Completing the theory of interfacial layering in the three-component polymer system A/A*/B: the three-species junction, short-time transients, and asymmetric chain lengths and mobilities

When a molten blend of two mutually indifferent polymers A and A* (Flory parameter $\chi_{AA^*}\approx 0$) is brought into contact with a pure melt of a third polymer B that is strongly attracted to A ($\chi_{AB}<0$) and strongly repelled by A* ($\chi_{A^*B}>0$), the interface spontaneously organizes into layered compartments separated by two moving fronts. Aradian, Saulnier, Raphaël and de Gennes (Macromolecules 37, 4664, 2004) solved this three-component moving-boundary problem in the self-similar regime for equal chain lengths and mobilities, under two deliberate bypasses: the thin three-species junction (region IV) was assumed 'permeable', with negligible chemical-potential drop across it, so that its microscopic 'inner problem' was left unsolved; and the very-short-time transients that establish the layered structure were only sketched qualitatively. The open problem is to complete this theory: (1) construct a self-consistent microscopic treatment of the three-species junction, including non-local square-gradient contributions and finite-resistivity (e.g. cross-linked) cases, and establish when the permeability assumption and the jump condition $\phi_1=\phi_2\exp[-(1-\phi_2)^2/(2\epsilon)]$ hold and what replaces them when they fail; (2) give a quantitative theory of the transient regime between first contact and the self-similar layered regime; and (3) extend the solution to unequal chain lengths and mobilities, incorporating the slow-/fast-mode (vacancy-flux and collective-flow) subtleties of asymmetric polymer interdiffusion. Later literature does not resolve these items; the closest later study (Miguel and Tomba, 2010) realizes a layered three-component interphase experimentally, but with mobility contrast in place of the required interaction contrast and only a simple diffusion model, and the recent two-component treatment of entangled interdiffusion with a polymeric source (Moriel and Stone, 2026) supplies only the asymmetric-transport machinery of item (3), not the three-species junction, transient or layering items.

Background

When two molten polymers are brought into contact, their chains interdiffuse across the contact plane. In the Flory-Huggins framework each pair of monomer species $i,j$ is characterized by an interaction (Flory) parameter $\chi_{ij}$: $\chi<0$ favors mixing (net attractive contacts), $\chi>0$ demixes, and the magnitude of $\chi N$ (with $N$ the polymerization index, i.e., the number of monomers per chain) decides miscibility. For two-component junctions of equal chain lengths the theory is classical. For a nearly neutral pair the interpenetration is entropic self-diffusion, each species moving with its self-diffusion coefficient $D_s$; in an entangled melt (chains longer than the entanglement threshold $N_e$) $D_s$ is set by reptation, the tube-constrained motion of a chain along its own contour, giving $D_s\propto N^{-2}$. For an attractive pair ($\chi_{AB}<0$) mutual diffusion is enthalpy-driven and strongly accelerated, with an effective diffusivity $D_f=2|\chi_{AB}|N\,D_s$. When one side is initially pure, interdiffusion generates profiles that terminate at sharp fronts moving as $t^{1/2}$ — a moving-boundary (Stefan-type) problem in which the front velocity is fixed by the diffusive flux across it, called 'class A' motion in the terminology adopted by the source paper. When the two species differ in chain length or mobility, a conceptual subtlety arises — whether mutual diffusion is controlled by the slow or the fast species ('slow-mode' versus 'fast-mode' theories, the latter requiring a compensating collective flow or vacancy flux); experiments favor the fast mode, and the issue remains delicate (see, e.g., Brochard, Jouffroy and Levinson, Macromolecules 16, 1638, 1983, and Composto, Mayer, Kramer and White, Phys. Rev. Lett. 57, 1312, 1986).

Against this backdrop, Aradian, Saulnier, Raphaël and de Gennes (Macromolecules 37, 4664-4675, 2004) posed and solved a minimal three-component problem: a semi-infinite blend of A and A* (mutually indifferent, $\chi_{AA^*}\approx0$, initial A volume fraction $\phi_0$) contacting a semi-infinite pure melt of B, with $\chi_{AB}<0$ and $|\chi_{AB}|N\gg1$, $\chi_{A^*B}>0$ and $\chi_{A^*B}N\gg1$, all chains having equal $N$ and equal monomeric mobility. They showed that the interface spontaneously structures into adjacent compartments: region I (A+A*, shrinking), a geometrically thin three-species junction (region IV), a growing A+B mixing layer (region II), and pure B beyond a front (region III). In the self-similar regime the left front sits at $x=-x_I(t)$ with $x_I(t)=2a(D_st)^{1/2}$ and the right front at $x=x_{II}(t)=2b(D_ft)^{1/2}$; the junction concentrations $\phi_1,\phi_2$ are time-independent, with a stationary composition jump of amplitude $\phi_2$; $a$ is fixed by $\sqrt{\pi}\,a\,e^{a^2}(1-\mathrm{erf}\,a)=\phi_0$; and in the small parameter $\epsilon\equiv D_s/D_f=1/(2|\chi_{AB}|N)\ll1$ one finds the scalings $a\sim\epsilon^0$, $b\sim\epsilon^{1/6}$, $\phi_2\sim\epsilon^{1/3}$, so the mixing layer grows very asymmetrically ($x_I/x_{II}\sim\epsilon^{1/3}$).

This solution rests on bypasses the authors state explicitly. (i) Region IV is treated as 'permeable': the mobility of A inside it is assumed no different from outside, so the chemical-potential drop across it is neglected and the matching condition reduces to $\phi_1=\phi_2\exp[-(1-\phi_2)^2/(2\epsilon)]$ (hence $\phi_1\approx0$). Solving the 'inner problem' of the interdependent diffusive process inside region IV 'would undoubtedly represent a complex task and require a study of its own'; the authors note that a densely cross-linked junction would instead demand the opposite, 'high resistivity' assumption, and that if region IV becomes as thin as a chain radius of gyration (the typical size of a coiled polymer chain) chemical potentials cannot even be defined locally within it, because non-local square-gradient terms — free-energy penalties on concentration gradients, as in Cahn-Hilliard theory — enter. (ii) The very-short-time transients that establish the layered configuration from the moment of contact are only sketched qualitatively (Appendix A of the source); 'a more precise description of the transient dynamics just sketched would nevertheless be desirable and remains ahead of us'. (iii) The whole solution assumes equal chain lengths and equal monomeric mobilities; the asymmetric case, which imports the slow/fast-mode subtleties, is set aside. The authors also flag that the required interaction triad is hard to realize experimentally and propose a related melt-versus-cross-linked-gel problem, which they themselves describe as 'a related, though different, problem'.

Problem Statement

Complete the theory of interfacial layering for the three-component polymer system of Aradian, Saulnier, Raphaël and de Gennes (2004) — a semi-infinite A+A* blend ($\chi_{AA^*}\approx0$, initial A volume fraction $\phi_0$) contacting a semi-infinite pure B melt with $\chi_{AB}<0$, $|\chi_{AB}|N\gg1$ and $\chi_{A^*B}>0$, $\chi_{A^*B}N\gg1$ — by removing the three bypasses of the published solution.

1. Inner problem of the three-species junction (region IV). Construct a self-consistent microscopic (boundary-layer) treatment of the region where A, A* and B coexist, including non-local square-gradient contributions and a finite junction resistivity, and determine from it: the internal concentration structure of the three-species overlap; the conditions under which the source's 'permeable junction' assumption and its consequence, the matching condition $\phi_1=\phi_2\exp[-(1-\phi_2)^2/(2\epsilon)]$ with $\phi_1\approx0$, are self-consistent; what replaces this condition when they fail (in particular for a high-resistivity junction such as a cross-linked network, or when the junction thickness is comparable to the chains' radii of gyration, where local chemical potentials are not defined); and how the outer front constants $a$, $b$, the jump amplitude $\phi_2$, and the scalings $a\sim\epsilon^0$, $b\sim\epsilon^{1/6}$, $\phi_2\sim\epsilon^{1/3}$ (with $\epsilon=1/(2|\chi_{AB}|N)$) are then modified.

2. Short-time transient regime. Provide a quantitative description of the dynamics between the instant of contact and the establishment of the self-similar layered regime: the initial formation and early evolution of the triple-overlap region, the crossover between the two early-time scenarios identified qualitatively in Appendix A of the source (governed by the sign of the effective Flory parameter experienced by B chains in the A+A* blend, i.e. by whether $\phi_0$ lies below or above a crossover value set by the competition between $\chi_{AB}$ and $\chi_{A^*B}$), and the crossover time and law into the $t^{1/2}$ self-similar regime.

3. Asymmetric chains and mobilities. Extend the moving-boundary layering solution to $N_A\neq N_{A^*}\neq N_B$ and/or unequal monomeric mobilities, incorporating the vacancy-flux/collective-flow (slow- versus fast-mode) subtleties of asymmetric polymer interdiffusion, and determine how the layered compartment structure, the two-front laws $x_I(t)\propto t^{1/2}$ and $x_{II}(t)\propto t^{1/2}$, and the stationary jump are modified.

The related problem of an A+A* melt contacting a loosely cross-linked gel of B, which the source proposes but explicitly distinguishes as different, is not part of this problem's completion boundary.

References

1. A. Aradian, F. Saulnier, E. Raphaël, P.-G. de Gennes, 'Interfacial Layering in a Three-Component Polymer System', Macromolecules 37, 4664-4675 (2004). DOI: 10.1021/ma0356008, https://doi.org/10.1021/ma0356008
2. A. Aradian, F. Saulnier, E. Raphaël, P.-G. de Gennes, 'Interfacial layering in a three-component polymer system', preprint (v1 22 Oct 2003, v2 5 Apr 2004). arXiv: cond-mat/0310538, https://arxiv.org/abs/cond-mat/0310538
3. María de la Paz Miguel, J. Pablo Tomba, 'Interphase development in a three-component polymer system with asymmetric mobility', Journal of Polymer Science Part B: Polymer Physics 48, 627-633 (2010). DOI: 10.1002/polb.21931, https://doi.org/10.1002/polb.21931
4. Francoise Brochard, Jacqueline Jouffroy, Paul Levinson, 'Polymer-polymer diffusion in melts', Macromolecules 16, 1638-1641 (1983). DOI: 10.1021/ma00244a016, https://doi.org/10.1021/ma00244a016
5. Russell J. Composto, James W. Mayer, Edward J. Kramer, Dwain M. White, 'Fast Mutual Diffusion in Polymer Blends', Physical Review Letters 57, 1312-1315 (1986). DOI: 10.1103/PhysRevLett.57.1312, https://doi.org/10.1103/PhysRevLett.57.1312
6. Avraham Moriel, Howard A. Stone, 'Polymer-polymer interdiffusion: effects of entanglements and a polymeric source', preprint (2026, submitted 18 Jun 2026). arXiv: 2606.20462, https://arxiv.org/abs/2606.20462
7. A. Zeinolebadi, N. Mohammadi, Hamedi H. Sangari, 'The Role of Polymer/Polymer Miscibility in Interfacial Healing Kinetics and Equilibrium Adhesion Energy: A Universal Approach', Journal of Adhesion Science and Technology 22, 1301-1311 (2008). DOI: 10.1163/156856108X319827, https://doi.org/10.1163/156856108X319827
8. H. Mohammadi, N. Mohammadi, 'Fracture of polymer blends: Effect of characteristic number of interfacial entanglements and matrix toughness', Polymer 53, 2769-2776 (2012). DOI: 10.1016/j.polymer.2012.04.034, https://doi.org/10.1016/j.polymer.2012.04.034
9. Haifeng Shi, Hua Jiang, Guoqiang Fan, Zhaohui Yang, Xiaohua Zhang, 'Interdiffusion behavior under a temperature gradient field in a poly(ethylene-co-hexene)/poly(ethylene-co-butene) system', RSC Advances 5, 60015-60023 (2015). DOI: 10.1039/C5RA08242K, https://doi.org/10.1039/C5RA08242K