Abstract

Polychlorinated biphenyl (PCB) formulation (Aroclor 1242 or 1248)–surfactant (Brij 97 or Triton CF54) emulsions (0.1–1.0%, v/v) were extracted with supercritical carbon dioxide (scCO₂). The declinations curve that resulted for each loading of PCB substrate was accurately modelled as a single-exponential decay, and the half-life [t_1/2, 10.76 min ± 2% (Brij 97) or 10.0 min ± 14% (Triton CF54)] was independent of the level of loading. When the extraction was combined with on-line dechlorination, t_1/2 was increased to 17.62 min ± 6% or 16.46 min ± 5%, respectively The dechlorination reactor contributed appreciably to the back-pressure of the system so that, under identical scCO₂ head pressures, the flow rate at the exit of the system was reduced from 1500 mL min⁻¹ (as decompressed gas) to 900 mL min⁻¹. When corrected for the difference in flow rates, the declination curves in the absence and presence of the reactor became virtually identical. For a surfactant suspension from a soil that was field-burdened with only 6 ppm PCBs, extended operation (extraction–dechlorination) for 15 h, during which the substrate suspension was replaced every 30 min, caused no loss of the dechlorination efficiency; however, for 1% (v/v) Aroclor 1242 or 1248 in surfactant suspension, the reactor gradually lost reactivity over 5 h of continued operation. However, the reactivity could be restored virtually completely by purging the reactor column with scCO₂ for 3 h or washing the column at ambient temperature with 30 mL of methanol–water (7 + 3, v/v). Mass spectrometry of the reaction products indicated that the principal products included biphenyl and toluene, but methylated benzenes and phenols were also present.