Analytical Challenges and Recent

Advances in the Determination of

Estrogens in Water Environments
Rodica Domnica Briciu1, Agata Kot-Wasik2, and Jacek Namiesnik2,*
1Babes-BolyaiUniversity Cluj-Napoca, Faculty of Chemistry and Chemical Engineering,
Department of Analytical Chemistry and
2Gdañsk University of Technology, Chemical Faculty, Department of Analytical Chemistry

Estrogens have been shown to be present in the water compartment, mainly due to the inefficient removal in
wastewater treatment plants (WWTP). The concentrations of these compounds, although very low (low ng/L), are
sufficient to induce estrogenic responses and alter the normal reproduction and development of wildlife organisms.
The compounds have been determined, by a variety of analytical procedures, in the influents and effluents of WWTP,
fresh waters, rivers, and even drinking waters. Determination of natural and synthetic estrogens and progestogens in
natural water is, however, a difficult analytical task, because of the very low detection limits required and the
complexity of the matrix. Thus, in general, complicated, timeconsuming extraction and purification processes, usually
based on the application of solid–liquid extraction, are performed before final determination by immunoassay, high-
performance liquid chromatography, or gas chromatography, very often coupled with mass spectrometry. This paper
reviews the analytical methods so far described for the analysis of estrogens, which are currently important
environmental pollutants presented in natural and wastewaters.
Discuss of the main steps, from sampling up to analysis, and the techniques most commonly used in the
determination is presented.

Introduction
The natural steroid hormones are generally synthesized from cholesterol in the gonads and
adrenal glands (1). A precise idea about what an endocrine-disrupting compound means should be
established in order to facilitate the identification of active compounds and to fulfill correct
regulatory control rules. On this goal, the International Programme on Chemical Safety and the US
EPA’s Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) in 1998
(EDSTAC, 1998), updated in 2000 (EDSTAC,
2000), proposed the following definition: “endocrine disrupter is an exogenous substance or
mixture that alters theory function the endocrine system causes adverse effects at the level of the
organism, its progeny, populations or suprapopulations of organisms, based on scientific
principles, data, weight-of-evidence, and the precautionary principle” (2). A variety of natural
compounds and anthropogenic chemicals are known or predicted to influence the endocrine
system, such as natural estrogens (e.g., 17-estradiol, estrone), natural androgens (e.g.,
testosterone), phytosteroids (e.g., 17-sitosterol), isoflavonoids (e.g., daidzeine), synthetic
estrogens (e.g., 17-ethinylestradiol), pesticides (e.g., atrazine), phthalates, alkylphenol ethoxylate
surfactants, etc. Steroid hormones are naturally occurring hormones like estrone (E1), estradiol
(E2), and estriol (E3) and synthetically prepared ones: ethynilestradiol (EE2). They are a family of
polycyclic ring structure chemicals containing a common carbon molecular framework. All sexual
hormones have at the base a steroidal struc ture (Figure 1A). The most important
physicochemical characteristics of these compounds are presented in Table I (3).
 Natural steroid hormones are released into the environment almost all the time by the urine
and excrement of all species, sexes, and types of farm animals. We can divide sources and release of
naturally produced steroid hormones into few groups: produced by humans, produced by
livestock, produced by wildlife, produced from treated sewage waste (4).
Data concerning non-domestic animals connected with the release of steroid estrogen
hormones are very poor. However, it was affirmed that they are released into the water bodies by
fish. This phenomenon was observed mainly before and during reproduction periods. In raw
sewage, the level of hormones varies due to the source and amount of rainfall (5).
According to water circulation systems, estrogens may be present in all water bodies. Even after
special cleaning treatment in municipal wastewater plants, trace amounts may penetrate into
drinking water. During recent years, the trace-level presence of these compounds, especially with
estrogenic properties, in the water of the environment has become a worldwide concern. Society
has become agitated because of the potential risk to human life and wildlife due to exposure to both
natural and synthetic chemicals, which may interfere with reproduction and development.

Reproduction (photocopying) of editorial content of this journal is prohibited without

publisher’s permission.
The endocrine disruptors may be released directly or indirectly into the aquatic environment.
The estrogenic chemicals end up in the aquatic environment through excretions of humans, farm
animals, and wildlife (6,7). In urinary excretion there is a presence of natural estrogens such as
estrone, estradiol, and estriol, which have medicine or veterinary applications. One of the major
sources of contamination of the aquatic environment are wastewater treatment plants (WWTP)
because several estrogens and related compounds are not totally removed or degraded by
biological treatments. According to this, steroid hormones are found in rivers with a high charge of
domestic and industrial wastewaters (8,9). In waterways, they may adsorb to solid particles such as
bed sediment or soil, where estrogens may persist for quite a long period (10,11). Endocrine-
disrupting compounds (EDCs), especially estrogens, are chemicals of an increasing public concern,
because the exposure of EDCs has caused adverse health impacts on wildlife (12,13). Many studies
have suggested that estrogens from treated wastewater are responsible for male fish feminization
and sexual disruption in some aquatic organisms (14–16). They have been detected in surface
waters and wastewaters at concentration levels of ng/L. Most notable, estrogens such as 17-
estradiol, estrone, and ethinylestradiol have been implicated in feminization of male fish
at concentrations as low as ng/L.
The entire world is paying attention to the research for finding the best method of changing the
impact of EDCs on human and animal health. It is essential to establish some new controllable
methods that allow complete elimination of estrogens from water. Due to an incomplete
elimination from WWTPs, both estrogens and residues of pharmaceutical products are found in
ground and surface waters (17–21).
The introduction of sewage treatment plant (STP) effluents, which contain various classes of
contaminants, into aquifers and sediments creates the potential for groundwater contamination.
The amount of EDCs in water can be estimated by analyzing aquifer materials that are
characterized by a great absorbing capacity. Therefore, depending on the environmental
conditions, the concentration of the estrogens can be determined by correlating the EDC amount in
aquifers with real concentrations in water (22).
Within the last 10–15 years, the increasing use of liquid chromatography–mass spectrometry
(LC–MS) has led to a revolution of environmental analysis, providing a new analytical tool that
enables the identification of highly polar organic pollutants without derivatization, down to
nanogram-per-liter levels in all kinds of water bodies (wastewater, surface water, groundwater,
and drinking water). The major innovation that enabled this involved the development of the
appropriate ionization interfaces to couple LC with MS. Currently, electrospray ionization (ESI) and
atmo spheric pressure chemical ionization (APCI) are the most commonly used LC–MS interfaces.
 Further innovations have been made in rapid on-line extraction, microextraction, and on-line
derivatization techniques in combination with gas chromatography (GC)–MS or GC–tandem MS
detection (23,24).
Many efforts have been undertaken for the accurate analysis of the estrogens in water down to
nanograms-per-liter and even subnanogram-per-liter concentrations. The first methods used
primarily GC–MS or GC–ion trap-MS–MS detection at the end of the 1990s. Today, an increasing
number of methods use LC–MS and LC–MS–MS (11,25,26). The main benefit of LC–MS–MS, in
comparison to GC–MS, is that LC–MS–MS shows lower statistical errors and unnecessary
derivatization. The lower sensitivity of the first-generation of LC–MS–MS methods is no longer a
disadvantage, because the new LC–MS–MS systems are competitive with GC–MS sensitivities. Only
when resolution is mandatory to separate isomers or congeners (such as for polychlorinated
biphenyls, dioxins, or brominated flame retardants), GC–MS–MS systems are still the method of
the choice (11,17–32).
In this paper, a discussion of the main steps, from sampling up to analysis, and the techniques
most commonly used in the determination of estrogens is presented.
Sample Preparation
The estrogens have been frequently investigated in environmental water; thus, several
analytical approaches have been reported in the literature. They usually comprise sample
preparation steps (starting from sampling and sample storage; the most critical step is the
extraction followed by the clean-up) and analysis (off-line and on-line solid-phase extraction
[SPE] followed by chromatographic separation with selective detection—MS).

Sample collection
The assessment of the estrogenic activity in environmental water samples begins with the
sample collection and some sort of storage until analysis. The sample collection must be significant
for an entire site. The best sample storage strategy consists in passing the field sample through the
SPE cartridge, washing the cartridge with methanol, and storing it at –18°C. Under these
conditions, which facilitate the storage of many samples for extensive monitoring, no significant
loss of the estrogens was observed after storage for 60 days. An alternative to this procedure is to
store the samples at 4ºC in amber bottles, previously preserved with 1% formaldehyde solution.
The addition of formaldehyde solution inhibits micro-bacterial growth (33). The storage of the
water in bottles led to no significant losses of the estrogens after 24 days when the samples were
preserved, but led to severe losses when the sample was not preserved (34).
The volume of the sample can vary from 50 mL to 20 L or even more depending on method
sensitivity. A volume higher then 5 L should be avoided, because it leads to an increase of the
humic acid concentration in the samples, and it creates high matrix effects (5).
Table I. Physicochemical Characterization of Steroidal Hormone

Molecular Meltin H2O Half-

Substance weight g point solubility life Hydrophobicit
(g/L) y
(g/mol) (°C) (h)
Estrone 270.3 254–256 19 4.6
6
Estradiol 272.3 173–179 3.6 36 3.5
8
Estriol 288.3 282 NA NA 2.3
9 *
Ethyniles 296.4 183 0.1 36 ± 3.7
tradiol 0 13
* NA = Not available.

A sample volume between 250 mL and 1 L is considered optimum for high-performance liquid
chromatography (HPLC)–MS–MS or GC–MS– MS analysis (20,35).
Within recent years, a new approach for sample collection has been presented. It involves the
extraction of estrogenic compounds directly on an adsorption medium, which is a permeable
membrane. The device used is the passive-sampler. The main passive sampling tool is known as
Polar Organic Chemical Integrative Sampler (POCIS). The absorption medium is sandwiched
between two disc-shaped semi-permeable plastic membranes held in place by two metal
compression rings, which are in turn mounted inside a protective perforated stainless steel
cylinder. This device is left in the water for a period of time (1 to 10 weeks), and the estrogenic
compounds are extracted directly on the membrane. The analytes were extracted by adding 100 mL
of analytical-grade dichloromethane to each 900-mL sample of water, and mixed in a clean glass
bottle. After this procedure, the next analytical steps can be performed (36–41).
Concerning the POCIS validation, Zhang and Hibberd (42) compared passive sampling with spot
sampling for some EDCs, including the estrogenic hormones. As adsorbtion materials, there were
two different membranes: polyethersulfone and polysulfone. The first one presents a greater
adsorbtion capacity. The kinetics of compound uptake, under laboratory conditions, were linear for
10 days. The results obtained by POCIS for the real samples were in good agreement with those from
the spot sampling.
Filtration
The filtration is usually the first step of the sample preparation. This step is particularly necessary
when the subsequent extraction of the sample is based on the SPE, because the suspended solids
could easily clog the adsorbent bed; and when the analysis is performed by immunochemical assay,
to avoid undesired adsorption on to antibodies.
The filtration step can be performed simultaneously with the sample collection and/or extraction, or
as a separate step. This will determine, in part, the type of the filter regarding to the physical form
(pad, filter aid powder, glass wool, etc.), diameter, and the filter holder. Most of the studies reviewed
employed the glass fiber filters with pore size between 0.22 and 1.20 µm (36–40).
Extraction
Extraction of the steroid sex hormones and the related synthetic compounds from the surface and
wastewaters is usually performed by SPE. The liquid–liquid extraction (LLE) has rarely been
reported (43–46).
For the extraction of the steroidal hormones, many SPE procedures are reported. Laganá et al. (47)
determined female hormones: 17-estradiol (17E2), estrone, estriol, and the synthetic
contraceptive additive 17-ethinylestradiol (17-EE2) that have a potency similar to natural
hormones; some alkylphenols, such as 4-nonylphenol and bisphenol A, the isoflavonoids daidzein,
genistein and biochanin A, and mycoestrogens, zearalenone, whose estrogenic properties were
well-recognized. The water samples were extracted (1 L) using HLB and Carbograph-4 cartridge.
They were conditioned before processing the sample. The cartridges were sequentially conditioned
with 10 mL of dichloromethane–methanol (50:50, v/v), 5 mL of methanol, and 10 mL of water. After
that, the sample was passed through the cartridge, followed by washing with 10 mL of water and 0.4
µL of methanol. The retained compounds were eluted with 7 mL of a dichloromethane–methanol
(50:50, v/v) solution. Chromatographic analyses were done using LC–MS–MS, which reduces the
sample preparation to a minimum. Therefore, the sample preparation step is a simple and
inexpensive extraction system based on SPE. Two of the often-used SPE sorbents are: Oasis HLB
(polystyrenedivinylbenzene-N-vinylpyrolidone terpolymer) and Carbograph-4 (graphitized carbon
black). The Carbograph-4 cartridges were chosen because of their well-known adsorption power,
based on anion-exchange and hydrophobic properties, especially toward polar compounds (48),
while Oasis HLB were chosen for their ability to retain a large number of compounds; basic, neutral,
and acidic (49). It is designed, in fact, for retaining both hydrophilic and hydrophobic compounds
with high capacity by means of both van der Waals and H-donor-H acceptor interactions. Contrary to
other cartridges, the hydrophilic-lipophilic balance of the Oasis HLB cartridges allows them to
become dry during the sample manipulation, the sorbent’s wet-ability being constant (50).
The Oasis HLB is one of the most important types of cartridge for the extraction of estrogenic
compounds (51–53). Other types, such as Envi+ or C18, gave adequate results for EDCs analysis
(54–57), but are less sensitive in comparison with the Oasis HLB. By combining the two cartridges,
the selectivity can be increased (58). An excellent comparison of the various types of SPE sorbents,
protocols, and devices has been made by Lopez de Alda and Barcelo (59). In this paper various
procedures for the determination of several estrogens (estriol, estradiol, ethynyl estradiol, estrone,
and diethylstilbestrol) and progestogens (progesterone, norethindrone, and levonorgestrel) in the
environmental matrices, including water and river sediment, are described. In all procedures, final
analysis of the target compounds is performed by the reversed-phase LC–diode array detection-MS,
whereas the sample preparation always includes an SPE step (Table II).
Table II. Summarization of SPE Possibilities for Estrogenic Compounds
from Water
Final Recove
SPE Matrix determinat ry Re
sorbent fs.
ion (%)
Oasis Natural waters; LC– > 47
HLB MS–MS 91
sewage treatment
plant
C18; Surface water; LC– > 60
NH2 wastewater MS 81
silica treatment plant;
bonded influent & effluent
C18; Wastewater and LC– > 61
NH2 MS 91
silica purified water
bonded
C18 Natural waters LC– > 62
UV 85
EN; STP; influent and LC– > 63
effluent water; MS 83
RP- river water; drinking
18e water
RP-18e River water LC– > 64
MS 90

A proper extraction method was reported by Filali-Meknassi et al. (60). The estrogens: E1, E2,
E2, EE2, and E3 were concentrated using 200–1000 mL aqueous matrix of interest. To extract the
studied compounds, two types of cartridges (C18 and NH2) were used, connected in series. The
C18 cartridge was applied as the first one and was rinsed with deionized water (2 mL), methan
(2mL), and dichloromethane (2 mL) prior to the extraction. The samples were passed through C18
cartridge followed by washing with 5 mL deionized water, drying, and finally washing with 5-mL
hexane. The elution of compounds was done with 5 mL of dichloromethane. The solvent must be
evaporated under nitrogen steam. Finally, the dry sample must be reconstituted with 3 mL of
methanol and passed through NH2 cartridge, which was first rinsed with 5 mL of methanol. The
estrogens were eluted from the NH2 cartridges with 5 mL methanol and evaporated under nitrogen
steam to approximately 100 µL and reconstituted with 900 µL methanol. The sample must be kept
at –25ºC until the analysis. The recoveries of the analytes by this method were greater then 90% in
all cases. The method has been shown to be accurate and precise.

Nowadays, hemimicelles and admicelles of sodium

The interactions between absorbents and estrogens are hydrophobic combined with cation-. The
recovery rate of the estrogens, investigated by HPLC–DAD, ranged between 85% and 105%, with a
standard deviation ranging from 3% to 8% (65).
Another new technique uses SPE discs. It appears to be a convenient means of extraction of the
estrogens from large volumes of water. Kelly reported good results for the extraction of estrone,
estradiol, and ethinyl estradiol applying polytetrafluoroethylene SPE discs impregnated with C
particles (66). A glass fiber filter paper and glass beads were placed on the top of the extraction discs
to prevent clogging with the particulate matter. The SPE discs impregnated with polymeric particles
seemed to provide a convenient procedure for the extraction of large volumes of river water
samples, giving good recoveries of the major estrogens: estrone, estradiol, estriol, and
ethinylestradiol, on the ng/L level. However, recoveries adapting discs were low for the sewage
effluents, possibly due to overloading by the large amount of the matrix with the presence of the
organic material. For these samples, the extraction applies large volume C cartridges and provides
an alternative procedure (56). The river water and the WWTP effluents were tested for the presence
of the pollutants E1, E3, 17E2, and 17EE2 using a new methodology that involves high-flow SPE and
LC–MS–MS (67). Without adjusting the pH, they extracted it from an 1-L sample with PolarPlus C18
Speedisks under a flow rate exceeding 100 mL/min, in which six samples could be done
simultaneously using an extraction station. The method was validated on spiked upstream river
water; precisions were mostly within 10% of the tested concentrations (10–100 ng/L), with a relative
standard deviation (RSD) lower than 10%. The limit of detection (LOD) of the environmental
matrixes was 0.78–7.65 ng/L. A pre-filtration step before SPE may significantly influence the
measurement of E1 and EE2 concentrations; disk overloading by water matrix may also impact the
analyte recoveries, along with ion suppression.
A simple, rapid, and sensitive method for the determination of five estrogens: estrone, 17-estradiol,
estriol, ethinyl estradiol, and diethylstilbestrol, was developed by using a fully automated method
consisting of in-tube solid-phase microextraction (SPME) coupled with LC–MS–MS (68). The
optimum in-tube SPME conditions were 20 draw/eject cycles of 40 µL of sample using a supel-q plot
capillary column as an extraction device. The extracted compounds were easily desorbed from the
capillary by passage of the mobile phase, and no carryover was observed. Applying the in-tube
SPMELC–MS–MS method, the limits of detection of the five estrogens ranged from 2.7 to 11.7 pg/mL.
The in-tube SPME method showed 34–90-fold higher sensitivity than the direct injection method. This
method was applied successfully to the analysis of environmental water samples without any other
pretreatment interference peaks. The details of the in-tube SPME technique and its applications have
also been summarized in a number of reviews (69–74).
Off-line SPE is the most frequently used method of extraction. However, during recent years, a new
on-line extraction method was developed. Rodriguez-Mozaz and co. (75) presented the advantages
and disadvantages of this method. A drawback of the off-line SPE procedures is that they can be time
consuming and cumbersome to perform, often requiring many steps before reaching an extract
concentration suitable for instrumental analysis, of which only a small portion is actually injected
onto the chromatographic column. The on-line SPE techniques have made it possible to develop a
faster method by reducing the sample preparation time and thus increasing the sample throughput.
The conditioning, washing, and elution steps can be performed automatically, and some systems also
permit one sample to be extracted, while another is being analyzed by LC (76). The on-line
procedures are particularly attractive in situations where, for example, large number of samples
and/or sample series have to be analyzed routinely with high sensitivity, or when hazardous or
highly infectious materials have to be processed; while the off-line procedures are favorable for their
applicability to on-site sampling and for the opportunity to inject the same extract several times. In
the following sections, an insight on the on-line configurations and sorbents applied for the
environmental analysis is provided, and the multi-residue methods using on-line SPE to extract
selected emerging contaminants from water are reviewed (77,78).
So far, the on-line configuration most widely used for the analytical determination of the emerging
contaminants in environmental water samples has been SPE coupled with LC–MS and LC–MS–MS
using quadrupole instruments. In such a configuration, the inherent advantages of on-line SPE and
LC–MS are put together in a single methodology with the following main features: automation,
sensitivity, selectivity, accuracy, reliability, high throughput, and minimal sample manipulation.
However, the most monitored methods, including those which rely on the use of the on-line
methodologies, are usually restricted to a few preselected compounds and very often are not
sufficient to assess the quality of the water, and there are still many unknown micro-contaminants
present that might be a threat to the environment and human health (79,80).
A novel, fully automated method is based on on-line SPE-LC–ESIMS–MS, which allows the
unequivocal identification and quantification of the most environmentally relevant estrogens in
natural and treated water at the level well below those of concern has been presented recently by
Rodriguez-Mozaz and co. (81). The RSD are between 1.43% and 3.89%, while recovery percentage is
higher than 74%, which suggests that the method is highly precise and accurate. This method was
used to investigate the presence of estrogenic compounds in waterworks before and after treatment
process.
Another attempt was proposed by Brossa et al. (82). Their method is based on on-line SPE
coupled with GC–MS through an oncolumn interface. The LOD for the method was at a level under
µg/L. The most recent review concerning on-line SPE coupled with LC–MS has been published by
Rodriguez-Mozaz and Lopez de Alda (75). They have described advanced technologies that have
found increasing application in the analysis of environmental contaminants, although their
application for the determination of emerging contaminants such as steroidal hormones
(previously unknown or unrecognized pollutants) has still been limited. So far, they indicate that
when using the on-line configuration of SPE coupled with LC–MS–MS, the SPE sorbents can be used
in both traditional (alkyl-bonded silicas and polymers) and novel restricted access materials,
molecularly imprinted synthetic polymers (MIPs), and immobilized receptors or antibodies
(immunosorbents) as materials. This technique can be used to investigate other contaminants such
as: alkylphenols, pesticides, and their derivatives, dezinfection by-products, mycotoxins, and so
on. The online SPE-LC–MS procedures in terms of accuracy, reproductibility, reliability
(confirmation) of results, and capacity for the multianalyte determination are highly superior of
those off-line and
even compared with the other techniques, such as biosensors.

Derivatization
To improve the stability of the compound and also the precision and the sensitivity of the GC
analysis, the sample extract is usually derivatized (83–87). The derivatization is carried out in the
case of thermolabile, polar, and low volatile compounds, such as estro-gens, to avoid thermal
decomposition and to improve chromatographic separation and the sensitivity of the analysis.
Unfortunately, there is sometimes a loss of the sample during the additional manipulation. LC–MS
and LC–MS–MS have some benefits over GC–MS analysis of the estrogens in the environmental
water. These methods may be coupled with on-line devices for sample preparation and pre
concentration techniques, such as SPE, and estrogens can be analyzed without derivatization (88).
In Table III, common derivatization agents used for the analyses of estrogen has been summarized.
Sometimes derivatization can be processed during SPE by modifying sorbent with a derivatization
agent. Such a method was developed by Salvador et al. (97), which used dansyl chloride for
derivatization. An Oasis HLB column was used for the sample clean-up and derivatization support.
The reaction between hydroxyl groups and reagent takes place in the on-line SPE column.
Dansylated estrogens are eluted further back and analyzed with HPLC–MS–MS. The method
allows determinations at 1 ng/L level in savage influent and effluent samples. Similar, Okeyo et
al. (98) indicate that SPME can be used for both derivatization and extraction. In this case, the
extraction fiber has a derivatization bis(trimethylsilyl) trifluoroacetamide headspace. The
detection limit can be decreased in this way to low ng/L levels. For the determination of estrogens,
it was shown that derivatization with pentafluorobenzyl (PFB) bromide is useful for more
sensitive determinations with GC–negative ion chemical ionization (NI-CI)-MS (99) or LC–NI-
APCI-MS (100). PFB derivatives were used as electron-capturing derivatives to produce an intense
(M–PFB) ion in both ionization systems (101). A highly sensitive and accurate method can be
obtained based on the derivatization of EE with different agents. The sensitivity may be improved
by using appropriate derivatization reagents [dansyl chloride, 2-fluoro-1-methylpyridinium
ptoluene sulfonate (FMPTS), pentafluorobenzoyl bromide (PFBBr)] to modify the structure of
estrogens so that their ionization efficiency is increased, making them more detectable by the LC–
MS. Lin and co. (102) tested how environmental matrices influence the detectability of the estrogens.
Both qualitative and semi-quantitative comparisons of the derivatization method were made. It was
found that dansyl chloride derivatives created signal intensities one or two orders of magnitude
greater than those normally found in the underived estrogen standards. The signal derived by
FMPTS were analytedependent. The PFBBr derivatives produced signals that were as much as 5.8
times those found in the underivatized estrogens.
 Table III. Derivatization Agents and Procedures Used
for Estrogen Analyses
Derivatization Detection Final
Refs.
agent limit (ng/L) determinat
ion
Pentafluorobenzo 0.2–17 GC–NCI-MS* 89
yl
bromide 2–70 GC–NCI-MS 90
Pentafluorobenzo 0.2–0.6 GC–NI-CI-MS 91
il
trimetylsilyl
p-Nitrobenzoyl 2.7–8.3† HPLC–FD‡ 92
chloride
MSTFA§ 1–100 GC–MS 93
Dansyl chloride 0.0025 LC–MS–MS 94
1 HPLC–ESI-MS– 95
MS
0.005–20 HPLC–ESI-MS– 96
MS
* NCI = negative chemical ionization.
† µg/L range.

‡ HPLC–FD = high-performance liquid chromatography with

fluorescence detection.
§ MSTFA = N-methyl-N-(trimethylsilyl)-trifluoroacetamide.

The results indicate that PFBBr is better to be used for estrogen analysis in complex samples such
WWTP or STP influent or effluent, while dansyl chloride should be used mainly for clean water
analysis.

Chromatographic Analysis
Up to now, the sensitive and uncomplicated methods for steroidal hormone analysis in the water
samples were developed using GC–MS, LC–MS, and MS–MS. These techniques are recognized as the
most sensitive and reliable instruments in environmental science. Biological methods like
immunoassay are among the most sensitive analytical methods, but they are very limited, on the one
hand. On the other hand, chromatographic techniques, which are not as sensitive as biological ones,
enable simultaneous screening of both steroids and conjugates, and the other compounds. Finally, LC
is not limited by the factors like non-volatility and high molecular weight, and enables the
determination of both conjugated and unconjugated estrogens without the need of the derivatization
(33,103).
GC and HPLC are commonly used for separation and determination of procedures, but the former
is applicable only to volatile estrogens and not to non-volatile ones, such as conjugates. GC–MS is
also widely used for the determination of the estrogen hormone, but LC–MS is recently considered
to be the most promising analytical method for the determination of steroids, including
conjugated steroids due to its sensitivity, specificity, and versatility. Chromatographic analyses of
steroidal hormone have been reviewed by Wolthers and Kraan (104).
The ability to provide timely, accurate, and reliable data is very important for any method to
perform an analysis of the chemicals. Therefore the method validation should be an integrated part
of the process of developing analytical methods. It is important to define the intended purpose of an
analytical method, and the method validation only needs to prove that the method is acceptable for
this purpose. In environmental analysis, this is particularly important, as the need for high precision
may be limited. For steroid estrogens specifically, the purpose of analytical methods can often be
limited to reliable detection of whether the substances are present above certain levels or not. Thus,
in this particular case, only limited method validation is needed. However, false positive or negative
results should be avoided because of the impact that can be created. Some researchers consider that
using an MS system in selected ion monitoring (SIM) mode, or MS–MS, is enough for confirmation,
but if there are interfering ions or only one transition is monitored, the results cannot be considered
valid. Moreover, there are great differences between MS interfaces [e.g. atmospheric pressure
ionization (API) spectra are more limited than electron ionization (EI) spectra]. Nowadays, the use
of high accuracy mass spectrometers, such as time of flight (TOF) and hybrid quadrupole (Q-TOF)
allows result confirmation (105–107).
Although sensitive LC–MS–MS methods using electrospray ionization (108) and APCI (53) are
available, both of these methods can be susceptible to response of a loss due to ion suppression
caused by matrix effects present in the complex samples (109–111). When ion suppression is
present, an additional preparative step including clean up or utilization of LC–MS methods may be
required. The use of MS–MS provides added specificity, which is necessary when analyzing the
samples of an increased matrix complexity. To aid in determining whether GC–MS–MS or LC–MS–MS
analysis should be done on a sample, preliminary screening of the groundwater samples for the
presence of estrogens can be done using enzyme-linked immunosorbent assay (ELISA) or other
assays.
Many other researchers investigated steroidal hormone in different matrices by HPLC–MS–MS
(112–120) and by GC–MS–MS (114–119,121,122). A good possibility to determine EDCs in water
is by determining the contaminants indirectly from fauna existing in the site of interest (120,123–
127).
GC methods
GC–MS is the most widely applied technique for the determination of estrogens and
progestogens in water. Generally, the detection of steroids by GC techniques including MS was
reviewed in 1999 (104). GC–MS is the most popular of all complex techniques for GC. GC–MS–MS
is the hyphenated technique combining GC with tandem MS. MS–MS is any general method
involving at least two stages of the mass analysis either in conjugation with a dissociation process or
a chemical reaction that causes a change in mass or an ion. Although various ionization methods
are available, EI, and chemical ionization (CI) are the most common for GC–MS analysis. In the most
common MS–MS, the first analyzer is used to isolate a precursor ion, which then undergoes of a
fragmentation, either spontaneously or by some activation, to yield product ions and neutral
fragments. A second spectrometer analyzes the product ions. By use of MS–MS instruments, the
selectivity of the analysis is increased not only by a specific mass quantitation, but this specific mass
can be related to a specific fragmentation of the product ions. The detection limits achieved by the
different methods employing GC–MS or GC–MS–MS as final analytical techniques were in the range
0.5–74 ng/L and 0.1–2.4 ng/L, respectively. An overview of GC–MS and GC–MS–MS determination
of estrogens is presented in Table IV.
GC was the first chromatographic method used for the determination of steroidal hormone. Kuch et
al. (135) analyzed a series of compounds with estrogenic activity (E1, 17-E2, 17-E2, 17-EE2, and
other EDCs) in surface and drinking water from southern Germany by GC–CI-MS. During this
analysis, the sample were extracted by SPE followed by derivatization of the phenolic compounds.
The recoveries of the steroids lay in the range between 71% and 79%, with the exception of the
estradiols (between 56% and 67%). The RSD varied from 9% to 15% and indicated a satisfactory
reproducibility and precision of the whole analytical protocol. Endogenous steroids such as E1, E2,
and E2, and the exogenous estrogen EE2 were determined almost unexceptionally in the lowest
ng/L range.
In the drinking water, E1 and EE2 were found at an average of 400 and 350 pg/L, respectively; E2
could only be determined in one sample at 300 pg/L, while E2 was found at an average of 700
pg/L. The investigated steroids were found in the lower ng/L range with mean concentrations of
2.5 ng/L E1, 1 ng/L E2 and E2, and 1.5 ng/L EE2. This correlates widely with the results of previous
investigations (136,137).
Table IV. Survey of GC–MS and GC–MS–MS Methods for Quantitative
Determination of Estrogens and Related Synthetic Compounds in
Aquatic Environmental Samples

Sample Detection LOD

Compound Matrix preparati method (ng/L)
on
E1, E2 ,E3, River Filtration/SPE ESI-MS 0.3–
water 9.0
EE2 + other STP (Oasis HLB +
NH2)/
steroidal Derivatization
hormone (MSTFA)
E1, 17-E2 STP Filtration/SPE ESI-MS < 100
,E3 effluent
s
17EE2 and WWTP (C18; Oasis ESI-MS–MS
HLB)
related /Derivatization
compounds (BSTFA)
E1, 17E2 ,E3 River Filtration/SDB- Quadrupole-MS
water XC 0.03–
0.50
17EE2 and STP extraction
effluent disc;
s
related C18/Derivatization
compounds (PFBO)
E1, 17E2, Wastew SPE ESI-MS 75
ater (Carbograph)/
17E2, E3 N2 Evaporation
(70°C)/
SPE
(C18
)/Der
ivatiz
ation
(BST
FA)
Steroidal Ground Filtration/SPE ESI-MS-MS 2–4
water
estrogens (Str
ataX
)/
Deri
vati
zati
on
(BST
FA)
E1, 17E2 Pearl Filtration/SPE ESI-MS 0.1–5
River (ENVI 18)
17E2, E3 Delta Deri
EE2 vati
zati
on
(BS
TFA;
MST
FA)
17- River Extraction (C18- ESI-MS 1
Oestradiol water discs)/
Oestrone, N2 MS–MS
17- evaporation/
Ethinyloestr Derivatization
adiol (MTBSTFA*)
E1, 17E2 Tama Filtration/SPE APCI-MS 0.1–
River (ENV/124) 0.28
17E2, E3 Japan /Derivatization
(PFBBr)
 E1, 17E2 Ground Filtration/SPE APCI-MS 0.2–
water (Oasis HLB) 0.6
17E2, E3 Derivatization
(PFBBr, TMSI†)
E1, 17E2 Elbe SPE (SDB-1, ESI-MS 1
River Chromabond
Germa HR-P and
ny EASY,
Nexus Isolute
ENV+,
LiChrolute EN,
Oasis HLB)
* MTBSTFA = N-methyl-N-tert.-butyldimethylsilyltrifluoroacetamide.
† TMSI = N-trimethylsilylimidasole.

The differences in the concentrations up to a factor of 10–20 for E1, E2, or EE2 may be caused by
different weather conditions (sunshine and dryness, dilution by rain, temperature) in the sampling
period, differences in the length of time through the treatment process, state of the art sewage
treatment plants, and composition of the influent water.
Soliman and co. (138) presented a rapid GC–MS method for routine measurement of steroidal
hormone and some other human pharmaceuticals in water. A short (12 m) column and steep
temperature programming ramp (18°C/min) allowed a rapid GC separation followed by sensitive
detection by MS in the SIM mode. Method detection limits between 8 and 85 ng/L were achieved
without sample clean-up or derivatization. A 40,000-fold concentration factor was achieved by on-line
continuous (O-C) LLE of the water samples with dichloromethane. Obviously, the GC–MS method
could be applied to the SPE and the SPME isolation procedures.
Chlorinated derivatives of E2 were reported to be produced by aqueous chlorination (139,140).
Nakamura and co. (141) determined estrone and its chlorinated estrones in drinking water. After the
water disinfection process with chloride, it is possible to change nontoxic compounds to toxic ones,
like these chlorinated estrones: 2-chloroestrone, 4-chloroestrone, and 2,4-dichloroestrone. These
estrones were determined by GC–EI-MS in SIM mode. They estimated the risk of these chlorinated E1
and related derivatives for human health and for aquatic wildlife; we need more detailed
quantification in the various water samples, including drinking water and studies of biological
activities, including estrogenicity of these compounds.
Like the rest of the world, New Zealand is affected by the presence of steroidal hormone as
contaminants in water (142). GC–MS analysis had shown that farm effluent samples contain high
level of estradiol (19–1360 ng/L) and its breakdown product estrone (41–3123 ng/L) compared with
piggery or goat farm effluents. The combined load for these estrogens (excluding h epimer) varied from
60 to > 4000 ng/L. The piggery effluent provided the lowest total estrogen load (46 ng/L), with
estrone accounting for nearly 60% of the measured estrogens in this sample, while the synthetic
analogue, 17-ethynylestradiol was detected only in one WWTP sample at trace level. An estrogen
receptor competitive binding assay was used to test the biological activity of the samples and
confirmed that most agricultural waste samples contain high levels of estrogenic compounds. The
potential of these wastes to cause endocrine disruption in the receiving ecosystem is unknown at the
present. This study showed that animal wastewaters are an emerging source of estrogenic
contaminants for natural waters.
Moreover, the distribution of female hormones, 17-estradiol and estrone, was determined in effluents
of 18 selected municipal treatment plants across Canada by Servos and co. (143). Estrogens (E1, E2, E3,
and EE2) have been often detected at a level of ng/L in STP influent and effluent water (144–146). Those
estrogens were also reported to be found at a similar level (ng/L) in coastal surface water (147). Zuo et al.
(148) reported that high concentrations, up to 4.7 ng/L, of EE2 were detected in estuary seawater,
where EE2 may affect lobster and fish abundance in the coastal seawater. Furthermore, E1, E2, and EE2
were detected at a level of pg/L, even in drinking water (135). Therefore, considering the
growing population and the concentrations of such estrogens in the effluent from STPs, they could play
significant roles as endocrine disruptors in aquatic wildlife.
In order to address the Austrian situation concerning endocrine disrupting compounds, a consortium
called Austrian Research Cooperation on Endocrine Modulators (ARCEM) was established in 1999
(149). Among several other issues that were investigated, ARCEM monitored more than 400 groundand
surface-water samples for selected estrogenic hormones and industrial chemicals. Appropriate
analytical methods were established using GC–MS for the detection of hormones. Since analytical results
were forwarded for toxicological assessments within the program, quantification limits below 0.1 ng/L
(ethinyl estradiol) and 10 ng/L (industrial chemicals) were required depending on the individual
compound. In this program, Hohenblum and co. (150) determined steroidal hormones by means of an
isotope dilution technique and GC with high-resolution MS.
All data gathered in the ARCEM program demonstrate that the concentrations are comparable to the
concentrations that were measured in Germany, Switzerland, and other countries (151–153). The
results indicate that both hormones occur in the selected groundand surface-water sites with
detectable concentrations.
LC methods
Due to limited sensitivity, it is not surprising that only a few reports exist on methods for
environmental analysis of estrogens by LC using detectors other than MSs. The use of
spectrophotometric techniques, including diode array detectors (DAD), is common in HPLC systems,
but a high sensitivity determination in a very low concentration range (ng/L) such as in the
environmental samples has not emerged. The adequate techniques of choice for analysis of the
described groups of emerging pollutants are LC–MS and LC–MS–MS. Before the advent of LC–MS,
many of these polar compounds were difficult and sometimes impossible to measure (154).
Table V. Survey of LC–MS and LC–MS–MS Methods for Quantitative
Determination of Steroid Sex Hormones and Related Synthetic
Compounds in Aquatic Environmental Samples

Sample LOD
Compound Matrix preparati (ng/ Re
L) fs.
on
E1, E2, E3, Baltic Filtration/SPE (Oasis 0.1– 1
17-EE2 Sea HLB) 3 70
and related
compounds
E1, E2, E3, STP Filtration/SPE 0.07 1
17-EE2 –0.18 71
E1, E2, E3, Tamagaw Filtration/ SPE 0.2– 1
EE2, a River, 34 72
and Japan (Shodex
sulfonated EDS-I)/Fluorisil
pufication
derivatives
E1, E2, E3, STP and SPE (Carbograph4) 0.00 1
EE2, and 3–15 73
sulfonated/ river
waters
glucuronated
derivatives
E1, E2, E3, STP SPE (Oasis HLB) 0.04 1
and EE2 –0.24 74
E1, E2, E3, River SPE (Superclean 5.1– 1
and EE2 water Envi+) 6.4 75
E1, E2, E3, River Filtration/SPE 2.3– 1
and EE2 water 10.6 76
E1, E2, E3, STP Filtration/SPE 2– 1
and EE2 200 77
E1, E2, E3, WWTP SPE 0.1– 1
EE2, (C18)/evaporation 0.2 78
 E1, and 3S
E1, E2, and New York Extraction (SDB-XC 4 1
EE2 waters discs) 79

In the last decades, LC–MS has experienced impressive progress, both in terms of technology
development and application. The interface designs have changed considerably and have become
much more sophisticated and efficient. Various forms of MS need to be considered for estrogen
analysis, because MS is a family of techniques. High accuracy is a well-known attribute of MS because
it is a very specific technique, and because it uses stable isotope internal standards. Nevertheless,
MS is not always accurate even when such standards are employed (155). The high specificity overall
in an MS-based method is achieved in one of the three ways: including a chromatographic,
electrophoretic, immunoextraction, or another resolving technique prior to MS detection; using a
high resolution form of MS such as a dual-sector, TOF, or ion cyclotron resonance instrument;
relying on one of several forms of tandem MS. These high-specificity options can be combined for
ultrasensitive estrogen detection.
Recent advances in MS for the measurement of estrogens and other EDCs in aquatic environmental
samples have been reviewed (156).
Today, the interfaces most widely used for the LC–MS analysis of steroids in the aquatic environment
are ESI and APCI. LC–MS and LC–MS–MS have been mostly applied in the SIM mode. LC–MS–MS offers
very good sensitivity and selectivity of the trace analysis of environmental pollutants (109,157). LC–
MS–MS is used widely to determine sexual hormones not only from environmental samples, but
also from biological (158–161). Therefore, in agricultural soil, a high concentration of estrogenic
compounds, which due to rainfall are going to contaminate groundwater, have been detected (162–
167). The presence of a steroidal hormone was detected in the plant leaves as well. The steroidal
hormones were filling the plants due to irrigation water (166).
The estrogen analysis by ESI-MS encounters three general problems (168). The first one, which is a
major shortcoming for trace ESI-MS, is that its response is often analyte-dependent. The second
problem is related to the first: the conditions in the overall system for highest sensitivity of each
analyte may be different. However, this problem can be minimized by instantly changing some of
the conditions during the analysis. Third, response can be very dependent on analyte purity (e.g., as
an HPLC peak), which gets worse with high throughput (fast HPLC). This is due to increased matrix
effects that suppress or enhance analyte signals. This general problem for ESI-MS has been recently
studied in more detail for drug analysis, and it was found that APCI can also have this problem
(169). Zhang and Henion (88) reported the analysis of three endogenous estrogen sulfates, E1-3S,
E3-3S, and E2-3S, along with two synthetic estrogens and a stable isotope internal standard, by LC–
ESI-MS–MS.
During the last 10 years, many methods for the determination of estrogenic compounds in water
have been developed. Various LC–ESI-MS methods used for the determination of estrogenic
compounds are presented in Table V.
MS detection shows significant selectivity when the target compound exists in a complex matrix. In
other words, MS detection is a reliable way for identifying the target compound, and HPLC is the
adequate tool for optimizing resolution before the MS detection. In order to overcome the potential
contamination problems associated with E2 determination, manual procedures should be excluded
as much as possible; therefore, an on-line autopretreatment system would be an effective
alternative.
Respecting these considerations, Watabe and co. (180) determined 17-estradiol in the river water
using a fully automated LC–MS method. They prepared a column of surface modified (SM) MIPs for
17-estradiol, utilizing 6-ketoecradiol as a pseudo template. Target compound (E2) is retained while
the interfering matrix constituents, with no affinity for MIPs, pass through the void volume into
waste.
MIPs for E2 were synthesized from 4-vinyl pyridine and ethylene dimethacrylate as a functional
monomer and cross-linking agent, respectively. MIPs selectively retain E2 and provide excellent
chromatographic resolution from interfering of inherent compounds in the river water sample
matrices. Therefore, freshly prepared MIPs were applied to quantitative MS (ESI–) detection of low
levels of E2 in the river water sample. In order to pre-concentrate the target compound for HPLC
analysis, column switching was coupled with a pretreatment column packed with the MIPs. The
repeatability of the actual determinations of the river water sample, in which background E2 was not
detected, spiked with 50 ng/L of E2, was 2.2% RSD with detection and quantitation limits of 1.8 and
5.4 ng/L, respectively. The surface modification of MIP particles packed in the pretreatment column
provided selective affinity and the on-line concentration of low levels of E2, while simultaneously
eliminating the sample matrix interference, resulting in a significant increase of sensitivity and
reproducibility for the LC–MS analysis of E2 in the riverwater samples.
However, it is quite difficult to make a direct determination. To avoid a derivatization step, which
can be annoying and difficult, a new LC–MS–MS method was developed. By using HPLC–ESIMS–MS,
Reddy and co. (174) presented a sensitive method (LOD ranged from 0.04 to 0.28 ng/L) to measure
steroid conjugates (glucuronide and sulphate) in the matrix-rich sewage influents and effluents. The
analyzed compounds were the glucuronide and sulfate conjugates of estrone (i.e., estrone-3-sulfate
and estrone-3-glucuronide) and estradiol (i.e., -estradiol-3-glucuronide, -estradiol17-glucuronide,
-estradiol-3-sulfate, and estradiol-17-sulfate). The MS was operated in the negative ESI mode using
MRM. The precision of the method was good as determined by the RSD of analyses of the three separate
samples. The RSD was < 10% for the most of the detected steroid sulfates; glucuronides had slightly
higher variability.
A highly sensitive and uncomplicated method of analyzing steroidal hormones in river and the
estuarine water samples was developed by Yamamoto and co. (175) using LC–MS–MS equipped with
an ESI source and an atmospheric pressure photoionization (APPI) source. Steroidal hormones
included not only estrogen but also androgen and conjugate forms of these two. APPI displayed
greater sensitivity than ESI for most of the examined unconjugated steroids, with very high sensitivity
for testosterone and 4-androstene3, 17-dione in particular. Contrarily, ESI was more effective for
conjugated hormones. The developed method was applied to the determination of hormones in the
rivers of Osaka City and their estuaries, where the detected hormones were affected by the effluent from
municipal WWTPs, and the hormone concentration values were comparable to those reported in
previous studies of such effluent. Because of the two-way flow and stagnancy of the streams and the
watercourses, the continuous input of steroidal hormones from WWTPs seems to bring about local
accumulation. The levels of androgen were 1 order of magnitude lower than those of estrogen. Estrone,
estrone 3-sulfate, and 4-androstene-3,17-dione were detected in almost all water samples, with
maxima of 51, 5.1, and 6.4 ng/L, respectively.
A method, which employs SPE and LC–MS–MS, using ESI in both positive and negative modes, has
been developed for the trace analysis of 15 pharmaceuticals, four metabolites of pharmaceuticals,
three potential endocrine disruptors, and one personal care product in various waters by Vanderford
and co. (181). Unlike many previous LC–MS–MS methods, which suffer from matrix suppression,
this method uses isotope dilution for each compound to correct the matrix suppression, as well as
SPE losses and instrument variability (182–186). The method was tested on five matrices, and the
results indicate that the method is very robust. The matrix spike recoveries for all the compounds
were between 88% and 106% in wastewater influent, 85% and 108% in wastewater effluent, 72%
and 105% in surface water impacted by wastewater, 96% and 113% in surfacewater, and 91% and
116% in drinking water. The method reports limits for all compounds between 0.25 and 1.0 ng/L,
based on 500 mL of the extracted samples and a final extract volume of 500 µL.
In terms of accuracy and repeatability, LC–MS, GC–MS–MS, and LC–MS–MS are generally
satisfactory, although the derivatization step used prior to GC, in addition to being time-consuming,
can constitute a source of inaccuracy. An advantage of GC–MS compared to LC–MS is the availability
of the data of mass spectra, which are useful for the identification of unknown peaks in
estrogenically active fractions. The recent introduction of MS–MS detection has essentially
improved the performance of chromatographic methods by reducing detection limits and
supporting analyte identification. In the future, the development of the equipment favors the choice
of LC-based methods as a strategy for analyzing estrogens. Also, some new ionization techniques
such as APPI and others have been developed for several types of highly sensitive triple quadrupole
instruments. These developments may mean that LC–MS–MS will become the first choice of
analytical methods due to increased sensitivity and higher selectivity.

Comparison with other determination methods

Different combinations of chemical and biological methods have been used for assessing EDCs
in environmental matrices. Generally, aqueous samples are concentrated via LLE or SPE to
enhance the detection of the target compounds, prior to bioassays and chemical detection by GC–
MS or LC–MS.
Experiments determining sexual hormone can be performed by bioassay techniques, both in
vivo and in vitro. In vivo experiments are avoided because they are often time-consuming and
expensive, and thus sophisticated analytical techniques for the measurement and assessment of
EDCs are highly valued. The low environmental concentrations, complicated sample matrices, and
the diversity of target compounds have generated a need for different robust in vitro bioassays. In
vitro yeast estrogen screening assays have successfully been used to assess estrogenic activity in
environmental samples (167), as they respond to all substances with receptor-mediated
estrogenic activity regardless of the chemical structure. Natural and synthetic estrogens seem to be
the most potent inducers of activity in in vitro studies (187). To identify the individual estrogenic
compounds in environmental samples, the results from biological assays must be combined with
GC or LC coupled with MS analyses. The results are highly sensitive and sometimes enough to avoid
bioassay techniques. GC or LC connected with single MS or the use of the immunochemical
techniques is the minimum necessity for providing sufficiently high quality results.
A comparison between ELISA, GC, and HPLC methods for the determination of estrogens is
presented by Li et al. (103). The results show that the ELISA method gave good correlation with GC
and HPLC results, although there was a slight tendency for ELISA values to be a little bit higher
than values of the GC and HPLC. It is possible that the clean-up and/or derivatization processes
required for GC and HPLC cause some loss of the target compounds, resulting in lower estimated
values. However, it seems that compared to LC or GC, ELISA gave better results but had some
disadvantages [e.g., it is not 100% specific, vulnerable to cross reactivity, requires independent
confirmation (e.g., HPLC–MS–MS or GC–MS–MS), which are not suitable for small sample loads,
synthesis of antibody can be difficult and expensive]. Another comparison was made by Sawaya
and co. (188). It was shown that for the determination of two xenobiotic estrogenic compounds,
namely diethylstilbestrol (DES), and ethinylestradiol, both methods, ELISA and GC–MS,
respectively, are proper to be used. The obtained data showed that the level of DES and
ethinylestradiol ranged from not-detected to 1.2 and not-detected to 0.90 ppb, respectively. In
view of the results obtained by ELISA, the employment of a cut-off value of 0.30 ppb would make
it reasonable to obtain low false-positive results, thus indicating that such a technique provides a
fast and reliable method for the detection and screening of the anabolic samples. All samples (both
negative and positive) were subjected to GC–MS analysis for confirmatory purposes. The results
obtained from the GC–MS analysis were found to be negative. These results show that the
activity seen and reported above was due to the matrix of the samples, but not due to the active
estrogenic compounds. Data on extraction recovery and coefficients of variation are also reported.
These results and some other similar indicate that GC–MS is a good alternative to ELISA (189–193).
The risk of increasing the concentration of EDC compounds in the environment, especially an aquatic
one, is very high. A screening of estrogenic activity on coastal surface water in the Baltic Sea has
been made already (194). As an analytical method for this purpose, LC–MS–MS has been selected.
The monitored compounds were E1, E2, E3, and EE2. The LC system was coupled with a triplestage
quadrupole MS (API 4000, Applied Biosystems/MDS Sciex; Foster City, CA) in MRM mode. By this
way, it was shown that the response in the yeast estrogen screen (YES) was expressed as measured
estradiol equivalents, which was in the range of 0.01 to 0.82 ng/L. Samples from stations
located in inner coastal waters showed higher estrogenic activities than those from outer located
stations. A comparison of measured estrogenicity (YES) and calculated estrogenicity (chemical
analysis) showed significant differences, probably due to the presence of anti-estrogenic
compounds and/or the estrogenic activity of unknown, not identified contaminants. The main
contributors to the overall estrogenic activity were synthetic and natural hormones.
Nowadays, the problem is to find an adequate method to remove the estrogenic hormones from
aqueous environment. The tendency is to use polypore mushrooms. The method efficiency can be
verified through both chromatographic and immunoassay methods. Such experiments were
realized by Auriol et al. (61), which indicated that both methods of determination are suitable and
sensitive to notice the estrogen concentration changes.

Future Trends
The analysis of steroid hormones in the environment constitutes a difficult task, firstly because of the
complexity of the environmental matrices, and secondly because of their very low, physiologically
active environmental concentrations. Thus, to achieve the sensitivity and selectivity for their analysis
at physiologically active concentrations (pg-ng/L in water), quite laborious and time-consuming
procedures are required. A typical analytical procedure includes, within the sample preparation,
various steps, such as filtration, extraction, purification, hydrolysis, derivatization, and evaporation.
For many years, the environmental determination of steroid sex hormones has been dominated by
the use of the biological techniques, such as immunoassays, and GC–MS. However, recently the
applications of LC–MS have experienced rapid growth, due to instrumental developments. The
introduction of the LC–MS–MS has largely improved the performance of the technique by reducing
the detection and quantitation limits and enhancing analyte identification.
GC–MS is an appropriate method for the separation of the noncomplex samples of estrogenic
compounds. In order to separate and determine a high number of steroidal hormones, with a
better resolution, new chromatographic techniques were developed by coupling GC and LC with
two mass spectrometers. Sensitivity, selectivity, and variability are only a few parameters which
are increased by using GC and LC tandem MS.
The next few years will show the general application of these advanced techniques, integrated
into completely automated, online systems. This will improve analytical performance, increase
sample throughput, and reduce operating costs and also contamination risk. Further advances in
the form of new extraction techniques, such as those based on the use, on-line or off-line, of
molecular-imprinting materials and immunoaffinity cartridges, which are yet under development,
and can be expected in the near future. These advances promise to simplify the detection and
measurements of steroidal hormones in environmental matrices.

Conclusions
As we saw previously, triple quadrupole instruments such as GC–MS–MS and LC–MS–MS are
superior to all other techniques regarding sensitivity. Moreover, new modes of ionization and
sample handling may improve the sensitivity even more. One example is the APPI technique, but
other ionization techniques such as ESI and APCI are more widely used. The research in sample
preparation has shown promising new tools that may be tried in this field. But under all
circumstances, if an LOD of less than 0.1 ng/L is needed, then further research is required. LC–
MS–MS seems superior to other methods. Single LC–MS and GC–MS methods are of limited value
without extraordinary sample clean-up efforts, as LODs are often below the sensitivity
level recommended for both effluent sewage and surface water analysis.
The advantage of using LC is that the enzymatic hydrolysis step required for the immunoassay
analysis of both conjugated and unconjugated estrogens, and derivatization step that normally
precedes a subsequent GC–MS analysis, can be obviated. Deconjugation techniques have been
developed that make it possible to detect the entire estrogen concentration level followed by GC–
MS analysis. Methods based on LC–MS are also available for direct analysis of conjugates. UV, GC–
FID, and HPLC are not generally recommended due to low sensitivity and reduced selectivity. GC or
LC connected with single MS or the use of immunochemical techniques are the minimum necessary
to provide sufficiently high quality results. Both single LC–MS and GC–MS can be used even for very
complicated matrices if certain identified quality criteria are fulfilled. The more advanced methods
like LC–MS–MS or GC–MS–MS are found most suitable for analysis, because these techniques
provide the highest sensitivity (LOD = 0.1 ng/L) and selectivity. Methods with an LOD of less than
0.1 ng/L of estrogen are not available on a commercial level. Such methods need further research
for sample preparation combined with the application of highly sensitive triple quadrupole
instruments. Immunochemical methods are also very sensitive (LOD = 0.05–850 ng/L), at least for
analyzing wastewater and STP effluents, but the selectivity is poor compared with the triple
quadrupole instruments. Immunochemical methods have the potential to provide useful data when
used in connection with chemical analysis, but today such strategies are not developed and
research is needed to develop an appropriate strategy combining immunochemical methods with
LC–MS–MS or GC–MS–MS or single MS.
In spite of 100 years of chromatography, during which thousands of articles have been published
on steroid analysis including estrogens, further advances are still needed for the analysis of
endogenous estrogens in different aqueous samples.

Acknowledgments
Rodica Domnica Briciu acknowledges Assoc. Prof. Costel Sarbu for ERASMUS scholarship and
Technical University of Gdansk, Poland. This work has been funded by the Ministry of Science and
Higher Education (Project KBN 2475/B/P01/2007/33 and N N404 027535).