Thermal Conductivity of Nanofluids: A Review and Physics-Based Multi-Mechanism Framework

Nanofluids, colloidal suspensions of nanoparticles dispersed in conventional base fluids, have been widely investigated as advanced heat transfer media because of reported thermal conductivity enhancements. Despite more than two decades of research, experimental results remain highly scattered and often irreproducible, and no transferable predictive framework has emerged. Experimental uncertainty contributes to this scatter, but it cannot fully explain the systematic dependence of measured conductivity on the preparation protocol, sonication energy, surfactant chemistry, aging time, thermal cycling, and dispersion stability. This review argues that many discrepancies reflect changes in suspension microstructure rather than measurement uncertainty alone. The review critically assesses experimental evidence and major model classes, including classical effective medium theories, semiempirical correlations, Brownian-based formulations, aggregation-inspired models, and data-driven approaches. A key finding is that thermal conductivity enhancement in nanofluids cannot be described as a single-valued function of the nanoparticle concentration and temperature. Instead, heat transport is governed by coupled factors including dispersion state, aggregation dynamics, interfacial structure, surface chemistry, thermal history, and aging. Aggregation is identified as a central missing physics because it modifies conductive pathways, suppresses Brownian motion, alters interfacial resistance, and disrupts nanolayer continuity. Building on these insights, the review introduces a stability-aware multi-mechanism framework. The framework interprets the nanofluid heat transport as a coupled resistance network. It incorporates static conduction, nanolayer effects, Kapitza interfacial resistance, Brownian-induced transport, and aggregation-driven penalties. Unlike prior hybrid models that add separate enhancement terms or absorb aggregation into empirical coefficients, the framework treats dispersion stability and the aggregation state as explicit variables that regulate mechanism activation, suppression, and coupling. The main contributions are to identify the physical origins of reported thermal conductivity scatter, evaluate why existing model classes remain nontransferable, introduce a stability-aware framework, and define regime-dependent validity domains in which existing models can be interpreted as limiting cases.