Thermal Degradation and Corrosion of Amines for CO2 Capture

This report examines the thermal degradation and corrosion of various amine solvents as they apply to amine scrubbing for CO2 capture. Amines were placed in stainless steel cylinders and heated in convective ovens to simulate the stripping conditions inside a scrubbing unit. Samples were measured for remaining amine concentration, to test for degradation, and metals concentration, to estimate corrosion of the cylinder. The maximum stripping temperature of a particular compound, a measure of resistance to thermal degradation, strongly correlated with amine chain length. The linear amines studied had the following max temperatures: EDA (116 °C), PDA (124 °C), DAB (126 °C), BAE (130 °C), HMDA (140 °C), MEA (116 °C), MPA (129 °C), and DGA® (134 °C). The SHA/PZ blends had the following weighted max temperatures: AMP (143 °C), AMPD (135 °C), TRIS (130 °C), tBuAE (150 °C), PM (97 °C), and PE (129 °C). The linear amines follow initial first-order degradation curves, consistent with literature mechanisms. EDA, PDA, BAE, and AMP degraded significantly more slowly under acid conditions, suggesting that the degradation mechanisms do not incorporate CO2. Acid loaded DAB degraded at a similar rate to CO2-loaded conditions. MEA corroded 15 times faster than MPA; MAE corroded 3 times faster than EAE; DMAE-PZ corroded qualitatively faster than DMAP-PZ. These three pairs support the hypothesis that two-carbon chains corrode more than three-carbon chains. EDA corroded 40 to 80 times more than PDA according to older studies, seen in Figure 38, but more recent tests show similar corrosion rates where EDA is only 1.2 times faster (Figures 32 and 33). Corrosion and amine concentration correlate strongly; corrosion does not correlate strongly with temperature or CO2-loading. Corrosion and formate generation appear to correlate, supporting corrosion mechanisms proposed in literature.