The adsorption of four wide-use pharmaceuticals (carbamazepine, diclofenac, ibuprofen, and ketoprofen) onto a porous silica was investigated under varied ionic strengths, different anions, divalent cations (Ca2+ and Mg2+), trivalent cations (Al3+ and Fe3+), and natural organic matter (NOM). The experiments demonstrated that at a given pH the adsorption was most affected by ionic strength, trivalent cations, and properties of pharmaceuticals. The increase of ionic strength resulted in an increase in the adsorption of ketoprofen, but a decrease in the adsorption of carbamazepine. Trivalent metal cations made intense increases in the adsorption of three acidic pharmaceuticals, which could be due to the formation of inner-sphere complex of the cations on the surface and/or complexation of the pharmaceuticals with both surface and aqueous metal species. It was found that the adsorption of carbamazepine was not affected by divalent and trivalent cations, whereas the adsorption of diclofenac was solely impacted by the presence of Al3+. Moreover, divalent cations at low concentration could slightly enhance the adsorption of ibuprofen and ketoprofen, whereas NOM caused a reduction in the adsorption of the tested pharmaceuticals except for diclofenac. These results suggest that ionic strength, divalent cations, trivalent cations, and NOM are notable factors affecting the adsorption of pharmaceuticals and thus the ultimate fate of pharmaceuticals in the aqueous environment.
Get your grade
or your money back
using our Essay Writing Service!
The ubiquitous presence of pharmaceuticals in the environment (Halling-Sorensen et al., 1998; Fent et al., 2006) has raised a number of critical concerns, including their fate in soil and water (Halling-Sorensen et al., 1998; Heberer, 2002; Loffler et al., 2005). The environmental fate of pharmaceuticals can be strongly influenced by their adsorption to soil and sediment; adsorption influences the distribution of substances between aqueous phase and solid surfaces, which in turn affects their transport in an aqueous environment and regulates their ultimate fate. Adsorption behavior of pharmaceuticals in water is controlled by adsorbent properties, pharmaceutical properties (e.g., water solubility, hydrophobicity, and acid/base properties), and water characteristics (e.g., pH, ionic strength, cations, anions, and natural organic matter (NOM)).
To date, several reports related to the adsorption of pharmaceuticals to natural materials or components of natural materials have been published. Soils (Figueroa and MacKay, 2005), clays (Figueroa et al., 2004; Pils and Laird, 2007), and hydrous oxides (Gu and Karthikeyan, 2005a) have been investigated for the adsorption of tetracycline. In addition, since silica is a major component of the coarse clay fraction of many soils, the adsorption of several pharmaceuticals onto silica has been explored (Goyne et al., 2005; Turku et al., 2007; Bui and Choi, 2009).
The investigation of the effect water characteristics on the adsorption of pharmaceuticals is essential in order to estimate their ultimate fate. To this end, except for pH, common in most reports, previous studies in this area have reported on the effects of ionic strength, calcium cations, and humic substances on pharmaceutical adsorption. An increase of ionic strength results in less overall sorption of oxytetracycline on clays (Figueroa et al., 2004) and the decreased sorption of tetracycline on silica (Turku et al., 2007). At low pHs, clay treated with Ca2+ showed lower sorption with oxytetracycline than clays treated with Na+ (Figueroa et al., 2004), though a reverse result was observed at a pH above 7. Humic acid at a concentration of 1 mg/L increased the sorption of oxytetracycline on clay, whereas 10 mg/L of humic acid decreased it (Kulshrestha et al., 2004). In addition, the adsorption of humic substances on clay suppressed tetracycline sorption (Pils and Laird, 2007). Nevertheless, the influence of all factors including pH, ionic strength, anions, cations, and NOM on the adsorption of pharmaceuticals has been rarely evaluated.
Carbamazepine (an antiepileptic), diclofenac, ibuprofen, and ketoprofen (nonsteroidal anti-inflammatory drugs) are four pharmaceuticals that are widely used and ubiquitously detected in aqueous environments (Halling-Sorensen et al., 1998; Fent et al., 2006). These pharmaceuticals are also considered compounds that have a high environmental risk (Risk Quotient > 1) (Hernando et al., 2006). Therefore, this study was initiated to determine the influence of water characteristics on the adsorption of these pharmaceuticals to silica. In our research, we aim to investigate the impact of different ionic strengths (0-50 mM), anions (nitrate, bicarbonate, sulfate, phosphate, and silicate), alkaline-earth cations (Ca2+, Mg2+), trivalent cations (Al3+, Fe3+), and NOM (Suwannee River fulvic acid (FA), Suwannee River humic acid (HA), Suwannee River NOM) on the adsorption of pharmaceuticals on silica.
2. Materials and Methods
2.1. Chemicals and adsorbent
Always on Time
Marked to Standard
All chemicals used in this study are reagent grade and purchased from Sigma-Aldrich, Junsei, Oriental Chemical Industries, and BASF. Suwannee River NOM (catalog no. 1R101N), Suwannee River FA (catalog no. 2S101F), and Suwannee River HA (catalog no. 2S101H) were purchased from the International Society of Humic Substances (ISHS) and used as received; the analytical data of these NOM can be found in the Supplementary Information (SI). In addition, high purity (>99%) carbamazepine, diclofenac, ibuprofen, and ketoprofen were purchased from Aldrich. The molecular structure and physicochemical properties of the pharmaceuticals are summarized in Table 1. Note that stock solutions of the pharmaceuticals were made in de-ionized water, with the addition of a small fraction of HPLC grade methanol (Fisher Scientific).
Mesoporous silica was prepared using a non-ionic template (Zhao et al., 1998). The synthesized material was then characterized by X-ray diffraction, transmission electron microscopy, N2 adsorption-desorption measurements, and point of zero charge (PZC) measurements; a more detailed characterization of the silica is described elsewhere (Bui and Choi, 2009). The Brunauer-Emmett-Teller (BET) specific surface area, pore volume, and pore size of the silica was estimated as 737 m2/g, 1.03 cm3/g, and ï¾8 nm, respectively, and the PZC of the silica was determined to be 4.0 Â± 0.1.
2.2. Batch adsorption experiments
Batch adsorption of the pharmaceuticals was conducted in 40-mL amber vials, in which working solutions were prepared with a known amount of inorganic salts and/or NOM in a buffer solution (2 mM 2-morpholinoethanesulfonic acid (MES) at pH 5.3). The solution pH could then be adjusted by adding HCl or NaOH 0.1 M as needed. Previously, MES was reported to have a minimal effect on the sorption of organic compounds to minerals (Nowack et al., 1996) and very weak complexing properties (Good et al., 1966). Next, stock solutions of pharmaceuticals were spiked to make an initial concentration of 100 Âµg L-1; to this solution 10 mg of silica was added to attain a solid:liquid ratio of 1 g/L. Note that blank solutions (pharmaceuticals without silica) were prepared in parallel with every sample. The samples and blanks were subsequently agitated in an incubator at 200 rpm for 24 h at 25 oC. Aliquots were taken at the end of the 24 h period, filtered through a 0.45-Âµm cellulose acetate membrane filter (MFS), and analyzed by LC-tandem MS. All experiments were performed in duplicate or triplicate.
Different working solutions were utilized to investigate the influence of chemicals in the solutions on the adsorption of pharmaceuticals on silica. Here, the effect of ionic strength on the pharmaceutical adsorption was determined using inert electrolyte KCl solutions (0-50 mM). The impact of different anions on the adsorption of pharmaceuticals was then investigated by using solutions containing 1 mM sodium salts of nitrate, bicarbonate, sulfate, phosphate, or silicate anions. In addition, chloride solutions of cations such as calcium, magnesium, aluminum, and ferric ions were prepared to estimate their effects on adsorption. Cation concentrations were selected such that they were consistent with those in natural water (Hem, 1985); 1-5 mM solutions were applied for divalent cations (Ca2+ and Mg2+) and a 0.1 mM solution was used for trivalent cations (Al3+ and Fe3+). To consider the NOM impact, solutions of 5 and 10 mg L-1 of Suwannee River NOM, Suwannee River FA, and Suwannee River HA at a total ionic strength of 10 mM (KCl) were prepared.
2.3. Sample preparation and analysis
Prior to analysis, the obtained filtrate was acidified using formic acid, spiked with corresponding surrogate standards (10,11-dihydrocarbamazepine (DHC) for carbamazepine, 2-(3-chlorophenoxy) propionic acid (cloprop) for the other pharmaceuticals) and extracted via solid phase extraction techniques (60 mg Waters Oasis HLB cartridges) to ensure that samples would be free of other organic compounds that could interfere the analysis or inorganic ions that could decay the column. Next, pharmaceuticals were analyzed on an LC-tandem MS system equipped with an HPLC Waters 2695 Separations Module accompanied with a Waters 2996 PAD, a mass spectrometer (Waters (Micromass) Quattro micro API), and an X-Terra MS C18 column (Waters, 50 x 2.1 mm I.D., 5 Âµm, 125 Å, endcapped C18 hybrid particles). The mass spectrometer was run in ESI positive mode for carbamazepine, ketoprofen, and DHC and in ESI negative for diclofenac, ibuprofen, and cloprop. Multiple reaction monitoring (MRM) was used to identify the compounds. Dwell times varied from 0.3 to 0.5 s depending on analytes. The experimental details of sample preparation and analytical methods are described elsewhere (Bui and Choi, 2009). Note that each sample was analyzed two times and the obtained values were averaged.
2.4. Data calculation and analysis
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
The percentage of adsorption of each pharmaceutical was calculated based on the difference between the concentration in the sample and the corresponding blank. Data sets were then analyzed and plotted using Sigma Plot version 11.0 (SPSS Inc., CA, USA) and Excel 2007 (Microsoft Corporation, WA, USA) software packages; differences between the samples at different ionic strengths were subsequently evaluated using a one-way analysis of variance (ANOVA) in conjunction with all pairwise multiple comparison procedures (Tukey Test). In addition, a one-way ANOVA based on multiple comparisons versus the control group (Dunnett's Method) was then used to estimate the difference between the samples and a control sample, which was applied to study the effects of anions, cations, and NOM. The samples in which the working solution contained only KCl and the buffer were used as 'controls', providing that the ionic strength of a sample and its control were close to each other. As such, the control for the samples containing 1 mM of anions was the sample of 1 mM of KCl; similarly, the control for samples containing 0.1 mM of trivalent cations, 1 mM of divalent cations, and 5 mM of divalent cations were the samples having 0, 1, and 10 mM KCl, respectively. The control for the sample containing NOM was the sample with 10 mM KCl.
3. Results and discussion
3.1. Effect of ionic strength
Fig. 1 shows the percentage of adsorption of four pharmaceuticals onto the porous silica at different ionic strengths (0-50 mM). The adsorption of diclofenac is seen to display a similar trend as the adsorption of ketoprofen, whereas the adsorption behaviors of carbamazepine and ibuprofen show similarity. The ANOVA test results indicate that ionic strength did not result any statistically significant change (p > 0.05) in the adsorption of diclofenac and ibuprofen. However, the increase of ionic strength from 0-1 to 20-50 mM caused a significant impact (p < 0.05) on the adsorption of carbamazepine and ketoprofen. The ANOVA results also demonstrated that variation of the ionic strength in smaller ranges did not produce a significant effect (p > 0.05) on the adsorption, thereby supporting the posit that small differences in ionic strength between the samples and controls presented in the next sections can be neglected.
Based on their pKa values (Table 1), carbamazepine is a neutral compound, though a large fraction of diclofenac, ibuprofen, and ketoprofen are deprotonated and negatively charged in the media (pH 5.3). As the ionic strength was increased from 0 to 20-50 mM, the adsorption of neutral carbamazepine slightly decreased (p < 0.05), whereas the adsorption of negatively charged ketoprofen increased somewhat (p < 0.05). Gao and Pedersen (2005) also demonstrated that the increase of ionic strength lowered the adsorption of uncharged sulfamethazine to montmorillonite but increased the adsorption of anionic sulfamethazine.
Adsorption behavior of silica in an aqueous solution is largely dependent on its ionization and surface charge (Papirer, 2000). Since the silica in this study has a PZC of 4.0, its surface is negatively charged in the media (pH 5.3); however, the surface charge of silica is low over a large pH range (2-6) due to its poor ionization (Persello, 2000). In addition, when the ionic strength increases, three processes may occur (Gao et al., 2008): 1) more surface silanols are ionized through the following equilibrium: ï‚ºSiOH + K+ï‚ºSiO-â€¦K+ + H+; 2) the solubility of pharmaceuticals could be influenced in such a way that a higher ionic strength may lower the solubility of neutral pharmaceuticals but induce the dissolution of ionizable acidic pharmaceuticals (Chan and Heng, 2005); and 3) a larger quantity of K+ is adsorbed as outer-sphere complexes in the ï¢ layer, resulting double-layer compression and a decrease in surface potentials.
Therefore, it was reasonable to ascribe most of the reduction in carbamazepine adsorption to both its lower solubility and the higher ionization of the silica surfaces. Indeed, lower carbamazepine solubility may favor aggregation among carbamazepine molecules through hydrophobic interactions (Turku et al., 2007), which potentially makes carbamazepine hard to access and adsorb in the silica pores. Additionally, the higher ionization of the silica surfaces implies that a smaller number of silanols remain, which are presumably responsible for the adsorption of carbamazepine (Bui and Choi, 2009), and consequently adsorption decreased. Besides, the increase of ketoprofen adsorption can be attributed to double-layer compression and a decrease of surface potentials; more negatively charged ketoprofen molecules could adsorb to the surface due to the reduction of anion repulsion.
3.2. Effect of anions
Since the concentration of anions such as nitrate, bicarbonate, sulfate, phosphate, and silicate in surface water is usually less than 1 mM (Hem, 1985), 1 mM anion solutions were used in these experiments. The pharmaceutical adsorption in the presence of different anions was then investigated and compared to adsorption in the presence of chloride ions (see Fig. S1, SI). Subsequent ANOVA test results indicate that the tested anions did not significantly impact (p > 0.05) adsorption of these pharmaceuticals to silica, as compared to chloride ions. This result can be attributed to the minimal adsorption of anions to silica surfaces, probably due to the electrostatic repulsion between the negatively charged silica surface and anions.
3.3. Effect of divalent cations
The adsorption of pharmaceuticals in the presence of divalent cations (Ca2+, Mg2+) was compared to adsorption in the presence of monovalent potassium ions (see Table S1, SI), with a summary of the results of the ANOVA test shown in Table 2. The results suggest that divalent cations did not significantly impact the adsorption of carbamazepine and diclofenac. However, divalent cations at low concentrations (1 mM) were found to significantly affect the adsorption of ibuprofen (both Ca2+ and Mg2+) and ketoprofen (only Mg2+), as the percentage of ibuprofen and ketoprofen adsorbed was higher in 1 mM of divalent cations than that in potassium ions (Table S1). Pils and Laird (2007) also reported that the sorption of tetracycline and chlortetracycline at pH 7 was higher on Ca-saturated soil components than on K-saturated soil components. The observed impacts can be results of the adsorption of divalent cations onto the silica surface. Below pH 7.5, divalent cations are physically adsorbed into the interface as counterions in place of potassium ions to form an outer-sphere complex ï‚ºSiO-â€¦M2+ (Tadros and Lyklema, 1969). And at low concentrations (1 mM), Ca2+ and Mg2+ are sorbed on silica to a greater extent than monovalent K+ (Tadros and Lyklema, 1969), which decreases the surface charge of silica, as compared to monovalent ions (Persello, 2000). Additionally, divalent cations are able to bridge between negative charge sites on silica surfaces and the negative charge of pharmaceuticals (Goyne et al., 2005; Pils and Laird, 2007); conversely, monovalent K+ can only satisfy one negative charge site at a time. Consequently, ketoprofen and ibuprofen, most of which are negatively charged in the media, are more favorable for adsorption on silica surfaces.
Accordingly, high concentrations (5 mM) of divalent cations were expected to incur a higher increase of pharmaceutical adsorption than monovalent K+. However, it was found that divalent cations at high concentration did not significantly affect the adsorption of pharmaceuticals (Table 2) except that the presence of 5 mM of Mg2+ facilitated a significant decrease in ibuprofen adsorption (Table S1), compared to when K+ was present. The indistinctive adsorption of pharmaceuticals in the presence of divalent cations and K+ suggests that the adsorption of both types of cations on silica surfaces reach a limited value and become equal, which can be a result of the low surface charge of silica (Persello, 2000). Nevertheless, it is not well understood why the adsorption of ibuprofen was reduced in the presence of 5 mM of Mg2+.
The insignificant impacts of divalent cations to the adsorption of diclofenac and carbamazepine are preferably discussed in the next section.
3.4. Effect of trivalent cations
In comparison with divalent cations, trivalent cations (Al3+ and Fe3+) displayed a more intense influence on pharmaceutical adsorption (Fig. 2) despite having a 10-50 times smaller concentration. Again, the ANOVA test results (Table 2) indicate that the adsorption of carbamazepine was not affected by the presence of trivalent cations, though the trivalent cations caused a significant impact on the adsorption of diclofenac, ibuprofen, and ketoprofen. Moreover, whereas the presence of Fe3+ resulted in a unique increase in the adsorption of ketoprofen (10.6%), the presence of Al3+ enhanced the adsorption of diclofenac (17.6%), ibuprofen (13.3%), and ketoprofen (18.6%) (Fig. 2). Meanwhile, the blank experiments (without silica) revealed that only in the presence of trivalent cations was there no change in pharmaceutical concentrations. These results are in agreement with previous reports (Westerhoff et al., 2005; Vieno et al., 2007). Therefore, the enhancement of pharmaceutical adsorption might be due to the adsorption of trivalent cations onto silica surfaces (Persello, 2000) and/or the formation of complexes between trivalent cations and the pharmaceuticals (Wang et al., 2008).
It is known that Al3+ and Fe3+ mainly adsorb as an inner-sphere coordination complex at the silica surface (Persello, 2000), forming positively charged surface groups (Houston et al., 2008) and resulting in a lower surface potential. Moreover, the specific adsorption of trivalent cations on the silica surface is able to reverse the electrophoretic charge (Wakatsuki et al., 1974). As a result, the silica surface becomes more active towards negatively charged acidic pharmaceuticals. Furthermore, it was reported that ciprofloxacin can form complexes with both surface metal species (Al, Fe) and also with dissolved metal cations via the coordination of carboxylate and/or keto groups (Gu and Karthikeyan, 2005b; Trivedi and Vasudevan, 2007). Hence, the three acidic pharmaceuticals (diclofenac, ibuprofen, and ketoprofen) containing carboxylate groups may form complexes with both Al3+ and Fe3+ adsorbed on the silica surface and also Al3+ and Fe3+ in the aqueous solution. Obviously, the formation of surface complexation between pharmaceuticals and adsorbed trivalent metals would enhance adsorption of the pharmaceuticals. In addition, Wang et al. (2008) reported that the adsorption of tetracycline on soil increased in the presence of Cu(II), which was partially attributed to the formation of water-soluble complexes between tetracycline and Cu(II). Accordingly, complexes formed between dissolved Al3+ and Fe3+ and the acidic pharmaceuticals would have less negative charge and thus are easily adsorbed on negatively charged surfaces of silica, rather than the pharmaceutical itself, thereby leading an increase in the adsorption of pharmaceuticals.
Ketoprofen, which contains both keto and carboxylate groups (see Table 1), may have a higher complexing affinity to Fe(III) (both surface and aqueous species) than the other acidic pharmaceuticals, possibly explaining why the presence of Fe3+ only significantly increases ketoprofen adsorption (Table 2). The table also shows that among the trivalent cations, a possibly stronger adsorption of Al3+ to the silica surface produces remarkable changes and a lower surface potential, compared with Fe3+ (Wakatsuki et al., 1974; Persello, 2000), resulting in a more significant impact to the adsorption of diclofenac and ibuprofen.
Neither divalent (Ca2+, Mg2+) nor trivalent (Al3+, Fe3+) cations impacted the adsorption of carbamazepine (Table 2). These results can be attributed to the fact that the adsorption of neutral carbamazepine is not affected by changes in the silica surface charge in the cases. To this end, it was previously proposed that the adsorption of carbamazepine is mainly responsible by the formation of hydrogen bonding between silica surfaces and carbamazepine (Bui and Choi, 2009).
The adsorption of diclofenac was only influenced by the presence of Al3+ (Table 2), while other factors including ionic strength, alkaline-earth cations, and Fe3+ did not cause any impact. This result indicates that Al3+ could strongly adsorb onto the silica surface to intensely alter the surface characteristics or that Al3+ can form a stronger complex with diclofenac. Another possible explanation stems from the desorption results of the pharmaceuticals in a pH 7 buffer (Bui and Choi, 2009), in which the percentage of desorption followed the order: diclofenac (ca. 20%) < carbamazepine < ibuprofen < ketoprofen. The lower desorption percentage of diclofenac suggests that a larger fraction of diclofenac is "nonlabile" adsorbed on silica, compared to the other pharmaceuticals. This nonlabile adsorbed fraction appears to be mostly unaffected by water characteristics, explaining why the adsorption of diclofenac was not significantly influenced by factors other than Al3+.
3.5. Effect of natural organic matters
The percentage of adsorption of four pharmaceuticals in the presence of NOM were investigated and compared with the controls (Table S2, SI). The ANOVA test results (Table 3) indicate that NOM did not make any statistically significant change in the adsorption of diclofenac on silica. Suwannee River HA solely affected the adsorption of ibuprofen, whereas Suwannee River FA significantly affected the adsorption of ketoprofen and ibuprofen. In addition, Suwannee River NOM significantly influenced the adsorption of carbamazepine, ibuprofen, and ketoprofen. Furthermore, the adsorption of pharmaceuticals was significantly decreased (Table S2) by the impact of NOM; a significant impact was also observed on the rejection of every pharmaceuticals in the presence of Suwannee River NOM (Comerton et al., 2009). Our preliminary experiments showed that the adsorption of NOM on silica was negligible, which could be a result of the negatively charged NOM interacting with the high hydrophilicity and negative charge of the silica surfaces (Yang et al., 2009). Hence, the decrease of pharmaceutical adsorption could be attributed to an association among pharmaceuticals and NOM, which can be formed via hydrophobic interactions (Kulshrestha et al., 2004; Tulp et al., 2009) and hydrogen bonding (Gu et al., 2007; Pils and Laird, 2007). Such an association of pharmaceuticals with supramolecular humic substances would confine the pharmaceuticals and consequently decrease their adsorption on silica.
The statistically insignificant impact of NOM on the adsorption of diclofenac (Table 3) may be due to the small amount of diclofenac in non-ionized form, compared to the rest of the pharmaceuticals (see Table 1). To this end, Tulp et al. (2009) demonstrated that neutral species sorbed to peat organic matter by a factor of ca. 40 times (for ibuprofen and ketoprofen) higher than their corresponding anionic species, due to the negative charge of the organic matter.
The different impacts of various NOM on pharmaceutical adsorption could be dependent on the ability in making an association between them, which in turn could be expected to depend on the hydrophobicity of both the pharmaceuticals and NOM (see Table S3, SI). Sorption of both neutral and anionic species to peat organic matter showed an increase with their hydrophobicity (logKow) (Tulp et al., 2009). Owing to the isolation procedures (IHSS, 2008), Suwannee River HA and Suwannee River FA contain only hydrophobic organic acids, with HA generally having a higher hydrophobicity than FA (Sutton and Sposito, 2005), whereas Suwannee River NOM contains both hydrophobic and hydrophilic acids as well as other soluble organic solutes. Note that the hydrophobicity of pharmaceuticals follows the order: ibuprofen > ketoprofen > carbamazepine (Table 1). Therefore, it could be reasonable to ascribe the unique impact of Suwannee River HA to ibuprofen adsorption (Table 3) to the match of hydrophobicity between ibuprofen and HA; this match may lead to form a considerable interaction between them. Similarly, the lower hydrophobicity of FA potentially makes it suitable to form an association with both ibuprofen and ketoprofen. Suwannee River NOM, which contains hydrophobic HA and FA, could apparently interact with ibuprofen and ketoprofen, thereby significantly affecting the adsorption of both (Table 3). The table also suggests that Suwannee River HA, Suwannee River FA, and thus hydrophobic HA and FA in Suwannee River NOM may not form considerable associations with carbamazepine, possibly due to the low hydrophobicity of carbamazepine. Hence, the impact of Suwannee River NOM to carbamazepine adsorption can be attributed to the association of the fraction of more hydrophilic acids and/or other soluble organic compounds (proteins, carbohydrates, etc.) with carbamazepine. It is not yet clear why a higher concentration of Suwannee River HA and Suwannee River FA resulted in such an insignificant impact on the adsorption of ibuprofen to silica (Table 3).
The adsorption of carbamazepine, diclofenac, ibuprofen, and ketoprofen on porous silica was investigated in the presence of different ionic strengths, anions, cations, and natural organic matter. Whereas the presence of anions did not cause any effect on the adsorption of pharmaceuticals, increases in ionic strength and the presence of divalent cations, trivalent cations, and NOM did significantly impact the adsorption of pharmaceuticals. The impact was found to be dependent on water characteristics (ionic strength, cations, NOM and their concentration) and the properties of pharmaceuticals (molecular charge, pKa, and hydrophobicity). As a result, this study suggests that the adsorption of pharmaceuticals to silica in an aqueous environment can be strongly influenced by pharmaceutical properties, trivalent cations, as well as the solution pH (Bui and Choi, 2009); moreover, the influence of ionic strength, divalent cations, and NOM could not be neglected. In further studies, the effects of a combination of these factors will be investigated and more pharmaceuticals considered.