Understanding and modelling ammonia partitioning and transport across reverse osmosis membrane

Reverse osmosis (RO) is increasingly applied for the reclamation of ammonia-rich wastewater. However, the mechanisms governing NH3/NH4+ transport across RO membranes remain unclear. In this study, we develop an ammonia partitioning and transport (APT) model to quantitatively describe NH3/NH4+ retention and transport, incorporating pH-dependent NH3/NH4+ partitioning and membrane charge variation. The theoretical model is validated through laboratory-scale RO experiments under varying pressures, feed concentrations, and pH conditions. In contrast to stable water permeability, the permeability coefficient of total ammonia nitrogen (TAN) is sensitive to pH value. TAN retention is optimal (89.1 ± 3.2%) near neutral pH, decreasing at higher pH due to increased NH3 fraction and at lower pH due to reduced membrane charge. By decoupling diffusion, advection, and electromigration, the APT model shows distinct NH4⁺ and NH3 flux contributions, revealing pH-dependent transport mechanisms governed by speciation and membrane charge. Molecular dynamics simulations further differentiate the transport pathways of NH3 and NH4⁺, showing that NH4+ experiences a significantly higher energy barrier. Validations with both synthetic and real manure streams demonstrate good predictive accuracy for advancing RO applications in ammonia-rich streams. This study presents a mechanistic RO model to predict ammonia transport that explicitly distinguishes NH3 and NH4 + transport and explains the pH dependence of RO performance, providing insights for the reclamation of ammonia-rich wastewater.