Ten peaks were detected in the feed solution by positive and negative ion modes (Figure 5 and Table 3). A mass spectrum of the 10 peaks ranging from 113 to 494 of molecular weight was obtained. However, the compositions and names of the chemicals could not be identified because of the lack of mass resolution of the LC-MS/MS. These 10 peaks were not detected in the draw solution and product water, and they were perfectly rejected in the FO side and FO–MD system. The high rejection of positively and negatively charged TOrCs was consistent with the previous studies using the FO or FO–RO systems (Hancock et al. 2011; Xie et al. 2013a; Coday et al. 2014). The rejection of charged TOrCs, such as pharmaceuticals in domestic wastewater, that are treated at the facility by a demonstration-scale sequencing batch membrane bioreactor system was greater than 80 and 99% in a bench-scale FO experiment and a hybrid FO–RO process, respectively (Hancock et al. 2011). Rejection of negatively charged TOrCs tends to increase because of enhanced electrostatic repulsion between the FO membrane and solutes (Xie et al. 2013a). Additionally, the formation of a larger hydrated layer around the ionic species may also affect the rejection of negatively charged TOrCs (Holloway et al. 2014). The rejection of positively charged compounds follows the general principle of size exclusion (Coday et al. 2014). The high rejection of positively charged TOrCs can be explained by a large hydrated radius of ionic species in an aqueous solution (Holloway et al. 2014). Such high rejection of TOrCs was also observed in a pilot-scale FO–RO system for 40 d (Hancock et al. 2011). In this pilot-scale experiment, TOrC rejection was greater than that observed in bench-scale experiments. This result was possibly due to membrane compaction, the establishment of a fouling layer, and optimized hydrodynamic conditions in the pilot-scale system (Hancock et al. 2011).