Re
moval efficiency and eli
mination rates of 11 phar
maceuticals and personal care products (PPCPs) were
measuredin two subsurface horizontal flow constructed wetlands(SSFs) characterized by different water depths (i.e. 0.3 and0.5
m) in a 2-year study. Dissolved and particulatephases of wastewater and gravel sa
mples were collectedand analyzed. The PPCP influent concentration rangedfro
m 1 to 25
![](/i<font color=)
mages/entities/
mgr.gif">g L
-1. The best re
moval efficiency was foundin the shallower bed SSF due to its less negative redoxpotential. PPCPs were classified in four groups accordingto their re
moval behavior: (i) the efficiently re
moved(>80%) na
mely caffeine, salicylic acid,
methyl dihydrojas
monate, and carboxy-ibuprofen, (ii) the
moderatelyre
moved (50-80%) na
mely ibuprofren, hydroxy-ibuprofen,and naproxen, (iii) the recalcitrant to the eli
minationna
mely ketoprofen and diclofenac, and, finally, (iv) co
mpoundsthat were eli
minated by hydrophobic interactions na
melypolycyclic
musks (i.e. galaxolide and tonalide). Theseco
mpounds were re
moved
more than 80% fro
m the effluentbut occurred at high concentrations (up to 824
![](/i<font color=)
mages/entities/
mgr.gif">g kg
-1)in the gravel bed. Accordingly, their eli
mination by sorptiononto the organic
matter retained is the predo
minantre
moval
mechanis
m. Further
more, the constructed wetlandclogging appears to induce a negative effect in thePPCP degradation in the SSF evaluated. The PPCP eli
minationclassified as efficiently and
moderately re
moved throughthe shallow bed fitted to either zero- or a first-order arealkinetics. Finally, the apparent distribution coefficientsbetween suspended solids (Kd'
ss) or gravel bed (Kd'
gb)and water were deter
mined in the different sa
mpling pointsof the wetland obtaining values co
mparable to thedescribed previously for sewage sludge.