The botanical biofiltration of VOCs with active airflow: is removal efficiency related to chemical properties?

• Published in: Atmospheric environment (1994) 2019-10, Vol.214, p.116839, Article 116839
• Main Authors: Pettit, T., Bettes, M., Chapman, A.R., Hoch, L.M., James, N.D., Irga, P.J., Torpy, F.R.
• Format: Article
• Language: English
• Subjects: Active green wall; Botanical biofilter; Green building; Living wall; Potted-plant; Sustainability
• ISSN: 1352-2310; 1873-2844
• DOI: 10.1016/j.atmosenv.2019.116839

Botanical biofiltration using active green walls is showing increasing promise as a viable method for the filtration of volatile organic compounds (VOCs) from ambient air; however there is a high level of heterogeneity reported amongst VOC removal efficiencies, and the reasons for these observations have yet to be explained. Comparisons of removal efficiencies amongst studies is also difficult due to the use of many different VOCs, and systems that have been tested under different conditions. The current work describes a procedure to determine whether some of these differences may be related to the chemical properties of the VOCs themselves. This work used an active green wall system to test the single pass removal efficiency (SPRE) of nine different VOCs (acetone, benzene, cyclohexane, ethanol, ethyl acetate, hexane, isopentane, isopropanol and toluene) and explored which chemical properties were meaningful predictor variables of their biofiltration efficiencies. Ethanol was removed most efficiently (average SPRE of 96.34% ± 1.61), while benzene was least efficiently removed (average SPRE of 19.76% ± 2.93). Multiple stepwise linear regression was used to determine that the dipole moment and molecular mass were significant predictors of VOC SPRE, in combination accounting for 54.6% of the variability in SPREs amongst VOCs. The octanol water partition coefficient, proton affinity, Henry's law constant and vapour pressure were not significant predictors of SPRE. The most influential predictor variable was the dipole moment, alone accounting for 49.8% of the SPRE variability. The model thus allows for an estimation of VOC removal efficiency based on a VOC's chemical properties, and supports the idea that system optimisation could be achieved through methods that promote both VOC partitioning into the biofilter's aqueous phase, and substrate development to enhance adsorption.' •Nine VOCs were tested for their removal efficiency through an active green wall.•The average removal efficiency for the VOCs ranged from 19.76 to 96.34%.•A model to estimate removal efficiency based on chemical properties was developed.•Dipole moment and molecular mass were significant predictors of VOC removal rate.•The model allows estimation of VOC removal based on a VOC's chemical properties.