One-step extraction and concentration of estrogens for an adequate monitoring of wastewater using ionic-liquid-based aqueous biphasic systems

Abstract

Ethinylestradiol (EE2) is a synthetic hormone that has been recognized as one of the most prominent endocrine disruptors found in the aqueous environment. Nevertheless, the low content of EE2 in wastewater makes its identification/quantification unfeasible – a major drawback for the evaluation of its persistence and environmental impact. In this context, a novel extraction/concentration method for EE2 from wastewater is proposed here based on aqueous biphasic systems composed of ionic liquids (ILs). Aqueous biphasic systems formed by several hydrophilic ILs and KNaC₄H₄O₆ were initially screened and optimized, with extraction efficiencies of EE2 for the IL-rich phase ranging between 92 and 100%. Remarkable results were obtained with systems that allow the complete extraction of EE2 in a single-step, and without loss of EE2 or the saturation of the extractive phase. Further, the concentration factors of EE2 attainable with these systems were investigated by a suitable manipulation of the composition of the phase-forming components and the corresponding volumes of the coexisting phases. An outstanding concentration of EE2 up to 1000-fold (from ng L⁻¹ to μg L⁻¹) in a single extraction and concentration step was achieved for the first time with IL-based aqueous biphasic systems. These systems are straightforwardly envisaged for the monitoring of wastewater as one-step extraction and concentration routes for a wide array of endocrine disrupting chemicals while allowing an adequate evaluation of their environmental impact.

Introduction

In the past few years, endocrine disrupting compounds (EDCs) have gained significant relevance due to their association with adverse human health effects and environmental concerns. The United States Environmental Protection Agency (USEPA), defines EDCs as “exogenous agents that interfere with the production, release, transportation, binding, action, or elimination of the natural hormones in the body, responsible for the maintenance of homeostasis and the regulation of the development process”.¹

17α-Ethinylestradiol, EE2 (Fig. 1), is a synthetic steroid hormone classified as an endocrine disruptor. This compound derives from 17β-estradiol (E2) and displays the most potent estrogenic activity amongst the estrogens found in sewage effluents.² EE2 is widely used for medical purposes, for instance, in hormone replacement therapy,³ in the treatment of prostate and breast cancers and in oral contraception,⁴ since it mimics the natural estrogens produced by humans (causing endocrine disruption) and is rapidly absorbed by the organism. The EE2 extensive consumption by humans and further excretion are responsible for its actual presence and persistence in effluents of different sewage treatment plants (up to 64 ng L⁻¹)^5,6 as well as in surface waters (up to 27 ng L⁻¹).^5,7

Fig. 1.

Chemical structure of EE2.

In 2012, the Legislative Commission of Water Framework Directive⁸ classified EE2 as a priority substance with a significant risk to or via the aquatic environment. Its high stability, low volatility and high octanol-water partition coefficient (K_OW) are responsible for its high resistance to degradation and broad bioaccumulation in aquatic organisms.⁹ Moreover, the concentration of EE2 has been increasing in sewage effluents, all around the world, due to its widespread use and resistance to degradation.^9–11 Risk assessment bioassays showed the high toxicity of EE2, even at a ng per L level, on a wide number of aquatic species.^12,13 All these claims pointed out the need for finding an effective treatment process for EE2. For instance, physical (sorption, membrane filtration, etc.), biological (activated sludge, etc.) and advanced oxidation (photolysis, strong oxidizers, etc.) processes have been investigated for such a purpose.^9,14 However, an accurate monitoring of the EE2 content in aqueous samples is crucial to evaluate the efficiency of these processes. Furthermore, a complemented identification and risk assessment of EE2 in the aquatic environment is of vital importance.^12,15 Several techniques to identify and quantify EE2, such as high performance liquid chromatography (HPLC),^4,16–18 liquid chromatography (LC)^19–21 and gas chromatography (GC),^22–24 combined with mass spectrometry (MS), have been used. Nevertheless, the presence of unknown EDCs in complex matrices of wastewater and the high detection limits of the equipment traditionally used are the major shortcomings in the EE2 identification and quantification.^20,23 Additional and complex pre-treatment processes, using volatile and hazardous organic solvents, are usually employed to purify and to concentrate EE2 from real samples. This pre-treatment stage is also time-consuming and expensive.²⁵ In this context, the development of alternative methodologies to concentrate EE2 from aqueous media while allowing their appropriate quantification is a challenging task.

Aqueous biphasic systems fall within the liquid–liquid extraction (LLE) techniques and involve the partitioning of molecules from one aqueous phase to another. Typically, these systems are formed by different pairs of solutes (polymer–polymer, polymer–salt or salt–salt) dissolved in water, and above specific concentrations the system undergoes a two-phase separation.²⁶ In addition to their large water content, the non-volatile nature of polymers and salts allows the phase-forming components to be recovered and recycled, and hence, aqueous biphasic systems are a more benign alternative to traditional liquid–liquid extraction routes which use volatile and hazardous organic compounds. Classical aqueous biphasic systems have already been investigated as extraction/concentration techniques.^27,28

In addition to the more conventional polymer-based systems, in 2003, Gutowski et al.²⁹ demonstrated that hydrophilic ionic liquids (ILs) can also form aqueous biphasic systems by the addition of inorganic salts. After this proof of principle, in the following years it was shown that IL-based aqueous biphasic systems can be formed with organic salts, amino acids, carbohydrates or polymers.³⁰ ILs belong to the molten salts category, and due to the large differences in size and shape of the constituting ions they cannot easily form an ordered crystalline structure, and thus, present melting points below a general temperature of 100 °C. Due to their ionic nature, most ILs present unique characteristics, such as a negligible vapour pressure, non-flammability, high thermal and chemical stabilities and a high solvation capacity.^30–33 Still, one of the most important features that has attracted both academia and industry is their aptitude as “designer solvents”, e.g. the capacity to be synthesized for a given task as a result of their plentiful cation and anion combinations.^34–39

IL-based aqueous biphasic systems were extensively explored for the extraction and purification of a wide variety of biomolecules.^30,40–42 The main advantage of using ILs in the formation of aqueous biphasic systems rests on the possibility of tailoring their phases’ polarities and affinities by an appropriate combination of their ions⁴³ and, therefore, exceptional results were already accomplished.³⁰ It was recently reported that they can be used for the extraction and concentration of alkaloids and bisphenol A from human fluids.^44,45 Yet, only concentration factors up to 100-fold were reported.⁴⁴ Furthermore, most of the previously reported systems are composed of ILs and high-charge density inorganic salts³⁰ which led to some environmental concerns. Taking this into consideration, and although no enrichment factors were investigated, in recent studies, biodegradable and more biocompatible organic salts were introduced in the composition of IL-based aqueous biphasic systems.^46,47

In this work, we propose the use of novel aqueous biphasic systems composed of ILs and a biodegradable organic salt as a concentration strategy for EE2 from wastewater. For this purpose, we initially determined the phase diagrams of IL-based aqueous biphasic systems to infer on their formation aptitude. After their evaluation, the ability of IL-based aqueous biphasic systems as a new alternative to the current concentration steps used in the EE2 identification and quantification was investigated.

Experimental section

Materials

EE2, (17α)-17-ethinylestra-1,3,5(10)-triene-3,17-diol, was supplied by Sigma, with a purity level of ≥98 wt% (Fig. 1). The ILs studied were: 1-butyl-3-methylimidazolium bromide, [C₄mim]-Br (99 wt%); 1-butyl-3-methylimidazolium trifluoroacetate, [C₄mim][CF₃CO₂] (>97 wt%); 1-butyl-3-methylimidazolium trifluoromethanesulfonate, [C₄mim][CF₃SO₃] (99 wt%); 1-butyl-3-methylimidazolium thiocyanate, [C₄mim][SCN] (>98 wt%); 1-butyl-3-methylimidazolium tosylate, [C₄mim][TOS] (98 wt%); 1-ethyl-3-methylimidazolium dicyanamide, [C₂mim][N(CN)₂] (>98 wt%); 1-butyl-3-methylimidazolium dicyanamide, [C₄mim]-[N(CN)₂] (>98 wt%); 1-hexyl-3-methylimidazolium dicyanamide, [C₆mim][N(CN)₂] (>98 wt%); tetrabutylammonium chloride, [Bu₄N]Cl (≥97 wt%); and tetrabutylphosphonium chloride, [Bu₄P]Cl (98 wt%). All imidazolium-based ILs were purchased from Iolitec. [Bu₄P]Cl was kindly offered by Cytec Industries Inc., and [Bu₄N]Cl was purchased from Sigma-Aldrich. The molecular structures of the investigated ILs are illustrated in Fig. 2. Ten more ILs were tested. However, it was not possible to form aqueous biphasic systems with these ILs – their names and acronyms are reported in the ESI.^† For the reduction of the water and volatile compounds content to negligible values, IL individual samples were dried under constant stirring under vacuum and at moderate temperature (~80 °C) for a minimum of 24 h. After this procedure, the purity of each IL was checked by ¹H, ¹³C and ¹⁹F (whenever possible) NMR spectra and deemed in accordance with the purity given by the suppliers. The organic salt potassium sodium tartrate tetra-hydrated, KNaC₄H₄O₆·4H₂O (>99 wt%) was acquired from Fluka. The water used was double distilled, passed through a reverse osmosis system and further treated with Milli-Q plus 185 water purification apparatus. HPLC grade methanol and acetonitrile, 99.9%, were from Fischer Chemical and HiPerSolv Chromanorm, respectively.

Fig. 2.

Chemical structures of the ILs used to form aqueous biphasic systems: (i) [C₄mim][CF₃SO₃]; (ii) [C₄mim]Br; (iii) [C₄mim][SCN]; (iv) [C₄mim]-[CF₃CO₂]; (v) [C₄mim][TOS]; (vi) [C₂mim][N(CN)₂]; (vii) [C₄mim][N(CN)₂]; (viii) [C₆mim][N(CN)₂]; (ix) [Bu₄N]Cl; (x) [Bu₄P]Cl.

Phase diagrams and tie-lines

The saturation (binodal) curve of each aqueous biphasic system, which represents the limit between the monophasic and biphasic regions, was determined through the cloud point titration method at (25 ± 1) °C and atmospheric pressure. Aqueous solutions of KNaC₄H₄O₆ at 35 wt% and aqueous solutions of the different hydrophilic ILs at variable concentrations (between 35 and 85 wt%) were prepared gravimetrically (±10⁻⁴ g) and used for the determination of the respective binodal curves. Repetitive drop-wise addition of the organic salt solution to each IL aqueous solution was carried out until the detection of a cloudy mixture. Then, repetitive drop-wise addition of double distilled water was carried out until the detection of a clear and limpid mixture. Whenever necessary, IL aqueous solutions were added to the salt solutions to complete the phase diagrams. Drop-wise additions were performed under constant stirring. The ternary system compositions corresponding to the description of the phase diagrams were determined by weight quantification of all components added to the mixture within ±10⁻⁴ g.

The tie-lines (TLs), which are straight lines that describe the composition of the coexisting phases of a given mixture point, were determined by a gravimetric method originally described by Merchuk et al.⁴⁸ A mixture composition at the biphasic region was gravimetrically prepared, vigorously stirred, and allowed to reach the equilibrium by the separation of both phases for at least 12 h at (25 ± 1) °C. After the separation step, both top and bottom phases were weighed. Finally, each individual TL was determined by the application of the lever-arm rule.

The experimental binodal curves were fitted according to eqn (1):

$$\left[\mathrm{IL}\right]=A\exp \left[\right(B\times {\left[\text{salt}\right]}^{0.5})-(C\times {\left[\text{salt}\right]}^3\left)\right]$$ (1)

where [IL] and [salt] correspond to IL and salt weight fraction percentages, respectively; and A, B and C are constants achieved by the least-squares regression of the experimental data. Their values and the corresponding standard deviations (σ) are provided in the ESI.^†

For the determination of TLs, the following system of four equations was used (eqn (2)–(5)) with four unknown variables ([IL]_IL, [salt]_IL, [IL]_salt, [salt]_salt):

$${\left[\mathrm{IL}\right]}_{\mathrm{IL}}=A\exp \left[\right(B\times {{\left[\text{salt}\right]}_{\mathrm{IL}}}^{0.5})-(C\times {{\left[\text{salt}\right]}_{\mathrm{IL}}}^3\left)\right]$$ (2)

$${\left[\mathrm{IL}\right]}_{\text{salt}}=A\exp \left[\right(B\times {{\left[\text{salt}\right]}_{\text{salt}}}^{0.5})-(C\times {{\left[\text{salt}\right]}_{\text{salt}}}^3\left)\right]$$ (3)

$${\left[\mathrm{IL}\right]}_{\mathrm{IL}}=\frac{{\left[\mathrm{IL}\right]}_{\mathrm{M}}}{\alpha }-\frac{1-\alpha }{\alpha }\times {\left[\mathrm{IL}\right]}_{\text{salt}}$$ (4)

$${\left[\text{salt}\right]}_{\mathrm{IL}}=\frac{{\left[\text{salt}\right]}_{\mathrm{M}}}{\alpha }-\frac{1-\alpha }{\alpha }\times {\left[\text{salt}\right]}_{\text{salt}}$$ (5)

where the indexes M, IL and salt correspond to the mixture, IL-rich phase and salt-rich phase, respectively. The parameter α is the ratio between the IL-rich phase and the total mixture weight. The solution of the referred system gives the concentration of IL and organic salt in the top and bottom phases. Some TLs were also analytically determined as described below. The tie-line length (TLL) denotes the distance, i.e., the difference in composition between the salt-rich phase and the IL-rich phase and was calculated according to eqn (6):

$$\mathrm{TLL}=\sqrt{{({\left[\text{salt}\right]}_{\mathrm{IL}}-{\left[\text{salt}\right]}_{\text{salt}})}^2+{({\left[\mathrm{IL}\right]}_{\mathrm{IL}}-{\left[\mathrm{IL}\right]}_{\text{salt}})}^2}$$ (6)

Extraction of ethinylestradiol

For the screening of improved IL-based aqueous biphasic systems for the one-step extraction and concentration of EE2, several ternary (biphasic) systems (IL + KNaC₄H₄O₆ + water) were prepared. The weight fraction percentage of each component was established taking into account a constant weight ratio between the coexisting phases of approximately 1. EE2, ca. 2.0 × 10⁻³ g, was added to each ternary mixture or individual experiments. The ternary mixture was vigorously stirred and left for achieving the equilibrium for at least 12 h, at (25 ± 1) °C, aiming at allowing the complete partitioning of EE2 between the two phases. After the separation of the top and bottom phases, the detection and quantification of EE2 was carried out through UV-spectroscopy, using a Shimadzu UV-1700, Pharma-Spec Spectrometer, at a wavelength of 284 nm. Blank controls of the ternary mixture were always prepared to eliminate possible interferences of the IL and KNaC₄H₄O₆. Three samples of each aqueous phase were analysed, in at least three individual systems, in order to determine the extraction efficiencies of EE2 and the respective standard deviations.

The percentage extraction efficiencies of EE2, EE_EE2%, were estimated by eqn (7):

$${\mathrm{EE}}_{\mathrm{EE}2}\%=\frac{{\mathrm{Abs}}_{\mathrm{EE}2}^{\mathrm{IL}}\times w_{\mathrm{IL}}}{{\mathrm{Abs}}_{\mathrm{EE}2}^{\mathrm{IL}}\times w_{\mathrm{IL}}+{\mathrm{Abs}}_{\mathrm{EE}2}^{\text{salt}}\times w_{\text{salt}}}$$ (7)

where ${\mathrm{Abs}}_{\mathrm{EE}2}^{\mathrm{IL}}$ and ${\mathrm{Abs}}_{\mathrm{EE}2}^{\text{salt}}$ are the absorbance values of EE2 in the IL-rich and KNaC₄H₄O₆-rich aqueous phases (taking into account the respective dilution factors – in weight) and w_IL and w_salt are the weights obtained for the IL-rich and KNaC₄H₄O₆-rich phases, respectively.

Concentration of ethinylestradiol in the [C₄mim][N(CN)₂]-based aqueous biphasic systems

The larger TL of the liquid–liquid aqueous biphasic systems composed of [C₄mim][N(CN)₂] + KNaC₄H₄O₆ + H₂O was determined by means of the preparation of several ternary systems and further addition of KNaC₄H₄O₆ until the detection of a solid phase. Once the equation commonly applied to describe the binodal data is not able to correctly describe the regions of the solubility curve for high IL and salt concentrations,⁴⁹ the concentration of each compound in both phases of this TL was analytically determined. [C₄mim][N(CN)₂] was quantified through UV-spectroscopy, using a Shimadzu UV-1700, Pharma-Spec Spectrometer, at a wavelength of 212 nm. Blank controls were prepared to eliminate interferences caused by the salt in the KNaC₄H₄O₆-rich phase. The water content in each phase was determined by evaporation, by means of an air oven at ~60 °C, until a constant weight of the non-volatile mixture [C₄mim][N(CN)₂] + KNaC₄H₄O₆ was achieved. The KNaC₄H₄O₆ amount was determined by the weight difference. This process was carried out in duplicate to ascertain on the associated standard deviations. In general, shorter TLs obtained by the quantification of all phase-forming components agree well with those obtained by the mass-balance method proposed by Merchuk et al.⁴⁸

The concentration factor of EE2 along the characterized TL was evaluated by the preparation of ternary systems at different compositions and at different weight ratios (weight of water added to the system per weight of the IL-rich phase): 100 and 1000. It should be noted that along the same TL the composition of each phase is maintained while varying only the volume or weight ratio of the phases. In these two situations, and where lower detection limits are required, EE2 was quantified using a Shimadzu High-Performance Liquid Chromatograph (HPLC) Prominence system equipped with a fluorescence detector. This device consists of a degasser DGU-20A5, a bomb LC-20AD, and a column oven CTO-10ASVP. An ACE® C18 column-PFP (5 μm, 150 mm × 4.6 mm) connected to an ACE® 5 C18 4.6 mm i.d. guard column was used for the separation. The mobile phase consisted of a water–acetonitrile mixture (55 : 45, v/v), at a flow rate of 0.8 mL min⁻¹, with an injection volume of 20 μL. The detection/quantification of EE2 was performed using a Shimadzu Prominence RF-20A XS fluorescence detector at an excitation wavelength of 280 nm and an emission wavelength of 310 nm.²⁵ Both the column and cell temperature were maintained at 25 °C. Three individual samples of the IL-rich phase were analyzed in order to determine the recovery of EE2 and the respective standard deviations. Individual standard stock solutions of EE2 were prepared in methanol at a concentration of 100 mg L⁻¹ and were further diluted at appropriate concentrations (between 2.5 and 100 μg L⁻¹) using ultrapure water to obtain the calibration curve. Water and acetonitrile used in the mobile phase were pretreated by filtration using 0.2 μm polyamide membrane filters from Whatman.

pH determination

The pH values (±0.02) of both the IL-rich and salt-rich phases were measured at (25 ± 1) °C using a Mettler Toledo Seven-Multi pH meter.

Results and discussion

Phase diagrams and tie-lines

New ternary phase diagrams were determined for several ILs + KNaC₄H₄O₆ + H₂O aqueous biphasic systems, at (25 ± 1) °C and at atmospheric pressure. The obtained liquid–liquid phase diagrams are depicted in Fig. 3 (the detailed weight fraction data are provided in the ESI^†). For all the phase diagrams, the biphasic region is localized above the solubility curve while the monophasic region is presented below. In general, the larger the biphasic region, the higher is the capacity of the IL to undergo liquid–liquid demixing and to form an aqueous biphasic system, i.e., the easier the IL is salted-out by the salt.

Fig. 3.

Evaluation of the (A) cation alkyl side chain length, (B) cation core and (C) anion nature in the ternary phase diagrams composed of IL + water + KNaC₄H₄O₆ at (25 ± 1) °C: (A) [C₂mim][N(CN)₂] (), [C₄mim]-[N(CN)₂] (), [C₆mim][N(CN)₂] (); (B) [Bu₄P]Cl (), [Bu₄N]Cl (), [C₄mim]-[N(CN)₂] (); (C) [C₄mim][CF₃SO₃] ()⁵⁷, [C₄mim][SCN] (), [C₄mim]-[N(CN)₂] (), [C₄mim][TOS] (-), [C₄mim][CF₃CO₂] (), [C₄mim]Br (). Adjusted binodal data by eqn(1) (—).

The ILs investigated allow addressing the effect of their chemical structures (cation alkyl side chain length and cation and anion nature) on the phase diagrams behaviour or aqueous biphasic systems formation aptitude. Fig. 3A depicts the effect of the imidazolium cation alkyl side chain length on the formation of aqueous biphasic systems, and follows the order: [C₂mim][N(CN)₂] ≪ [C₄mim][N(CN)₂] ≪ [C₆mim][N(CN)₂]. Longer aliphatic chains at the cation enhance the IL hydrophobicity and lead to a wider biphasic region.^50,51 Fig. 3B presents the influence of the cation core, with the [Bu₄N]Cl and [Bu₄P]Cl ILs, on the liquid–liquid demixing. The phase diagram for [C₄mim][N(CN)₂] is also depicted as one imidazolium-based fluid reference. Both [Bu₄N]- and [Bu₄P]-based ILs have an identical ability for the formation of aqueous biphasic systems due to their similar chemical structures (four butyl chains at the cation which are responsible for their high hydrophobicity). The effect of the IL anion nature on the aqueous biphasic system formation is shown in Fig. 3C, where the following order was observed: Br⁻ ≈ [CF₃CO₂]⁻ < [TOS]⁻ < [N(CN)₂]⁻ < [SCN]⁻ ≪ [CF₃SO₃]⁻. This rank is in close agreement with previous studies using other inorganic or organic salts^46,52–55 and it is related to the ability of the IL anion to hydrogen-bond with water: IL anions with lower hydrogen bond basicity values (β) present a higher ability to form aqueous biphasic systems.⁵⁶

Besides its biodegradable and biocompatible features, KNaC₄H₄O₆ presents a significant salting-out effect. This is an important characteristic while foreseeing the use of aqueous biphasic systems for the one-step extraction and concentration of a target analyte since it will permit to obtain considerably long tie-lines, and thus, high concentration factors.

The experimental TLs, i.e., the composition of the phases for a given mixture point, along with their respective lengths (TLL), are reported in Table 1. The pH values of both phases in each aqueous biphasic systems are also shown in Table 1. The pH values of the coexisting phases range from neutral to slightly alkaline (6.23–9.24) and are useful to explore the possibility of using these aqueous biphasic systems as extractive platforms for specific compounds and/or compounds that may suffer speciation.

Table 1.

Experimental data for TLs and TLLs of IL + KNaC₄H₄O₆ aqueous biphasic systems and respective pH values of the coexisting phases

	Weight fraction composition/wt%
IL	[IL]IL	[salt]IL	pHIL	[IL]M	[salt]M	[IL]salt	[salt]salt	pHsalt	TLL
[C4mim]Br	35.41	12.98	6.93	34.13	13.92	26.80	19.27	7.11	10.67
[C4mim][CF3SO3]	62.09	1.16	6.23	40.82	4.57	19.28	8.03	6.29	43.36
[C4mim][CF3CO2]	41.56	9.95	9.10	27.12	19.87	8.75	32.50	8.99	39.82
	37.86	11.87	9.24	25.72	19.98	13.16	28.37	8.88	29.71
[C4mim][SCN]	59.72	1.98	6.63	27.25	15.19	2.55	25.22	6.68	61.71
	55.66	2.45	6.64	23.05	14.99	5.11	21.88	6.70	54.15
	49.31	3.37	6.40	31.23	10.28	8.29	19.05	6.44	43.92
[C4mim][TOS]	57.71	4.24	7.21	38.20	14.93	1.72	34.90	7.34	63.84
	49.84	6.06	7.10	32.91	14.98	4.49	29.94	6.96	51.25
	45.17	7.40	7.10	30.38	15.05	6.96	27.16	7.11	43.02
	41.37	8.64	7.01	29.64	14.59	9.83	24.66	6.93	35.37
[C2mim][N(CN)2]	45.89	6.39	8.69	33.90	15.04	4.22	36.46	8.71	51.39
	43.05	7.64	8.78	32.00	14.98	9.97	29.61	8.68	39.71
[C4mim][N(CN)2]	61.75	1.32	8.64	19.91	30.01	2.39	36.17	8.69	68.84
	54.78	2.55	8.76	29.86	15.09	3.37	28.41	8.63	57.55
	50.69	3.28	8.52	29.92	13.50	5.66	25.41	8.55	50.18
	46.37	4.22	8.61	29.88	12.29	8.23	22.87	8.74	42.44
[C6mim][N(CN)2]	49.51	2.87	7.99	22.86	2.87	7.00	14.25	8.30	44.01
	44.99	3.33	8.04	20.04	10.06	8.66	13.12	7.99	37.62
[Bu4P]Cl	41.58	4.75	6.60	28.09	14.93	2.44	34.29	6.72	49.03
	37.16	6.30	6.65	24.84	15.58	3.93	31.34	6.73	41.61
	35.20	7.10	6.65	24.80	15.20	4.40	30.56	6.82	38.72
[Bu4N]Cl	46.95	2.58	8.06	30.01	15.04	2.41	35.34	7.94	55.29
	36.61	5.87	8.11	25.17	15.03	4.12	31.88	8.24	41.62

Extraction of ethinylestradiol

For a successful extraction, an appropriate manipulation of the phases’ properties, which control the selectivity and the partition of the solute of interest, is always required. This approach was taken into consideration by scanning systems with different ILs and by manipulating their phases’ compositions and volumes (cf. the ESI^†), as presented and discussed below.

Effect of the IL type

The extraction efficiencies of EE2, and respective standard deviations, in aqueous biphasic systems formed with different ILs, KNaC₄H₄O₆ and H₂O are depicted in Fig. 4 (cf. the ESI^† with the detailed values). To avoid dissimilarities in the extraction efficiencies that could result from the differences in the composition of the coexisting phases, all the partitioning experiments were carried out at a fixed TLL (≈ 43) – Table 1.

Fig. 4.

Extraction efficiencies of EE2, EE_EE2%, in several aqueous biphasic systems at (25 ± 1) °C: IL-rich phase pH (); salt-rich phase pH ().

EE2 preferentially migrates towards the IL-rich phase in all investigated aqueous biphasic systems, with extraction efficiencies higher than 92%. The higher affinity of EE2 for the IL-rich phase correlates well with its octanol–water partition coefficient (K_OW) value. The log(K_OW) of EE2 is 4.15,⁹ meaning that the synthetic hormone has a preferential affinity for more hydrophobic phases, which, in these systems, corresponds to the IL-rich phase. Since the pH values of each phase range between 6.2 and 8.6, and EE2 is mostly present in its neutral form, possible electrostatic interactions between the salt or IL ions and the charged EE2 species are neither significant nor responsible for the solute preferential migration.

As depicted in Fig. 4, the EE2 extraction efficiency is strongly related to the hydrophobicity of the IL that forms a given aqueous biphasic system. The extraction efficiencies of EE2 decrease in the order: [C₄mim][N(CN)₂] > [Bu₄P]Cl ≈ [Bu₄N]Cl > [C₄mim][CF₃SO₃] > [C₄mim][SCN] > [C₄mim][TOS]. Outstandingly, with the aqueous biphasic systems composed of [C₄mim]-[N(CN)₂] and KNaC₄H₄O₆ it was possible to achieve a complete extraction of EE2 for the IL-rich phase in a single-step.

TLL effect

Fig. 5 depicts the impact of TLL on the EE2 extraction with [C₄mim][N(CN)₂]-based aqueous biphasic systems (detailed data in the ESI^†). [C₄mim][N(CN)₂] was chosen since this IL leads to the best extraction efficiency (100% of extraction achieved in a single-step) and allows working with long TLs that further provides the highest concentration factors.

Fig. 5.

Evaluation of the TLL in the extraction efficiencies of EE2, EE_EE2%, in the [C₄mim][N(CN)₂] + KNaC₄H₄O₆ + H₂O aqueous biphasic systems, at (25 ± 1) °C: binodal curve data (); TL data (●); TLL values ().

From Fig. 5, it is possible to conclude that the TLL (at least in the range studied) does not affect the extraction capacity for EE2 and a complete extraction is always attained. The TLL values range from 46.7 to 68.8, which means that it is possible to optimize the process to be more economical and environmentally benign by decreasing the concentration of IL without losing the complete extraction efficiency.

Concentration of ethinylestradiol in [C₄mim][N(CN)₂]-based aqueous biphasic systems

The main characteristic of aqueous biphasic systems to be used as concentration platforms comprises the presence of very long tie-lines. Indeed, the longer the TL, the higher the concentration factor that can be achieved. For a one-stage extraction–concentration step, the complete extraction of the target analyte for the IL-rich phase is required. At this stage, we should also guarantee that the IL-rich phase does not saturate with the extracted biomolecule; otherwise, the accurate quantification of EE2 in the IL-rich phase will be not accomplished leading to underestimated results. The aqueous biphasic system composed of [C₄mim][N(CN)₂] + KNaC₄H₄O₆ + H₂O was selected here since it led to a single-step complete extraction of EE2, without saturation of the IL-rich phase, and allowed working with long TLs.

Fig. 6 depicts the solubility curve of the [C₄mim][N(CN)₂] + KNaC₄H₄O₆ + H₂O system, as well as the TL and mixture points investigated. The extraction efficiency values and respective standard deviations for different initial compositions along the same and largest TL are also shown in Fig. 6.

Fig. 6.

Extraction efficiencies of EE2, EE_EE2%, for different initial compositions along the same TL in the [C₄mim][N(CN)₂] + KNaC₄H₄O₆ + H₂O aqueous biphasic systems, at (25 ± 1) °C: binodal curve data (); TL data (●); initial compositions (▲); EE2 final concentration in the IL-rich phase, [EE2]_IL ().

Different initial compositions (cf. the ESI^†) along the same TL, and with a TLL value circa 69, lead to different weight ratios of the coexisting phases (IL-rich and salt-rich phases). Nevertheless, since the aim of this work is the concentration of an aqueous sample containing EE2 in wastewater, it should be taken into account that the amount of the real sample added to the system is the same as the amount of water required to create the initial mixture point. Thus, in this situation, the concentration factor is equal to the total amount of water added to the mixture point divided by the IL-rich phase amount (phase for which EE2 is completely extracted). Yet, even if the IL-rich phase and water present similar densities ([C₄mim][N(CN)₂]-rich phase density at 25 °C ≈ 1.04 g cm⁻³ as experimentally determined by us), for higher concentration factors this difference becomes significant and should be considered. Therefore, in the following results regarding the concentration factors, volumes are considered instead of the weights of water and of the coexisting phases.

The complete extraction of EE2 was always attained for all the mixture compositions evaluated – Fig. 6. The weight ratio between the water (containing EE2) added in the mixture and the IL-rich phase ranges from 1.12 to 10.03, which means that it is possible to concentrate the synthetic hormone by reducing the volume of the IL-rich phase without losing complete extraction performance. Indeed, this possibility of concentrating EE2 may overcome the main problem on its detection derived from its low concentrations in sewage treatment plants and wastewater. For the mixture point with a concentration factor of 10.03, it was experimentally possible to concentrate EE2 up to 200-fold regarding its saturation solubility in water (4.8 mg L⁻¹).⁹

In a wastewater real sample, EE2 is present in concentrations in the order of ng L⁻¹. Therefore, when dealing with real samples, a concentration factor of 10 is still not sufficient. Thus, the [C₄mim][N(CN)₂]-based aqueous biphasic system was further tested for higher concentration factors using an HPLC with a fluorescence detector for that purpose. Concentration factors of 100 and 1000 can be obtained in the largest TL of the ternary system (Fig. 6). The mixture points required to create these conditions are: 2.69 wt% of [C₄mim][N(CN)₂] + 36.00 wt% of KNaC₄H₄O₆ and 2.43 wt% of [C₄mim][N(CN)₂] + 36.15 wt% of KNaC₄H₄O₆, respectively.

Standard samples with initial concentrations of ~450 and 90 ng L⁻¹ were used to evaluate the EE2 extraction efficiency and the concentration factors of 100 and 1000 afforded by the [C₄mim][N(CN)₂]-based aqueous biphasic systems. The major goal is to reach a final concentration of EE2 at the IL-rich phase higher than the HPLC lower limit detection (3.1 μg L⁻¹). The EE2 recovery results obtained were 98 ± 8% and 102 ± 10% for a concentration factor of 100 and 1000, respectively. Therefore, the concentration of EE2 in wastewater can be increased at least up to 1000-fold, in a single-step, without losing EE2 and without saturating the IL-rich phase, simply by the tuning of the mixture point composition for a minimum IL-rich phase volume. Furthermore, the phase-forming components of the phase containing EE2, mainly IL and water, do not interfere with the HPLC quantification. In summary, the proposed methodology allows to increase the EE2 concentration by, at least, three orders of magnitude (from ng L⁻¹ to μg L⁻¹), and enables its appropriate identification and quantification.

The economical and sustainable viability of the proposed method to detect EE2 in sewage treatment plants is also ensured. The amount of IL used for aqueous biphasic system formation is inversely proportional to the concentration factor required. For instance, for a real water sample of ≈ 61 mL, (aqueous biphasic systems with a total volume of 100 mL) only 2.69 g or 2.43 g of [C₄mim][N(CN)₂] is required to reach a concentration factor of 100 and 1000, respectively.

The analysis of effluent samples requires analytical techniques with high sensitivity because of their complex matrices and low concentration of marker pollutants. LC-MS and LC-MS/MS, due to high sensitivity and selectivity, have been selected as the techniques of choice for environmental analysis of steroid hormones.⁵⁸ However, they display several drawbacks, namely, the use of expensive devices, high maintenance costs and require skilled analysts. Therefore, these techniques are unaffordable for many analytical laboratories. Compared with LC-MS and LC-MS/MS, HPLC coupled either to UV or a fluorescence detector is a simpler, cheaper, easy-to-use and extensively available technique. Yet, due to its higher detection limits, a pre-concentration step is always required. To this end, HPLC conjugated with solid-phase extractions (SPE) or liquid– liquid extractions (LLE) are regularly used.^59–61 Both these pre-concentration techniques require large quantities of toxic and volatile organic solvents and are time-consuming. Low-cost dispersive liquid–liquid micro extraction (DLLME) has also been proposed^44,45 and successfully applied for the concentration of steroid hormones, with an enrichment factor of 178 and an extraction efficiency of 89% for EE2.²⁵ Compared to these methods, IL-based aqueous biphasic systems do not require the use of volatile organic solvents and allow obtaining a complete extraction of EE2 (with no loss of solute) and significantly higher concentration factors. Furthermore, it is possible to tune the aqueous biphasic system properties through the change of the IL nature creating, therefore, a new plethora of concentration systems for other target analytes present in wastewater. The concentration factors here obtained are also far superior to those previously reported with IL-based aqueous biphasic systems with maximum enrichment factors of 10 in the extraction of opium alkaloids or steroid hormones from biological fluids,^62,63 and up to 100 in the extraction of bisphenol A from human fluids.⁴⁴

Conclusions

Aiming at overcoming one of the major limitations in the analysis and monitoring of wastewater, a novel methodology is proposed here to concentrate EE2 by the application of IL-based aqueous biphasic systems. As a first approach, their ternary phase diagrams were determined and the single-step extraction capacity of several IL-based aqueous biphasic systems was addressed. Extraction efficiencies ranging between 92% and 100% were obtained for the IL-rich phase revealing the high affinity of EE2 to the most hydrophobic phase. In particular, outstanding extraction efficiencies, i.e., the complete extraction of EE2 in a single-step, were attained with [C₄mim][N(CN)₂]-based aqueous biphasic systems. Therefore, these types of systems were further tested to optimize the concentration factors of EE2. It was found that by tuning the mixture point composition for a minimum IL-rich phase volume, the EE2 concentration in wastewater can be increased at least up to 1000-fold in a single-step. The proposed methodology allows the increase of the EE2 concentration by three orders of magnitude (from ng L⁻¹ to μg L⁻¹), thus overcoming the detection limits of conventional analytical equipment commonly used in the analysis and monitoring of wastewater. These systems are straightforwardly envisaged for the monitoring of wastewater as potential one-step extraction and concentration routes for a wide array of endocrine disrupting chemicals or other trace pollutants.