Development and Evaluation of a Novel Method for Determining Absorbent Composition in Aqueous Ammonia-Based CO2 and SO32- and SO42- Loaded Capture Process Solutions via FT-IR Spectroscopy

CO, capture using aqueous ammonia is a potentially attractive option for emissions reductions from energy production and industrial processes. From an operational perspective, the capture absorbent must be monitored continuously to maintain the maximum efficiency of the capture process. In practice the composition of the absorbent is typically evaluated offline and retrospectively via wet chemistry methods, delaying any necessary variations to the process conditions to maintain maximum efficiency. Online absorbent monitoring methods incorporating spectroscopy via Raman or Fourier transform infrared (FT-IR) are attractive options due to their rapid response times and flexibility of the resulting output to be incorporated directly into process control packages. The present study outlines an evaluation of the FT-IR spectroscopic technique with analysis via partial least squares regression (PLSR) for a range of dilute to concentrated aqueous ammonia absorbents from similar to 0.3-6.0 M and over a range of CO2 loadings from similar to 0.0-0.6 mol CO2/mol NH3. The water concentration in the samples ranges from, similar to 35.2-55.2 M. The effect of interfering SOx species on the FT-IR method has been evaluated by incorporating dissolved SO32- and SO42- components into the solutions from 0.0-1.5 M. The analysis results in accurate concentrations for all analytes. The robustness of the analysis results has been evaluated and discussed. Additionally, FT-IR spectroscopy with PLSR was compared with conventional titration methods for a selected series of mixed NH3/CO2 standard solutions and a series of liquid samples from a bench-scale CO2 absorption process. At low concentrations where the total NH3 concentration is less than 4.0 M and the total CO2 concentration is less than 1.5 M, both the combined PLSR with FT-IR method and the conventional potentiometric titration methods were suitable for the evaluation of the liquid compositions. However, at concentrations out of the low concentration range, the combined PLSR and FT-IR method was proven to have a robustness and accuracy greater than those of the conventional potentiometric titration methods. Therefore, given the simplicity and rapid turnaround of FT-IR spectroscopy in combination with PLSR, we consider this to be a superior and flexible technique for monitoring of CO2 loaded aqueous ammonia solutions.