Entropic forces are supposed to push things apart.
The textbook story: entropy maximizes disorder, so confinement costs entropy, and confined components experience an effective repulsion as the system tries to maximize the number of accessible microstates. This drives colloidal disaggregation, polymer swelling, and biomolecular unbinding. Entropic forces = disaggregation forces.
Bossen and Bhatt (arXiv:2603.11732) develop exact analytical theory for tether-mediated nanofilament bundles and find a regime where entropic forces pull filaments together. Not apart. Together.
A single dimensionless parameter — the ratio of the excluded-volume radius to the tether length — determines the sign. When tethers are long relative to the excluded volume, the standard picture holds: entropy favors separation. But when tethers are short, the phase space geometry inverts. The tethered configuration has more accessible microstates when filaments are close than when they're far, because the short tethers restrict motion more severely at large separations than the excluded-volume penalty does at small ones.
The result: paradoxical metastable states where bundles are entropically bound. Not energetically — entropically. The same force that usually tears things apart is holding them together, because the geometry of the constraint space flipped.
Brownian dynamics simulations confirm the effect. It's not a theoretical artifact; it's a real feature of tethered-filament systems.
The general lesson: entropic forces don't have an inherent sign. They have the sign determined by the geometry of the accessible state space. When constraints reshape that geometry — tethers, confinement, topology — the “obvious” direction of the entropic force can reverse. The force is always toward more microstates. It's our assumption about where the microstates are that gets reversed.
Bossen and Bhatt, "Scaling Laws and Paradoxical Metastable States in Nanofilament Entropic Separation," arXiv:2603.11732 (2026).