Quantifying OH radical generation in hydrodynamic cavitation via coumarin dosimetry: Influence of operating parameters and cavitation devices

•Monitored hydroxylation products to quantify generation of OH radicals.•Use of sulphuric acid for adjusting pH was nearly 3 times more effective than hydrochloric acid.•Vortex based cavitation device outperformed orifice and venturi based devices.•Scale-up by factor 50 was not found to be detrimental to hydroxylation. Hydrodynamic cavitation (HC) has been extensively investigated for effluent treatment applications. Performance of HC devices or processes is often reported in terms of degradation of organic pollutants rather than quantification of hydroxyl (OH) radicals. In this study, generation of OH radicals in vortex based cavitation device using coumarin dosimetry was quantified. Coumarin was used as the chemical probe with an initial concentration of 100 µM (15 ppm). Generation of OH radicals was quantified by analysing generated single hydroxylated products. The influence of operating parameters such as pH and type of acid used to adjust pH, dissolved oxygen, and inlet and outlet pressures was investigated. Acidic pH was found to be more conducive for generating OH radicals and therefore subsequent experiments were performed at pH of 3. Sulphuric acid was found to be more than three times effective than hydrochloric acid in generating OH radicals. Effect of initial levels of dissolved oxygen was found to influence OH radical generation. Performance of vortex based cavitation device was then compared with other commonly used cavitation devices based on orifice and venturi. The vortex based cavitation device was found to outperform the orifice and venturi based devices in terms of initial per-pass factor. Influence of device scale (nominal flow rate through the device) on performance was then evaluated. The results presented for these devices unambiguously quantifies their cavitational performance. The presented results will be useful for evaluating computational models and stimulate further development of predictive computational models in this challenging area.