NMR Spectroscopic Investigation of the Reactive Absorption of Carbon Dioxide in Aqueous Amine Solutions

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Removing sour gases such as CO₂ and H₂S from gas streams is industrially important and has become increasingly relevant for climate mitigation, especially for flue gases from fossil-fuel power plants. These separations are typically carried out by reactive absorption using aqueous amine solutions, which requires property models that accurately capture both vapor–liquid equilibrium and liquid-phase reaction chemistry. Despite decades of use, the detailed reaction network among amines, water, and CO₂—including byproduct formation and degradation—is still not sufficiently understood, particularly for blended amine systems. Quantitative online NMR spectroscopy is presented as a key method because it can directly determine true liquid-phase speciation (e.g., amine vs. carbamate, molecular CO₂ vs. bicarbonate) under industrially relevant conditions. This work introduces an online NMR apparatus featuring a thermostated, pressurizable cell coupled to the NMR probe via a capillary loop, enabling non-invasive, in situ speciation measurements. Using this setup, equilibria in aqueous MEA, DEA, MDEA, and piperazine solutions were studied over 20–80 °C, 20–40 wt% amine, and CO₂ loadings up to 1.3 mol/mol, and byproducts were observed in all cases. The measured speciation data, together with literature VLE data, were used to derive new interaction parameter sets for the extended Pitzer model and to incorporate byproduct reactions; several equilibrium constants were also newly determined or updated. Overall, the study demonstrates that simultaneous correlation of chemical and phase equilibria is feasible, and suggests the extended Pitzer model as a particularly strong option, while emphasizing that further improvements depend on more reliable VLE datasets.

Article activity feed