Inadequacy of carbamazepine-spiked model wastewaters for testing photocatalysis efficiency

The study was performed in order to clarify whether carbamazepine-spiked solutions used as model wastewaters are suitable for the assessment of carbamazepine removal from real secondary municipal effluents by photocatalytic oxidation in the presence and absence of activated carbon. Therefore, carbamazepine (10 mg L(-1)) was dissolved in deionized water or in secondary municipal effluent. Photocatalytic oxidation of these model wastewaters was carried out with TiO2 "P25" (100 mg L(-1)) and UV-A lamps in the absence and in the presence of 20 mg L(-1) powdered activated carbon (PAC). Carbamazepine was analyzed photometrically. In deionized water at pH 5.5, carbamazepine was nearly completely removed with a UV dose of 6.48 kJ L(-1). A similar efficiency of photocatalytic oxidation of carbamazepine added to secondary effluent was observed when the suspension pH was 2.7, while at pH 8 and 10.6, carbamazepine removal from spiked secondary effluent with the same UV dose was only 40 and 60%, respectively. Although PAC addition resulted in an initial adsorptive carbamazepine reduction of 20 to 35% from the model wastewaters, it did not lead to markedly enhanced carbamazepine removal in the subsequent photocatalysis phase. During photocatalytic oxidation of unspiked secondary effluent (initial carbamazepine concentration: 133 ng L(-1)) at pH 7.3 with and without PAC, carbamazepine concentrations were analyzed by HPLC/MS/MS. While PAC addition resulted in the adsorption of about 90% of the initial carbamazepine, photocatalysis did not lead to any carbamazepine removal at all. This indicates that the experiments with spiked model wastewaters – even in a secondary effluent matrix – are absolutely inadequate for predicting photocatalytic carbamazepine removal under real conditions.