Abstract
The removal of per- and polyfluoroalkyl substance (PFAS) pollution from the environment is a globally pressing issue, due to some PFAS' recalcitrant, bioaccumulative, and carcinogenic nature. Destruction via ultrasonic waves (sonolysis) is a promising contender for industrialisation due to; moderate power consumption, applicability to several PFAS and sample types, and limited by-products. Liquid flow rate through an ultrasonic reactor can affect the size, shape, and spatial distribution of ultrasonic cavities and hence their chemical activity. Such effects have not been studied during PFAS sonolysis, and temporal effects have not been studied much beyond the reactant concentration. Here, the effects of varying recirculating flow rate on the ultrasonic defluorination of perfluorooctane sulfonic acid (PFOS) and implications for industrial scale up are presented. Under the ultrasonic power (200 W L(-1), 2.27 W cm(-2)) and frequency (410 kHz) used, flow rates of 79 and 214 ml min(-1) enhanced defluorination up to 14 % during 30 min of treatment. However, these effects were temporal and most significant in the initial minutes of treatment. This indicated a dynamic bubble size distribution which stabilised after around 15 min. Defluorination rates of PFOS were compared with measured potassium iodide dosimetry, calorimetry, sonoluminescence (SL), and sonochemiluminescence (SCL). Flow rates which enhanced defluorination correlated moderately with enhanced SCL and negatively impacted SL, calorimetry, and dosimetry. Effects were attributed to perturbed cavity surfaces, leading to asymmetric cavity collapse, and the possibility of enhanced solvated electron production/interaction. SL, SCL, dosimetry, and calorimetric measurements were also temporal, and each showed different times to equilibrate. Flow rates of 439 and 889 ml min(-1) returned all sonochemical measurements to the levels without flow, likely due to continued collapse temperature quenching by furthered bubble asymmetry. Flow also enhanced reactor cooling, which is significant for industrial temperature control. The pump energy consumed was small (≈1.9 %) compared to that of the amplifier and chiller, hence, PFOS defluorination was more cost-effective using flow. However, the effect may be limited for the longer treatment times needed for environmental remediation.