wetland plant species commonly used in CWs (Typha latifolia, Phragmites
australis, Iris pseudacorus and Juncus effusus) were capable of removing ibuprofen
and iohexol from spiked culture solutions. The pharmaceuticals were taken up by the
roots and translocated to the aerial tissues. However, at the end of the experiment,
plant accumulation constituted only up to 1.1 and 5.7% of the amount ibuprofen and
iohexol spiked initially [89]. In addition, these observations have been confirmed at
mesocosm scale: Hijosa-Valsero et al. [85] observed that plants removed pharmaceuticals not only by uptake but also by the adsorption to plants roots. Pharmaceuticals, such as ibuprofen, salicylic acid and caffeine, could be attached on the root of
P. australis and Typha angustifolia. These compounds were also present in plant
biomass. Therefore, there has been the interest in predicting the uptake of pharmaceuticals. In this regard, the role of Log K ow has also been explored to predict both
uptake and phytodegradation efficiency [57]. However, contrary to classical organic
contaminants like PAHs or PCBs, for which, Log K ow is a good proxy, it does not
work as predictor for pharmaceuticals [50]. Compounds with Log K ow in the range
(1.8–3.1) are expected to be uptake by plants. For ionizable pharmaceuticals,
pH-dependent Log D ow can be calculated, but still fails to support predictions
[26]. Pharmaceuticals Log K ow can vary widely, e.g. À3.05 for iohexol (x-ray
contrast media), 3.5 for ibuprofen or 4.5 for diclofenac. However, the corresponding
Log D ow at pH 8 is À3.11, 0.5, and 0.7, respectively [73, 89], falling outside the
optimal uptake range. For ionizable organic compounds, ionization is expected to
reduce their uptake to the shoots owing to a decrease in their membrane permeability. Nevertheless, these compounds have been documented to be uptake by different
plants. The best modelling efforts are being used to understand the underlying
processes of uptake and transport in plant, as well as for the analyses of the most
relevant parameters and processes, but cannot be expected to predict exact concentrations for ionisable compounds [90].
Micropollutants are thought to be simply driven by diffusion since no specific
transporters for synthetic organic micropollutants exist in the cell membranes of
plants [76]. Therefore, phytotranspiration is of relevance for pharmaceuticals
removal. Some studies have investigated the amount of water loss during their
campaigns, finding that evaporation and evapotranspiration may greatly influence
the results. In planted beds this water loss is greatly increased by plant transpiration,
and evapotranspiration rates depend heavily on the plant type and their vegetative
stage. Naturally, loss of water due to evapotranspiration can cause artefacts in
determining the removal efficiency due to the “up concentration” of compounds in
the effluent, while on a mass flow calculation such error will be avoided. In addition,
evapotranspiration will also affect the water balance and can result in increased HRT
instead of the expected theoretical HRT [27]. Moreover, ibuprofen removal for
different CW mesocosms was positively correlated with the rate of evapotranspiration indicating that plant uptake may be an important process for the ibuprofen
removal [91]. Even though the majority of evidence for uptake and translocation has
been found in laboratorial conditions, these were recently confirmed for full-scale
systems. Analysis of plant tissues, from P. australis from HSSF systems, evidenced
402
P. N. Carvalho
australis, Iris pseudacorus and Juncus effusus) were capable of removing ibuprofen
and iohexol from spiked culture solutions. The pharmaceuticals were taken up by the
roots and translocated to the aerial tissues. However, at the end of the experiment,
plant accumulation constituted only up to 1.1 and 5.7% of the amount ibuprofen and
iohexol spiked initially [89]. In addition, these observations have been confirmed at
mesocosm scale: Hijosa-Valsero et al. [85] observed that plants removed pharmaceuticals not only by uptake but also by the adsorption to plants roots. Pharmaceuticals, such as ibuprofen, salicylic acid and caffeine, could be attached on the root of
P. australis and Typha angustifolia. These compounds were also present in plant
biomass. Therefore, there has been the interest in predicting the uptake of pharmaceuticals. In this regard, the role of Log K ow has also been explored to predict both
uptake and phytodegradation efficiency [57]. However, contrary to classical organic
contaminants like PAHs or PCBs, for which, Log K ow is a good proxy, it does not
work as predictor for pharmaceuticals [50]. Compounds with Log K ow in the range
(1.8–3.1) are expected to be uptake by plants. For ionizable pharmaceuticals,
pH-dependent Log D ow can be calculated, but still fails to support predictions
[26]. Pharmaceuticals Log K ow can vary widely, e.g. À3.05 for iohexol (x-ray
contrast media), 3.5 for ibuprofen or 4.5 for diclofenac. However, the corresponding
Log D ow at pH 8 is À3.11, 0.5, and 0.7, respectively [73, 89], falling outside the
optimal uptake range. For ionizable organic compounds, ionization is expected to
reduce their uptake to the shoots owing to a decrease in their membrane permeability. Nevertheless, these compounds have been documented to be uptake by different
plants. The best modelling efforts are being used to understand the underlying
processes of uptake and transport in plant, as well as for the analyses of the most
relevant parameters and processes, but cannot be expected to predict exact concentrations for ionisable compounds [90].
Micropollutants are thought to be simply driven by diffusion since no specific
transporters for synthetic organic micropollutants exist in the cell membranes of
plants [76]. Therefore, phytotranspiration is of relevance for pharmaceuticals
removal. Some studies have investigated the amount of water loss during their
campaigns, finding that evaporation and evapotranspiration may greatly influence
the results. In planted beds this water loss is greatly increased by plant transpiration,
and evapotranspiration rates depend heavily on the plant type and their vegetative
stage. Naturally, loss of water due to evapotranspiration can cause artefacts in
determining the removal efficiency due to the “up concentration” of compounds in
the effluent, while on a mass flow calculation such error will be avoided. In addition,
evapotranspiration will also affect the water balance and can result in increased HRT
instead of the expected theoretical HRT [27]. Moreover, ibuprofen removal for
different CW mesocosms was positively correlated with the rate of evapotranspiration indicating that plant uptake may be an important process for the ibuprofen
removal [91]. Even though the majority of evidence for uptake and translocation has
been found in laboratorial conditions, these were recently confirmed for full-scale
systems. Analysis of plant tissues, from P. australis from HSSF systems, evidenced
402
P. N. Carvalho