(97c) Deploying an in Situ Raman Methodology for Online, Live Solvent Composition Analysis for Process Monitoring of CO2 Capture

Solvent-based carbon capture continues to represent a key pathway to meet climate change mitigation goals. For both, the continued adoption of these capture processes in industrial process and for development work by researchers, the optimization and testing of this process would be significantly aided by integrating in situ solvent monitoring capability.[1] Real-time analysis for feedback on varying water, carbon dioxide, and dissolved gas contaminants offers the operator the potential to respond more quickly in maintaining a balance of steady operation. It also allows insight into solvent integrity and life cycle analysis. Using unique data acquisition techniques targeted for this purpose, Raman spectra of a single-component water-lean solvent, N-(2-ethoxyethyl)-3-morpholinopropan-1-amine (EEMPA) were collected at known and controlled CO₂, H₂O, SO₂, and NO concentrations via a custom-built vapor-liquid equilibrium cell coupled with an in situ Raman cell. The quantification of CO₂, H₂O, SO₂, and NO loadings in EEMPA was performed by principal component regression and partial least squares methods with analysis of uncertainties. An algorithm was developed using this training set data that calculates and outputs the 5-component composition of any unknown EEMPA-based solvent sample from the capture process. This model has been deployed and demonstrated for use in an industrial scale capture process at the National Carbon Capture Center in Alabama, US. We describe and showcase the development of this capability and the results from its deployment in realistic conditions.

[1] Lines et al. Anal. Chem. 2023, 95, 15566-15576. DOI: 10.1021/acs.analchem.3c02281