The effects of blending diisopropanolamine and 2-amino-2-methyl-1-propanol for carbon dioxide capture process: Comparative study using thermodynamic modeling and artificial neural network predictions

Amine-based absorption processes represents a promising strategy for CO₂ capture. This study systematically examined the CO₂ solubility, absorption behaviors, and regeneration energy (Q_regen) of aqueous blended-amine solutions of diisopropanolamine (DIPA) and 2-amino-2-methyl-1-propanol (AMP) solutions. CO₂ solubility was experimentally measured across five blending ratios (30:0:70 to 0:30:70 (w/w)), four temperatures of (313–383 K), and CO₂ partial pressure up to 450 kPa. The data were well-correlated with both the electrolyte non-random two-liquid (e-NRTL) thermodynamic model and the artificial neural network (ANN) model. The model-physical alignment score was suggested as a novel metric to ensure consistency with fundamental absorption behaviors. While the thermodynamic model provided mechanistic insights, the ANN model offered a simpler, computationally efficient approach. Both models successfully generalized with a reduced dataset, reinforcing their applicability in absorbent screening. AMP exhibited superior CO₂ solubility, whereas DIPA offered benefits of lower pH and reduced heat of absorption. The absorption behavior of aqueous blended-amine solutions was characterized by prioritized AMPH⁺ formation, leading to an AMP-dominated system at low CO₂ loading ratios. The analysis of Q_regen, including sensible, reaction, and latent heats, revealed that the lowest Q_regen was achieved from the single AMP system. The addition of 5 wt% DIPA resulted in only an 8% increase in Q_regen, still achieving a 33.4% reduction compared to MEA. These findings highlight aqueous DIPA-AMP solutions as energy-efficient and chemically stable absorbents.