Singularly isostatic and geometrically unstable rigidity of metal-organic frameworks

Metal-organic frameworks (MOFs) combine high porosity with structural fragility, raising important questions about their mechanical stability. We develop a rigidity-based framework in which spring networks parameterized by UFF4MOF are used to construct rigidity and dynamical matrices. Large-scale analysis of 5,682 MOFs from the CoRE 2019 database shows that most frameworks are formally over-constrained yet cluster sharply near the isostatic threshold, revealing accidental geometric modes and placing many MOFs near mechanical instability. In the representative case of UiO-66, we show that auxiliary long-range constraints introduced by tuning the neighbor cutoff lift these modes into soft, flat, finite-frequency bands. The results show that rigidity-matrix analysis can rapidly identify MOFs likely to remain mechanically stable. This near-criticality mirrors behavior known from topological mechanics and points to a deeper design principle in porous crystals.

💡 Research Summary

This paper introduces a rigidity‑based computational framework for assessing the mechanical stability of metal‑organic frameworks (MOFs). By representing each MOF as a periodic network of atoms connected with harmonic springs whose force constants are derived from the UFF4MOF force field, the authors construct both a rigidity matrix (R) and a dynamical matrix (D). The rigidity matrix encodes linear bond‑stretching and angular bending constraints; its transpose combined with the spring‑constant matrix (K) yields the dynamical matrix via D = M⁻¹ᐟ² Rᵀ K R M⁻¹ᐟ², where M is the diagonal mass matrix. Using the Maxwell‑Calladine relation ν = N₀ − N_ss = dN_s − N_c, the global index ν, the number of zero‑energy modes (N₀), and the number of states of self‑stress (N_ss) are obtained for each structure.

The methodology is applied to 5,682 MOFs from the CoRE 2019 database. For each crystal the authors evaluate ν/N_s (N_s = number of atoms) and find that the majority are formally over‑constrained (ν/N_s < 0). Strikingly, however, the distribution clusters tightly around the isostatic point ν ≈ 3 (ν/N_s ≈ 0 for large cells), indicating that most frameworks sit at the brink of mechanical instability. This near‑isostatic behavior is attributed to “accidental geometric modes” that arise when symmetry or local geometry makes some constraints redundant, creating extra zero‑energy motions despite an overall surplus of constraints.

Three mechanical regimes are identified:

1. Generic Rigidity (ν < 3, N₀ = 3) – all constraints are independent; only global translations/rotations are zero modes, and the framework is robust.
2. Geometrically Unstable (ν < 3, N₀ > 3) – symmetry‑induced redundancy yields additional zero modes; the structure is globally over‑constrained but locally soft.
3. Singular Isostatic (ν = 3, N₀ > 3, N_ss > 0) – zero modes coexist with self‑stress states; the framework is marginally stable and can be driven rigid or floppy by tiny geometric perturbations.

Representative MOFs illustrate these categories. ABIXOZ exemplifies generic rigidity with a strongly negative ν/N_s (‑1.89) and no non‑trivial zero modes; its phonon dispersion is smooth and its inverse participation ratio (IPR) indicates highly delocalized vibrations. IKEBUV01 lies in the singular isostatic regime (ν/N_s ≈ +0.075) and possesses twelve zero modes exactly cancelled by twelve self‑stress states. The low‑frequency bands are flat, and the IPR map shows that the soft motions are dominated by hydrogen atoms on the linkers, leaving the metal‑oxide backbone essentially rigid.

UiO‑66 is a case of geometric instability. Although globally over‑constrained (ν/N_s = ‑1.14), it exhibits 238 zero modes at the Γ point. These modes are overwhelmingly associated with hydrogen (≈ 83 % participation) and a smaller carbon component (≈ 16 %); the metal nodes move negligibly. By tuning the neighbor‑cutoff parameter τ to include longer‑range auxiliary bonds, the authors demonstrate a systematic reduction of zero modes (from 238 to 215, then to zero) and the emergence of soft, flat phonon bands at finite frequency. This illustrates how adding realistic long‑range interactions can lift accidental geometric modes into low‑energy collective vibrations, thereby stabilizing the framework.

The authors also compute a local Maxwell index ν_i for each atom (ν_i = 3 − r_i, where r_i is the local constraint density). Histograms of ν_i reveal that hydrogen atoms often experience under‑constraint (positive ν_i), while metal and oxygen atoms cluster near ν_i ≈ 0 (isostatic) or negative values (over‑constrained). This atom‑resolved picture provides a quantitative tool for identifying which parts of a MOF are prone to flexing and which contribute to overall rigidity.

Overall, the study demonstrates that (i) a spring‑network rigidity analysis can be performed at high throughput for thousands of MOFs, (ii) most MOFs naturally reside near an isostatic critical point, making them sensitive to subtle structural changes, and (iii) the interplay of topology, geometry, and long‑range constraints mirrors phenomena known from topological mechanics, suggesting a deeper design principle for porous crystals. The framework offers a fast screening metric to flag mechanically fragile candidates before costly DFT or molecular dynamics simulations, and it opens avenues for rational design of robust MOFs by engineering constraint redundancy or reinforcing long‑range interactions. Future work should integrate more accurate electrostatic models, explore temperature‑dependent effects, and validate predictions against experimental mechanical testing.