Influence of bubble size on perfluorooctanesulfonic acid degradation in a pilot scale non-thermal plasma treatment reactor

• Published in: Chemical engineering journal (Lausanne, Switzerland : 1996) Vol. 489; p. 151349
• Main Authors: Alam, David, Lee, Samiuela, Hong, Jungmi, Fletcher, David F., Liu, Xinying, McClure, Dale, Cook, David, le Nepvou de Carfort, Johan, Krühne, Ulrich, Cullen, P.J., Kavanagh, John M.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2024
• Subjects: Bubble transport; Non-thermal plasma; PFAS; Remediation; Scale up; Scale up; Non-thermal plasma; Bubble transport; Remediation; PFAS
• ISSN: 1385-8947; 1873-3212
• DOI: 10.1016/j.cej.2024.151349

[Display omitted] •Smaller bubble size increases the rate of PFAS transport to the plasma discharge.•Decreasing the bubble size from 2 to less than 1 mm improved the rate by 150 %.•Mathematical model was developed based on mass transfer and adsorption isotherms.•Modelling predicted removal rates were comparable with experimental data. A 25L working volume non-thermal plasma-based treatment reactor was trialled to destroy per- and polyfluoroalkyl substances (PFAS) utilising argon bubbles to transport PFAS to the surface to be destroyed with plasma interaction at the argon-liquid interface. The breakdown rate of PFAS and the system's overall energy efficiency could be improved while minimising gas usage by utilising small bubbles (0.6–0.7 mm d32) to maximise the transport of PFAS to the plasma discharge for destruction. Vertically scaling the treatment reactor dimensions increases the overall liquid height and dwell time for bubbles to contact and transport PFAS molecules to the surface. The removal rate of perfluorooctane sulfonate (PFOS) correlated with the total surface area of the gas. Significant concentration gradients of PFOS could be observed when sampling from different liquid heights within the 25 L reactor. A one-dimensional model of mass transfer to the surface of rising bubbles was developed and gave good predictions of the overall rates of PFOS breakdown with modelled time constants of 0.14–0.18 min−1 versus 0.16 ± 0.01 min−1 for the fine bubble diffuser, and 0.048–0.053 min−1 versus 0.06 min−1 for the medium bubble diffuser. The time constant compared favourably with similar experiments at the 2 L scale of 0.11 min−1.