the loop length is written \(L=2\pi R\), and the ratio \(q=n/p\) appears as the effective wavenumber of the small-amplitude solution [1503.01023]. This suggests a common structural theme: closed-loop polymer models are often indexed by discrete topological data that survive continuous deformations.

## 3. Topology-preserving augmentations and reducibility

The combinatoric core of the rod model is the statement that extra arc segments may be inserted without changing the winding number, provided the insertion corresponds to a **Reidemeister type-II move**. Two elementary augmentation procedures are allowed for \(w=1\) [1509.03528].

First, one may **insert a pair of same-side arcs**,
\[
T_c \longrightarrow T_c\,T_s\,T_s.
\]
Second, one may **pull a same-side arc across the rod**,
\[
T_s \longrightarrow T_c\,T_s\,T_c.
\]
The paper also records compound forms,
\[
T_x \longrightarrow T_x\,T_s\,T_s,\qquad
T_s \longrightarrow (T_c)^n\,T_s\,(T_c)^n.
\]
These operations do not break the strand, do not permit passage through the rod, and add only local topologically trivial structure, so the winding class is preserved [1509.03528].

For \(w=1\), admissible strings must be closed, have matching first and last indices, connect properly between adjacent arc factors, contain an even total number of \(T\)'s, contain even numbers of both \(T_c\) and \(T_s\), and be **algorithmically reducible** to the basic loop. The reduction scheme introduces
\[
G_n=T_c(T_s)^{2n},\qquad U_n=T_c(T_s)^{2n+1},
\]
with
\[
G_n\leftrightarrow G_0,\qquad U_n\leftrightarrow U_0,
\]
together with the relations
\[
X\,U_0^{2n}\,Y \leftrightarrow XY,\qquad
G_0^{m}U_0G_0^{m}\leftrightarrow U_0.
\]
Repeated application of these rules reduces a valid string to \(G_0G_0\), i.e. to the basic \(w=1\) loop [1509.03528].

The appendix relates this construction to the two-strand braid group \(B_2\). The braid-language winding number is
\[
w=\frac{\#(\sigma)-\#(\sigma^{-1})}{2},
\]
but braid relations alone are not sufficient because the polymer partition function depends on geometric placement and on the distinction between \(T_c\) and \(T_s\). The \(T\)-sequence formalism therefore supplements topological equivalence with position-dependent statistical weights [1509.03528].

## 4. Statistical weights and topologically constrained partition functions

For a flexible polymer, each arc is modeled as a **Gaussian random walk confined to a half-space** with an absorbing boundary at the plane. The model assumes a minimal length scale \(\epsilon\), takes each sub-arc to begin and end at distance \(\epsilon\) from the \(xy\)-plane, approximates the trans-plane connector as normal to the plane, and treats each arc as an independent random walk in a half-space [1509.03528].

The half-space Green’s function is
\[
T(\vec r,\vec r_0,s,\epsilon)=(2\pi s)^{-3/2}
\left[e^{-(\vec r-\vec r_0)^2/2s}-e^{-(\vec r+\vec r_0)^2/2s}\right].
\]
With \(\vec r_0\) and \(\vec r\) both at height \(\epsilon\), this becomes
\[
T(\vec r,\vec r_0,s,\epsilon)=(2\pi s)^{-3/2}
e^{-[(x-x_0)^2+(y-y_0)^2]/2s}\left[1-e^{-\epsilon^2/s}\right].
\]
For \(\epsilon\ll1\), the arc weight used in the model is
\[
T_p(\vec r,\vec r_0,s,\epsilon)=
\frac{\epsilon^2}{\sqrt{(2\pi)^3 s^5}}
e^{-[(x-x_0)^2+(y-y_0)^2]/2s},
\]
with parity distinguishing crossing and same-side sectors [1509.03528].

For a valid configuration \(\chi\) with \(N\) piercings, the constrained partition function is
\[
Z^{(w=1)}_\chi(L)=
\int_0^\infty dX \int_{-\infty}^\infty dY \int_\epsilon^\infty dS\;
\delta(x_0-x_N)\delta(y_0-y_N)
\prod_{j=1}^N T_{p_j}(x_{j-1},x_j;y_{j-1}-y_j;s_j;\epsilon)\,
\delta\Big(\sum_{k=1}^N s_k-L\Big).
\]
After Laplace transformation in \(L\) and Fourier transformation in the \(y\)-differences,
\[
\tilde Z^{(w=1)}_\chi(t)=
\int_0^\infty dX\;\delta(x_0-x_N)\int_{-\infty}^\infty dk\;
\prod_{i=1}^N T^{L,F}_{p_i}(x_{i-1},x_i;k;t).
\]
The transformed weights \(T^{L,F}\) are the basic building blocks of the topologically constrained partition function [1509.03528].

The full fixed-winding partition function is then approximated from above and below. The lower-bound approximation is
\[
Z^{(w)}_{\mathrm{appx1}}=
\left\{
\frac{T_c}
{\left(1-T_sT_s\right)
\left[
1-T_sT_c\left(1-T_sT_s\right)^{-1}
T_sT_c\left(1-T_sT_s\right)^{-1}
\right]}
\right\}^{2w},
\]
and satisfies
\[
Z^{(w)}_{\mathrm{appx1}}\le Z^{(w)}.
\]
An upper-bound construction introduces effective weights \(T_c^{\mathrm{eff}}\) and \(T_s^{\mathrm{eff}}\) through nonlinear recursion relations and yields
\[
Z^{(w)}_{\mathrm{appx2}}=
\left[T_c^{\mathrm{eff}}T_c^{\mathrm{eff}}\right]^w,
\]
with
\[
Z^{(w)}_{\mathrm{appx1}}\le Z^{(w)}\le Z^{(w)}_{\mathrm{appx2}}.
\]
The lower bound is emphasized because the upper-bound recursion is harder to evaluate explicitly [1509.03528].

Observable counting is implemented by source fields,
\[
T_s\to e^{g_s}T_s,\qquad T_c\to e^{g_c}T_c,
\]
so that
\[
\langle N_c\rangle=
\left[\frac{\partial}{\partial g_c}\log Z^{(w)}(g_c,g_s)\right]_{g_c=g_s=0},
\qquad
\langle N_s\rangle=
\left[\frac{\partial}{\partial g_s}\log Z^{(w)}(g_c,g_s)\right]_{g_c=g_s=0}.
\]
These derivatives count the mean numbers of crossing and same-side arc types in the topologically constrained ensemble [1509.03528].

## 5. Two-slit confinement, force, and arc composition

The combinatoric model is made explicit in a geometry where the wound loop passes between **two slits** separated by distance \(d\) and width \(\Delta\). In this case the admissible \(x\)-domain is
\[
x_i\in\left[\frac d2,\frac d2+\Delta\right].
\]
This geometry allows direct calculation of average length, free energy, force on the slit, and the relative abundance of arc types [1509.03528].

For zero slit width, \(\Delta=0\), the transformed same-side and crossing weights simplify to
\[
T_s^{F}(d;k;s_i)=\frac{\sqrt{3}\,\epsilon^2 e^{-k^2 s_i/6}}{2s_i^2},
\]
\[
T_c^{F}(d;k;s_i)=\frac{\sqrt{3}\,\epsilon^2 e^{-3d^2/(2s_i)} e^{-k^2 s_i/6}}{2s_i^2}.
\]
The Laplace-transformed same-side term is
\[
T_s^{L,F}(d;k;t)=
\int_\epsilon^\infty ds_i\;e^{-s_it}\,
\frac{\sqrt{3}\epsilon^2 e^{-k^2 s_i/6}}{2s_i^2},
\]
and for the crossing term the lower integration limit is approximated by \(d^2\) rather than \(\epsilon\), because the dominant contribution comes from \(s_i\gtrsim d^2\) [1509.03528].

The average polymer length in the lower-bound scheme is
\[
\langle L\rangle(t)=
-\frac{\partial}{\partial t}
\log\left[\tilde Z^{(w=1)}_{\mathrm{appx1}}(t,d)\right].
\]
Large \(t\) corresponds to short typical polymer length. In this limit the undressed basic configuration \(T_cT_c\) dominates, and increasing slit separation \(d\) raises the minimal allowed length [1509.03528].

The probability that the loop is exactly the basic winding state is
\[
P(T_c^2)=
\frac{\int_{-\infty}^{\infty} dk\;(T_c^{L,F})^2}
{\tilde Z^{(w=1)}_{\mathrm{appx1}}(t,d)}.
\]
This probability tends to \(1\) for sufficiently large \(t\), so the short-chain limit is dominated by the bare \(T_cT_c\) loop. Correspondingly, \(\langle N_c\rangle\to2\) and \(\langle N_s\rangle\to0\) at large \(t\) [1509.03528].

The free energy is
\[
F=-\log Z.
\]
As the polymer approaches its minimal allowed length, the free energy rises sharply because entropy is reduced. For finite slit width, the Fourier-transformed arc weights acquire explicit \(\Delta\)-dependence, and the slit force is defined by
\[
f(t,d,\Delta)=
-\frac{\partial}{\partial \Delta}
\left(
-\log\left[\tilde Z^{(w=1)}_{\mathrm{appx1}}(t,d,\Delta)\right]
\right)
=
\frac{\partial}{\partial \Delta}\log \tilde Z.
\]
At small separation \(d\), the slit compresses the polymer and the force is compressive; at larger \(d\), the polymer becomes stretched and the force changes sign. The ratio
\[
\frac{\langle N_s\rangle}{\langle N_c\rangle}
\]
is nonmonotonic in \(d\): it is small for large \(d\), rises as same-side segments become more common at smaller \(d\), and decreases again when additional crossing opportunities reappear [1509.03528].

## 6. Related closed-loop polymer and ring architectures

A second major line of closed-string polymer modeling is the **Gaussian Edwards continuous-chain** treatment of polymers with a closed backbone. In this framework the polymer is represented by a continuous trajectory \(\vec r(s)\), \(s\in[0,L]\), with connectivity Hamiltonian
\[
H^0=\frac12\int_0^L\left(\frac{d\vec r(s)}{ds}\right)^2.
\]
Two topologies are compared: an **open backbone** generalized bottlebrush and a **closed backbone** decorated ring, each carrying \(f_c\) linear branches and \(f_r\) closed loops at \(n\) branching points [2201.09053].

The universal size ratio is
\[
g=\frac{\langle R_g^2\rangle_{\rm branched}}
{\langle R_g^2\rangle_{\rm chain}}.
\]
Exact path-integral expressions are obtained for both architectures, and the central structural conclusion is that **closed loops reduce the effective size more strongly than linear side chains**, while **decorated rings are more compact than open bottlebrushes** at the same molecular weight and branch content. The compactification strengthens as \(n\) increases, and in the special case \(f_c=0, f_r=1\), the model resembles loop-extrusion-type compaction in chromatin [2201.09053].

A mechanically distinct closed-loop model is the **closed semiflexible polymer bound to a cylinder**, where equilibrium states are classified by the winding number \(p\) and oscillation number \(n\). Closed completely bound states exhibit conserved axial torque and contact force, and for \(p\ge2\) a topological obstruction prevents complete unbinding. In particular, the short-loop state with \(p=2\) and \(n=1\) provides a stable constriction of the cylinder and partially unbinds as the length is increased [1503.01023]. This model shares with the rod-winding construction the central role of conserved topology, but replaces Gaussian sub-arcs by Euler-elastica bending energy,
\[
H_B=\frac12\int ds\,\kappa(s)^2.
\]

An exactly solvable analogue of a heterogeneous closed one-dimensional medium is the **closed piecewise uniform string**, composed of \(2N\) alternating uniform segments of equal length \(a=L/(2N)\). The transverse displacement field \(\phi(\sigma,\tau)\) satisfies the \(1+1\)-dimensional wave equation, the wave speed is kept constant across junctions, and the spectrum is obtained through transfer matrices and Chebyshev polynomials [2012.14301]. The model is explicitly described as analogous to polymer-like closed systems and periodic structures such as \(\delta\)-rings or Kronig–Penney lattices. Its exact solvability yields explicit eigenfrequencies, vacuum energy, free energy, and entropy, with the notable property that the vacuum energy has **no ultraviolet divergence** because the heat-kernel coefficient \(a_1\) vanishes when the wave speed is constant [2012.14301].

## 7. Terminological distinctions

The expression “closed string” is used in several non-equivalent ways, and the polymer usage should be distinguished from at least two others that appear in adjacent literatures. In polymer crystallization, a **string reaction coordinate** is a non-linear string in order-parameter space, constructed from observables such as \(U\) and \(Q_6\); it is explicitly a mathematical string used to parametrize transition progress, not the polymer itself and not a topologically closed ring [1501.03936].

The term is also distinct from **closed-string theory** in high-energy physics. Papers on Regge closed-string scattering, open/closed topological string duality, Hosotani symmetry breaking in closed-string theory, and closed-string viewpoints on ABJM matrix models concern string-theoretic amplitudes, matrix integrals, or background fields rather than statistical polymer loops [1001.4843], [1907.02410], [1307.3352], [1607.06414]. In that setting, “closed string” refers to left-right factorized fundamental strings or topological-string partition functions, not to a polymer ring or a combinatorially constrained loop.

Within polymer and soft-matter usage, therefore, a closed-string polymer model is most precisely understood as a model of a **closed polymeric contour** whose admissible conformations are shaped by closure, topology, and geometric environment. The rod-winding model provides a direct topological partition-function construction [1509.03528]; decorated rings provide exact Gaussian size comparisons [2201.09053]; cylinder-bound elastica show how winding topology stabilizes constriction [1503.01023]; and the closed piecewise uniform string supplies an exactly tractable analogue of a periodic heterogeneous ring-like medium [2012.14301].

Source: https://www.emergentmind.com/topics/closed-string-polymer-model