Poly-GRACE: Analytical Coarse-Graining for Polymers

• Poly-GRACE is a family of analytical approaches that construct multi-resolution coarse-grained models for polymers, preserving key structural and thermodynamic properties.
• It employs liquid-state theory, specifically PRISM and Ornstein–Zernike equations, to derive state-specific potentials without ad hoc parameter fitting.
• The framework ensures consistency in compressibility, free energy, and equation of state across scales, achieving quantitative agreement with atomistic simulations.

Poly-GRACE refers to a family of frameworks, models, and methodologies unified by the principle of constructing generalized, multi-resolution coarse-grained models, particularly for polymers and complex molecular systems, that consistently preserve both structural and thermodynamic properties across all levels of representation. While the term itself does not denote a single canonical algorithm, it synthesizes several influential approaches to analytical, algorithmic, and computational coarse-graining—in particular, those designed to achieve state-dependent, structure-preserving, and thermodynamically-consistent coarse-grained potentials using analytically grounded, transferable strategies rather than ad hoc parameter fitting. Poly-GRACE-type methodologies are directly inspired by and conceptually aligned with the analytical integral-equation-based coarse-graining framework presented in "An analytical coarse-graining method which preserves the free energy, structural correlations, and thermodynamic state of polymer melts from the atomistic to the mesoscale" (McCarty et al., 2014).

1. Conceptual Foundation and Motivation

Poly-GRACE frameworks address the intrinsic challenge of constructing multiscale models for polymer melts (or more generally, soft condensed-matter systems) that retain fidelity to both structure (e.g., pair distribution functions, correlation functions) and thermodynamics (e.g., equation of state, compressibility, excess free energy) across coarse-graining hierarchies. Traditional numerical approaches to coarse-graining, such as variational force matching and iterative Boltzmann inversion, typically rely on fitting effective interactions to atomistic data at a single state point, often leading to models that are not transferable to different thermodynamic conditions or levels of resolution. Moreover, most such protocols fail to guarantee the preservation of fundamental thermodynamic quantities or the minimization of relative entropy between atomistic and coarse-grained ensembles except through post hoc parameter searches.

Poly-GRACE models, in contrast, derive the effective coarse-grained interaction analytically from liquid-state theory principles—specifically, from the Ornstein–Zernike integral equation formalism, as specialized to polymers by the PRISM (Polymer Reference Interaction Site Model) theory. This approach achieves state specificity, structure preservation, and thermodynamic consistency in a single analytical construction, explicitly parameterized by a molecular-level quantity—the $k\to 0$ limit of the monomer direct correlation function, $c_0$—that encodes essential thermodynamic information.

2. Analytical Coarse-Graining via PRISM and State-Specific Potentials

At the core of the Poly-GRACE analytical strategy is the mapping of the complex, high-dimensional configuration space of atomistic polymer solutions onto a lower-dimensional space of coarse-grained (CG) coordinates, such as centers of mass or "blobs." For monomer-level structure, the PRISM equation for the total correlation function $\hat h^{mm}(k)$ is

$$\hat{h}^{mm}(k)=\hat{\omega}^{mm}(k)\hat{c}^{mm}(k)\left[\hat{\omega}^{mm}(k)+\rho \hat{h}^{mm}(k)\right],$$

where all quantities are defined in Fourier space and parameterized by the polymer architecture and thermodynamic state.

The crucial unifying feature of the Poly-GRACE paradigm is that the resulting effective CG interaction is not an arbitrary pairwise fit, but rather the analytically determined potential of mean force, expressed in terms of $c_0$: $c_0 \equiv \hat c^{mm}(k=0).$ This scalar provides a thermodynamically consistent parameterization bridging all levels of the coarse-graining hierarchy. The interaction at the mesoscale, for example in the single-site ("soft sphere") limit, is given by the hypernetted-chain closure as

$$\frac{v^{cc}(r)}{k_BT} = -\ln[1 + h^{cc}(r)] + h^{cc}(r) - c^{cc}(r).$$

Thermodynamic consistency is achieved because this form incorporates both structural correlations and the excess free energy, such that the CG model exactly reproduces the compressibility, equation of state, and free energy known from the reference atomistic melt.

3. State Parameterization and Matching: $c_0$ and $d_{HS}$

A defining operational aspect of Poly-GRACE-like frameworks is the non-variational, state-specific determination of the critical coarse-grained parameter $c_0$. Rather than optimizing force fields to trajectory matching or structural agreement, the approach solves the PRISM equation for a model polymer (e.g., semiflexible chain with hard-core repulsion and attractive tail), adjusting an effective hard-sphere diameter $c_0$0 until the predicted pressure (from the mean-field equation of state)

$c_0$1

matches the reference simulation or experimental value. Once $c_0$2 (and thus $c_0$3) is fixed for a given thermodynamic condition, all structural and thermodynamic CG predictions (correlation functions, compressibility, free energies) are determined analytically with no further parameter fitting.

This workflow guarantees that the same macroscopic state point is correctly represented at any resolution in the model's hierarchy, and the effective CG potential remains thermodynamically consistent by construction.

4. Preservation of Structure, Thermodynamics, and Free Energy

The principal technical advance of the Poly-GRACE genus of models is that the CG interaction is not only structure-preserving (reproducing, e.g., $c_0$4 and $c_0$5) but also preserves the correct free energy landscape in the reduced coordinates. This is achieved analytically:

• Compressibility: The isothermal compressibility computed at atomistic and CG levels is identical, as $c_0$6.
• Equation of state: The mean-field EOS and its Carnahan-Starling-like corrections are analytically propagated to all levels of CG.
• Free energies: The Helmholtz and Gibbs free energies per monomer/chain, as well as excess free energies, admit explicit expressions as functions of $c_0$7, the packing fraction, and other state variables.
• Entropy: The mapping entropy (loss due to reduction in degrees of freedom) and the entropy of the CG model can be analytically distinguished, with the mapping entropy arising solely from the representational reduction, not the effective potential.

The crucial result is that the pairwise CG interaction potential, unlike an empirical fit, is a genuine free-energy surface in the reduced coordinate space, correctly accounting for the integrated-out monomer-level fluctuations and preserving the excess free energy of the melt.

5. Quantitative Validation and Relative Entropy Minimization

Empirical validation reported in (McCarty et al., 2014) demonstrates that the analytical CG framework achieves quantitative agreement with united-atom (UA) molecular dynamics for polyethylene melts across multiple chain lengths and densities, at both the structural and thermodynamic level (e.g., $c_0$8, compressibility, equation of state, free energy). The procedure does not rely on variational optimization, force matching, or iterative Boltzmann inversion; rather, it implicitly minimizes the relative entropy between atomistic and CG ensembles by ensuring that the reduced description matches the reference structural distributions exactly as encoded by PRISM.

The approach thereby guarantees that:

• No iterative parameter search is necessary.
• The structural relative entropy is minimized (in the sense of the PRISM formalism), so that the CG model samples the correct macroscopic state.
• The method generalizes across different thermodynamic states and system parameters, provided $c_0$9 is computed consistently.

6. Generalization and Position within the Multiscale Modeling Landscape

Poly-GRACE methodologies are distinguished from purely empirical or simulation-based coarse-graining primarily by their analytical character and by their structural and thermodynamic fidelity across all levels of model resolution. Their reliance on integral equation theory and state-parameterized effective potentials establishes a rigorous foundation for multiscale polymer modeling that is transferrable and robust. This general approach is applicable to a broad class of polymeric systems under different thermodynamic or chemical conditions, so long as appropriate PRISM-type closure relations and state-specific matching protocols are employed.

Poly-GRACE thus stands as a reference paradigm in polymer coarse-graining: models that (i) preserve excess free energy, compressibility, and equation of state exactly; (ii) promote analytic tractability and predictive power; and (iii) are rooted in a liquid-state-theoretic framework that ensures consistency, transferability, and thermodynamic soundness, all without variational optimization artifacts. This framework has proven effective in producing agreement with both simulation and experiment while providing interpretability and generalization unavailable to traditional fitted coarse-grained models.

---

Principal Reference:

"An analytical coarse-graining method which preserves the free energy, structural correlations, and thermodynamic state of polymer melts from the atomistic to the mesoscale" (McCarty et al., 2014).