Study on the Mechanical Properties and Interfacial Interaction Mechanism of Nano-SiO2-Modified Expanded Polystyrene Lightweight Concrete

Expanded polystyrene (EPS) foam concrete is attractive for lightweight building applications, yet its practical use is often limited by weak EPS–cement interfacial bonding, which promotes interfacial debonding and crack propagation and thereby compromises mechanical performance. Although nano-SiO2 (NS) has been reported to improve EPS–cement compatibility, the interfacial strengthening mechanism is still not fully clarified across scales, especially the molecular-level interactions that govern the formation of a robust interfacial transition zone (ITZ). Herein, EPS particles were modified with NS and a multi-scale framework (macro tests, micro-characterization, and molecular dynamics (MD) simulations) was employed to establish a mechanistic linkage between interfacial chemistry/structure and macroscopic performance. The results show that an optimal NS dosage of 9% (by cement mass) increases the 28-day compressive strength and flexural strength of EPS concrete by up to 18.3% and 11.2%, respectively, compared with the unmodified system. SEM, XRD, and FTIR collectively indicate a denser interfacial microstructure, increased hydration-product accumulation near the EPS surface, refined interfacial porosity, and the occurrence of condensation-related reactions involving NS. MD simulations further reveal that NS facilitates the formation of molecular bridges between EPS and C–S–H through hydrogen bonding and ionic interactions, which enhances interfacial adhesion and contributes to improved ITZ thermal stability. This study provides a cross-scale mechanistic understanding for designing high-performance EPS foam concrete via targeted interfacial engineering. MD simulations further suggest that NS enhances interfacial bonding by increasing the occurrence of hydrogen-bond networks and ionic associations at the EPS/C–S–H interface, as evidenced by the intensified interaction-related distributions and peaks in the simulation outputs.