Reverse Diffusion¶
Uphill transport in which a component moves toward higher concentration during an unstable or coupled mixture evolution because the chemical-potential gradient, rather than concentration gradient alone, drives flux.
Core Idea¶
Reverse diffusion names an apparent inversion of ordinary smoothing: a component accumulates where it is already richer. In phase-separating mixtures, this is thermodynamically possible because chemical potential, not bare concentration, is the correct driving field.
The name is dangerously broad. Bulk advection, reaction, pressure-driven reverse osmosis, and data artifacts can also produce concentration increases, so flux must be defined relative to the medium and mechanism.
Cross-Domain Echoes¶
See how this entry connects to another domain.
Scope of Application¶
- Phase separation. Describes spinodal amplification of composition fluctuations.
- Materials science. Analyzes alloy and polymer domain evolution.
- Nonideal transport. Uses chemical potentials and cross-coupled fluxes.
- Terminology audit. Separates membrane, convective, and phase-field usages.
Clarity¶
State transported component, reference frame, concentration field, measured flux, temperature and composition regime, free-energy and mobility model, boundary conditions, advection or reaction controls, and whether the phrase means phase separation or forced membrane transport. Inclusion test: Require observed or modeled component flux toward locally higher concentration with an identified nonideal chemical-potential or externally forced mechanism. Exclusion test: Exclude ordinary down-gradient Fickian diffusion, convective transport mistaken for diffusion, osmosis described without a membrane and pressure balance, and time-reversed animation of mixing. Nearest boundary: Reverse osmosis is pressure-driven solvent transport through a semipermeable membrane; uphill diffusion in phase separation is chemical-potential-driven component transport within a mixture. Exit condition: The label should be rejected when velocity advection, reaction source, measurement artifact, or coordinate choice explains the concentration change better than a constitutive diffusive flux. Common misclassifications: It is not diffusion literally running backward in time. Uphill concentration flux does not by itself violate the second law. Reverse osmosis is not the same mechanism. Advection toward a concentrated region is not diffusion. Nearest named distinctions: Reverse osmosis: Uses pressure to drive solvent across a semipermeable membrane. Advection: Carries material with bulk flow. Spinodal decomposition: Is the phase-separation regime in which uphill diffusion can occur. Negative diffusion coefficient: Is an informal local description and can omit the stabilizing higher-order term.
Manages Complexity¶
A simple concentration gradient can point opposite the actual thermodynamic force in a nonideal system. Distinguishing state change from material flux requires conservation, frame choice, and constitutive modeling.
Abstract Reasoning¶
- Define concentration and flux relative to the material or laboratory frame.
- Rule out advection, reaction, and measurement artifacts.
- Specify free-energy functional, chemical potential, mobility, and conservation law.
- Test whether the state lies in an unstable or cross-diffusive regime.
- Use mechanism-specific language rather than merging phase separation with reverse osmosis.
Knowledge Transfer¶
Chemical-potential-driven uphill flux transfers to other nonideal mixtures, but the Cahn–Hilliard model, mobility, and order parameter must fit the material. The loose phrase alone does not transfer a mechanism.
Relationships to Other Abstractions¶
Current abstraction Reverse Diffusion Domain-specific
Parents (1) — more general patterns this builds on
-
Reverse Diffusion is a kind of Diffusion Prime
Reverse Diffusion is a strict kind of Diffusion: it is diffusive transport driven by chemical potential even when concentration rises along the flux direction.
Hierarchy paths (3) — routes to 3 parentless roots
- Reverse Diffusion → Diffusion → Gradient
- Reverse Diffusion → Diffusion → Propagation
Neighborhood in Abstraction Space¶
Reverse Diffusion sits in a moderately populated region (43rd percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Thermodynamic & Transport Processes (34 abstractions)
Nearest neighbors
- P-Laplacian — 0.89
- Jellium — 0.89
- Thermodynamic System — 0.87
- Microrheology — 0.86
- Path Integral Formulation — 0.86
Computed from structural-signature embeddings · 2026-10-08