Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/231611832

Catalyticoxidativedegradationof17α-ethinylestradiolbyFeIII-TAML/H2O2:Estrogenicitiesofthe...

ArticleinWaterResearch·September2012

DOI:10.1016/j.watres.2012.09.012·Source:PubMed

CITATIONS

8

READS

197

5authors,including:

JianLinChen

CityUniversityofHongKong

18PUBLICATIONS192CITATIONS

SEEPROFILE

L.JamesWright

UniversityofAuckland

166PUBLICATIONS3,727CITATIONS

SEEPROFILE

NareshSinghal

UniversityofAuckland

99PUBLICATIONS835CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyNareshSinghalon01December2016.

Theuserhasrequestedenhancementofthedownloadedfile.Allin-textreferencesunderlinedinblue

arelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.

ww.sciencedirect.com

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 8

Available online at w

journal homepage: www.elsevier .com/locate/watres

Catalytic oxidative degradation of 17a-ethinylestradiol byFeIII-TAML/H2O2: Estrogenicities of the products of partial,and extensive oxidation

Jian Lin Chen a, Shanthinie Ravindran a, Simon Swift b, L. James Wright c,Naresh Singhal a,*aDepartment of Civil & Environmental Engineering, University of Auckland, Auckland 1142, New ZealandbMolecular Medicine and Pathology, University of Auckland, Auckland 1142, New ZealandcSchool of Chemical Sciences, University of Auckland, Auckland 1142, New Zealand

a r t i c l e i n f o

Article history:

Received 9 April 2012

Received in revised form

29 August 2012

Accepted 3 September 2012

Available online 12 September 2012

Keywords:

FeIII-TAML

FeIII-B*

EE2Estrogenicity

Catalysis

Oxidation

Daughter products

* Corresponding author. Department of CiviRoad, Auckland 1142, New Zealand. Tel.: þ6

E-mail address: n.singhal@auckland.ac.n0043-1354/$ e see front matter ª 2012 Elsevhttp://dx.doi.org/10.1016/j.watres.2012.09.012

a b s t r a c t

The oxidative degradation of the oral contraceptive 17a-ethinylestradiol (EE2) in water by

a new advanced catalytic oxidation process was investigated. The oxidant employed was

hydrogen peroxide in aqueous solution and the catalyst was the iron tetra-amido macro-

cyclic ligand (FeIII-TAML) complex that has been designated Na[Fe(H2O)(B*)] (FeIII-B*). EE2(10 mM) was oxidised rapidly by the FeIII-B*/H2O2 (5 nM/4 mM) catalytic oxidation system at

25 �C, and for reactions at pH 8.40e11.00, no unchanged EE2 was detected in the reaction

mixtures after 60 min. No oxidation of EE2 was detected in blank reactions using either

H2O2 or FeIII-B* alone. The maximum rate of EE2 loss occurred at pH 10.21. At this pH the

half-life of EE2 was 2.1 min and the oxidised products showed around 30% estrogenicity

removal, as determined by the yeast estrogen screen (YES) bioassay. At pH 11.00, partial

oxidation of EE2 by FeIII-B*/H2O2 (5 nM/4 mM) was studied (half-life of EE2 was 14.5 min) and

in this case the initial intermediates formed were a mixture of the epimers 17a-ethynyl-

1,4-estradiene-10a,17b-diol-3-one (1a) and 17a-ethynyl-1,4-estradiene-10b,17b-diol-3-one

(1b) (identified by LC-ToF-MS and 1H NMR spectroscopy). Significantly, this product

mixture displayed a slightly higher estrogenicity than EE2 itself, as determined by the YES

bioassay. Upon the addition of further aliquots of FeIII-B* (to give a FeIII-B* concentration of

500 nM) and H2O2 (to bring the concentration up to 4 mM assuming the final concentration

had dropped to zero) to this reaction mixture the amounts of 1a and 1b slowly decreased to

zero over a 60 min period as they were oxidised to unidentified products that showed no

estrogenicity. Thus, partial oxidation of EE2 gave products that have slightly increased

estrogenicity, whereas more extensive oxidation by the advanced catalytic oxidation

system completely removed all estrogenicity. These results underscore the importance of

controlling the level of oxidation during the removal of EE2 from water by oxidative

processes.

ª 2012 Elsevier Ltd. All rights reserved.

l & Environmental Engineering, The University of Auckland, Private Bag 92019, 3 Grafton4 9 923 4512; fax: þ64 9 373 7462.z (N. Singhal).ier Ltd. All rights reserved.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 86310

1. Introduction peroxide is a readily manufactured oxidant that is free of both

Endocrine disrupting compounds (EDCs) are subclass of

organic contaminants that have been widely detected in

wastewater and surface waters (Kolpin et al., 2002). Among

them, the synthetic estrogen 17a-ethinylestradiol (EE2),

a component of oral contraceptives since 1970s has been well

documented in the literature to cause endocrine disruption in

non-target species (Sumpter and Johnson, 2005). In vivo tests

show that EE2 is approximately 10-fold more potent than the

female sex hormone 17a-estradiol (E2) in inducing vitellogenin

in male sheepshead minnow (exposed for 16 days) (Thorpe

et al., 2003). On the basis of in vivo estrogenic potency on the

life-cycle of the fathead minnow (Pimephales promelas) (Lange

et al., 2001) EE2 was suggested as one of the most important

EDCs in effluents from wastewater treatment plants (WTPs)

(Johnson and Sumpter, 2001). The predicted no effect

concentration (PNEC) is proposed at 0.1 ng L�1 (0.1 ppt) for EE2based upon the fertilization success in full life cycle study

with zebra fish in a report commissioned by the Environment

Agency of England and Wales (Young et al., 2002). Effluents

from conventional WTPs often contain EE2 in excess of its

PNEC, e.g. typical sewage effluent EE2 concentrations were

found to be ca. 0.5 ng L�1 (Hashimoto et al., 2007).

Several studies have shown that both ozonation and

chlorination are useful technologies for the oxidation of

pharmaceuticals and EDCs during water treatment (Lee et al.,

2008; Lee and Gunten, 2008; Ternes et al., 2002). Ozone treat-

ment would be effective but expensive, whereas chlorination

would be cheaper but can potentially produce various unde-

sirable chlorinated disinfection by-products (DBPs)

(Crittenden et al. 2005). It has been demonstrated by E-screen

assays that showed oxidation with ozone resulted in the rapid

transformation of EE2, reaching a stabilized estrogenic level

(proliferative effect ¼ 2.5) in 10 min, whereas for chlorination

it tookmore than 120 min for elicited estrogenicity to stabilize

(proliferative effect¼ 8.2) (Alum et al., 2004). Potassium ferrate

(Fe(VI)) is an emerging water treatment chemical that can be

dosed directly into water and its effectiveness for the oxida-

tive removal of phenolic EDCs has been confirmed in both

natural water and wastewater (Lee et al., 2005). As a stoichio-

metric oxidant, potassium ferrate oxidizes reduced sulfur-

containing compounds, pharmaceuticals, and endocrine dis-

ruptors (He et al., 2009; Lee et al., 2009). However, because of

its low selectivity, for example it is also highly effective for

phosphate removal, Fe(VI) shows low efficiency overall for the

removal of organic pollutants when applied in the real

wastewater matrices (Lee et al., 2009). More recently it has

been demonstrated that iron complexes of specially designed

tetra-amido macrocyclic ligands (FeIII-TAMLs) can catalyse

oxidations by H2O2 in water and that the resulting oxidation

systems can deeply oxidize and significantly mineralize water

contaminants such as pentachlorophenol, trichlorophenol

(Gupta et al., 2002) and toxic thiophosphate pesticide hydro-

lyzates (Chanda et al., 2006a). Importantly, the degradation

products of these reactions do not appear to present toxicity

concerns. The FeIII-TAML/H2O2 oxidation system is regarded

as “green” because the catalysts themselves are not toxic and

they contain no inherently toxic elements. Also, hydrogen

chlorine and toxic heavy metals.

The FeIII-B*/H2O2 oxidation system has also been shown to

effectively eliminate estrogenicity fromwater contaminated by

natural and synthetic estrogens, which are environmental

persistent and resistant to bacterial degradation (Khunjar et al.,

2011; Snyder et al., 1999). In particular, treatment of an aqueous

solution of EE2 (80 mM) with the FeIII-TAML designated FeIII-B*

(83 nM) and H2O2 (4 mM) for 60 min at pH 10.0 has been shown

to result in >95% removal of EE2 with accompanying loss of

estrogenicity (ca. 95% reduction). In this study we have inves-

tigated the oxidative removal of EE2 by this method in more

detail, specifically the partial EE2 oxidation by FeIII-TAML to

intermediate products with the aim of assessing the change in

overall estrogenicity in sub-optimal degradation environments.

2. Experimental

2.1. Materials and reagents

EE2 (>98%) was obtained from Sapphire Bioscience (New

Zealand). The molecular structure of EE2 is shown in Fig. 1a.

Solvents were high-performance liquid chromatography

grade from Ajax FineChem (New Zealand). H2O2 (30%) was

purchased from Ajax FineChem (New Zealand). The synthesis

of FeIII-B* has been reported in the previous study (Horwitz

Colin et al., 2006) and the structure is shown in Fig. 1b.

2.2. Methodology

Triplicate experiments were performed for each reaction. The

initial conditions of reactions were: 10 mM EE2 dissolved in the

0.01 M Na2CO3/NaHCO3 buffer, 5 nM FeIII-B*, 4 mM H2O2,

shaking at 120 cycles min�1 in an incubation shaker at 25 �C.The degradation of EE2 in the FeIII-B*/H2O2 systemwas studied

at pH 8.40e11.00. The reactions were carried out for each time

point (0, 2, 5, 10, 15, 30 and 60 min) in 50 mL buffer in different

amber glass reagent bottles. Previous studies have shown that

the FeIII-TAML catalysts undergo self-destruction at pH 11 and

25 �C with half-lives of less than 30 min (Chanda et al., 2006b),

suggesting that the FeIII-TAML catalyst was exhausted after

1 h of reaction. Therefore, after the reaction at pH 11.00 had

continued for 60min (designated reaction I) an extra aliquot of

H2O2 and FeIII-B* in the same buffer solution was added to the

reaction in succession to make the [FeIII-B*/H2O2] either 5 nM/

4 mM (designated reaction II), 50 nM/4 mM (reaction III) or

500 nM/4 mM (reaction IV) (assuming the concentrations of

these reactants were zero before the addition). Samples were

taken from reactions II, III and IV at total times 65, 70, 90 and

120min. Triplicate reactionswere carried out for each of these

sampling time points. At the termination of reactions, cata-

lase (12,000 units which is 60 times the concentration capable

of destroying 4 mM H2O2 in 1 min) was added, and then the

sampleswere acidified to pH 3.5 by addition of 4MHCl and left

to stand a further 1 min. This destroyed residual hydrogen

peroxide and deactivated the FeIII-B* through acid-induced

demetalation (Shappell et al., 2008). Control reactions were

carried out in the absence of either H2O2 or FeIII-B* or both to

OH

C C

HO

12

3

45

67

8

910

1112

1314

15

16

17 H

H3C

H

H

H

a b

Fig. 1 e Chemical structures of a: EE2 and b: FeIII-B*.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 8 6311

quantify the background decay of EE2 and a blank reaction

(FeIII-B* plus H2O2 in the buffer but without EE2) was also

evaluated.

2.3. Analytical methods and characterization

2.3.1. Liquid chromatographyemass spectroscopy (LCeMS)Liquid chromatography and analysis were performed on

a Shimadzu Series LCeMS 2020 system equipped with

a degasser, a binary pump, an autosampler, and a column

oven. The detector of this system was a quadrupole mass

spectrometer. The ionization method used was electrospray

ionization (ESI). The analysis was carried on a C12 polar

reversed-phase column (Synergi MAX-RP, Phenomenex) with

the particle size of 4 mm, the length of 150 mm and inner

diameter of 2.0 mm. The column was guarded with a pre-

column with 4 mm � 3.0 mm C12 security guard cartridge

(Synergi MAX-RP, Phenomenex). The column oven tempera-

ture was set to 30 �C. The mobile phase was 65% organic

solvent (acetoniltrile/methanol, 2/3) in deionized water at

a flow rate of 0.2 mL min�1. Ten microliter samples were

infused into the LC column. Samples were analyzed by MS

under full scanmode (m/z 50e500). Based on the results of full

scans, negative selected ion monitoring (SIM) of m/z 295 and

311 were studied for the residual EE2 and the initially oxidised

intermediate, respectively. Meanwhile, the samples from the

control and blank reactions were also analyzed under the

negative SIM mode of m/z 295 and 311.

2.3.2. EstrogenicityThe estrogenicity of the samples from reactions I, II, III and IV,

as well the sample from the end of reaction at pH 10.21 were

assessed by a four-hour yeast estrogen screen (YES) bioassay

(Balsiger et al., 2010; Liu and Picard, 1998; Picard et al., 1990).

All samples and the blank reaction were evaluated by

measuring the hormone-induced chemiluminescent signal on

Xenogen IVIS-200 optical in vivo imaging system. A sequence

of EE2 solutions at different concentrations (ranging from

8 � 10�5 to 8 mM) was prepared to study the response of

estrogenicity to different EE2 concentration.

2.3.3. Liquid chromatographyetime of flight-massspectroscopy (LCeToF-MS)The solution at the end of reaction 1 and after addition of

catalase and acid as described above was filtered by GF 2 glass

fiber filter, and concentrated by solid phase extraction with

a 500-mg hydrophilicelipophilic balance (HLB) cartridge from

Waters Corp. The cartridges were preconditioned with 5 mL

methanol followed by 5 mL Milli-Q water. The acidified and

filtered solution was loaded onto the preconditioned cartridge

at a flow rate of 10.0 mL min�1. The cartridge was dried under

high vacuum and then eluted with 5 mL methanol/acetone (1/

1, v/v) at a flow rate of 4 mL min�1. The eluate was collected

and the organic solventwas evaporated to dryness by exposure

to a slow stream of nitrogen gas. Methanol (0.5mL) was used to

dissolve the residue and the solution made up to 1 mL with

deionized water. Ten microliters of this concentrated sample

was introduced using flow-injection (FIA) into an ESI source

(positive ionizationmode, capillary voltage�4500V) using pure

nitrogen as drying/nebulizer gas (180 �C, 6 L min�1 drying gas,

nebulizer pressure 3 bar) in a Bruker micro-ToF-QII (Bruker

Daltonics, Bremen, Germany) coupled with a Dionex Ultimate

3000 HPLC with autosampler (Dionex, Germering, Germany)

and the same column used in previous LCeMS analysis. The

mobile phase was 65% organic solvent (acetoniltrile/methanol,

2/3) in deionized water containing 0.1% formic acid at a flow

rate of 0.2 mL min�1. Data obtained for 1a/b are as follows: MS

(ESI, DCM) (m/z): calc. for C20H24NaO3þ, 335.1618, found

335.1616; calc. for C20H25O3þ, 313.1798, found 313.1811.

2.3.4. Nuclear magnetic resonance (NMR)Further structure confirmation of initial intermediates was

performed by Bruker AVIII-400 MHz spectrometer (Bruker

BioSpin), which provided 1H NMR analysis of pure EE2 dis-

solved in CDCl3. A portion of the same sample (dissolved in

methanol) that was used in the LC-ToF-MS analysis was

evaporated to dryness by a slow stream of nitrogen gas, dis-

solved in CDCl3 and the 1H NMR spectrum obtained. 1H NMR

(CDCl3, 400 MHz) data for 1a are: d (pprn) ¼ 7.17 (d, 1H, C1-H);

6.36 (d.d., lH, C2-H); 6.16 (t, lH, C4-H); for 1b are: d (pprn) ¼ 7.10

(d, 1H, C1-H); 6.16 (d.d., lH, C2-H); 6.01 (t, lH, C4-H).

2.4. Calculation of the frontier electron density

The Frontier Electron Density of different EE2 atoms was

calculated using Gaussian 03 program (Gaussian, Inc.) Opti-

mization of the EE2 structure was carried out with the STO-3G

basis set at level of Unrestricted Hartree�Fock (UHF) (Yi and

Harper, 2007). Based on this optimized structure, the highest

occupied molecular orbital (HOMO) was calculated using the

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 86312

samemethod and basis set. The frontier electron density�c4r

can be calculated as � c4r ¼ ½2 �PðCriHOMOÞ2�, where r is the

number of carbon atoms in i: 2s, 2px, 2py, and 2pz.

3. Results and discussion

3.1. Effect of pH on EE2 degradation by FeIII-B*/H2O2

There were minimal changes in the EE2 concentrations in the

control reactions (0.5e0.9%), indicating that the degradation

of EE2 by either H2O2 or FeIII-B* alone was negligible. The time-

dependent degradation kinetics of EE2 when treated with the

FeIII-B*/H2O2 system was described using the first-order decay

kinetics Shappell et al., 2008, C/C0 ¼ S/S0 ¼ e�kt, where C0

(mg L�1) is the initial concentration, C is the concentration at

time t (min), S0 is the initial peak area corresponding to the

initial concentration C0, S is the peak area corresponding to

the concentration C at time t (min), and k (min�1) is the

apparent decay coefficient. Plots of residual EE2 (determined

by LCeMS) vs time at different pH values are shown in Fig. 2a

with first order decay curves fitted to the data. Since R2 ranges

from 0.94 to 0.99, first-order decay kinetics describes EE2degradation in the FeIII-B*/H2O2 system well at pH 8.40e11.00.

A plot of the EE2 decay coefficients vs pH is presented in

Fig. 2b, and this shows that the decay coefficient of EE2 reaches

a maximum around pH 10.21 and then declines at higher pHs.

The half-life (t½) of EE2 when treated with the FeIII-B*/H2O2

oxidising system at pH 10.21 is 2.1 min.

The rates of oxidation by FeIII-B*/H2O2 are pH dependent in

a manner that has been rationalized mechanistically (Ryabov

and Collins, 2009). The resting state of the catalyst in water is

the bis(aqua)iron(III)-TAML complex [Fe(H2O)2B*]� and this

complex is a polyprotic acid (Ghosh et al., 2003). Reaction of

a deprotonated form of this complex with hydrogen peroxide

to form the strongly oxidising and highly reactive interme-

diate complex is usually a kinetically important step in cata-

lytic oxidation reactions catalysed by Fe-TAMLs. Although an

Fe(V) ¼ O complex has been characterized as the active

a

0 20 40 60

0

20

40

60

80

100

pH 8.40

pH 9.15

pH 9.51

pH 10.21

pH 11.00

pH 10.67

Time (min)

EE

2 R

esid

ual (%

)

Fig. 2 e a: Degradation of EE2 at pH 8.40e11.00 by FeIII-B*/H2O2,

10 mM, FeIII-B* was 5 nM, and H2O2 was 4 mM. Mean percent of

and results were presented by fitting to the first-order decay ki

intermediate in non-aqueous media (de Oliveira et al., 2007)

the precise nature of the strongly oxidising active interme-

diate complex(es) in water is still not known with certainty. It

has been determined previously that for the concentration

ranges of the catalyst and hydrogen peroxide that were used

in this study no important catalytic decomposition of

hydrogen peroxide to oxygen and water occurs (i.e. catalase-

like activity is unimportant) (Ghosh et al., 2008).

3.2. Results of LCeMS analysis

Samples of reaction Iwere analysed using a single-quadrupole

LCeMS under full scan monitoring in the range m/z 50e500.

Two main peaks were observed in the selected ion chro-

matogram at m/z 311 (3.6 min) and m/z 295 (6.8 min) (Fig. 3).

The chromatograms were obtained under negative SIMmode,

detecting the molecular ions M-H�, which were denoted M312

andM296, for the peaks at 3.6min and 6.8min, respectively, to

reflect their nominal molecular weights. To show that M296

corresponded to EE2, standard EE2 was infused into the LCeMS

system under the same LC program and MS conditions. The

results depicted in Fig. 3 reveal that the signal intensity of the

peak at 3.6 min (M312) increased as the reaction time

increased, indicating this material was being accumulated.

The peak at 6.8 min which corresponded to EE2 (M296) had

disappeared by time 60 min, indicating that the concentration

of EE2 at this time was under the detection limit (5 pg abso-

lute). The concentration and estrogenicity profiles of EE2 and

M312 of reaction at pH 10.21 were studied (Fig. 4a). In this case,

the LCeMS results show that M312 concentration reaches the

maximum at 5 min and >99% EE2 is removed. YES assay

shows that at the end of reaction w70% estrogenicity

remained. To slow down the reaction speed, reaction I was

then repeated, but with additional aliquots of FeIII-B* and H2O2

being added after 60 min to bring the FeIII-B*/H2O2 concen-

trations nominally up to 5 nM/4 mM (reaction II), 50 nM/4 mM

(reaction III) and 500 nM/4 mM (reaction IV). The amounts

added were calculated assuming that all the FeIII-B* and H2O2

had been consumed in reaction I after 60 min. The addition

b

analyzed by LCeMS. Nominal t0 concentration of EE2 was

time 0 concentration, n [ 3 reactions per time point for EE2netics. b: the decay coefficient at pH 8.40e11.00.

Fig. 3 e Chromatogram of the samples collected from reaction 1 (EE2, 10 mM; FeIII-B*/H2O2, 5 nM/4 mM) from 0 to 60 min,

showing selected ion monitoring (SIM) of parent EE2 (peak 2) and intermediates M312 (peak 1). Ten microliter sample is

infused into LC. The mobile phase is 65% organic solvent (acetoniltrile/methanol, 2/3) in deionized water at a flow rate of

0.2 mL minL1.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 8 6313

of more catalyst and hydrogen peroxide led to significant

decreases in the intensity of the signal at 3.6 min (M312) over

60 min (Fig. 4b). In reaction IV which had the highest amount

of added catalyst, this signal disappeared completely within

25 min. These results suggest M312 is an intermediate

degradation product that is formed by partial oxidation of EE2and this product is oxidatively degraded by further treatment

with the FeIII-B*/H2O2 system.

The peak areas of EE2 and compound M312 observed in

each of the reactions I, II, III and IV are plotted in Fig. 4 (left

vertical axis) and the decreases in peak areas of these

compounds were fitted to first-order decay kinetics. The

a

Fig. 4 e The change of peak area (left vertical axis) of EE2 and th

vertical axis) determined by YES assay under different conditio

peak area. The blue curve is the estrogenicity. a: Reactions in 60

11.00. Red and green dot curves are the peak areas of EE2 and M3

estrogenicity of sample collected at 60 min of reaction at pH 10.2

M312 at pH 11.00, respectively. Blue curve is the estrogenicities

addition of FeIII-B* and H2O2 at 60 min: reaction II (re-adding FeI

50 nM/4 mM), and reaction IV (reading FeIII-B*/H2O2, 500 nM/4 m

colour in this figure legend, the reader is referred to the web ve

values of the apparent decay coefficients (k) for EE2 and M312

were determined and these are summarized in Table 1 with

corresponding R2 and t½ values. Clearly, with constant

hydrogen peroxide concentrations the degradation rate of

M312 accelerates as the FeIII-B* concentrations become higher.

3.3. Estrogenicity of residuals

By YES assay, the total estrogenicities of the residuals from

reactions I, II, III and IV at different sampling times are pre-

sented in Fig. 4 (right vertical axis). The residuals from reac-

tion I showed a slight increase in the total estrogenicity, even

b

e intermediate product M312 and the estrogenicities (right

n. The red curve is EE2 peak area. The green curve is M312

min with initial [FeIII-B*/H2O2][ 5 nM/4mM at pH 10.21 and

12 at pH 10.21, respectively. The blue diamond point is the

1. Red and green solid curves are the peak areas of EE2 and

at pH 11.00. b: The continue reactions at pH 11.00 withII-B*/H2O2, 5 nM/4 mM), reaction III (re-adding FeIII-B*/H2O2,

M), respectively. (For interpretation of the references to

rsion of this article.)

Table 1 e The first-order decay coefficients and half life of EE2 and M312 in FeIII-B*/H2O2 system at pH 11.00.

EE2 (0e60 min) M312 (60e120 min)

Reactiona Reaction I Reaction II Reaction III Reaction IV

k (min�1) b 0.048 � 0.0065 0.014 � 0.0013 0.040 � 0.010 0.167 � 0.016

R2 0.97 0.98 0.93 0.99

t½ (min) 14.5 49.5 17.3 4.2

a Reaction I: initial [FeIII-B*]/[H2O2], 5 nM/4 mM; reaction II: extra aliquots of FeIII-B*/H2O2 added to reaction II at time 60 min to bring

concentrations nominally to 5 nM/4 mM; reaction III: extra aliquots of FeIII-B*/H2O2 added to reaction I at time 60 min to bring concentrations

nominally to 50 nM/4 mM; reaction IV: extra aliquots of FeIII-B*/H2O2 added to reaction I at time 60min to bring concentrations nominally to 500

nM/4 mM.

b Mean � SD of triplicate experiments.

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 86314

though more than 99% of the EE2 had been removed by the

treatment with FeIII-B*/H2O2. By calculating the t-statistics in

the LINEST function, the absolute t-observed value of M312 is

1.16, which is larger than 0.65, the absolute t-observed value of

EE2, suggesting that M312 contributes to estrogenicity more

than EE2. A similar phenomena of prolonged elicitation of

estrogenicity at an elevated level using E-screen assay was

also reported during a study of EE2 chlorination (Alum et al.,

2004). The residuals from reactions II and III showed only

slightly reduced estrogenicity after 60 min, even though these

treatments reduced the initial levels of M312 by 54% and 77%

respectively. Clearly, the combination of oxidized products

(unidentified) that were formed during these treatments still

retained significant levels of estrogenicity that were not much

reduced from that of the EE2 originally present in solution. In

contrast, on addition of 500 nM/4 mM FeIII-B*/H2O2 at the end

of reaction I (i.e. reaction IV) more than 99% of M312 was

removed and the estrogenicity of the residuals decreased

sharply with the final estrogenicity similar to that of EE2 at

a concentration of only approximately 0.08 nM. Meanwhile,

no estrogenicity was found in the blank reaction (FeIII-B* plus

H2O2 in the buffer but without EE2). These results indicate that

the FeIII-B*/H2O2 oxidation system can effectively remove

estrogenicity from solutions containing EE2. However, a key

finding is that in oxidative treatments aimed at destroying EE2the level of oxidation is of crucial importance. Significantly, at

low levels of oxidation, measurements of the EE2 levels

remaining after oxidative treatment provide no indication of

the estrogenicity of the residual products in solution.

3.4. Initial oxidation products formed on treatment ofEE2 with FeIII-B*/H2O2

The high resolution LCeToF-MS results show the single

molecular ion for M312 which corresponds to a molecular

formula of C20H24O3 and therefore this product contains just

one more oxygen atom than EE2 (C20H24O2). A sample of M312

was obtained by chromatographic purification of the products

formed on carrying out reaction I on a semi-preparative scale.

The 1H NMR spectrum of M312 was obtained and the 5e8 ppm

region is shown in Fig. 5b. For comparison the 1H NMR spec-

trumofEE2 is presented inFig. 5a.Thebiggest changes fromthe

spectrum of EE2 are observed in the chemical shifts of the

protons in Ring A. Furthermore, two sets of closely similar

signals areobserved forRingAprotons indicating that twovery

similar products are present in the sample. The larger set of

signals in this region (at 7.10 (d, 1H, C1-H), 6.16 (d.d., lH, C2-H)

and 6.01 (t, lH, C4-H)) correspond almost exactly to those re-

ported for 17a-ethynyl-1,4-estradiene-10b,17b-diol-3-one (1b)

(Sedee and van Henegouwen, 1985), as do the other signals in

the 1HNMRspectrum.This, in conjunctionwith theMS results,

confirms the identity of the larger of the two main products

present as 1b. The smaller set of coupled signals (at 7.17 (d, 1H,

C1-H), 6.36 (d.d., lH, C2-H) and 6.16 (t, lH, C4-H)) is tentatively

assigned to the epimer, 17a-ethynyl-1,4-estradiene-10a,17b-

diol-3-one (1a) on the basis that the chemical shifts of these

signals are similar but distinct from thoseof the corresponding

signals in 1b while the coupling constants remain the same,

and the high resolution MS results indicate the molecular

formula of this product is identical to that of 1b.

The electron density values for the different C atoms of EE2were calculated and the results are summarized in Table 2. EE2has a single aromatic ring (A), and it also contains the fused

saturated cyclic rings B, C, and D (Fig. 1a). Ring D carries

a hydroxyl group and an acetylide group located at C17. As

expected, it was found that the electron density associated

with Ring A is significantly higher than that in the other rings

based on the frontier electron density (FED) calculation. The

electron density at C10 of EE2 is 0.376, which is the highest

value for all the atoms in EE2. Given that there is relatively little

steric hindrance to attack by electrophiles at Ring A, it can be

anticipated that C10 should therefore be the most favorable

site for attack. Although we have no evidence of the mecha-

nism by which EE2 is oxidized by the FeIII-B*/H2O2 system, the

formation of 1a and 1b as initial products is not inconsistent

with attack at C10 of EE2 by an oxygen-based electrophile in the

first step. It is noteworthy that in a related study the FEDs for

the E2 molecule have been evaluated and it was found that the

FEDwas higher at the phenol moiety, especially at the C10 and

C3 atoms (the values were 0.508 and 0.364, respectively). Based

on these results Ohko et al. (Ohko et al., 2002) proposed that

the C10 or C3 atoms of E2 should be the sites at which the first

electron is extracted upon oxidation. The EE2 molecular

structure is similar to E2, and the ethynyl group does not affect

the frontier electron densities much, suggesting that EE2should behave in a similar manner. Using electronic theory to

model the biotransformation of EE2 under aerobic conditions,

Barr et al. also proposed that Ring A is cleaved before Ring B

(Barr et al., 2011). The present study uses the UHF method to

calculate the FED of EE2. It has been reported that in the

calculation of the theoretical momentum profiles the use of

simple basis sets at the HartreeeFock (HF) level with the

Fig. 5 e 1H NMR spectra of a: pure 0.5 mM EE2 and b: the sample at the end of reaction I (EE2, 10 mM; FeIII-B*/H2O2, 5 nM/4 mM)

treated by solid phase extraction. The solvent is CDCl3. The1H NMR data in b are: compound 1a, d (pprn)[ 7.17 (d, 1H, C1-H);

6.36 (d.d., lH, C2-H); 6.16 (t, lH, C4-H); compound 1b, d (pprn) [ 7.10 (d, 1H, C1-H); 6.16 (d.d., lH, C2-H); 6.01 (t, lH, C4-H). NB I

suggest you place this data in the experimental section and include all the proton signals you observed for themixture of 1a

and 1b.

Table 2 e The electron density values of Different C Atomof EE2 by Gaussian 03 program.

Atom no. Electron density

C1 0.089

C2 0.146

C3 0.264

C4 0.099

C5 0.087

C10 0.376

All of other electron density values <0.01

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 8 6315

simple and restricted minimum basis STO-3G and the 4e31G

split valence basis sets was in qualitative agreement with

experimental data and with the general trends of the observed

momentum profiles, while the shapes of the intensity distri-

butions and the locations of the maxima were significantly in

error; neither of these simple basis sets was saturated, and nor

did they include the diffuse functions that have been shown

to be essential for accurately calculating momentum distri-

butions (Brion et al., 2001). Although UHF theory is the most

common molecular orbital method for open shell molecules

where the numbers of electrons of each spin are not equal, it

has its drawbacks, such as a single Slater determinant of

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 86316

different orbitals for different spins is not a satisfactory

eigenfunction of the total spin operator-<S2> and that the

electronic state can be contaminated by excited states. Alter-

native methods for calculating the electronic spin density

ratios include the density functional theory (DFT) and second-

order MøllerePlesset perturbation (UMP) method. The unre-

stricted DFT methods (generalized gradient approximation or

hybrid functionals alike) are less affected by the phenomenon

of spin contamination than the HartreeeFock theory and also

give better predictions of the electron spin resonance spectra

of organic radicals (Boone et al., 2003).

Structure-activity relationships for natural, synthetic and

environmental estrogens indicate that the H-bonding ability

of the phenolic ring and the presence of a ring structure as

being the primary contributors to estrogenicity of compounds

(Fang et al., 2001). The results of 1H NMR in the present study

(Fig. 5) show that except for C3 and C10 there is no difference

between themain ring structure of EE2 and the intermediates.

The phenolic ring (Ring A) is therefore not destroyed and the

intermediates would be expected to show estrogenicity.

Investigations of the oxidation of EE2, e.g. by ozone (O3),

oxygen under photosensitizing conditions, hydroxyl radical

($OH), chlorine dioxide, ferrate (Fe(VI)), cytochrome P450,

chlorination and bromination, have all confirmed that reac-

tion initially occurs at Ring A (Alum et al., 2004; Lee et al., 2008;

Lee and Gunten, 2008; Skotnicka-Pitak et al., 2008; Wang et al.,

2004). Oxidative biodegradation of EE2 also proceeds via initial

oxidation of Ring A. Thus, in the biodegradation of EE2 by

Sphingobacterium sp. JCR5 oxidation to estrone (E1) initially

occurs via oxidation of Ring A and this is then followed by ring

opening oxidation reactions on Ring B (Haiyan et al., 2007; Ren

et al., 2007). Also, in studies of the link between nitrification

and biotransformation of EE2 it was shown that Ring A is the

site of electrophilic initiating reactions including conjugation

and hydroxylation (Yi and Harper, 2007). This appears to be

the first time that the epimers 1a and 1b have been isolated as

products obtained directly from the oxidation of EE2 although

compound 1b has been obtained from chemical reduction of

the corresponding hydroperoxide that in turnwas obtained by

the photosensitized oxidation of EE2 by oxygen (Sedee and van

Henegouwen, 1985). Compound 1a/b has been proposed as an

intermediate in the mineralization of EE2 by hydroxyl radicals

generated by the autocatalytic decomposition of ozone (Zhang

et al., 2006) and by microalgae, Scenedesmus quadricauda

(Della Greca et al., 2008).

It has been established that the FeIII-B* catalysed oxida-

tions by H2O2 do not proceed via $OH radicals, i.e. they are

“non-Fenton” oxidations (Ghosh et al., 2008). The Fenton

reaction (Kiwi et al., 1994; Yaping and Jiangyong, 2008) and

other catalysed oxidations by H2O2 (Szabo-Bardos et al., 2011;

Timofeeva et al., 2011) produce �OH radicals. Instead it is

proposed that in the FeIII-B*/H2O2 oxidation system, the active

oxidant is an iron-oxo-based intermediate that is formed

through the interaction of FeIII-B* with H2O2 in the presence of

base (Collins, 2002). This could react with EE2 via amechanism

that involves oxygen atom transfer in a key step, although

other mechanisms that involve electron transfer are also

possible. From our results it would appear that the species

that are formed very early on in the oxidation of EE2 by FeIII-B*/

H2O2 are 1a and 1b. Further oxidation of these intermediates

by this catalytic system probably leads to cleavage of Ring A

(and/or Ring B) with subsequent loss of estrogenicity. Under

the appropriate conditions the FeIII-B*/H2O2 oxidation system

could ultimately lead to mineralization of EE2.

3.5. Significance of FeIII-TAML

An essential green chemistry criterion for catalysts used in

water purification is they must not display any endocrine

disrupting properties themselves. A range of different FeIII-

TAML activators have been demonstrated not to alter tran-

scription mediated by mammalian thyroid, androgen, or

estrogen hormone receptors. This suggests that the FeIII-

TAMLs do not bind to these receptors, thereby reducing

concerns that these catalysts might have endocrine disrupt-

ing activity (Ellis et al., 2010). In the present study, the results

of the YES assay for the blank reaction with FeIII-B* confirmed

that this catalyst has no endocrine disrupting activity. Also,

FeIII-TAML catalysts are not persistent in water, at least under

catalytic operating conditions (Chanda et al., 2006b).

Although this study is conducted at the optimal pH of

10.21, the removal efficiency of EE2 by FeIII-B*/H2O2 system

continues to be high at pH w9 with a half-life of w30 min

(Fig. 2). While the present FeIII-B* functions best at high pH,

there are FeIII-TAML variants being developed that can func-

tion at neutral pH. These activators do not alter the general

mechanism of catalysis by FeIII-TAML activators of peroxides

(Ellis et al., 2010), which involves the interaction of FeIII-TAML

with hydrogen peroxide (H2O2) to form strongly oxidising,

high-valent iron intermediate complexes as the active

oxidants. Differences between the different FeIII-TAML vari-

ants lie in the “oxidising strengths” of the high-valent iron

intermediates, the rates at which these react with substrates

and the longevity of the catalysts during catalytic oxidations.

It is therefore expected that most of the FeIII-TAML variants

follow the same initial oxidative degradation pathways for

substrates such as EE2. The current findings will provide

valuable information relating to the optimization and opera-

tion of practical systems using such catalysts.

The present study highlights the need for integrating the

determination of intermediate metabolites and estrogenicity

assessment in treatment studies. Their significance is illus-

trated via the partial oxidation of EE2 with the FeIII-B*/H2O2

system which results in the formation of a mixture of prod-

ucts that show increased estrogenicity. The intermediates

1a/1b have been previously reported (without identification

of specific isomers) during EE2 degradation by hydroxyl radi-

cals produced via ozone decomposition (Zhang et al., 2006), by

microalgae S. quadricauda (Della Greca et al., 2008), and by

photosensitized decomposition (Sedee and van Henegouwen,

1985); however, these studies did not monitor the change in

estrogenicity of the treated solution.

4. Conclusion

It has been found that the FeIII-B*/H2O2 catalytic oxidation

system under suitable conditions (pH 10.21, [FeIII-B*] 5 nM,

[H2O2] 4 mM) can rapidly oxidize EE2 in aqueous solution (t½for loss of EE2 is 2.1 min) to products that after 1 h show

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 8 6317

around 30% estrogenicity removal, as determined by the YES

bioassay. However, under low levels of oxidation (partial

oxidation) the epimers 17a-ethynyl-1,4-estradiene-10a,17b-

diol-3-one (1a) and 17a-ethynyl-1,4-estradiene-10b,17b-diol-3-

one (1b) are the dominant products formed (as determined by

LCeMS). Under such a condition (exact conditions and pH

perhaps not being that important) the mixture of products

shows increased estrogenic activity, despite the reduction in

the level of EE2; the estrogenicity of the solutions correlated

with the amount of these compounds present. Based on these

observations it is highly likely that partial oxidation by other

oxidation systems might also produce the same products and

hence also result in increased estrogenic activity. These

results emphasize the importance of avoiding low levels of

oxidation (partial oxidation) in any oxidative EE2 removal

protocols and also that reduction in EE2 levels in solution

during oxidative treatments does not necessarily correlate

with a reduction of estrogenic levels.

Acknowledgment

This study was supported by a grant from the University of

Auckland Faculty Research Development Fund. Yeast strains

for the YES assay were kindly provided by Dr Marc B. Cox,

University of Texas at El Paso. The authors thank Brendan

Harvey for assistance with the NMR analysis experiments and

helpful discussions; and Dr. Siouxsie Wiles and Priscila

Dauros for assistance with luminometry for the YES bioassay.

r e f e r e n c e s

Alum, A., Yoon, Y., Westerhoff, P., Abbaszadegan, M., 2004.Oxidation of bisphenol A, 17b-estradiol, and 17a-ethynylestradiol and byproduct estrogenicity. EnvironmentalToxicology 19 (3), 257e264.

Balsiger, H.A., de la Torre, R., Lee, W.-Y., Cox, M.B., 2010. A four-hour yeast bioassay for the direct measure of estrogenicactivity in wastewater without sample extraction,concentration, or sterilization. Science of the TotalEnvironment 408 (6), 1422e1429.

Barr, W.J., Yi, T., Aga, D., Acevedo, O., Harper, W.F., 2011. Usingelectronic theory to Identify metabolites present in 17a-Ethinylestradiol biotransformation pathways. EnvironmentalScience & Technology 46 (2), 760e768.

Boone, A.J., Chang, C.H., Greene, S.N., Herz, T., Richards, N.G.J.,2003. Modeling the spin-dependent properties of open-shellFe(III)-containing systems: towards a computationaldescription of nitrile hydratase. Coordination ChemistryReviews 238, 291e314.

Brion, C.E., Young, J.B., Litvinyuk, I.V., Cooper, G., 2001. Aninvestigation of the HOMO frontier orbital electron densitydistributions of NH(3), the methylamines and NF(3) using DFTand electron momentum spectroscopy. Chemical Physics 269(1e3), 101e106.

Chanda, A., Khetan, S.K., Banerjee, D., Ghosh, A., Collins, T.J.,2006a. Total degradation of fenitrothion and otherorganophosphorus pesticides by catalytic oxidationemploying Fe-TAML peroxide activators. Journal of theAmerican Chemical Society 128 (37), 12058e12059.

Chanda, A., Ryabov, A.D., Mondal, S., Alexandrova, L., Ghosh, A.,Hangun-Balkir, Y., Horwitz, C.P., Collins, T.J., 2006b. Activity-stability parameterization of homogeneous green oxidationcatalysts. Chemistry e A European Journal 12 (36), 9336e9345.

Collins, T.J., 2002. TAML oxidant activators: a new approach to theactivation of hydrogen peroxide for environmentallysignificant problems. Accounts of Chemical Research 35 (9),782e790.

Crittenden, J.C., Trussell, R.R., Hand, D.W., Howe, K.J.,Tchobanoglous, G., 2005. Water Treatment e Principles andDesign, second ed. John Wiley & Sons.

de Oliveira, F.T., Chanda, A., Banerjee, D., Shan, X.P., Mondal, S.,Que, L., Bominaar, E.L., Munck, E., Collins, T.J., 2007. Chemicaland spectroscopic evidence for an Fe-V-Oxo complex. Science315 (5813), 835e838.

Della Greca, M., Pinto, G., Pistillo, P., Pollio, A., Previtera, L.,Temussi, F., 2008. Biotransformation of ethinylestradiol bymicroalgae. Chemosphere 70 (11), 2047e2053.

Ellis, W.C., Tran, C.T., Roy, R., Rusten, M., Fischer, A.,Ryabov, A.D., Blumberg, B., Collins, T.J., 2010. Designinggreen oxidation catalysts for purifying environmentalwaters. Journal of the American Chemical Society 132 (28),9774e9781.

Fang, H., Tong, W.D., Shi, L.M., Blair, R., Perkins, R., Branham, W.,Hass, B.S., Xie, Q., Dial, S.L., Moland, C.L., Sheehan, D.M., 2001.Structure-activity relationships for a large diverse set ofnatural, synthetic, and environmental estrogens. ChemicalResearch in Toxicology 14 (3), 280e294.

Ghosh, A., Mitchell, D.A., Chanda, A., Ryabov, A.D., Popescu, D.L.,Upham, E.C., Collins, G.J., Collins, T.J., 2008.Catalase�peroxidase activity of Iron(III)�TAML activators ofhydrogen peroxide. Journal of the American Chemical Society130 (45), 15116e15126.

Ghosh, A., Ryabov, A.D., Mayer, S.M., Horner, D.C., Prasuhn, D.E.,Sen Gupta, S., Vuocolo, L., Culver, C., Hendrich, M.P.,Rickard, C.E.F., Norman, R.E., Horwitz, C.P., Collins, T.J., 2003.Understanding the mechanism of Hþ-induced demetalationas a design strategy for robust iron(III) peroxide-activatingcatalysts. Journal of the American Chemical Society 125 (41),12378e12379.

Gupta, S.S., Stadler, M., Noser, C.A., Ghosh, A., Steinhoff, B.,Lenoir, D., Horwitz, C.P., Schramm, K.-W., Collins, T.J., 2002.Rapid total destruction of chlorophenols by activatedhydrogen peroxide. Science 296 (5566), 326e328.

Haiyan, R., Shulan, J., ud din Ahmad, N., Dao, W., Chengwu, C.,2007. Degradation characteristics and metabolic pathway of17[alpha]-ethynylestradiol by Sphingobacterium sp. JCR5.Chemosphere 66 (2), 340e346.

Hashimoto, T., Onda, K., Nakamura, Y., Tada, K., Miya, A.,Murakami, T., 2007. Comparison of natural estrogen removalefficiency in the conventional activated sludge process andthe oxidation ditch process. Water Research 41 (10),2117e2126.

He, C., Li, X.Z., Sharma, V.K., Li, S.Y., 2009. Elimination of sludgeodor by oxidizing sulfur-containing compounds withferrate(VI). Environmental Science & Technology 43 (15),5890e5895.

Horwitz Colin, P., Collins Terrence, J., Spatz, J., Smith Hayden, J.,Wright, L.J., Stuthridge Trevor, R., Wingate Kathryn, G.,McGrouther, K., 2006. Feedstocks for the Future. AmericanChemical Society, pp. 156e169.

Johnson, A.C., Sumpter, J.P., 2001. Removal of endocrine-disrupting chemicals in activated sludge treatment works.Environmental Science & Technology 35 (24), 4697e4703.

Khunjar, W.O., Mackintosh, S.A., Skotnicka-Pitak, J., Baik, S.,Aga, D.S., Love, N.G., 2011. Elucidating the relative roles ofammonia oxidizing and heterotrophic bacteria during thebiotransformation of 17 alpha-ethinylestradiol and

wat e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 3 0 9e6 3 1 86318

Trimethoprim. Environmental Science & Technology 45 (8),3605e3612.

Kiwi, J., Pulgarin, C., Peringer, P., 1994. Effect of Fenton and photo-Fenton reactions on the degradation and biodegradability of 2and 4-nitrophenols in water treatment. Applied Catalysis BEnvironmental 3 (4), 335e350.

Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M.,Zaugg, S.D., Barber, L.B., Buxton, H.T., 2002. Pharmaceuticals,hormones, and other organic wastewater contaminants in u.s.streams, 1999�2000: a national reconnaissance.Environmental Science & Technology 36 (6), 1202e1211.

Lange, R., Hutchinson, T.H., Croudace, C.P., Siegmund, F.,Schweinfurth, H., Hampe, P., Panter, G.H., Sumpter, J.P., 2001.Effects of the synthetic estrogen 17a-ethinylestradiol on thelife-cycle of the fathead minnow (Pimephales promelas).Environmental Toxicology and Chemistry 20 (6), 1216e1227.

Lee, Y., Escher, B.I., von Gunten, U., 2008. Efficient removal ofestrogenic activity during oxidative treatment of waterscontaining steroid estrogens. Environmental Science &Technology 42 (17), 6333e6339.

Lee, Y., Gunten, U.v, 2008. Transformation of 17a-ethinylestradiolduring water chlorination: effects of bromide on kinetics,products, and transformation pathways. EnvironmentalScience & Technology 43 (2), 480e487.

Lee, Y., Yoon, J., von Gunten, U., 2005. Kinetics of the oxidation ofphenols and phenolic endocrine disruptors during watertreatment with ferrate (Fe(VI)). Environmental Science &Technology 39 (22), 8978e8984.

Lee, Y., Zimmermann, S.G., Kieu, A.T., von Gunten, U., 2009.Ferrate (Fe(VI)) application for municipal wastewatertreatment: a novel process for simultaneous micropollutantoxidation and phosphate removal. Environmental Science &Technology 43 (10), 3831e3838.

Liu, J.W., Picard, D., 1998. Bioactive steroids as contaminants ofthe common carbon source galactose. Fems MicrobiologyLetters 159 (2), 167e171.

Ohko, Y., Iuchi, K.-i., Niwa, C., Tatsuma, T., Nakashima, T.,Iguchi, T., Kubota, Y., Fujishima, A., 2002. 17b-Estradioldegradation by TiO2 photocatalysis as a means of reducingestrogenic activity. Environmental Science & Technology 36(19), 4175e4181.

Picard, D., Khursheed, B., Garabedian, M.J., Fortin, M.G.,Lindquist, S., Yamamoto, K.R., 1990. Reduced levels of hsp90compromise steroid receptor action in vivo. Nature 348 (6297),166e168.

Ren, Y.-X., Nakano, K., Nomura, M., Chiba, N., Nishimura, O.,2007. Effects of bacterial activity on estrogen removal innitrifying activated sludge. Water Research 41 (14), 3089e3096.

Ryabov, A.D., Collins, T.J., 2009. Advances. In: Rudi van, E.,Colin, D.H. (Eds.), Inorganic Chemistry. Academic Press,pp. 471e521.

Sedee, A., van Henegouwen, G.B., 1985. Photosensitizeddecomposition of contraceptive steroids: a possibleexplanation for the observed (photo)allergy of the oralcontraceptive Pill. Archiv Der Pharmazie 318 (2), 111e119.

Shappell, N.W., Vrabel, M.A., Madsen, P.J., Harrington, G.,Billey, L.O., Hakk, H., Larsen, G.L., Beach, E.S., Horwitz, C.P.,

Ro, K., Hunt, P.G., Collins, T.J., 2008. Destruction of estrogensusing Fe-TAML/peroxide catalysis. Environmental Science &Technology 42 (4), 1296e1300.

Skotnicka-Pitak, J., Garcia, E.M., Pitak, M., Aga, D.S., 2008.Identification of the transformation products of 17 alpha-ethinylestradiol and 17 beta-estradiol by mass spectrometryand other instrumental techniques. Trac-trends in AnalyticalChemistry 27 (11), 1036e1052.

Snyder, S.A., Keith, T.L., Verbrugge, D.A., Snyder, E.M.,Gross, T.S., Kannan, K., Giesy, J.P., 1999. Analytical methodsfor detection of selected estrogenic compounds in aqueousmixtures. Environmental Science & Technology 33 (16),2814e2820.

Sumpter, J.P., Johnson, A.C., 2005. Lessons from endocrinedisruption and their application to other issues concerningtrace organics in the aquatic environment. EnvironmentalScience & Technology 39 (12), 4321e4332.

Szabo-Bardos, E., Somogyi, K., TorT, N., Kiss, G., Horvath, A., 2011.Photocatalytic decomposition of l-phenylalanine over TiO2:identification of intermediates and the mechanism ofphotodegradation. Applied Catalysis B: Environmental 101(3e4), 471e478.

Ternes, T.A., Meisenheimer, M., McDowell, D., Sacher, F.,Brauch, H.-J., Haist-Gulde, B., Preuss, G., Wilme, U., Zulei-Seibert, N., 2002. Removal of pharmaceuticals during drinkingwater treatment. Environmental Science & Technology 36 (17),3855e3863.

Thorpe, K.L., Cummings, R.I., Hutchinson, T.H., Scholze, M.,Brighty, G., Sumpter, J.P., Tyler, C.R., 2003. Relative potenciesand combination effects of steroidal estrogens in fish.Environmental Science & Technology 37 (6), 1142e1149.

Timofeeva, M.N., Hasan, Z., Orlov, A.Y., Panchenko, V.N.,Chesalov, Y.A., Soshnikov, I.E., Jhung, S.H., 2011. Fe-containing nickel phosphate molecular sieves asheterogeneous catalysts for phenol oxidation andhydroxylation with H2O2. Applied Catalysis B Environmental107 (1e2), 197e204.

Wang, B., Sanchez, R.I., Franklin, R.B., Evans, D.C., Huskey, S.-E.W., 2004. The involement of CYP3A4 and CYP2C9 in themetabolism of 17a-ethinylestradiol. Drug Metabolism andDisposition 32 (11), 1209e1212.

Yaping, Z., Jiangyong, H., 2008. Photo-Fenton degradation of 17b-estradiol in presence of a-FeOOHR and H2O2. AppliedCatalysis B Environmental 78 (3e4), 250e258.

Yi, T., Harper, W.F., 2007. The link between nitrification andbiotransformation of 17a-ethinylestradiol. EnvironmentalScience & Technology 41 (12), 4311e4316.

Young, W.F., Whitehouse, P., Johnson, I., Sorokin, N., 2002.Proposed Predicted-No-Effect-Concentrations (PNECs) forNatural and Synthetic Steroid Oestrogens in Surface Waters.Environment Agency R & D Technical Report P2-T04/1;England and Wales Environment Agency, Bristol, U.K., pp.93e95.

Zhang, X., Chen, P.Y., Wu, F., Deng, N.S., Liu, J.T., Fang, T., 2006.Degradation of 17 alpha-ethinylestradiol in aqueous solutionby ozonation. Journal of Hazardous Materials 133 (1e3),291e298.