Trends in Analytical Chemistry, Vol. 22, No.

10, 2003

Trends

Analysis and removal of emerging

contaminants in wastewater and
drinking water
Mira Petrovic, Susana Gonzalez, Damia` Barcelo
The occurrence of trace organic contaminants in wastewaters, their
behavior during wastewater treatment and production of drinking water
are key issues in the re-use of water resources. Elimination of dierent
classes of emerging contaminants, such as surfactant degradates, pharmaceuticals and polar pesticides in wastewater-treatment plants (WWTPs)
was found to be rather low, so sewage euents are one of the main
sources of these compounds and their treatment-resistant metabolites.
This article reviews the state-of-the-art in the analysis of several groups of
emerging contaminants (acidic pharmaceuticals, antibacterial agents,
acidic pesticides and surfactant metabolites) in wastewaters. It also discusses the elimination of emerging contaminants in WWTPs applying
conventional activated sludge treatment (AST) and advanced treatment
processes, such as membrane bioreactors (MBRs) and advanced oxidation
processes (AOPs), as well as during production of drinking water.
# 2003 Published by Elsevier B.V.
Keywords: Acidic pesticides; Acidic pharmaceuticals; Advanced treatment; Emerging
contaminants; Surfactant degradates; Wastewater treatment

1. Introduction
Mira Petrovic*,
Susana Gonzalez,
Damia` Barcelo
Department of Environmental
Chemistry, IIQAB-CSIC,
c/Jordi Girona 18-26, E-08034
Barcelona, Spain

*Corresponding author.
Tel.: +34 93 400 6172;
Fax: +34 93 204 5904;
E-mail: mpeqam@cid.csic.es

Until the beginning of the 1990s, nonpolar hazardous compounds (i.e. persistent organic pollutants (POPs) and heavy
metals) were a focus of interest and
awareness as priority pollutants, so were
part of intensive monitoring programs.
Today, these compounds are less relevant
for the industrialized countries, since a
dramatic reduction of emissions has been
achieved through the adoption of appropriate measures and the elimination of
the dominant sources of pollution.
However, the emission of so-called
emerging or new unregulated contaminants has become an environmental
problem, and there is widespread consensus that this kind of contamination
may require legislative intervention. This
group mainly comprises products used in

large quantities in everyday life, such as

human and veterinary pharmaceuticals,
personal care products, surfactants and
surfactant residues, plasticizers and various industrial additives. The characteristic of these contaminants is that they do
not need to be persistent in the environment to cause negative eects, since their
high transformation and removal rates
can be oset by their continuous introduction into the environment. One of the
main sources of emerging contaminants
are untreated urban wastewaters and
WWTP euents (Fig. 1). Most current
WWTPs are not designed to treat these
types of substance and a high portion of
emerging compounds and their metabolites can escape elimination in WWTPs
and enter the aquatic environment via
sewage euents.
The partial or complete closure of water
cycles is an essential part of sustainable
water-resource management, and the
increasing scarcity of pristine waters for
drinking water supply and the growing
consumption of water by industry and
agriculture should be countered by the
ecient, rational utilization of water
resources. One of the options is to
increase the re-use of euents for various
purposes, especially in industrial and
agro/food production. However, because
of the high cost of the end-of-pipe
approach (i.e. drinking water treatment),
indirect potable re-use requires ecient
treatment of wastewaters prior to their
discharge. Thus, the occurrence of trace
organic contaminants in wastewaters,
their behavior during wastewater treatment and production of drinking water
are key issues that require further study.

0165-9936/$ - see front matter # 2003 Published by Elsevier B.V. doi:10.1016/S0165-9936(03)01105-1

685

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

Figure 1. Components of a (partially) closed water cycle with indirect potable re-use.

Many believe that, of all emerging contaminants,

antibiotics are of greatest concern, since their emission
in the environment can increase the occurrence of
resistant bacteria in the environment [1]. However,
other emerging compounds, especially polar ones, such
as acidic pharmaceuticals, acidic pesticides and acidic
metabolites of non-ionic surfactants, also deserve particular attention. Because of their physico-chemical
properties (high water solubility and often poor degradability), they are able to penetrate through all natural
ltration steps and man-made treatments, thus presenting a potential risk in drinking water supply [2,3].
Dierent classes of emerging contaminants, mainly
surfactant degradates, pharmaceuticals and personal
care products (PPCPs) and polar pesticides were found
to have rather low elimination rates and have been
detected in WWTP euents and in the receiving surface
waters. However, for most emerging contaminants,
occurrence, risk assessment and ecotoxicological data
are not available, and it is dicult to predict their fate in
the aquatic environment. Partly, the reason for this is a
lack of analytical methods for their determination at
trace concentrations. Analysis of emerging contaminants is a real analytical challenge, not only
because of the diversity of chemical properties of these
compounds, but also because of generally low concentrations (usually at part per billion (ppb) or part per
trillion (ppt) levels) and the complexity of matrices.

686

http://www.elsevier.com/locate/trac

This article reviews the state-of-the-art in the analysis of several groups of emerging contaminants (acidic
pharmaceuticals, antibacterial agents, acidic pesticides
and surfactant metabolites) in wastewaters. It discusses
various aspects of current liquid chromatography (LC)
mass spectrometry (MS)-(MS) [LC-MS-(MS)] and gas
chromatography (GC)-MS [GC-MS] methodology, including sample preparation. It also surveys the elimination
of emerging contaminants in WWTPs by AST and
applying advanced treatment processes, such as MBRs
and AOPs. In addition, it discusses the elimination in
treatment processes at plants for treating drinking water.

2. Analysis of emerging contaminants in

wastewater
One of the major limitations in the analysis of emerging
contaminants remains the lack of methods for quantication of low concentrations. The prerequisite for
proper risk assessment and monitoring of the quality of
waste, surface and drinking waters is the availability of
multiresidual methods that permit measurement at the
low ng/l level (or even below that). However, these
compounds have received little attention because they
are not on regulatory lists as environmental pollutants.
However, today analytical methodology for dierent
groups of emerging contaminants is being developed

[54]
20
MS
GC

[53]
0.42.6
MS

Ibuprofen, naproxen, ketoprofen, tolfenamic acid,

diclofenac, meclofenamic acid

SPE (HLB)

BSTFA
(bis (trimethylsilyl)-triuoroacetamide)
MTBSTFA
(N-methyl-N-(tert-butyldimethylsilyl)
triuoroacetamide)
SPE (polar Empore disk)

GC

[52]
0.620
MS
Pentaorobenzyl bromide
SPE (C18)

GC

[50]
[51]
0.0162.18
0.34.5
MSMS
MS

Diazomethane
SPE (MCX or polymeric sorbent)
SPE (HLB)

LC
GC

[49]
520
MSMS
LC

SPE (C18)

556
LC

MS

2
MS
LC

Sequential SPE
(C18+polymeric sorbent)
SPE (polymeric sorbent)

Bezabrate, diclofenac, ibuprofen, gembrozil,

carbamezapine
Salicylic acid, ibuprofen, ketoprofen, naproxen,
bezabrate, diclofenac
Bezabrate, clobric acid, diclofenac, fenoprofen,
gembrozil, ibuprofen, inomethacin, ketoprofen,
naproxen
Bezabrate, clobric acid, ibuprofen
Ibuprofen, clobric acid, ketoprofen, naproxen,
diclofenac
Clobric acid, diclofenac, ibuprofen, phenazone,
propyphenazone
Clobric acid, naproxen, ibuprofen

LOD
(ng/l)
Detection
Chromatographic
method
Derivatization
Extraction
Compounds

Table 1. Methods for the analysis of acidic pharmaceuticals in wastewaters

2.1. Acidic pharmaceuticals

Dierent methods, mainly based on LC-MS and GC-MS,
in combination with either polymer or C18-based solidphase extraction (SPE), are being developed for the analysis of pharmaceutical compounds. However, most
methods are tailored for neutral compounds (e.g. antibiotics) and less complex matrices (surface and groundwater), while only a limited number of papers describe
procedures applicable to the analysis of polar drugs in
wastewater. A survey of analytical methods for the
quantication of regularly used polar pharmaceuticals
in wastewater matrices is given in Table 1.
A typical analytical method includes the use of octadecylsilica, polymeric, or hydrophilic-lipophilic balanced
(HBL) supports for o-line SPE of water samples, with
either disks or, most frequently, cartridges at low pH
(typically pH=2).
Separation techniques include GC and LC, while, for
detection, MS is the technique most widely employed.
Because of the low volatility of polar pharmaceuticals,
GC-MS analysis requires an additional derivatization,
which makes sample preparation laborious and time
consuming, and also increases the possibility of contamination and errors. Moreover, some compounds are
thermolabile and decompose during GC analysis (e.g.
carbamazepine forms iminostilben as degradation product) [4].
As a result, use of LC-MS and LC-MS-MS is increasing.
When reviewing the principal methods for the analysis
of pharmaceuticals in aqueous environmental samples,
Ternes [4] indicated that LC-MS-MS is the technique of
choice for assaying polar pharmaceuticals and their
metabolites. However, he pointed out the diculty in
the enrichment step, as well as the low resolution and
signal suppression in the electrospray (ESI) interface
because of matrix impurities.
Farre et al. [5] compared LC-(ESI)-MS and GC-MS
(after derivatization with BF3-MeOH) for monitoring
some acidic and very polar analgesics (salicylic acid,
ketoprofen, naproxen, diclofenac, ibuprofen and gembrozil) in surface water and wastewater. Results
showed a good correlation between methods, except for
gembrozil, for which derivatization was not completely achieved in some samples.
In general, the limits of detection (LODs) achieved
with LC-MS-(MS) methods were slightly higher than
those obtained with GC-MS methods (see Table 1); however, LC-MS methodology showed advantages in terms
of versatility and sample preparation being less complicated (i.e., derivatization was not needed).

Reference

and an increasing number of methods is reported in the

literature. Still, analysis of this group of contaminants
requires further improvement in terms of sensitivity
and selectivity, especially for very complex matrices,
such as wastewater.

[5]

Trends

[48]

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

http://www.elsevier.com/locate/trac

687

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

Table 2 summarizes the quantitation and qualier

ions used by the various authors for the determination
of polar drugs in wastewaters using selected ion monitoring (SIM) or multiple reaction monitoring (MRM).
The use of triple-quadrupole MS in LC analysis has substantially increased the selectivity and the sensitivity of
the determination, resulting in LODs better than those
in single-quadrupole LC-MS. Acidic drugs were usually
detected using an ESI interface under negative ionization conditions and deprotonated molecules were
chosen as precursor ions. Typical fragmentation patterns obtained with LC-MS-(MS) showed a loss of CO2
(or loss of the acidic moiety), with a limited number of
other fragments. For example, for diclofenac, ibuprofen

and ketoprofen, the product ions generated by loss of

CO2 were the only fragment ions formed.
2.2. Acidic pesticides
Chlorinated phenoxy acid herbicides account for the
majority of pesticides used worldwide, and their presence in environmental waters is well documented.
However, their behavior during wastewater treatment
has rarely been studied. This group includes, for example, mecoprop (MCPP), MCPA, 2,4-D, 2,4-DP 2,4,5-T,
2,4-DB. These compounds are characterized by high
polarity and thermal lability. For these reasons, LC is
generally more suitable for their analysis. However, the
methods used to determine chlorinated phenoxy acid

Table 2. Quantitation and diagnostic ions (m/z) used for the LC-MS and GC-MS, and base peaks of precursor and product ions used for LC-MS-MS analysis
of acidic pharmaceuticals in wastewaters. Data compiled from references listed in Table 1
Compound

Analytical method Ionization mode MS

MS-MS

SIM ions
Ibuprofen

LC-MS
LC-MS-MS
GC-MS

NI
NI
Positive EI

LC-MS
LC-MS-MS
GC-MS

NI
NI
Positive EI

Clobric acid LC-MS

LC-MS-MS
GC-MS

NI
NI
Positive EI

213, 127

Benzabrate

LC-MS
LC-MS-MS
GC-MS

NI
NI
Positive EI

360, 274

Gembrozil

LC-MS
LC-MS-MS

NI
NI

249, 121

Ketoprofen

LC-MS
LC-MS-MS
GC-MS

NI
NI
Positive EI

253, 209, 197

LC-MS
LC-MS-MS
GC-MS

NI
NI
Positive EI

229, 185, 173, 170

Diclofenac

Naproxen

Diazomethane derivative.
Pentauorobenzyl derivative.
c
Trimethylsilyl derivative.
d
Tert-butyldimethylsilyl derivative.
b

688

http://www.elsevier.com/locate/trac

205, 161

Precursor (m/z) Product 1 (m/z)

Product 2 (m/z)

205M-H

161M-H-CO2 

294M-H

205M-H-CO2 

213M-H

127M-H-CO2 

85C4 H5 O2 

360M-H

274M-H-C4 H6 O2 

154M-H-C12 H14 O3 

249M-H

121M-H-C7 H12 O2 

253M-H

209M-H-CO2 

229M-H

185M-H-CO2 

177, 220
161, 343, 386b
263, 278, 234c
294, 250, 232
a

214, 309
214, 216, 475b
352/354/356d

128, 228
128, 130, 394b
128, 143, 286c

128, 228a
128, 130, 394b

209, 268a
311d

185, 244
243, 302, 185c
287d

170M-H-C2 H3 O2 

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

herbicides are still dominated by GC with either electron capture detection (ECD) or MS detection. The main
disadvantage of GC analysis is that it requires prior derivatization step, usually using highly toxic and carcinogenic diazomethane or, less frequently used, acid
anhydrides, benzyl halides and alkylchloroformates.
The injection-port derivatization with an ion-pair
reagent has been successfully applied [6], as well as in
situ derivatization prior to solid-phase microextraction
(SPME) [7].

Trends
Alternative methods based on LC-ESI-MS have been
proposed. When using LC-MS-(MS), phenoxy acid herbicides are detected under negative ionization conditions, typically yielding [M-H]- ion and one abundant
fragment formed by the loss of the acidic moiety [8,9],
as shown in Fig. 2 for MCPP, 2,4DP and 2,4,5T.
Recently, in-tube SPME followed by LC-MS was
applied for the determination of six chlorinated phenoxy acid herbicides [10]. With river water, LODs were
5^30 ng/l, but wastewater was not tested.

Figure 2. MS-MS spectra and the proposed fragmentation pattern of: (a) MCPP (precursor ion m/z 213); (b) Dichlorprop (2,4-DP) (precursor ion m/z 233);
and, (c) 2,4,5-T (precursor ion m/z 253).

http://www.elsevier.com/locate/trac

689

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

2.3. Antiseptics
Several methods have been proposed for the determination of triclosan (5-chloro-2-[2,4-dichlorophenoxy]
phenol), which is used as an antiseptic agent in a vast
array of personal care (e.g. toothpaste, acne cream,
deodorant, shampoo, toilet soap) and consumer products (childrens toys, footwear, kitchen cutting
boards).
A method based on diazomethane derivatization and
GC-ECD was applied for quantication of triclosan in
the wastewater of a slaughterhouse [11].
Lindstrom et al. [12] detected triclosan and methyl
triclosan in lakes and in a river in Switzerland applying
either SPE (macroporous polymeric adsorbent), diazoethane derivatization, silica clean-up and GC-MS
analysis or passive sampling with semi-permeable
membrane devices (SPMDs). LODs varied with the
sample matrix and were < 0.4 ng/l for surface water,
< 1^2 ng/l for wastewater and < 1^2 ng/g for the
SPMD samples.

McAvoy et al. [13] used C18 SPE for isolation of

triclosan from wastewater and supercritical uid CO2
for sludge, followed by derivatization to trimethylsilylethers and quantitation by GC-MS. Recovery from
spiked wastewater was in the range 79^88% for inuent,
from euent 36^87% and from primary sludge 70^90%.
Recently, Aguera et al. [14] compared the use of GCMS, in electron impact ionization (EI) and negative chemical ionization (NCI) modes, and LC-ESI-MS-MS for
the analysis of triclosan in urban wastewater and
marine sediments. Mass spectra obtained are shown in
Fig. 3. GC-NCI-MS was found to represent the best
approach for triclosan allowing LODs at ng/l levels with
an adequate identication criterion (three ions). However, the LC-MS-MS spectrum obtained for triclosan did
not show fragmentation at all under the dierent fragmentation conditions assayed, only the molecular
cluster being present. The method yielded higher
LODs M 0.35 mg/l for spiked urban wastewater and
3.5 mg/kg for marine sediment.

Figure 3. Full scan mass spectra of triclosan (in EI and NCI) (reproduced with permission from [14]. # 2003 Elsevier).

690

http://www.elsevier.com/locate/trac

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

2.4. Alkylphenolic compounds
The trace analysis of alkylphenol ethoxylates (APEOs)
and their acidic metabolites by LC-MS or LC-MS-MS
using atmospheric pressure chemical ionization (APCI)
and ESI was recently reviewed by Petrovic et al. [15,16]
and the analytical performances for oligomeric mixtures of APEOs discussed. Generally, an ESI interface is
used for the analysis of alkylphenolic compounds
because of its higher sensitivity, especially for alkylphenols and carboxylated compounds.
Alkylphenoxy carboxylates (APECs) were detected in
both NI and PI modes. In the NI mode, using ESI, APECs
give two types of ions:
 one corresponds to the deprotonated molecule
[MH] (m/z 277, 321, 263 and 307, corresponding to nonylphenol carboxylate (NPE1C),
nonylphenol ethoxycarboxylate (NPE2C), octylphenol carboxylate (OPE1C), and octylphenol
ethoxycarboxylate (OPE2C), respectively); and,
 the other corresponds to deprotonated alkylphenols [17].
The relative abundance of these two ions depends on
the extraction voltage. In the presence of ammonium
acetate and when using an APCI under PI conditions,
NPE1C gave [M+NH4]+ ions at m/z 296, while NPE2C
gave [M+NH4]+ ion at m/z 340 [18].
LC-ESI-MS was also applied for the analysis of the
dicarboxylated breakdown products [carboxylated
alkylphenoxy carboxylates (CAPECs)] in wastewaters
[19,20]. However, the identication of these compounds using LC-MS, under conditions giving only
molecular ions, is dicult, since CAnPEmCs have the
same molecular mass as APECs but have one ethoxy
unit less and a shorter alkyl chain (An1PEm1C).
Moreover, since some compounds partly co-elute, the
unequivocal assignment of the individual fragments
can be accomplished only by using LC-MS-MS. Typical
fragmentation patterns obtained with LC-ESI-MS-MS
showed the formation of the carboxy-alkylphenoxy
fragment, with the additionally loss of CO2 or an acetic
acid group, in the case of CA5PE12C leading to m/z 149
and 133 fragments [19].
MS-MS spectra of APECs [19,21,22] show intense
signals at m/z 219 (for NPECs) and m/z 205 (for OPECs),
which are produced after the loss of the carboxylated
(ethoxy) moiety, while sequential fragmentation of the
alkyl chain resulted in ions with m/z 133 and 147.
To overcome the problem of low volatility of acidic
alkylphenolic compounds, various o-line and on-line
derivatization protocols have been developed. O-line
derivatization to corresponding triemethylsilyl ethers,
methyl ethers, acetyl esters, pentauorobenzoyl or heptauorobutyl esters, respectively, was applied as a
common approach in GC-MS.

Trends
On-line direct GC injection-port derivatization using
ion-pair reagents (tetraalkylammonium salts), has
been also reported [23]. The most signicant ions in
GC-(EI)-MS of methylated NPECs were fragments produced by rupture of the benzylic bond in the branched
nonyl side-chain [23^25]. GC-CI-MS spectra of the
NPECs with isobutane as reagent gas showed characteristic hydride ion-abstracted fragment ions shifted
by 1 Da from those in the corresponding EI mass spectra
[22]. When using ammonia as reagent gas, intense
ammonia-molecular ion adducts of the methyl esters,
with little or no secondary fragmentation, were reported for NPECs [26]. The ions selected were as follows:
m/z 246, 310, 354 and 398 for NPE1C, NPE2C,
NPE3C and NPE4C, respectively.

3. Elimination by AST
The present state-of-the-art of wastewater treatment
involves the AST process preceded by conventional
physico-chemical pre-treatment steps. Table 3 summarizes data on the elimination of emerging contaminants in WWTPs.
3.1. PPCPs
Daughton and Ternes [1] reviewed the occurrence of
over 50 individual PPCPs (metabolites from more than
10 broad classes of therapeutic agents or personal care
products in environmental samples) mainly in WWTP
euents, surface and ground water and much less frequently in drinking water. Acidic drugs comprised the
major group of PPCPs detected in municipal WWTPs
and, among them, bezabrate, naproxen, and ibuprofen were most abundant (concentrations up to 4.6 mg/l
in German municipal WWTPs).
Tixier et al. [27] found that carbamezapine presented
the highest daily load from the WWTP into Lake Greifensee (Switzerland), followed by diclofenac and
naproxen. Their elimination during passage through a
municipal sewage treatment in most cases was found to
be quite low (see Table 3), in the range 35^90%, and
some compounds, such as carbamazepine, exhibited an
extremely low removal (only 7%) [28]. Consequently,
through sewage euents, PPCPs can enter receiving
surface waters and thus present a risk in the production
of drinking water. For example, clobric acid, a metabolite of three lipid regulating agents (clobrate, etobrate
and fenobrate), has been identied in river and ground
water and even in drinking water at concentrations up
to 165 ng/l [29,30].
3.2. Acidic pesticides
Chlorinated phenoxy acids are widely used in agriculture, but also as herbicides on lawns, algicides in
paints and coatings and roof-protection agents in
http://www.elsevier.com/locate/trac

691

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

Table 3. Elimination at WWTPs (AST). Data complied from references [12,28,36,51,5457]

Average
elimination (%)a

Euent
concentrations
(g/l)

Main degradation
products

Observation

Non-ionic surfactants
Alkylphenol ethoxylates

9099

<0.1350

APEC, CAPEC, AP

Primary degradation fast; ultimate

degradation less than 40%,
with metabolites being potential
endocrine disruptors

Pharmaceuticals
Ibuprofen
Diclofenac
Clobric Acid

6590
6975
3451

0.370.60 (3.4)b
0.060.81 (2.1)
0.120.36 (1.6)

83
4566
69
4669
7

1.12.2 (4.6)
0.270.61 (2.6)
0.020.38 (0.87)
0.310.40 (1.9)
0.302.1 (6.3)

Antiseptics
Triclosan

4492

0.0700.650

Methyl triclosan

Possible photodegradation

Pesticides
MCPP and MCPA

20400

2Methyl-4-Cl-phenol

Application period (mid-March

until mid-May)

<20
<20

2,4-Dichlorphenol
2,4-D; 2,4-dichlorphenol

Compound

Benzabrate
Naproxen
Ketoprofen
Gembrozil
Carbamazepine

2,4-D
2,4,5-T
a

Rapid photodegradation
Degradation product of
lipid-regulating agents

Low removal rate

Primary elimination of the parent compound.

Range of average values detected (in parentheses: maximum concentration detected).

at-roof sealants. As a result, residues of these substances are introduced into the aquatic system through
dierent pathways. For example, in the catchment area of
Lake Greifensee, 65% of MCPP originated from WWTPs
and the remaining 35% from diuse sources [31].
Degradation of acidic pesticides under laboratory
conditions is well studied, but there are few publications dealing with their behavior in real WWTPs. Generally, AST was found to be ineective in removing
chlorinated phenoxy acid herbicides from settled sewage. However, under laboratory conditions MCPP
proved to be biodegradable (nearly 100%); however,
this requires a long adaptation time (lag-phase) of activated sludge [32]. In real WWTPs, this presents a major
diculty since, like the majority of herbicides, MCPP is
applied only during a short growth period of plants,
which means that WWTPs that contain a non-adapted
activated sludge, receive shock-loads of herbicides,
which will not be eliminated.
A long acclimatization period (about 4 months) was
also observed in a bench-scale study using sequencing
batch reactors before 2,4-D biodegradation was established [33]. Subsequently, at steady-state operation,
all reactors achieved practically complete removal
( > 99%) of 2,4-D.

692

http://www.elsevier.com/locate/trac

3.3. Alkylphenolic surfactants

Although their environmental acceptability is strongly
disputed, APEOs are still among the most widely used
non-ionic surfactants. Currently, under optimized conditions, more than 90^95% of these surfactants are
eliminated by conventional biological wastewater
treatment (normally AST). Even if such high elimination rates are achieved, the principal problem is the formation of treatment-resistant metabolites out of the
parent surfactants. The widespread incidence of APEOderived compounds in treated wastewaters and the subsequent disposal of euents into aquatic system raise
concerns about the impact of these compounds on the
environment. Studies have shown that their neutral
(alkylphenols and short ethoxy chain ethoxylates) and
acidic treatment-resistant metabolites (APECs) possess
the ability to mimic natural hormones by interacting
with the estrogen receptor.
It was estimated that 60^65% of all nonylphenolic
compounds introduced into WWTPs are discharged
into the environmentM 19% as carboxylated derivatives, 11% as lipophilic nonylphenol ethoxylate
(NP1EO) and nonylphenol diethoxylate (NP2EO), 25%
as nonylphenol (NP) and 8% as non-transformed
nonylphenol ethoxylates (NPEOs) [34].

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

However, contrary to the general belief that NPECs
are the refractory metabolites, Di Corcia et al. [35]
found that CAPECs are the dominant products of NPEO
biotransformation. By averaging data relative to the
treated euents of ve major activated sludge WWTPs
of Rome (Italy) over 4 months, relative abundances of
NPEOs (nEO =1 and 2), NPECs and CAPECs were found
to be respectively 10  2%, 24  5% and 66  7%.
The concentrations of the acidic metabolites, NPECs
and CAPECs typically are in the low mg/l range. However, high values (up to several hundred mg/l) are detected in euents of WWTPs receiving industrial
wastewaters, especially from tannery, textile, pulp and
paper industries [36].

4. Elimination by modern WWTPs

Although adopted as the best available technology; biological treatment eects only partial removal of a wide
range of emerging contaminants, especially polar ones,
which are discharged into the nal euent. Thus, it has
become evident that the application of more enhanced
technologies may be crucial to fulll the requirements
to recycle municipal and industrial wastewaters as
drinking water. In recent years, there have been studies
of new technologies for not only wastewater treatment
but also production of drinking water. Among them
membrane treatment, using both biological (MBRs) and
non-biological processes (reversed osmosis, ultraltration, nanoltration), and AOPs are most frequently
considered as they may be appropriate for removing
trace concentrations of emerging polar contaminants.
4.1. Membrane processes
MBR technology is considered the most promising
development in microbiological wastewater treatment.
Now, when economic reasons no longer limit the application of MBRs in industrial and municipal WWTPs [37],
and new requirements are being set for wastewater treatment, MBRs may be key in direct or indirect recycling of
wastewaters, because of two of their characteristics:
(a) the low sludge load in terms of BOD, so that the
bacteria are forced to mineralize poorly degradable organic compounds; and
(b) the long life of the sludge gives the bacteria time
to adapt to the treatment-resistant substances
[38,39].
However, although many articles have reported on
the application of MBRs to the treatment of urban and
industrial wastewaters, there are few papers reporting
on the behavior of emerging contaminants during MBR
treatment, and all of those deal with nonylphenolic
compounds.

Trends
Using an MBR unit that comprises three bioreactors
and an external ultraltration unit followed by granular activated carbon (GAC) adsorption, Witgens et al.
[40] reported the removal of more than 90% of NP in
wastewater from a waste-dump leachate plant. The use
of a set of laboratory nanoltration membranes resulted in the retention of more than 70% of NP and this
process was regarded as an alternative for the nal
treatment of MBR euents.
Li et al. [41] used GC-MS and LC-MS-MS to assess the
elimination eciency in a membrane-assisted biological WWTP. The results showed that, compared to conventional WWTPs, membrane-assisted biological
treatment with biomass concentrations of about 20 g/l
could improve the eciency of eliminating NPEOs (and
other ionic and non-ionic surfactants), but could not
entirely stop the discharge with the permeates.
4.2. Treatment by AOPs
There have been studies of AOPs, which use a combination of ozone with other oxidation agents (UV radiation, hydrogen peroxide, TiO2) to enhance the
degradation of polar pharmaceuticals [42^44] and
NPEO metabolites [45].
Ternes et al. [44] used a pilot plant for ozonation and
UV disinfection of euents from a German municipal
WWTP containing antibiotics, beta blockers, antiphlogistics, lipid-regulator metabolites, musk fragrances and iodinated X-ray contrast media. When
10^15 mg/l ozone was used (contact time, 18 min), no
pharmaceuticals were detected. However, the ionic
iodinated X-ray contrast compounds exhibited removal
eciencies no higher than 14%.
In another study [43], ozonation was demonstrated to be a suitable tool for carbamazepine abatement, even under the process conditions usually
adopted in drinking water facilities. However, despite
good primary elimination, a low degree of mineralization was observed and there was no proper total carbon
balance, even after prolonged ozonation, which
indicated the presence of unidentied degradation
products.
However, the degradation eciency of an AOP is limited by the radical scavenging capacity of the matrix of
the treated water. Thus, for sucient degradation of the
pharmaceuticals ( > 90%) from wastewater, the ozone
concentration has to be equal to the dissolved organic
carbon (DOC) value [42], which means that economic
considerations have to underpin the feasibility of the
process for wastewater treatment.
Recently, using a laboratory-scale reactor, Ike et al.
[45] determined that the eectiveness of ozone treatment in the degradation of NPEO metabolites follows
the order: NPE1C > > NP > NP1EO. Acidic metabolites
were completely degraded within 4^6 min (initial concentration, 0.4^1.0 mg/l), the NP concentrations were
http://www.elsevier.com/locate/trac

693

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

reduced by 75^80% in 6 min, while only 25^50% of

NP1EO was eliminated in the same time.

5. Elimination in drinking water-treatment

plants
The occurrence of organic micro-contaminants in raw
water and their removal in the course of production of
drinking water and possible formation of disinfection
by-products are key issues in relation to the quality of
drinking water. Although, compounds discussed in this
review are currently not regulated in drinking water
directives, precautionary principles should be
employed, and the removal of all organic micro-contaminants should be as high as possible. However, several studies have shown that the removal of emerging

polar contaminants during drinking water treatment is

incomplete.
The elimination of selected pharmaceuticals (clobric
acid, diclofenac, carbamezapine, bezabrate) during
drinking water treatment was investigated in the
laboratory, at the pilot-plant scale and in real waterworks in Germany [46]. Sand ltration under aerobic
and anoxic conditions, as well as occulation using
iron(III) chloride, did not signicantly eliminate the
target pharmaceuticals, while ozonation was quite
eective in eliminating these polar compounds. Diclofenac and carbamezapine were reduced by more than
90%, bezabrate was eliminated by 50%, while clobric
acid was stable even at high ozone doze. Filtration with
granular activated carbon (GAC) under waterworks conditions was very eective in removing pharmaceuticals,
apart from clobric acid, less of which was adsorbed.

Figure 4. Fate of nonylphenolic compounds during production of drinking water. (a) Total concentration of nonylphenolic compounds and their elimination
during dierent treatment steps at waterworks Sant Joan Despi` (Barcelona, Spain); (b) Average composition (calculated on a molar basis) of nonylphenolic
compounds in raw water; and, (c) Average composition (calculated on a molar basis) of nonylphenolic compounds in pre-chlorinated water.

694

http://www.elsevier.com/locate/trac

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

The behavior of polar alkylphenolic compounds during processing of contaminated water in waterworks
and their possible occurrence in treated water was
rarely considered to be of interest and there are hardly
any data available for drinking water.
The elimination of neutral and acidic nonylphenolic
compounds and their brominated and chlorinated derivatives during drinking water-treatment processes at
the waterworks that supply drinking water to Barcelona (Spain) was investigated utilizing a very sensitive
LC-MS-MS method [47]. The concentration of total
nonylphenolic compounds M NPECs (nEO=0^1), NPEOs
(nEO=0^1) and NP M in raw water from the Llobregat
river entering the waterworks was in the range
8.3^21.6 mg/l, with NPE2C being the most abundant
compound. Pre-chlorination reduced the concentration of short-ethoxy chain NPECs and NPEOs by
25^35%, and of NP by almost 90%. However, this
reduction was partly because of their transformation
to halogenated derivatives. After pre-chlorination,
halogenated nonylphenolic compounds represented
approximately 13% of the total metabolite pool, of
which 97% were in the form of brominated acidic
metabolites. The eciency of further treatment steps to
eliminate nonylphenolic compounds (calculated for the
sum of all short-ethoxy chain metabolites, including
halogenated derivatives) was as follows:
 settling and occulation followed by rapid sand
ltration (7%);
 ozonation (87%);
 GAC ltration (73%); and,
 nal disinfection with chlorine (43%), resulting
in overall elimination in the range 96^99%
(mean 98% for four sampling dates), as shown in
Fig. 4.

6. Conclusions
The application of advanced LC-MS and GC-MS technologies to environmental analysis has allowed the
determination of a broader range of compounds and
thus permitted more comprehensive assessment of
environmental contaminants. Among the various compounds considered as emerging pollutants, acidic pharmaceuticals, surfactant degradates and acidic
pesticides are of particular concern, because of both
their ubiquity in the aquatic environment and health
concerns.
Elimination of these emerging contaminants during
wastewater and drinking water treatment is not satisfactory, so control of improved treatment has to be
strict to ensure that the proportion of these micro-contaminants removed is as high as possible. Thus, in view

Trends
of the possible re-use of WWTP euents, more research
is needed to evaluate the fate of emerging contaminants
and their eects in the aquatic environment. Moreover,
as disinfection processes (either chlorination or ozonation) potentially shift the assessment of the risk of
human consumption of the parent compound to its
degradation products, generic analytical protocols will
have to be developed for the simultaneous determination of parent compounds and their metabolites.

Acknowledgements
The work described in this article was supported by the
EU Project P-THREE (EVK1-CT-2002-00116) and by
the Spanish Ministerio de Ciencia y Tecnologia
(PPQ2002-10945-E). M. Petrovic acknowledges the
Ramon y Cajal contract from the Spanish MCyT. S.
Gonzalez acknowledges the grant from the Spanish
MCyT (PPQ2001-1805-CO3-01).

References
[1] C.G. Daughton, A.T. Ternes, Environ. Health Perspect. 107
(1999) 907.
[2] T.P. Knepper, F. Sacher, F.T. Lange, H.J. Brauch,
F. Karrenbrock, O. Roeden, K. Linder, Waste Manage. 19
(1999) 77.
[3] I. Janssens, T. Tanghe, W. Verstraete, Water Sci. Technol. 35
(1997) 12.
[4] T.A. Ternes, Trends Anal. Chem. 20 (2001) 419.
[5] M. Farre, I. Ferrer, A. Ginebreda, M. Figueras, L. Olivella,
L. Tirapu, M. Vilanova, D. Barcelo, J. Chromatogr. A 938
(2001) 187.
[6] W.H. Ding, C.H. Liu, S.P. Yeh, J. Chromatogr. A 896 (2000)
111.
[7] T. Henriksen, B. Svensmark, B. Lindhardt, R.K. Juhler,
Chemosphere 44 (2001) 1531.
[8] O. Pozo, E. Pitarch, J.V. Sancho, F. Hernandez, J. Chromatogr.
A 923 (2001) 75.
[9] R. Bossi, K.V. Vejrup, B.B. Morgensen, W.A.H. Asam,
J. Chromatogr. A 957 (2002) 27.
[10] M. Takino, S. Daishima, T. Nakahara, Analyst (Cambridge,
U.K.) 126 (2001) 602.
[11] M. Graovac, M. Todorovic, M.I. Trtanj, M.M. Kopecni,
J.J. Comor, J. Chromatogr. A 705 (1995) 313.
[12] A. Lindstrom, I.J. Buerge, T. Poiger, P.A. Bergqvist,
M.D. Muller, H.R. Buser, Environ. Sci. Technol. 36 (2002)
2322.
[13] D.C. McAvoy, B. Schatowitz, M. Jacob, A. Hauk, W.S. Eckho,
Environ. Toxicol. Chem. 21 (2002) 1323.
[14] A. Aguera, A.R. Fernandez-Alba, L. Piedra, M. Mezcua,
M.J. Gomez, Anal. Chim. Acta 480 (2003) 193.
[15] M. Petrovic, D. Barcelo, J. Mass. Spectrom. 36 (2001) 1173.
[16] M. Petrovic, E. Eljarrat, M. Lopez de Alda, D. Barcelo,
J. Chromatogr. A. 974 (2002) 23.
[17] M. Petrovic, A. Diaz, F. Ventura, D. Barcelo, Anal. Chem. 73
(2001) 5886.
[18] A. Di Corcia, J. Chromatogr. A 794 (1998) 165.
[19] N. Jonkers, T.P. Knepper, P. de Voogt, Environ. Sci. Technol.
35 (2001) 335.

http://www.elsevier.com/locate/trac

695

Trends

Trends in Analytical Chemistry, Vol. 22, No. 10, 2003

[20] A. Di Corcia, A. Constantino, C. Crescenzi, E. Marinoni,

R. Samperi, Environ. Sci. Technol. 32 (1998) 2401.
[21] M. Petrovic, A. Diaz, F. Ventura, D. Barcelo, J. Am. Soc. Mass
Spectrom. 14 (2003) 516.
[22] C. Hao, T.R. Croley, R.E. March, B.G. Koenig, C.D. Metcalfe,
J. Mass. Spectrom. 35 (2000) 818.
[23] W.H. Ding, C.T. Chen, J. Chromatogr. A 862 (1999) 113.
[24] M. Ahel, T. Conrad, W. Giger, Environ. Sci. Technol. 21
(1987) 697.
[25] E. Stephanou, M. Reinhard, H.A. Ball, Biomed. Environ. Mass
Spectrom. 15 (1988) 275.
[26] J.A. Field, R.L. Reed, Environ. Sci. Technol. 30 (1996) 3544.
[27] C. Tixier, H.P. Singer, S. Ollers, S.R. Muller, Environ. Sci.
Technol. 37 (2003) 1061.
[28] T.A. Ternes, Water Res. 32 (1998) 3245.
[29] H.J. Stan, M. Linkerhanger, Vom Wasser 83 (1994) 57.
[30] T. Heberer, H.J. Stand, Vom Wasser 86 (1996) 19.
[31] A.C. Gerecke, M. Scharer, H.P. Singer, S.R. Muller,
R.P. Schwarzenbach, M. Sagesser, U. Ochsenbein, G. Popow,
Chemosphere 48 (2002) 307.
[32] L. Nitscheke, A. Wilk, W. Schussler, G. Metzner, G. Lind,
Chemosphere 39 (1999) 2313.
[33] S.S. Mangat, P. Elefsiniotis, Water Res. 33 (1999) 861.
[34] M. Ahel, W. Giger, M. Koch, Water Res. 28 (1994) 1131.
[35] A. Di Corcia, R. Cavallo, C. Crescenzi, M. Nazzari, Environ. Sci.
Technol. 34 (2000) 3914.
[36] M. Petrovic, D. Barcelo, Concentrations of surfactants in
wastewater treatment plants, in: T. Knepper, P. de Vooght,
D. Barcelo (Editors), Analysis and Fate of Surfactants in the
Aquatic Environment, Elsevier, Amsterdam, The Netherlands,
2003, pp. 655^673.
[37] M. Gander, B. Jeerson, S. Judd, Sep. Purif. Technol. 18 (2000)
119.
[38] T.A. Peters, R. Gunther, K. Vossenkaul, Filtr. Sep. 2000
(2000) 18.

696

http://www.elsevier.com/locate/trac

[39] P. Cote, H. Buisson, C. Pound, G. Arakaki, Desalination 113

(1997) 189.
[40] T. Wintgens, M. Gellenkemper, T. Melin, Desalination 146
(2002) 387.
[41] H.Q. Li, F. Jiku, H.F. Schroder, J. Chromatogr. A 889 (2000)
155.
[42] C. Zwiener, F.H. Frimmel, Water Res. 34 (2000) 1881.
[43] R. Andreozzi, R. Marotta, G. Pinto, A. Pollio, Water Res. 36
(2002) 2869.
[44] T.A. Ternes, J. Stuber, N. Herrmann, D. McDowell, A. Ried,
M. Kampmann, B. Teiser, Water Res. 37 (2003) 1976.
[45] M. Ike, M. Asano, F.D. Belkada, S. Tsunoi, M. Tanaka,
M. Fujita, Water Sci. Technol. 46 (2002) 127.
[46] T.A. Ternes, Environ. Sci. Technol. 36 (2002) 3855.
[47] M. Petrovic, A. Diaz, F. Ventura, D. Barcelo, Environ. Sci.
Technol. 37 (2002) 4442.
[48] R. Loos, G. Hanke, S.J. Eisenreich, J. Environ. Monit. 5 (2003) 384.
[49] X.-S. Miao, B.G. Koenig, C.D. Metcalfe, J. Chromatogr. A 952
(2002) 139.
[50] D. Calamari, E. Zuccato, S. Castiglioni, R. Bagnati, R. Fanelli,
Environ. Sci. Technol. 37 (2003) 1241.
[51] S. Ollers, H.P. Singer, P. Fassler, S.R. Muller, J. Chromatogr. A
911 (2001) 225.
[52] V. Koutsouba, Th. Heberer, B. Fuhrmann, K. Schmidt-Baumler, D. Tsipi, A. Hiskia, Chemosphere 51 (2003) 69.
[53] G.R. Boyd, H. Reemtsma, D.A. Grimm, S. Mitra, Sci. Total
Environ. 311 (2003) 135.
[54] I. Rodriguez, J.B. Quintana, J. Carpintero, A.M. Carro,
R.A. Lorenzo, R. Cela, J. Chromatogr. A 985 (2003) 265.
[55] B. Ferrari, N. Paxeus, R.L. Giudice, A. Pollio, J. Garric, Ecotoxicol. Environ. Safety, 55 (2003) 359.
[56] M. Stumpf, T.A. Ternes, R.D. Wilken, S.V. Rodrigues,
W. Baumann, Sci. Total Environ. 225 (1999) 135.
[57] R.K. Juhler, S.R. Sorensen, L. Larsen, Water Res. 35 (2001)
1371.