Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
a r t i c l e i n f o a b s t r a c t
Article history: Adsorption of levofloxacin (LEV), one of the extensively used antibiotics, onto goethite was investi-
Received 5 April 2013 gated using batch experiments. The adsorption of LEV on goethite was pH-dependent. A maximum
Received in revised form adsorption was reached at pH 6. Above or below pH 6, the adsorption decreased. In the presence of
18 September 2013
calcium (Ca2+ ), a decrease in adsorption was observed, due to probably formation of Ca2+ –LEV com-
Accepted 28 September 2013
plexes in solutions. Phosphate also showed a significant inhibition on LEV adsorption over a pH range
Available online 12 October 2013
of 3–10. Phosphate competed with LEV for binding sites on the surface of goethite, and the electro-
static competition between LEV and phosphate on goethite surface might be another reason for the
Keywords:
Adsorption
decrease in adsorption. These results indicated that Ca2+ and phosphate have a great influence on the
Goethite distribution of LEV in soils and waters, which will eventually affect its antibacterial activity in the
Levofloxacin environment.
Fluoroquinolone © 2013 Elsevier B.V. All rights reserved.
Phosphate
Calcium
0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.colsurfb.2013.09.056
592 X. Qin et al. / Colloids and Surfaces B: Biointerfaces 116 (2014) 591–596
2. Materials and methods (LC-20AT, Shimadzu) system with an Inertsil® ODS-SP column
(4.6 mm × 250 mm, 5 m particle size) coupled to a UV detector.
2.1. Materials LEV was detected at the peak wavelength of 293 nm, which was
determined beforehand offline by measuring light absorbance
Goethite was prepared according to the method of Hiemstra of LEV solution in the range of wavelength of 200–700 nm using
et al. [18] and Filius [19]. The material prepared was confirmed a UV-Vis spectrophotometer (HP 8453, USA). The mobile phase
to be pure ␣-FeOOH through X-ray diffraction (XRD) measure- contained a mixture of 0.025 M phosphate buffer at pH 3.0 (77%)
ment (Rigaku, D/MAX 2500). The specific surface area of goethite and acetonitrile (23%). The flow rate is 1 mL/min. These conditions
is 83.5 m2 /g determined by N2 gas adsorption method. The pristine were the same as those used for antibiotics measurements of
point of zero charge (PZC) of goethite prepared following the same MacKay and Seremet [8]. The concentration of adsorbed LEV on
recipe is around 9.2 as reported by Filius [19] and Weng et al. [20], goethite was calculated from the difference in the concentrations
etc. of LEV in solutions before and after adsorption, according to
Levofloxacin (>98.0%) was purchased from Tokyo Chemical
C0 − C
Industry Co. (Japan), and no further purification was carried out. q= (1)
m·S
One LEV molecule contains one carboxyl group and one piperazinyl
group. Their protonation constants (pKa1 and pKa2 ) are 6.05 (or where q (mol/m2 ) is the concentration of adsorbed LEV, C0 and
6.02) and 8.22 (or 8.15), respectively [21,22]. Acetonitrile (HPLC C (M) are the initial and equilibrium concentrations of LEV in
grade) was obtained from Honeywell Burdick & Jackson (USA). solutions, m (g/L) and S (m2 /g) are the concentration and specific
Other chemicals were analytical reagent grade or higher. MilliQ surface area of goethite, respectively. The initial concentration (C0 )
water was used throughout the experiments. of LEV equals to the LEV concentration measured in the control (no
goethite). No significant losses are observed in control samples.
2.2. Adsorption experiments Adsorption of LEV on goethite in the presence of Ca2+ was fitted
using the linearized expression of Freundlich model, which can be
Adsorption envelopes of LEV were obtained from results of batch expressed as
experiments. In the batch experiments, 50 mL polyethylene cen- 1
trifuge tubes were used, and they were flushed with ultrapure N2 ln q = ln kF + ln C (2)
N
during the experiments to minimize dissolution of CO2 , to avoid the
influence of (bi)carbonate in solutions on the adsorption of LEV to where kF ((mol/m2 )(M)−1/N ) and 1/N are Freundlich constants,
goethite. 10 mL LEV stock solution of 20 M and 10 mL goethite sus- which are related to the adsorption capacity and surface site het-
pension of 1.6 g/L (or MilliQ water in control samples) were added erogeneity of the adsorbent [23].
into the tubes. Both the LEV and goethite stock solutions were pre- The speciation of LEV in solutions under different pH and
pared in a 0.01 M NaCl background. The final total concentration of Ca2+ concentrations was calculated with the speciation software
LEV was 10 M, and the final goethite concentration was 0.8 g/L. ECOSAT [24]. The constants for Ca2+ –LEV complexes can be found
Acid or base (0.01 M HCl and NaOH) was added to adjust pH to a in Table 1.
chosen value in the range of 3–10. The concentration of goethite
in the stock solution was determined by weighing subsamples in 3. Results and discussion
triplicates after being dried at 105 ◦ C for 4 h.
Effects of phosphate on LEV adsorption to goethite were 3.1. Effect of pH
investigated using adsorption envelopes obtained from batch
experiments. The suspensions were prepared in a similar way as The structure and speciation of LEV as a function of pH are shown
above. The final total concentration of LEV was 10 M, and the in Fig. 1. LEV molecules are primarily cationic (protonated piper-
final goethite concentration was 0.8 g/L. Phosphate (NaH2 PO4 ) was azinyl group, H2 L+ ) below pH 5, anionic (deprotonated carboxyl
added to a final concentration of 0.1 and 0.2 mM. Acid or base was group, L− ) above pH 8.5, and zwitterionic (or neutral, HL0 ) at pH
also used to adjust pH to a chosen value in the range of 3–10. between its pKa1 and pKa2 values. The speciation of LEV molecule
Effects of Ca2+ on LEV adsorption was studied in three batch under different pH influences its sorption on minerals greatly.
experiments with, respectively, no Ca2+ , 0.005 M Ca2+ and 0.01 M Adsorption of LEV onto goethite at different pH is shown in
Ca2+ . Each batch experiment contains 8 suspensions with an initial Fig. 2. The adsorbed concentration of LEV reaches its maximum
concentration of LEV ranging from 1 to 10 M. The final goethite (0.034 mol/m2 , about 25% of LEV added was adsorbed) at a pH
concentration is the same as above (0.8 g/L) and the background value around 6, and then decreases when pH is lower or higher.
electrolyte is 0.01 M NaCl as well. Acid and base were added to the This pH dependency is consistent with previous reports on the
suspension to obtain a nominal pH of 5.0. The amount of acid or adsorption of other antibiotics with a similar structure [13,14,26].
base added was based on experience when no Ca2+ was present. The maximum concentration of LEV adsorbed is close to the value
The presence of Ca2+ leads to a higher pH than aimed at. The end reported for ciprofloxacin adsorbed on goethite [8].
pH is between 5 and 6. The pH difference between samples at the At pH < 5, both the goethite surface and LEV in solution are posi-
same Ca2+ level is very small. Adsorption isotherms were conducted tively charged. The adsorption will then be due to specific chemical
from these data. adsorption. The positive charge on both LEV and goethite increases
All suspensions were shaken for 24 h in the dark. Then the with the decrease in pH values. Less LEV was adsorbed with a
end pH of each sample was measured. The suspensions were cen- decreasing pH due to increased electrostatic repulsion between the
trifuged at 12 000 rpm for 20 min, and the supernatant was stored protonated amine groups on LEV molecules (adsorbed or free ones)
at 4 ◦ C in the dark for later analyses. All experiments were carried and the positive charged goethite surface (pH < PZC). The presence
out in triplicates. of piperazinyl group in antibiotics inhibited their adsorption onto
goethite, especially at low pH [8,27]. The maximum adsorption of
2.3. Measurements and data analyses LEV appeared at pH 6 (Fig. 2), and this pH value is close to the
pKa1 of LEV, which means that a significant fraction (about 50%) of
The concentration of LEV in solutions was determined the carboxylic group on LEV is deprotonated under this pH. Such
with a high performance liquid chromatography (HPLC) adsorption behavior that a maximum adsorption around a pH close
X. Qin et al. / Colloids and Surfaces B: Biointerfaces 116 (2014) 591–596 593
Table 1
The calculated stability constants (log K) of Ca2+ –LEV complexes.a
to the pKa value of protonation in the functional groups has also the positive charge on goethite surface also decreased, therefore
been observed for the sorption of some organic acids to goethite the concentration of LEV absorbed onto goethite decreased with
[19]. increasing pH due to decreased electrostatic attraction compared
Fourier transform infrared (FTIR) measurements were usually to pH around 6. At pH > 9, LEV molecules and goethite surface were
performed in order to investigate the mechanism of antibiotics both negatively charged. Therefore less LEV (<5%) was adsorbed on
adsorption on minerals. Trivedi and Vasudevan [28] demonstrated goethite than at lower pH because of electrostatic repulsion.
that ciprofloxacin adsorption on goethite was through bidentate The LEV adsorption envelope is asymmetric on the two sides
chelates involving both oxygen atoms in the carboxylate group. The of the peak around pH 6. The decrease of adsorption is stronger at
complexation with both carboxylate and ketone groups were also the high pH side compared to that at the low pH side (Fig. 2). This
reported when studying the sorption of antibiotics on hydrous fer- indicates that LEV adsorption is not only dependent on solution
ric oxides (HFO) [7], nonporous Al2 O3 [17], nanophase TiO2 [26], speciation of LEV, which shows a symmetric pattern (Fig. 1b). The
and nano-sized Fe3 O4 [23]. The interactions between antibiotics change of charge on goethite surface with pH is asymmetric, which
and goethite were attributed to the complexation of their carboxyl will contribute partly to the asymmetric adsorption envelope of
or/and ketone groups with the hydroxyl groups on goethite surface, LEV. More importantly, chemical complexation between the func-
so their adsorption was greatly influenced by the density and spe- tional groups on antibiotics and minerals consumes protons [26],
ciation of the functional groups on both antibiotics and goethite. At which favors adsorption at low pH.
pH above 7, most of carboxyl groups of LEV were deprotonated, and
3.2. Effect of Ca2+
Fig. 1. Molecular structure (a) and pH-dependent speciation (b) of LEV. The pKa Fig. 2. Adsorption of LEV onto goethite as a function of pH. Error bars (±1 standard
values of LEV were from Sousa et al. [22]. deviation, n = 3) are shown in the figures.
594 X. Qin et al. / Colloids and Surfaces B: Biointerfaces 116 (2014) 591–596
Fig. 3. Effect of Ca2+ on LEV adsorption onto goethite. The equilibrium pH values at
three Ca2+ levels are in the ranges of 5.10–5.20, 4.97–5.04, and 5.76–5.82, respec-
tively. The lines are the Freundlich isotherm with the fitted parameters listed in
Table 2. Error bars (±1 standard deviation, n = 3) are shown in the figure.
Table 2
Freundlich isotherm parameters for LEV adsorption on goethite.a
4. Conclusions
3.3. Effect of phosphate
In the natural environment, adsorption of antibiotics to mineral
The shape of the adsorption envelopes for LEV on goethite in
surfaces and natural organic matter influences their distribution,
the presence of phosphate is similar to that without phosphate
mobility, degradability and their effects on microorganisms. Metal
(Fig. 2). The maximum adsorption is observed around pH 6 as well
(hydr)oxides, including mainly iron, aluminum and manganese
(Fig. 5). Comparing with that without additions of phosphate, less
(hydr)oxides, are an important group of minerals in soil, water and
LEV was adsorbed in the presence of 0.1 mM and 0.2 mM phos-
sediments, which are rather reactive in terms of interactions with
phate, and the maximum amounts of LEV adsorbed decreased by
a wide range of anions, cations and organic molecules. This work
33% and 67%, respectively. In the presence of 0.2 mM phosphate,
studied LEV adsorption on goethite and the concentration and pH
concentrations of adsorbed LEV are less than 0.005 mol/m2 over
dependency of the adsorption, and effects of the presence of Ca2+
the whole pH range except for that around pH 6. As also shown in
and phosphate on the adsorption. As expected, the LEV adsorp-
Fig. S2, LEV adsorption to goethite decreases with the increasing
tion increases with the increase of concentration. The results of
concentrations of phosphate. The strong inhibition is observed at
this study showed that adsorption of LEV on goethite was greatly
acidic conditions (pH < 7).
influenced by pH in solutions, and the maximum adsorption was
Phosphate had a very high affinity for the surface of iron
observed around pH 6. The LEV adsorption envelope is asymmet-
(hydr)oxides, and decreased the adsorption of small organic acids
ric on the two sides of the peak, showing a steeper decrease on the
and natural organic matter [33–35]. It has been found that at the
high pH side compared to the low pH side. Both Ca2+ and phosphate
very low loading of phosphate (<0.25 mol/m2 ), phosphate had
decreased LEV adsorption on goethite at different concentrations
a negligible effect on ciprofloxacin sorption onto nano-size mag-
of LEV and pH values, and different mechanisms were proposed.
netite [23]. However, the loading in this study is much higher
Adsorption isotherms at different levels of Ca2+ were well fitted to
(>0.93 mol/m2 ). As a result, the inhibition of LEV adsorption on
the Freundlich model. Less LEV was adsorbed in the presence of
goethite by phosphate was very significant. This observed decrease
Ca2+ , due to probably formation of Ca2+ –LEV complexes in solu-
of LEV adsorption in the presence of phosphate can probably be
tions, which consequently decreased the free concentration of LEV
attributed to competition between LEV and phosphate during their
in solution, and affected the solid–liquid equilibrium of LEV. The
adsorption on goethite surface.
presence of phosphate inhibited the adsorption of LEV in the range
Site competition can be one of the mechanisms. There were 6.15
of pH 3–9. A stronger inhibition was observed at acidic conditions
sites/nm2 singly ( FeOH0.5− ) and triply ( Fe3 O0.5− ) coordinated
compared to that at higher pH. Competition for binding sites and
surface groups on the surface of goethite [19], which were equal
electrostatic repulsion between LEV and phosphate for adsorption
to 10 mol/m2 . It was suggested that depending on the solution
to goethite were proposed as the major mechanisms explaining
pH and phosphate loading, phosphate mainly forms unprotonated
the competition effects. pH is a key solution chemistry parameter,
bidentate ( Fe2 O2 PO2 ) and singly protonated monodentate com-
whereas Ca2+ and phosphate are, respectively, a major cation and
plexes ( FeOPO2 OH) on the surface of goethite [36]. In our study,
anion in most environmental systems. The results obtained in this
the calculated concentration of adsorbed phosphate on goethite
study help us to better understand the behavior of antibiotics in the
was relatively low (<1.5 mol/m2 ), therefore, the unprotonated
environment.
bidentate phosphate–goethite complex ( Fe2 O2 PO2 ) was domi-
nant over the whole pH range. One phosphate ion was complexed
with two hydroxyl groups on goethite surface, and one antibiotic Acknowledgments
molecule formed complexes with one hydroxyl group [7,28]. In
the absence of phosphate, there were enough binding sites for This work was supported by the Fundamental Research Funds
the adsorption of LEV molecules on goethite surface. The maxi- for the National Natural Science Foundation of China (41172226),
mum concentration of LEV adsorbed observed in this study is only the Fundamental Research Funds for the Central Universities
0.034 mol/m2 , which is less than 1% of the total sites. It indicates (2652013024 and 2652013089), and the project (1212011121171)
that LEV adsorption to goethite is very weak. In the presence of from the China Geological Survey. We appreciate BeiShiDe Instru-
0.1 mM and 0.2 mM phosphate, respectively 18–24% and 21–29% ment S&T (Beijing) Co. Ltd. for determining surface area of goethite.
596 X. Qin et al. / Colloids and Surfaces B: Biointerfaces 116 (2014) 591–596
We thank Yang Wang and Juan Xiong for their help in some exper- [16] W. Yang, Y. Lu, F. Zheng, X. Xue, N. Li, D. Liu, Chem. Eng. J. 179 (2012) 112–118.
iments and advice on ECOSAT. [17] K.W. Goyne, J. Chorover, J.D. Kubicki, A.R. Zimmerman, S.L. Brantley, J. Colloid
Interface Sci. 283 (2005) 160–170.
[18] T. Hiemstra, J.C.M. De Wit, W.H. Van Riemsdijk, J. Colloid Interface Sci. 133
Appendix A. Supplementary data (1989) 105–117.
[19] J. Filius, Modeling the adsorption of weak organic acids on goethite. The ligand
and charge distribution model, Ph.D. Thesis, Wageningen University, NL, 2001.
Supplementary data associated with this article can be [20] L.P. Weng, W.H. Van Riemsdijk, L.K. Koopal, T. Hiemstra, Environ. Sci. Technol.
found, in the online version, at http://dx.doi.org/10.1016/j.colsurfb. 40 (2006) 7494–7500.
2013.09.056. [21] J.A.O. González, M.C. Mochón, F.J.B. de la Rosa, Talanta 52 (2000)
1149–1156.
[22] I. Sousa, V. Claro, J.L. Pereira, A.L. Amaral, L. Cunha-Silva, B. de Castro, M.J. Feio,
References E. Pereira, P. Gameiro, J. Inorg. Biochem. 110 (2012) 64–71.
[23] S. Rakshit, D. Sarkar, E.J. Elzinga, P. Punamiya, R. Datta, J. Hazard. Mater. 246–247
(2013) 221–226.
[1] A.K. Sarmah, M.T. Meyer, A.B.A. Boxall, Chemosphere 65 (2006) 725–759. [24] M.G. Keizer, W.H. Van Riemsdijk, ECOSAT: Equilibrium Calculation of Spe-
[2] E.M. Golet, I. Xifra, H. Siegrist, A.C. Alder, W. Giger, Environ. Sci. Technol. 37 ciation and Transport. Technical Report of the Soil Quality Department,
(2003) 3243–3249. Wageningen University, Wageningen, 1998.
[3] D. Fatta-Kassinos, S. Meric, A. Nikolaou, Anal. Bioanal. Chem. 399 (2011) [25] B. Urbaniak, Z.J. Kokot, Anal. Chim. Acta 647 (2009) 54–59.
251–275. [26] T. Paul, M.L. Machesky, T.J. Strathmann, Environ. Sci. Technol. 46 (2012)
[4] C. Tong, X. Zhuo, Y. Guo, J. Agric. Food Chem. 59 (2011) 7303–7309. 11896–11904.
[5] A.L. Córdova–Kreylos, K.M. Scow, ISME J. 1 (2007) 585–595. [27] H. Zhang, C. Huang, Chemosphere 66 (2007) 1502–1512.
[6] D. Nasuhoglu, A. Rodayan, D. Berk, V. Yargeau, Chem. Eng. J. 189–190 (2012) [28] P. Trivedi, D. Vasudevan, Environ. Sci. Technol. 41 (2007) 3153–3158.
41–48. [29] R.P.J.J. Rietra, T. Hiemstra, W.H. Van Riemsdijk, Environ. Sci. Technol. 35 (2001)
[7] C. Gu, K.G. Karthikeyan, Environ. Sci. Technol. 39 (2005) 9166–9173. 3369–3374.
[8] A.A. MacKay, D.E. Seremet, Environ. Sci. Technol. 42 (2008) 8270–8276. [30] L.P. Weng, L.K. Koopal, T. Hiemstra, J.C.L. Meeussen, W.H. Van Riemsdijk,
[9] Y. Zhao, J. Geng, X. Wang, X. Gu, S. Gao, J. Colloid Interface Sci. 361 (2011) Geochim. Cosmochim. Acta 69 (2005) 325–339.
247–251. [31] D.L. Ross, C.M. Roley, Int. J. Pharm. 87 (1992) 203–213.
[10] R.A. Figueroa, A. Leonard, A.A. MacKay, Environ. Sci. Technol. 38 (2004) [32] J. Al-Mustafa, Acta Chim. Slov. 49 (2002) 457–466.
476–483. [33] J.S. Geelhoed, T. Hiemstra, W.H. Van Riemsdijk, Environ. Sci. Technol. 32 (1998)
[11] C.J. Wang, Z. Li, W.T. Jiang, Appl. Clay Sci. 53 (2011) 723–728. 2119–2123.
[12] Z. Li, H. Hong, L. Liao, C.J. Ackley, L.A. Schulz, R.A. MacDonald, A.L. Mihelich, S.M. [34] X.P. Qin, F. Liu, G.C. Wang, Chem. Eng. J. 209 (2012) 458–463.
Emard, Colloids Surf. B 88 (2011) 339–344. [35] L.P. Weng, W.H. Van Riemsdijk, T. Hiemstra, Environ. Sci. Technol. 42 (2008)
[13] R.A. Figueroa, A.A. Mackay, Environ. Sci. Technol. 39 (2005) 6664–6671. 8747–8752.
[14] E. Tanis, K. Hanna, E. Emmanuel, Colloids Surf. A 327 (2008) 57–63. [36] R. Rahnemaie, T. Hiemstra, W.H. Van Riemsdijk, Langmuir 23 (2007)
[15] L. Xu, J. Dai, J. Pan, X. Li, P. Huo, Y. Yan, Chem. Eng. J. 174 (2011) 221–230. 3680–3689.