Adsorption isotherms of 17β-estradiol on granular activated carbon (GAC)

Adsorption isotherms of 17β-estradiol on granular activated carbon (GAC)

Chemosphere 44 (2001) 1573±1579 Adsorption isotherms of 17b-estradiol on granular activated carbon (GAC) Maria F...

300KB Sizes 1 Downloads 15 Views

Chemosphere 44 (2001) 1573±1579

Adsorption isotherms of 17b-estradiol on granular activated carbon (GAC) Maria Fuerhacker a


, Astrid D urauer b, Alois Jungbauer


Institute for Water Provision, Water Ecology and Waste Management, University of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria b Institute of Applied Microbiology, University of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria Received 28 June 2000; accepted 7 October 2000

Abstract The adsorption characteristics of three types of activated carbon for 17b-estradiol were studied by long term experiments to assess the time which is necessary to reach equilibrium between the solid and the liquid phase. The adsorption kinetics were measured by liquid scintillation counting using radio-labelled 17b-estradiol at various concentrations of 17b-estradiol in the ppt range. 17b-estradiol is quickly adsorbed and conditions close to equilibrium were reached after 50±180 min. The equilibrium concentrations were calculated to be at 49±81% of the initial concentration in the concentration range between 1 and 100 ng/l, with 0.51 ng/l for a 1 ng/l and between 5.9 and 14.6 ng/l for 100 ng/l initial concentration. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Adsorption properties; Granular charcoal; Activated carbon; 17b-estradiol

1. Introduction The ability of chemicals to mimic estrogenic activity has drawn attention to a group of substances called ``environmental estrogens''. These chemicals are found in environmental samples and are a potential threat to human and wildlife, showing endocrine disrupting activity such as increased vitellogenin concentrations in ®sh (Fawell and Wilkinson, 1994; Sumpter and Jobling, 1995). Endocrine-disrupting compounds can alter hormone pathways that regulate reproductive processes. Ecological implications of exposure to endocrine disrupting compounds are known and di€erent impacts could be observed (Colborn and Clement, 1992). Very potent compounds with endocrine disruptive activity are

* Corresponding author. Tel.: +43-1-36006-5821; fax: +43-1368-9949. E-mail address: [email protected] (M. Fuerhacker).

the natural female sex hormones 17b-estradiol, estriol and estrone. The no-observed adverse e€ect level (NOAEL) of 17b-estradiol inducing vitellogenin synthesis in male rainbow trout, a severe malfunction, is for ethinylestradiol and 17b-estradiol in the range of 0.3±10 ng/ L, respectively (Sheahan et al., 1994; Environment Agency, 1996). The daily production of estrogens in a premenopausal woman is in the microgram range. These hormones are metabolised through the entero-hepatic circuit and excreted through urine and faeces in conjugated forms as glucuronides and glucosides exhibiting no, or strongly reduced, endocrine activity. These conjugates are cleaved in sewage treatment plants (K orner et al., 1999; Ternes et al., 1999). To which extent the conjugates are hydrolysed in groundwater or contaminated surface water is not known. The concentrations of endocrine disrupting compounds in drinking water or drinking water sources have not been adequately investigated yet. Nevertheless, there is a potential of contamination with endocrine disrupting compounds in groundwater (Kalbfus, 1998). Rurainsky et al. (1977)

0045-6535/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 5 4 3 - 9


M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579

found 17b-estradiol (up to 0.94 ng/l) and ethinylestradiol (up to 22.5 ng/l) in wells. Should drinking water sources be contaminated with endocrine disruptive compounds, they have to be eciently removed. Activated carbon is commonly used for the removal of micro-pollutants in drinking water treatment (e.g., Zytner, 1992). For design of cleaning processes adsorption isotherms are rough guidelines. They describe the thermodynamics of adsorption (i.e., adsorption equilibria) and are often used to estimate carbon dosages for achieving adsorbate removal. Knowing the parameters of adsorption isotherms it is possible to predict the extent of removal. The equilibrium concentration of the free compound is the ultimate reduction level that can be achieved. For practical considerations a sucient contact time has to be ensured. The e€ective contact time for granular activated carbons (GAC) in water treatment plants is usually in a range from minutes to hours. Therefore it is important to know the time to reach adsorption equilibrium. Insucient contact time may result in non-equilibrium conditions and reduced removal rates. Thus, kinetic limitations may be important for slow adsorbing compounds. The aim of this work was to study the adsorption kinetics of natural estrogen 17b-estradiol with di€erent activated charcoals in a low dose typical for a real contamination. To ®nd the minimum contact time, kinetic experiments were performed and equilibrium concentrations were measured for di€erent initial concentrations of the hormone. Since the environmental relevant concentration range of 17b-estradiol is very low we investigated its GAC adsorption by spiking water samples with 0.5±100 ng/l hormone.

2. Methods and material

Fig. 1. Chemical structure of [2,4,6,7-3 H]-17b-estradiol.

shire, UK) with an activity of 3111.7 GBq/mmol at a concentration of 37 MBq/ml was used (structure see Fig. 1). This concentration corresponds to 3.2 lg/ml 17bestradiol. The scintillation liquid, Optiphase Hi Safe 2, was purchased from Wallace Scintillation products formulated by Fison Chemicals, Loughborough Laes, UK. The radioactivity was determined in a liquid scintillation counter (Wallace 1416, Helsinki, Finland) after appropriate pre-treatment. 2.2. Charcoal properties The GAC in this study were Filtrasorb 200 (F200) (Chemviron Carbon), DonauCarbon GC830 and DonauCarbon GK50 (Donauchemie, Krems, Austria). The F200 and the GC830 are prepared from pit-coal and the GK50 from coconuts shells. All of them are used for drinking water processing. F200 was used for kinetic and equilibrium experiments. Adsorption capacities and equilibrium concentrations were also investigated with other charcoals (Table 1). 2.3. Batch experiments

2.1. Materials The experiments were carried out with radio-labelled 17b-estradiol. [2,4,6,7-3 H]-17b-estradiol (NEN Herfor-

Prior to use, the carbon was washed several times with deionised water, until the ®nes (i.e., particles ¯oating on the surface of the wash water) were removed.

Table 1 Properties of GAC used


Type of GAC

Particle Size 90% (mm)

Weight (kg/m3 )

Hardness (%)

Moisture (%)

Ash (%)

Speci®c surface (m2 /g)

F200 Chemviron carbon Donau carbon GC830 Donau carbon GK50



75 min

2 max




425  10%

90 min

5 max

12 max

1000 min

0.6±2.36 or 2.36±4.75

425  10%

95 min

5 max

3 max

1000 min

This value was not found in the literature.

M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579

The washed carbon was dried in an oven at 110°C for two days and stored in the desiccator until use. A known quantity of GAC was suspended in water and spiked by various concentrations of radio-labelled 17b-estradiol. 2 g of granulated charcoal were used per litre. The radio-labelled stock solution was diluted 640fold. This solution was used for the preparation of the spiked water samples. 50 ml of deionised water samples containing 100 mg charcoal were spiked with standard solution to get di€erent concentrations (0.5, 1, 5, 10, 50 and 100 ng/l) of radio-labelled 17b-estradiol. During the whole experiments the charcoal was kept suspended by a magnetic stirrer (700 rpm) or a horizontal shaker (250 rpm). Aliquots of 200 ll were taken out after di€erent time intervals and the GAC was removed by centrifugation. For scintillation measurements, 3 ml scintillation liquid were pipetted into a standard plastic vial and 100 ll of the supernatant sample were added. Scintillation was counted for 240 s. The scintillation liquid with the activated carbon was used to measure background radiation. For kinetic experiments samples were taken after 0.5, 2, 4, 6, 8, 10, 30 min, 1, 2, 3, 24, 48 h and 6 days. Based on the kinetic experiments 180 min were chosen as an end point of the equilibrium time for the adsorption isotherm. For comparison we used two di€erent GAC: DonauCarbon GC830 and DonauCarbon GK50 in addition to the F200. The isotherms were done with 6 concentrations (0.5, 1, 5, 10, 50 and 100 ng/l) of 17bestradiol. All experiments were made in duplicates, to compare the results. Also, blanks (deionised water without 17b-estradiol) and standards were analysed. For long term behaviour, the two di€erent GAC (DonauCarbon GC830 and DonauCarbon GK50) were incubated for 6 days and analysed. 2.4. Theory For the batch uptake studies the adsorption kinetic of 17b-estradiol was approximated by a kinetic equation describing a reaction to an equilibrium. Assuming a pseudo-®rst order reaction rate for the adsorption process in the low concentration range we can write the reaction equation k



where A is the concentration of free estradiol, A the adsorbed estradiol and k is the net reaction rate constant. The reaction is ®rst order in both directions. The di€erential equation for description of the rate of change is dA ˆ dt


kr †…A

A †;



with k are the forward and reverse reaction constant and A as the concentration of free estradiol in equilibrium. The integral form of Eq. (2) is A ˆ A ‡ …A

A †  e




The time (t50 ) where 50% of the initial material is adsorbed is calculated as t50 ˆ

ln 2 : k


Assuming a Langmuir type adsorption characteristic the amount of bound estradiol A can be described as A ˆ Amax

kA A ; 1 ‡ kA A


with Amax the maximal estradiol concentration bound to charcoal and kA the equilibrium binding constant. The underlying hypothesis of this model is an interaction where the binding sites for 17b-estradiol do not interfere with each other during the association and dissociation phases. When co-operative interaction takes place, the relationship between concentration bound to the surface and the equilibrium concentration of free solute can be described in the simplest fashion by a Langmuir± Freundlich adsorption isotherm. n

A ˆ Amax

kA A : 1 ‡ kA An


The data obtained from the batch uptake experiments were approximated by Eq. (3). The curve ®tting program Table Curveâ 2D from SPSS (Erkrath, Germany) was used for this purpose.

3. Results and discussion In future a possible problem of drinking water supply might be contaminated water sources. Contamination by steroid hormones is very likely. We wanted to examine the capability of activated carbon to remove steroid from water. For that purpose tritium labelled 17b-estradiol, a prominent representative of steroid hormones was dissolved in water and incubated with activated carbon. Samples were drawn in increasing time intervals. The selected concentration range for adsorption experiments was between 0.5 and 100 ng/l, to simulate a realistic range of contaminated ground water or surface water as source for drinking water. This is a very low level and the driving force for reaching the equilibrium between water and charcoal is very weak. To ®nd out the minimum contact time for the experiments in a ®rst step GAC suspension in water was incubated with radio-labelled 17b-estradiol for


M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579

9600 min. to measure long term kinetic reactions. At the beginning of the adsorption phase samples were drawn more frequently than in the late phase since from preliminary studies a fast initial adsorption was expected. These experiments are shown in Fig. 3 and the fast initial kinetic was con®rmed. The results of the long term kinetic experiments (Fig. 3) showed that 17b-estradiol is quickly adsorbed and a free hormone concentration close to the equilibrium concentration was reached within 50 to 180 min. Approximation with Eq. (3), assuming a pseudo-®rst order reaction rate con®rmed the observation. This adsorption periods were used for further experiments. In addition kinetic experiments at concentration ranges of 0.5±100 ng/l studying charcoal F200 are performed (Fig. 2). The

results could be approximated with Eq. (3). Goodness of ®t dated indicated that the assumed pseudo-®rst order kinetic is correct and a higher order reaction rate does not occur. The estimated parameters are shown in Table 2. Single solute adsorption equilibria can be predicted by approximating the data with an appropriate model describing the liquid/solid phase equilibrium. Two most commonly applied isotherms for approximating the adsorption equilibria of small molecules on charcoal are the Langmuir and the Freundlich isotherms. These equations give good estimations in the concentration close to saturation, but are not appropriate for estimations at very low concentrations. As our experiments should provide information for the low concentration

Fig. 2. Kinetic test results for adsorption of 17b-estradiol for di€erent concentrations.

M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579


The predicted equilibrium concentrations, t50 , when 50% of the 17b-estradiol is adsorbed and the adsorption capacities and rates are listed in Table 3. 4. Comparison between di€erent charcoals In Fig. 3 and Tables 4 and 5 experimental and predicted data for GC830 and GK50 are given in comparison to the activated carbon F200. The batch uptake of 17b-estradiol was approximated by Eq. (3) and again a pseudo-®rst-order adsorption kinetic was observed. The speci®c adsorption properties suggest a signi®cant di€erence in terms of equilibrium concentration between the F200 in the long term experiment and the short term experiments of 5.9 ng/l and 27.2 ng/l (Tables 3 and 4). A lower activity at the beginning in the short term experiment was measured by LSC compared to the long term experiments. As we are working in the pptconcentration-range a lower activity of the stock standard solution or a dilution error of the stock may count for this di€erence. While comparing the graphs and the kinetic results, it is important to note that in all cases the same kinetics were observed. It can be concluded, that non of the charcoal will be able to remove the 17b-estradiol by batch adsorption in a way to reach a level which could be considered appropriate for drinking water purposes, assuming the same toxicological criteria as applied for wildlife ®sh (10 ng/l 17b-estradiol) with a safety margin of at least one magnitude. Due to the afore-mentioned problem occuring in the sub-ppb range we do not consider that the three di€erent charcoals do not exhibit di€erent adsorption properties. The adsorption capacities at the investigated level of 100 ng/l 17bestradiol was found to be less than 50 ng/g activated charcoal. Fig. 3. Kinetic results of the long term tests for three di€erent charcoals.

5. Conclusions

range, we did not try to reach saturation of the charcoal at all. For that reason we measured kinetic behaviour at di€erent concentrations and selected a simple equation based on physical-chemical reaction adsorption kinetic to ®t the experimental data. The parameters for the equations describing our experimental data are given in Table 2. The equilibrium concentration and the 50% adsorption time (t50 ) is calculated using Eqs. (3) and (4). The predicted adsorption rates are in the range between 49% and 81% in the respective range 1±100 ng/l. For the 0.5 ng/l concentration the variability of the data is too high to ®t the data with an equation. The range below 0.5 ng/l is close to the detection limit.

Our data provide a basic information about the adsorption of 17b-estradiol in deionized water measured as single component. The design of a large scale adsorption column requires such information. In addition to adsorption isotherms, information on other factors a€ecting adsorption are necessary. These include the physical characteristics of the natural mixture, the properties of the adsorbates and the background matrix. For an adsorbent such as GAC, surface area, pore and particle size distribution and surface chemistry are major factors a€ecting adsorption. Nevertheless, our results show, that it will be hardly possible to reduce the concentration of 17b-estradiol below the NOAEL (no-observed adverse e€ect level) for aquatic wildlife of 10 ng/l of 17b-estradiol; though the properties of the charcoals were slightly di€erent.


M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579

Table 2 Parameters for the selected equation on F200 at 20°C Parameters


Std error

t Value

95% Conf. Lim.

95% Conf. Lim.


1 ng/l A A k

1.01 0.510 0.059

0.0585 0.0442 0.0258

17.27 11.54 2.30

0.889 0.419 0.0059

1.13 0.60 0.110


5 ng/l A A k

4.71 1.19 0.067

0.118 0.0853 0.0079

39.94 13.94 8.43

4.47 1.013 0.0500

4.95 1.36 0.082


10 ng/l A A k

9.35 2.75 0.074

0.357 0.248 0.0135

26.21 11.05 5.48

8.61 2.23 0.046

10.08 3.25 0.102


50 ng/l A A k

46.13 9.43 0.058

0.837 0.638 0.0050

55.14 14.80 11.71

44.4 8.11 0.048

47.86 10.75 0.068


100 ng/l A A k

96.34 27.22 0.074

1.74 1.21 0.0063

55.27 22.44 11.75

92.74 24.71 0.061

99.94 29.72 0.087


Table 3 Predicted equilibrium concentrations, adsorption capacities, adsorption rates and t50 for 17b-estradiol at di€erent initial concentrations Parameters (ng/l)

Equilibrium concentration (Ng/l)

Adsorption capacity (ng/g)

Adsorption rate (%)

t50 (min)

1 5 10 50 100

0.51 1.19 2.75 9.43 27.2

0.25 1.91 3.63 20.3 36.4

49 76 73 81 73

11.7 10.3 9.4 12.0 9.4

Table 4 Parameters for the selected equation for GC830, GK50 and F200 for long term experiments at a concentration of 100 ng/l Parameters


Std Error

t Value

95% Conf. Lim.

95% Conf. Lim.


GC830 A A k

98.0 11.3 0.034

1.12 0.971 0.0019

87.6 11.6 17.6

95.7 9.30 0.030

100 13.3 0.038


F200 A A k

96 5.9 0.355

2.53 1.103 0.0264

38.1 5.3 13.5

90.7 3.47 0.298

101.9 8.3 0.414


GK50 A A k

99.7 14.6 0.048

1.69 1.27 0.0038

58.9 11.59 12.7

96.2 12.0 0.040

103 17.2 0.056


M. Fuerhacker et al. / Chemosphere 44 (2001) 1573±1579


Table 5 Equilibrium concentrations, adsorption capacity and t50 for three di€erent charcoals at a concentration of 100 ng/l Parameters

Equilibrium concentration (ng/l)

Adsorption capacity (ng/g)

t50 min

GC830 F200 GK50

11.3 5.9 14.6

44.4 47.1 42.7

20.4 2.0 14.4

Acknowledgements The authors gratefully acknowledge the work of Susana da Assuncao Bessa Alves who carried out the sample preparation. Her work is highly appreciated. References Colborn, T., Clement, C., 1992. Chemically induced alterations in sexual and functional development. In: The Wildlife/ Human Connection. Princeton Scienti®c Publishing, Princeton, NJ. Environment Agency, 1996. The identi®cation and assessment of estroenic substances in sewage treatment works e‚uents, Rep. No. P 38. Fawell, J.K., Wilkinson, M.J., 1994. Oestrogenic substances in water: A review. AQUA 43, 219±221. Kalbfus, W., 1998. Exposition und Wirkung endokriner Substanzen im aquatischen System. Wiener Mitteilungen 153, 33±44. K orner, W., Hanf, V., Schuller, W., Kempter, C., Metzger, J., Haganmaier, H., 1999. Development of a sensitive E-screen

assay for quantitative analysis of estrogenic activity in municipal sewage plant e‚uents. Sci. Total Envorin. 225, 33±48.  Rurainsky, R.D., Theiss, H.J., Zimmermann, W., 1977. Uber das Vorkommen von nat urlichen und synthetischen  Ostrogenen im Trinkwasser. GWF-Wasser/Abwasser 118, 288±291. Sheahan, S.A., Bucke, D., Matthiessen, P., Sumpter, J.P., Kirby, M.F., Neall, M., Waldock, M., 1994. The e€ects of low level 17-a ethynylestradiol upon plasma vitellogenin levels in male and female rainbow trout. In: M. R. L. R. (Ed.). Sublethal and Chronic E€ects of Pollutants on Freshwater Fish. FAD, Fishing News Books, Oxford. Sumpter, J.P., Jobling, S., 1995. Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ Health Perspect 103, 173±178. Ternes, T.A., Kreckel, P., Mueller, J., 1999. Behaviour and occurence of estrogens in municipal sewage treatment plants ± II. Aerobic batch experiments with activated sludge. Sci. Total Environ. 225, 91±99. Zytner, R.G., 1992. Adsorption±desorption of trichloroethylene in granular media. Water Air and Soil Pollution WAPLAC 65, 245±255.