Rapini-Papoular Surface Energy

• The Rapini–Papoular surface energy model is a quadratic formulation that penalizes deviations in liquid crystal alignment from a preferred orientation at surfaces.
• It produces Robin-type boundary conditions that interpolate between free (Neumann) and rigid (Dirichlet) anchoring by varying the anchoring strength, W.
• Extensions include homogenization in composites and chiral adaptations for cholesteric systems, supporting analysis of defect morphology and phase transitions.

The Rapini–Papoular surface energy describes the energetic cost associated with deviations of a liquid crystal's orientational order (typically the director or order-parameter tensor) from a preferred “easy axis” at a confining surface. This quadratic, symmetry-based interfacial term has become the foundational model for weak and strong anchoring phenomena in molecular, colloidal, and composite nematic and cholesteric materials, and underlies a wide range of theoretical, numerical, and experimental studies. Its canonical form leads to Robin-type boundary conditions that interpolate between Dirichlet (rigid) and Neumann (free) anchoring, plays a central role in homogenization and defect morphology, and admits generalization to tensorial frameworks and chiral materials.

1. Mathematical Formulation and Physical Principles

The Rapini–Papoular (RP) surface anchoring energy density penalizes deviations of the orientational state at a surface from a favored direction or order parameter, with quadratic dependence. In director-based models (e.g., for nematic or cholesteric LCs), the RP energy at a substrate is given by

$$V_{\rm RP} = \tfrac{W}{2} \sin^2(\theta - \theta_0),$$

where $W$ denotes the anchoring strength (units: J m$^{-2}$), $\theta$ is the angle between the director $\mathbf{n}$ and the substrate “easy axis” $\mathbf{n}^0$, and $\theta_0$ accounts for any pretilt or azimuth (Guo et al., 2019).

In $Q$-tensor based (Landau–de Gennes) descriptions, the RP term generalizes as: $f_s(Q, \nu) = W\,| Q - Q_0 |^2,$ where $Q$ is the order parameter at the boundary, $Q_0$ is the prescribed preferred tensor (e.g., $Q_0 = n_0 \otimes n_0 - \tfrac{1}{3} I$ for uniaxial planar or homeotropic anchoring), and $\nu$ is the surface unit normal (Canevari et al., 2019, Calderer et al., 2013, Ceuca, 2020). For cholesterics, necessary chiral invariants may supplement this form (Guo et al., 2019).

Anchoring strength $W$ determines the competition with elastic bulk tendencies; $W \to 0$ corresponds to “free” anchoring and $W \to \infty$ recovers the strong Dirichlet condition.

2. Role in Boundary Value Problems and Variational Formulations

The RP energy is incorporated as a boundary integral in total free energy functionals, appearing alongside bulk terms (Frank–Oseen, Landau–de Gennes, or elastic energies):

$$F[Q] = \int_\Omega \left[ f_{\rm bulk}(Q, \nabla Q) \right] dV + \frac{W}{2} \int_{\partial \Omega} |Q - Q_b|^2 dS$$

where $Q_b$ prescribes the easy-axis state on $\partial \Omega$ (Rajamanickam, 2 Jan 2026).

Variation yields mixed (Robin-type) boundary conditions: $$L\,\partial_\nu Q + W(Q - Q_b) = 0 \quad \text{on } \partial\Omega,$$ with $L$ the appropriate elastic coefficient, $\partial_\nu$ the surface normal derivative. For director fields, the RP form produces: $$\text{Torque balance:} \quad K (\hat N \cdot \nabla) \hat n + \partial_{\hat n} E_{\rm RP} = 0$$ (Guo et al., 2019). This interpolates between Neumann (free) and Dirichlet (rigid) anchoring conditions depending on $W$.

In Landau–de Gennes–type models for elastomers and composites, the RP term may act on the pull-back order parameter $\tilde Q(x) = Q(\varphi(x))$ defined on the undeformed reference configuration, ensuring energetic consistency with mixed kinematic settings (Calderer et al., 2013).

3. Quantification and Extracted Parameters

Anchoring strengths $W$ vary widely across substrates, surface treatments, and material systems:

• Typical $W$ values: $10^{-7}$–$10^{-4}$ J m$^{-2}$ for nematic colloids and coated substrates (Canevari et al., 2019).
• Dimensionless anchoring number: $\chi = W R / K$ (with $R$ particle radius, $K$ elastic constant); $\chi \ll 1$ indicates weak anchoring, $\chi \gtrsim 1$ strong anchoring (Canevari et al., 2019).
• Experimental extractions: e.g., $W\approx 1\times 10^{-7}$ J m$^{-2}$ for untreated glass (two orders weaker than coated, which reach $10^{-5}$–$10^{-4}$ J m$^{-2}$) (Orlova et al., 2013).

In confined cells, the RP mechanism robustly describes equilibrium pitch selection, twist parameter discontinuities, and metastable branches. For cholesterics, the measured pitch parameter $\nu = 2D/P$ relates to the equilibrium (free) pitch $\nu_0 = 2D/P_0$ via

$$\nu_0 = \nu + \frac{w}{\pi} \sin 2(\pi\nu - \phi_e), \quad w = \frac{W D}{2 K_t}$$

with $K_t$ the Frank twist elastic constant, $D$ cell thickness, and $\phi_e$ the easy-axis azimuth (Orlova et al., 2013).

4. Extensions: Homogenization, Generalized and Chiral Surface Energies

The RP form admits direct extension to structured or multi-component systems and chiral media.

Homogenization in Colloidal and Scaffold Composites

In dilute colloid or microstructure-LC composites, the quadratic RP term on colloidal surfaces scales with inclusion area and generates effective “homogenized” bulk corrections in the macroscopic limit (Canevari et al., 2019, Ceuca, 2020). For inclusions of size $\varepsilon^\alpha$ distributed with spacing $\varepsilon$, the total RP surface energy

$$W \int_{\partial \Omega_\varepsilon} |Q - Q_\nu|^2 \, d\sigma$$

converges (as $\varepsilon\to 0$) to an effective bulk term, e.g.,

$$\mathcal{F}_0[Q] = \int_\Omega \big\{ f_{\rm elastic}(\nabla Q) + f_{\rm bulk}(Q) + W_{\rm eff}\text{tr}(Q^2) \big\} dx,$$

where $W_{\rm eff}$ is set by inclusion shape, distribution, and RP parameters (Ceuca, 2020). This enables tunability of the phase transition temperature and bulk order via engineered anchoring.

Symmetry Extensions in Chiral (Cholesteric) Liquid Crystals

In cholesteric systems, broken inversion symmetry allows additional pseudoscalar (chiral) terms in the interfacial energy that are disallowed in the achiral RP form (Guo et al., 2019). The most general quadratic-in-order parameter interfacial energy for cholesterics adds the term: $$E_c = \frac12 q_0 W_c \varepsilon_{\alpha\beta\gamma} Q_{\mu\beta} N_\gamma Q^0_{\mu\alpha}$$ where $q_0$ is the chiral pseudoscalar (inverse pitch), $Q$, $Q^0$ are fluid and "frozen" (substrate) order tensors, $N$ is the surface normal, $W_c$ the chiral anchoring constant, and $\varepsilon$ the Levi–Civita tensor. In the uniaxial limit, this reduces to a term proportional to $\sin^2\theta \sin 2\phi$, generating preferred azimuthal tilt and possible anchoring transitions.

5. Asymptotic Regimes, Defect Morphology, and Numerical Aspects

The role of RP anchoring is particularly pronounced in confined geometries and near topological defects.

Asymptotic Regimes in Confined Domains

Two length scales govern the balance between bulk and surface energies:

• Coherence length $\xi = \sqrt{L/|A|}$ (core size of defects)
• Extrapolation length $l_{\rm ex} = L/W$ (anchoring penetration depth) (Rajamanickam, 2 Jan 2026)

For domain size $h$ much less than $\xi$ (small-domain regime), the RP energy dominates, leading to averaged boundary alignment and even isotropic melting under symmetric frustration. For $h \gg \xi$ (Oseen–Frank regime), the RP term imposes a boundary layer of width $l_{\rm ex}$. Director fields show $O(1/\epsilon)$ tilt away from the easy axis, as opposed to $O(1/\epsilon^2)$ in the Dirichlet limit (Rajamanickam, 2 Jan 2026).

Defect Cores and Boundary Layer Structure

Incorporating the RP term leads to Robin-type BCs at boundaries and corners. This results in physically realistic defect cores with continuous order parameter profiles, in contrast to flat or over-constrained “cores” produced by rigid Dirichlet conditions (Rajamanickam, 2 Jan 2026). The strength and form of RP anchoring significantly modulate both interior and boundary defect morphologies, with implications for switching, bistability, and response to external fields.

6. Limitations, Generalizations, and Experimental Realizations

While the Rapini–Papoular form captures the dominant energetic mechanism for orientational anchoring at surfaces in a wide range of nematic and cholesteric systems, it has known limitations and generalizations:

• In cases with degenerate planar anchoring, higher-order or modified (e.g., quartic) potentials may be required (Ceuca, 2020).
• For diluted colloidal dispersions and composite lattices, the effective (homogenized) bulk corrections and phase behavior depend on microgeometry, with the RP form as a tunable input (Canevari et al., 2019, Ceuca, 2020).
• In chiral systems, symmetry-allowed terms coupled to the chiral pseudoscalar can shift equilibrium orientations and drive interfacial transitions not present in the achiral case (Guo et al., 2019).

Experimentally, extracted $W$ values enable tuning of anchoring from weak to strong, and comparison of pitch jumps, twist parameter branches, and director alignments with theoretical predictions validates the RP formalism across substrates and treatments (Orlova et al., 2013).

7. Summary Table: Rapini–Papoular Surface Energy in Liquid Crystals

System/Model	Mathematical Form	Anchoring Strength Units
Director, single easy axis	$V_S = \frac{W}{2}\sin^2(\phi_s-\phi_e)$	J m$^{-2}$
Landau–de Gennes, $Q$ tensor ($3D$)	$f_s(Q,\nu) = W\,| Q - Q_0 |^2$	J m$^{-2}$
2D $Q$-tensor	$(W/2)\int_{\partial\Omega}|Q-Q_b|^2\,ds$	J m$^{-1}$ (per length)
Chiral extension (cholesteric)	$$E_c = \frac12 q_0 W_c\, \varepsilon_{\alpha\beta\gamma} Q_{\mu\beta} N_\gamma Q^0_{\mu\alpha}$$	J m$^{-2}$
Homogenized composite	$$W\int_{\partial \Omega}|Q-Q_\nu|^2 d\sigma\to W_{\rm eff}\,\mathrm{tr}(Q^2)$$ (bulk term)	J m$^{-2}$

The Rapini–Papoular model remains central in the theoretical, numerical, and experimental study of weak and strong anchoring in nematic and cholesteric liquid crystals, underpinning defect physics, phase transitions, boundary-driven reorientations, and engineered composite responses (Orlova et al., 2013, Calderer et al., 2013, Canevari et al., 2019, Ceuca, 2020, Rajamanickam, 2 Jan 2026, Guo et al., 2019).