---
title: Multi-Chain Sliplink Model in Polymer Rheology
url: https://www.emergentmind.com/topics/multi-chain-sliplink-model
type: topic
---

# Multi-Chain Sliplink Model in Polymer Rheology

Searching arXiv for the cited slip-link and primitive-chain-network papers to ground the article in the current literature.
The multi-chain sliplink model is a coarse-grained description of entangled polymer systems in which topological constraints are represented by slip-links: junctions that connect chain segments, permit contour sliding, and couple chain conformations to stress, relaxation, and topology renewal. In the literature considered here, the term covers equilibrium weakly slip-linked Gaussian chains, Brownian multi-chain melt models built from Rouse chains and mobile slip-links, and primitive-chain-network formulations in which sliplinks, chain ends, and permanent cross-links define a dynamical network of Gaussian strands [1104.2952], [1306.2107], [2011.03225], [2508.01526].

## 1. Conceptual scope and model architectures

The literature suggests that the multi-chain sliplink model is best understood as a family of related formulations rather than a single unique equation set. The common element is the replacement of many-body topological constraints by explicit slip-links or sliplink nodes, while retaining Gaussian-chain elasticity and stochastic dynamics.

In the equilibrium formulation of Uneyama and Horio, one considers weakly slip-linked Gaussian polymer chains and treats the slip-link count in a grand ensemble with an effective chemical potential $\epsilon$ per slip-link, while chains are also placed in a grand ensemble with chemical potential $\mu$ per chain [1104.2952]. In the numerical melt model studied in 2013, each polymer is represented as a Rouse chain of $N_m$ beads of size $b$, and each entanglement is represented by a “slip-link,” that is, a small ring which can slide along a given chain but is elastically tethered to a fixed “anchor” point in space [1306.2107]. In primitive-chain-network simulations, entangled polymer melts are represented as dynamical networks of Gaussian strands connected by slip-links and permanent cross-links; each slip-link enforces transient pairing and allows sliding of polymer contour length, whereas each cross-link fixes connectivity and forbids any sliding [2011.03225].

| Formulation | Representation | Principal emphasis |
|---|---|---|
| Weakly slip-linked Gaussian chains [1104.2952] | $M$ Gaussian chains with weak slip-link constraints | partition function, free energy, pressure |
| Brownian melt model [1306.2107] | $N_p$ Rouse chains with $Z=N_m/N_e$ slip-links | Langevin dynamics, renewal, rheology |
| Primitive-chain-network model [2011.03225], [2508.01526] | Gaussian strands, sliplinks, chain ends, and possibly cross-links | contour transfer, topology change, nonlinear flow |

A central distinction among these versions concerns how multi-chain character is introduced. In the 2013 model there is no explicit inter-chain spring; coupling among chains enters through slip-link destruction and re-creation on randomly chosen chain ends so that the overall density of slip-links in the box remains uniform [1306.2107]. In the primitive-chain-network model, by contrast, each polymer chain is a path from one dangling end to another, alternating strands and sliplink nodes, and both node positions and strand contour lengths are dynamical variables [2508.01526].

## 2. Equilibrium statistical mechanics of weak slip-linking

For $M$ non-interacting Gaussian chains of contour length $N$, the single-chain partition function is
$$
\mathcal Z_1=\frac{V\,\mathcal Q}{\Lambda^3},
$$
so the canonical partition function of the ideal system is
$$
\mathcal Z^{(0)}=\frac{1}{M!}\bigl(\mathcal Z_1\bigr)^M
=\frac{V^M\mathcal Q^M}{M!\Lambda^{3M}}.
$$
Weak slip-linking is introduced through the lowest-order non-trivial cluster of two chains joined by one slip-link, implemented by the constraint $\mathbf R_1(s_1)=\mathbf R_2(s_2)$ via a Dirac delta. Under the hypothesis that slip-links themselves form an ideal “gas,” the total slip-link count is treated in a grand ensemble with chemical potential $\epsilon$, and the chain count is treated in a grand ensemble with chemical potential $\mu$ [1104.2952].

The resulting Helmholtz free energy admits a low-density expansion in the average number of slip-link points per chain,
$$
\widetilde Z\equiv-\frac{2}{M}\frac{\partial\mathcal F}{\partial\epsilon},
$$
for which
$$
\mathcal F=\mathcal F^{(0)}-\tfrac12\,k_BT\,M\,\widetilde Z+O(\widetilde Z^2),
\qquad
\mathcal F^{(0)}=k_BT\,M\Bigl[\ln\!\frac{M\Lambda^3}{V\mathcal Q}-1\Bigr].
$$
Thus, to leading order, the excess free energy from slip-links is
$$
\Delta\mathcal F\equiv\mathcal F-\mathcal F^{(0)}
=-\tfrac12\,k_BT\,M\,\widetilde Z.
$$
This establishes that the equilibrium statistics of a slip-linked system is different from one of the corresponding ideal chain system without any constraints by slip-links [1104.2952].

The same treatment fixes the regime of validity. The assumptions are: chains are ideal Gaussian, slip-links bind two chain-segments in real space, slip-links do not interact among themselves, only low-order slip-link clusters are retained, the treatment is grand-canonical in both chains and slip-links, self-slip-links are shown to vanish and so are omitted, and any repulsive correction is treated in a virial expansion valid for weak added interactions [1104.2952].

## 3. Pressure reduction, effective attraction, and repulsive correction

The pressure derived from the free energy is
$$
P=-\Bigl(\frac{\partial\mathcal F}{\partial V}\Bigr)_{T,M}
=\frac{k_BT\,M}{V}
-\frac12\,\frac{k_BT\,M}{V}\,\widetilde Z
+O(\widetilde Z^2).
$$
Equivalently,
$$
P-P^{(0)}=-\tfrac12\,k_BT\,\frac{M}{V}\,\widetilde Z+O(\widetilde Z^2),
\qquad
P^{(0)}=\frac{k_BT\,M}{V}.
$$
The pressure of a slip-linked system therefore decreases compared with the ideal system [1104.2952].

In this formulation, the pressure decrease implies that slip-linked chains spontaneously form aggregated cluster like compact structures. The paper identifies the origin as entropic: each slip-link reduces the number of available configurations for two chain-segments, hence an entropic loss that mimics an attractive potential. This is expressed as a negative second virial and is formally identical to an effective attraction between chains. In the strong-linking limit, chains collapse into a single “macromolecule,” driving $P\to0$ and hence full aggregation [1104.2952].

Because this attraction is artificial from the standpoint of static thermodynamics, the model introduces an explicit repulsive potential $U_{\rm rep}(\mathbf r)$ between either monomers or slip-link sites. Denoting its second virial by $B_2$ (or $\widetilde B_2$ for slip-link centers), the total pressure in a weak-coupling expansion is
$$
\widetilde P\approx P+k_BT\,B_2\Bigl(\frac{N\,M}{V}\Bigr)^2.
$$
One then tunes $U_{\rm rep}(r)$ so that its virial exactly cancels the negative slip-link virial at order $(M/V)^2$. Provided $U_{\rm rep}$ is short-range and weak, higher virials remain small, and one approximately recovers the ideal-gas equation of state up to desired order [1104.2952].

## 4. Brownian multi-chain implementation for melts

In the numerical study of the Likhtman slip-link model, each polymer is represented as a Rouse chain of $N_m$ beads connected by entropic springs, and each slip-link is specified by a curvilinear coordinate $x_j\in[0,N_m]$ on its host chain and an anchor point $\vec a_j\in\mathbb R^3$. The total potential is
$$
U=U_{\rm Rouse}+U_{\rm SL},
$$
with
$$
U_{\rm Rouse}(\{\vec r_i\})
=\frac{3k_BT}{2b^2}\sum_{i=1}^{N_m}(\vec r_i-\vec r_{i-1})^2,
$$
and
$$
U_{\rm SL}(\{\vec s_j\})
=\frac{3k_BT}{2N_s b^2}\sum_{j=1}^Z(\vec a_j-\vec s_j)^2.
$$
The slip-link stiffness is
$$
k_{\rm sl}=\frac{3k_BT}{N_s b^2}.
$$
Monomer positions obey Langevin dynamics with friction $\xi$, and each slip-link coordinate slides along the chain according to
$$
\xi_s\,\frac{dx_j}{dt}
=-\frac{\partial U_{\rm SL}}{\partial x_j}+g_j(t),
$$
with
$$
\langle g_j(t)\rangle=0,\qquad
\langle g_j(t)g_k(t')\rangle
=2\,\xi_s\,k_BT\,\delta_{jk}\,\delta(t-t').
$$
The choice $\xi_s\ll\xi$ is made so that slip-link diffusion causes little dissipation [1306.2107].

The instantaneous total stress tensor is written as
$$
\sigma_{\alpha\beta}^{\rm T}
=\sigma_{\alpha\beta}^{\rm Rouse}
+\sigma_{\alpha\beta}^{\rm SL},
$$
and the shear relaxation modulus is obtained from the Green–Kubo formula
$$
G(t)=\frac{V}{k_BT}\,\frac{1}{3}
\Big\langle
\sum_{\alpha=1}^2\sum_{\beta>\alpha}^3
\sigma^{\rm T}_{\alpha\beta}(0)\,
\sigma^{\rm Rouse}_{\alpha\beta}(t)
\Big\rangle.
$$
The mean slip-link density is
$$
\rho_{\rm sl}=\frac{N_p Z}{V},
$$
and enters the plateau value of $G(t)$ in the same way that an entanglement density would in the tube model [1306.2107].

The numerical implementation uses a simple Euler–Maruyama scheme with time step $\Delta t\ll\tau_0\equiv\xi b^2/(3\pi^2k_BT)$, typically $\Delta t\sim10^{-4}\tau_0$. A static binary correspondence is set up among the $N_pZ$ slip-links in the box in pairs. Whenever a slip-link reaches a chain end, it is destroyed and re-created at a chain-end position on a randomly chosen chain, and its companion in the binary pair is simultaneously destroyed and re-created at a chain-end position on another random chain. This mimics creation and annihilation of entanglements while keeping the total slip-link number fixed, and ensures that slip-link sites remain uniformly distributed along the chains. Typical parameters are $N_m=32$–$128$, $N_e=2$–$8$, $N_s\approx0.5$–$2$, and $\xi_s/\xi=0.1$; convergence is checked by varying $\Delta t$ by a factor $2$ and confirming that $G(t)$ and steady-state viscosities change by $<5\%$ [1306.2107].

## 5. Linear viscoelasticity, nonlinear flow, and inhomogeneous extensions

The 2013 numerical study reports a crossover from Rouse-like decay to the emergence of a plateau as $Z$ increases or $N_s$ decreases. Fitting the long-time tail of the simulated $G(t)$ to the single-chain reptation form yields a plateau modulus $G_N^{(0)}$ and a reptation time $\tau_d$. Over the range studied,
$$
\tau_d\propto N_e^{-1.0\pm0.1},\qquad
G_N^{(0)}\propto N_e^{-0.6\pm0.1},
$$
and
$$
\tau_d\propto N_s^{-0.55\pm0.05},\qquad
G_N^{(0)}\propto N_s^{-0.08\pm0.05}.
$$
These relations quantify how slip-link density and slip-link stiffness control long-time relaxation [1306.2107].

Under steady shear $\dot\gamma$, the model displays stress overshoot at strain $\gamma_{\max}\sim\dot\gamma^{\,0.5}$, shear-thinning viscosity $\eta(\dot\gamma)\sim\dot\gamma^{-x}$ with $x\approx0.65$–$0.7$, first normal-stress coefficient $\psi_1\sim\dot\gamma^{-1.2}$, and second normal-stress coefficient $\psi_2\sim-\dot\gamma^{-1.5}$. These exponents are in reasonable agreement with experiment, where $x\approx0.85$, $\psi_1\sim\dot\gamma^{-1.0\text{–}1.5}$, and $\psi_2\sim\dot\gamma^{-1.6}$ [1306.2107].

The same framework is extended to inhomogeneous systems by adding a lattice-discretized compressibility term
$$
\mathcal H_{\rm hom}
=\frac{\kappa_0\delta^3}{2\rho_0}
\sum_{\vec c}\bigl[\rho(\vec c)-\rho_0\bigr]^2,
$$
with local density
$$
\rho(\vec c)
=\frac1{\delta^3}\sum_{n=1}^{N_p}\sum_{i=1}^{N_m}
W(\vec r_i^{\,(n)}-\vec c).
$$
This produces the correct compressibility without altering the single-chain Gaussian statistics. Fixed spherical fillers of effective radius $\sigma_{\rm eff}$ can then be introduced through a short-range repulsive force, and the filler-polymer stress is added to the total stress in the Green–Kubo formula. In a well-dispersed nanocomposite the zero-shear viscosity follows
$$
\eta(\phi)=\eta_0\Bigl(1+\tfrac52\phi+\beta\phi^2\Bigr),
$$
with $\beta\approx3$ in semi-quantitative agreement with suspension theory [1306.2107].

## 6. Primitive-chain-network formulation and cross-linked slip-link networks

In primitive-chain-network simulations used by Masubuchi, the instantaneous network free energy can be written as
$$
H[\{R\},s]
=\sum_{\langle ij\rangle}\frac{3}{2N_{ij}b^2}|{\bf R}_{ij}|^2
-\sum_{\alpha\in{\rm slip}}\ln\Omega_{\rm slip}(s_\alpha)
+\sum_{\mu\in{\rm cross}}U_{\rm cross}.
$$
Here ${\bf R}_{ij}$ is the end-to-end vector of the Gaussian strand linking nodes $i$ and $j$, $N_{ij}$ is the number of Kuhn segments in that strand, the slip-link term is the entropic weight for distributing $s_\alpha$ Kuhn segments on either arm of slip-link $\alpha$, and $U_{\rm cross}\to\infty$ if any contour transfer is attempted at a permanent cross-link. The model parameters include $N_e$, $Z$, the fraction $q_c$ of slip-links converted to permanent cross-links, the active fractions $P_{ac}$ and $P_{as}$ in the percolated network, and the continuum-theory moduli $G_c$ and $G_e$ [2011.03225].

The network evolves by coupled Langevin equations for node positions and contour-transfer equations for slip-link variables. If $s_\alpha\to0$ or $s_\alpha\to N_{\max}$, the slip-link is destroyed. Whenever a dangling chain end accumulates more than $N_{\max}$ segments, a new slip-link is created linking that segment to a randomly chosen nearby segment on another chain; this enforces a roughly constant entanglement density. Under an imposed homogeneous deformation gradient ${\boldsymbol\lambda}$, the instantaneous Cauchy stress tensor is computed from the virial of spring forces,
$$
\sigma_{\alpha\beta}
=\frac1V\Bigl\langle
\sum_{\langle ij\rangle}
\frac{3k_BT}{N_{ij}b^2}\,
R_{ij,\alpha}R_{ij,\beta}
\Bigr\rangle,
$$
with a subtracted isotropic pressure to enforce incompressibility in steady-state. For uniaxial extension, one defines $\sigma(\lambda)\equiv\sigma_{11}(\lambda)-\sigma_{22}(\lambda)$ and often plots the Mooney–Rivlin form
$$
T_M(\lambda)=\frac{\sigma(\lambda)}{\lambda-\lambda^{-2}}.
$$
The long-time plateau stress is read off as the network “equilibrium” elastic response [2011.03225].

The simulated stress–strain curves were compared to the theories of Ball et al. and Rubinstein–Panyukov. In the Ball fits, the slippage parameter was fixed at $n=0.2$. If one chooses $G_c=P_{ac}/2$ and $G_e=4P_{as}/7$, these moduli underpredict the simulated plateau stresses when $Z\ge20$; by allowing $G_c$ to increase slightly while keeping $n=0.2$ and fitting $G_e$ freely, Ball’s formula captures both the zero-strain modulus and the magnitude of strain-softening quite well. In the Rubinstein–Panyukov comparison, fixing $G_c=P_{ac}/2$ and $G_e=4P_{as}/7$ over-softens at large $\lambda$, whereas treating $G_e$ as a free fit parameter while leaving $G_c$ set by $P_{ac}$ allows the curve to pass through the simulation data at both small and large $\lambda$ [2011.03225].

A recurrent interpretive issue concerns the meaning of fitted moduli. The fitted values differ from simple counting arguments because the prepared networks inherit an exponential distribution of strand lengths, a cutoff $N_{ij}\ge N_{\min}$ artificially stiffens the shortest strands, and clusters of cross-links “cage” nearby slip-links, suppressing full slippage. Hence fitting experimental data to the theories does not provide a fraction of entanglements in the system unless the network only consists of Gaussian strands and it correctly reaches the state of free-energy minimum [2011.03225].

## 7. Stress-controlled creep, disentanglement, and recurrent limitations

The 2025 extension of the primitive-chain-network model addresses creep, that is, stress-controlled flow. In this formulation an entangled polymer melt or solution is represented as a network of strands and nodes. Nodes come in two types: sliplinks, at which exactly two different chains interpenetrate, and chain ends, which are free to appear or disappear by sliding off or hooking onto a strand. Each strand $i$ contains $n_i$ Kuhn segments of length $b$ and has end-to-end vector ${\bf r}_i$, with segment free energy
$$
U_{{\rm seg},i}=\frac{3k_BT}{2n_i b^2}|{\bf r}_i|^2,
$$
and tension
$$
{\bf f}_i=\frac{3k_BT}{n_i b^2}{\bf r}_i.
$$
A weak penalty for local fluctuations of the sliplink density $\phi({\bf R})$ around the mean density $\phi^\ast$ is imposed through
$$
A[\phi]=\int d^3R\,W\,[\,\phi(R)\ln(\phi(R)/\phi^\ast)-(\phi(R)-\phi^\ast)\,]
\quad {\rm for}\ \phi>\phi^\ast,
$$
and $A[\phi]=0$ for $\phi\le\phi^\ast$ [2508.01526].

The sliplink-position Langevin equation includes solvent drag, chain-tension forces, osmotic force from $A[\phi]$, and Brownian noise, while chain sliding is governed by a one-dimensional Langevin equation along the chain contour. Network topology changes occur through end-constraint release: whenever the number of Kuhn segments at a dangling chain end, $n_e$, drifts outside the interval
$$
\tfrac12\,n_0\le n_e\le\tfrac32\,n_0,
$$
that chain end is removed from or attached to the nearest strand, thus eliminating or creating a sliplink. The instantaneous microscopic shear stress follows from the Kramers expression
$$
\sigma_{xy}(t)
=\frac{1}{V}\sum_i f_{i,x}(t)\,r_{i,y}(t)
=\frac{3k_BT}{b^2V}\sum_i\frac{r_{i,x}(t)\,r_{i,y}(t)}{n_i(t)}.
$$
To perform a creep test, the imposed shear rate is adjusted at each discrete time step so that the measured stress stays close to the target stress:
$$
\dot\gamma(t+\Delta t)
=\dot\gamma(t)+a\,\frac{\sigma_{xy}(t)-\sigma^{\rm set}}{\sigma^{\rm set}},
$$
with feedback-control coefficient $a=16$ [2508.01526].

The molecular signatures differ markedly between rate-controlled and stress-controlled start-up. Rate-controlled start-up produces stress overshoot, chain-orientation and stretch overshoots, and a concomitant undershoot in the creep rate if mapped to stress control. Stress-controlled start-up shows only monotonic approach of orientation and stretch to their steady values, and a milder reduction of entanglement, measured by the surviving fraction $U_s(t)$ of the original sliplinks, because the feedback destroys the coherence of the tumbling motion. Simulations were compared to a literature dataset of an entangled polybutadiene solution, and qualitative agreement was found in the nonlinear range, under large stresses [2508.01526].

Across the formulations summarized here, two limitations recur. First, weak slip-linking by itself produces a negative slip-link virial and an artificial entropic attraction, so static equilibrium thermodynamics may require a short-range repulsive correction to restore ideal statistics [1104.2952]. Second, fitted rheological moduli need not coincide with simple counts of active cross-links and entanglements unless every strand is Gaussian and every slip-link can slide until the network truly finds its global free-energy minimum under the applied deformation [2011.03225]. A plausible implication is that equilibrium, nonlinear rheology, and topology-renewal rules must be interpreted as coupled components of the model rather than as separable approximations.

Source: https://www.emergentmind.com/topics/multi-chain-sliplink-model