1. Introduction
Steroids occur naturally in microorganisms, animals and plants. They primarily contain three cyclohexanes, one pentagonal carbon ring attached to different functional groups and side chains [
1]. The steroid hormones are all derivatives of cholesterol and are also low-molecular-weight lipophilic compounds [
2,
3]. Hormones can be broadly classified as either natural or synthetic [
3]. Furthermore, they are divided into different families, including mineralocorticoids, glucocorticoids, androgens, estrogens and progesterone [
4]. Glucocorticoids have been used as anabolic growth promoters due to their metabolic properties [
5]. Hydrocortisone is a corticosteroid belonging to the glucocorticoid family. These types of hormones have found use in human medicine because of their immunosuppressive and anti-inflammatory properties [
6,
7]. Hydrocortisone is naturally produced by the adrenal cortex as a stress exposure response [
8]. Hydrocortisone also takes part in the metabolism of fat, protein and carbohydrates [
8]. Estrogens, such as estrone (E1), 17-β-estradiol (E2) and estriol (E3), are steroid hormones that are responsible for the upkeep of the health of breast, skin, brain and reproductive tissues [
9]. Physiologically, progesterone (PRO) works as an estrogen hormonal balancer and is involved in the menstrual cycle, pregnancy and embryogenesis [
10]. However, high levels of these hormones can increase the risk of osteoporosis, have neurotoxic effects on the central nervous system [
5], decrease reproductive performance and the rate of fertilization in fish [
11] and have potential carcinogenic effects on humans [
3]. They are also considered as an endocrine disruptor, which can cause adverse effects on wildlife and humans by interacting with the endocrine system [
12].
In recent years, traces of some steroid hormones have been found in food; these can be traced back to the use of hormones as growth promoters of animals in both industry and farm animals because of their anabolic effects [
13,
14]. In addition, some steroids (such as progesterone) are used as human and veterinary pharmaceuticals, and as a result, both humans and animals have become sources of environmental pollution. This is of great concern, as the presence of these compounds in food and water negatively affects some organs and systems, such as the cardiovascular system, tissues and the nervous system [
15]. Thus, the EU and other countries such as China prohibit the use of hormones because of their potential to be endocrine disruptors in both wildlife and humans [
13,
15]. While the use of hormones has been banned, they can at times be found in compound mixtures to escape surveillance, thus resulting in residues of such chemicals being found in the environment [
16]. As a result, highly sensitive and selective analytical methods are required for the determination of these hormones.
The most common methods for the determination of hormones include chromatographic methods, such as gas chromatography (GC) and liquid chromatography (LC), combined with mass spectrometry detectors for example quadrupole [
17]; orbitrap mass spectroscopy [
3]; and electrospray ionization mass spectroscopy amongst others [
15]. However, due to the complexity of sample matrices, sample preparation is necessary to determine trace hormone residues [
16]. Therefore, different sample preparation techniques have been used for the extraction of hormones from various sample matrices. These include, but are not limited to, solid-phase extraction [
18], liquid–liquid extraction and hollow fiber liquid-phase microextraction [
19]. Solid-phase extraction is one of the commonly used sample preparation techniques [
20]. Because of its advantages, such as ease of operation, high enrichment factors and various sorbent materials, SPE is an attractive sample preparation technique for analytical chemists [
21].
Activated carbon (AC) is one of the most used materials for the removal and extraction of organic pollutants [
22]. Powdered activated carbon can be made from a range of organic materials and is reported to have functional groups that can absorb organic compounds, well-defined porosity and high specific surface areas [
23]. Generally, AC is light in weight such that separation from aqueous solutions is a challenge. Thus, for practical application, combining AC powder with magnetic nanoparticles provides simple and quick use and the reuse of AC [
24]. Βeta-cyclodextrin (β-CD) is a cyclic oligosaccharide with hydrophobic inner cavities that can form host–guest interactions with the aid of supramolecular interaction between β-CD and the analyte of interest [
25,
26]. Due to their attractive features, such as unique cavity, cost effectiveness, non-toxicity, biodegradability and renewable properties, β-CD has proven to be an exceptional adsorbent [
27,
28,
29,
30]. However, β-CD is highly soluble in water. For this, researchers have resolved this challenge by incorporation into a composite [
27,
28] or by crosslinking or immobilization [
29].
In this study, the combination of magnetic activated carbon and β-CD as an adsorbent for the solid-phase extraction of selected steroid hormones in water samples prior to high-performance liquid chromatography–diode array detector (HPLC-DAD) was reported for the first time. The incorporation of magnetic activated carbon and β-CD solved the problems of water solubility of β-CD and the difficulties of recovering activated carbon after use. In addition, the composite combined the benefits of the hydrophobic cavity of β-CD, the high specific surface area of activated carbon and the magnetic properties of magnetite nanoparticles. The HPLC technique was chosen as an analytical method of choice due to its ease of operation and because no derivation step is required prior to analysis [
31]. However, at trace levels, the HPLC technique encounters limitations in the detection of some organic compounds. As a result, a modified yet simple version of SPE was introduced as a sample preparation technique, where SPE was primarily selected for its inclusive nature for the use of modified adsorbents to increase the selectivity of the method [
32]. The adsorbent was characterized by X-ray diffraction, scanning electron microscopy and transmission electron microscopy. The parameters affecting the extraction and preconcentration of the steroid hormones were investigated using response surface methodology.
2. Materials and Methods
2.1. Materials and Reagents
Hydrocortisone (HYD), β-estradiol (E2), progesterone (PRO) and estrone (E1) were obtained from Sigma-Aldrich (St. Louis, MO, USA). The physical properties and chemical structure of the analytes are presented in
Table S1. Sodium hydroxide, acetic acid (99.9%), methanol (HPLC grade), ferrous sulphate, ferric (III) chloride and β-cyclodextrin were all purchased from Sigma-Aldrich. A 20 mg L
−1 mixed hormone stock solution was prepared by dissolving an appropriate amount of the analyte in HPLC-grade methanol and was kept chilled at 2 °C. Standard solutions were prepared daily by diluting the stock solution with ultra-pure water (Direct-Q
® 3UV-R purifier system, Millipore, Merck).
2.2. Instrumentation
All pH measurements were carried out using an OHAUS ST series pen pH meter. The adsorption studies were carried out using the Branson 5800 Ultrasonic Cleaner (Danbury, CT, USA). A scanning electron microscopy (SEM, TESCAN VEGA 3 XMU, LMH instrument, Tescan Company, Brno, Czech Republic) coupled with energy dispersive X-ray spectroscopy (EDS) was used to study the morphology and elemental composition of the adsorbent at an accelerating voltage of 20 kV. The transmission electron microscopic image was captured using transmission electron microscopy (TEM, JEM-2100, JEOL, Tokyo, Japan).
An Agilent high-performance liquid chromatography (HPLC) 1200 Infinity series, equipped with a photodiode array detector (Agilent Technologies, Waldbronn, Germany), was used for all analyses. The separation was carried out using an Agilent Zorbax Eclipse Plus C18 column (3.5 μm × 150 mm × 4.6 mm) (Agilent, Newport, CA, USA) operated at an oven temperature of 25 °C. The chromatograms were recorded using a 1.00 mL min−1 flow rate and solvent mixture of 55% mobile phase A (water) and 45% mobile phase B (acetonitrile), and adsorption wavelengths of 230, 260, 280 and 288 nm using an isocratic elution system were used.
2.3. Collection of Samples
Wastewater samples (influent and effluent) were collected from a nearby urban wastewater treatment plant (Gauteng, South Africa) between January 2018 and December 2019. The river water samples were obtained from a river that receives WWTP effluent water. During the study period, 50 influent and effluent samples, as well as 50 downstream river water samples, were collected using the grab sampling technique. All samples were collected using clean 200 mL glass bottles and kept at 2 °C until use.
2.4. Preparation of β-Cyclodextrin-Decorated Magnetic Activated Carbon
Iron solutions of 0.75 mol L
−1 FeCl
3 and 0.5 mol L
−1 FeSO
4.7H
2O were prepared separately. The iron solutions were dissolved at a Fe
3+/Fe
2+ ratio of 2:1 and stirred for 5 min. Then, 4 g of β-cyclodextrin and activated carbon from waste tires (already prepared by Dimpe and colleagues [
33]) were mixed via vigorous stirring and heated at 70 °C. Subsequently, 5 mol L
−1 NaOH (50 mL) was added to the above solution and heating continued at 70 °C for 30 min. A magnet was used to remove the black precipitate that formed during the changing of the color of the mixture. The resulting magnetic material was washed repetitively with a mixture of ethanol and water (50% v/
v) solution to eliminate impurities (unreacted materials). Finally, the obtained magnetic material was dried in an oven at 80 °C for 24 h before use.
2.5. Ultrasound-Assisted Magnetic Solid-Phase Microextraction Procedure
An appropriate amount of adsorbent (4.3–55.7 mg) was measured accurately into a glass vial with a cap; 5 mL of the sample (with pH adjusted accordingly from 4–9) was then added into the adsorbent. Thereafter, the adsorbent was dispersed using ultrasound in an ultrasonic water bath for the specified amount of time (7–32 min). After the time had elapsed, the supernatant was discarded, with separation achieved with the aid of an external magnet. The analyte was then eluted using an accurate volume (313–1086 µL) of HPLC-grade methanol before analysis.
2.6. Optimization of Extraction Procedure
A multivariate optimization approach was used for the optimization of the sample preparation procedure. The optimization strategy was based on a 2
4−1 fractional factorial design and central composite design to determine the experimental parameters that are significant for the preconcentration of steroid hormones.
Table 1 shows the summary of the experimental design conditions for the fractional factorial and central composite designs.
2.7. Adsorption and Reusability
The adsorption experiments were carried out according to the method described by Mashile and colleagues [
34]. Briefly, 15.9 mg of adsorbent was weighed and transferred into nine sealable glass containers. Then, 5 mL of stock solutions with varying concentrations (namely, 1–10 mg L
−1) was then added into the adsorbent. Agitation of the solution by an ultrasonic water bath was performed for 20 min; thereafter, the adsorbent and supernatant were separated with the aid of an external magnet. Analysis of the supernatant (1 mL) was then carried out using 1 mL of the HPLC-DAD. In addition, the reusability of the adsorbent was also investigated by appropriate modification of a method described by Nyaba and coworkers [
35].
2.8. Method Validation
After the determination of the optimum conditions, the developed method was validated according to Hoga and colleagues [
36]. Under optimum conditions, the analytical figures of merit of the described method were evaluated using the limits of detection and quantification (LOD and LOQ) using the expressions LOD = (3 × Sd)/s and LOQ = (10 × Sd)/s, where Sd and s are the standard deviation of 10 replicate measurements at the lowest calibration concentration and slope, respectively. The linear ranges (LRs), correlation coefficient (R2) and enrichment factor were also determined using prepared sample solutions in the range of 50–500 µg L
−1. The precision of the method was also investigated in terms of repeatability and reproducibility using the intraday precision (repeatability) (
n = 15) and interday precision (reproducibility,
n = 7 working days). Validation of the developed method was carried out using spiked real sample recoveries due to the absence of certified reference materials (CRM). Briefly, the water samples were spiked at two levels (1 and 5 µg L
−1). The developed method was then applied to the unspiked and spiked water samples.