Understanding the corrosion of CO2-loaded 2-amino-2-methyl-1-propanol solutions assisted by thermodynamic modeling

Liangfu Zheng, Naser S. Matin, Jesse Thompson, James Landon, Nicolas E. Holubowitch, Kunlei Liu

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


Corrosion of A106 carbon steel in a naturally aerated 30 wt.% 2-amino-2-methyl-1-propanol-based solution (AMP, a sterically hindered primary amine) with 0.43 molCO2/molAMP was evaluated at 80 °C. Substantial decrease in corrosion rate, i.e., over two orders of magnitude, was observed over the initial 70 h, which is the result of formation of a protective FeCO3 layer followed by passivation of the A106 surface. Mechanisms for formation of these protective layers are discussed with comparison to corrosion in a 30 wt.% monoethanolamine solution as well as with the help of thermodynamic modeling of the AMP-H2O-CO2 system. Experimental solubility data from literature were employed to extract a thermodynamic model for aqueous solutions of AMP with concentrations ranging from 17.8 to 36.6 wt.% at various CO2 loadings. Liquid phase speciation was determined by employing an electrolyte-NRTL model. The AMP carbamate stability constant, molecule-ion pair, and molecule–molecule interaction parameters in the studied concentrations were obtained. The determined CO2 equilibrium properties are in agreement with previously reported experimental data.

Original languageEnglish
Pages (from-to)211-218
Number of pages8
JournalInternational Journal of Greenhouse Gas Control
StatePublished - Nov 1 2016

Bibliographical note

Publisher Copyright:
© 2016 Elsevier Ltd


  • AMP
  • Corrosion
  • Passivity
  • Steel
  • Sterically hindered amine
  • Thermodynamic modeling

ASJC Scopus subject areas

  • Pollution
  • General Energy
  • Management, Monitoring, Policy and Law
  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Understanding the corrosion of CO2-loaded 2-amino-2-methyl-1-propanol solutions assisted by thermodynamic modeling'. Together they form a unique fingerprint.

Cite this