Formation of Chloroform and
Chlorinated Organics by
Free-Chlorine-Mediated Oxidation of
K R I S T A L . R U L E ,
V I R G I N I A R . E B B E T T , A N D
P E T E R J . V I K E S L A N D *
Department of Civil and Environmental Engineering,
418 Durham Hall, Virginia Polytechnic and State University,
Blacksburg, Virginia 24060-0361
The widely used antimicrobial agent triclosan (5-chloro-
2-(2,4-dichlorophenoxy)phenol) readily reacts with free
chlorine under drinking water treatment conditions. Overall
second-order kinetics were observed, first-order in free
chlorine and first-order in triclosan. Over the pH range of
4-11.5, the kinetics were pH sensitive as a result of
the pH dependent speciation of both triclosan and free
chlorine. Using a Marquardt-Levenberg routine, it was
determined that this pH effect indicates that the dominant
reaction in this system is between the ionized phenolate
form of triclosan and hypochlorous acid (HOCl). The overall
second-order rate coefficient was determined to be
kArO- ) 5.40 ((1.82) × 103M-1s-1. Three chlorophenoxy-
phenols and two chlorophenols were identified by gas
chromatographic-mass spectroscopic analysis. The
chlorophenoxyphenol compounds include two monochlori-
nated triclosan derivatives (5,6-dichloro-2-(2,4-dichloro-
phenoy)phenol and 4,5-dichloro-2-(2,4-dichlorophenoxy)-
phenol) and one dichlorinated derivative (4,5,6-trichloro-
(2,4-dichlorophenoxy)phenol); these species form via
bimolecular electrophilic substitution of triclosan. 2,4-
Dichlorophenol was detected under all reaction conditions
and forms via ether cleavage of triclosan. In experiments
with excess free chlorine, 2,4,6-trichlorophenol was
formed via electrophilic substitution of 2,4-dichlorophenol.
Chloroform formation was observed when an excess of
free chlorine was present. A Hammett-type linear free-
(σ+) was established to correlate the reactivity of HOCl
(log kArO- ) -(10.7 ( 2.2)Σσ+o,m,p+ 4.43). This LFER
was used to obtain estimates of rate coefficients describing
the reactivity of the intermediates 5,6-dichloro-2-(2,4-
dichlorophenoy)phenol (kArO-≈ 6 × 102), 4,5-dichloro-2-(2,4-
(2,4-dichlorophenoxy)phenol (kArO- ≈ 4 × 101).
Triclosan (5-chloro-2-(2,4 dichlorophenoxy)phenol) is a
as toothpastes, acne creams, deodorants, and hand soaps at
concentrations that range from 0.1 to 1%. Although intro-
duced over 30 years ago, the application of triclosan has
increased dramatically over the last 10 years. Currently, it is
toothbrush handles, and athletic clothing, among other
Triclosan is used in many products because it exhibits
antibacterial as well as antifungal and antiviral properties
(1). Until recently, the compound was believed to solely act
as a nonspecific biocide that disrupts bacterial membrane
functionality (2). Since 1998, however, several studies have
indicated that triclosan can act as a site-specific biocide.
These studies, which examined the effects of triclosan on E.
coli (3), Mycobacterium smegmatic (4), and M. tuberculosis
(5), concluded that triclosan preferentially reacts with enoyl
reductase, an enzyme essential to fatty acid synthesis. The
develop resistance, which would render the compound
that the high doses employed in antibacterial goods result
in cell lysis from several simultaneous effects (6). Neverthe-
are similar to those associated with antibiotic resistance (7),
could promote microbial immunity.
quantities of the compound are washed down household
drains and enter sewage systems. Surveys have measured
levels ranging from 0.062 to 21.9 µg/L (8-12). Triclosan
removal within WWTPs varies with the type of secondary
and tertiary treatment employed, with reported removal
percentages between 0 and 100% (9-11, 13, 14). Activated
sludge processes generally have high triclosan removal
percentages of >90% (9, 11, 14), while attached growth
processes have lower removal percentages and are less
consistent (12). A relatively high octanol-water partition
coefficient (log Kow) of 4.8 (15) indicates the tendency of the
compound to sorb to organic material, and thus wasted
(11, 13, 14). Reported WWTP effluent concentrations range
from 0.042 to 22.1 µg/L (12, 13, 16).
The incomplete removal of triclosan via wastewater
treatment, the land-application of triclosan laden biosolids
(11, 13, 14), and leaking sewer pipes (17) result in the
continued release of the compound into the aquatic envi-
ronment. Recent studies have detected triclosan in natural
States Geological Survey surveyed 139 U.S. streams consid-
ered highly susceptible to contamination by numerous
(18). In this study, triclosan was detected in 57.6% of the
streams at a median concentration of 0.14 µg/L and a
maximum concentration of 2.3 µg/L. Reports suggest that
triclosan is readily removed from natural waters via bio-
are in the range of 0.21-0.33 h-1for Mag Brook in England
(12) and 0.06 h-1in Cibolo Creek in Texas (23). These rates,
phase and do not account for its potential accumulation in
sediments. In fact, triclosan has been measured in marine
sediments near a WWTP effluent pipe at levels between 0.27
and 130.7 µg/kg (16).
The fate of triclosan in the environment is significantly
influenced by its pH dependent speciation. The pKa of
* Correspondingauthorphone: (540)231-3568;fax: (540)231-7916;
e-mail address: email@example.com.
Environ. Sci. Technol. 2005, 39, 3176-3185
31769ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 200510.1021/es048943+ CCC: $30.25
2005 American Chemical Society
Published on Web 04/02/2005
triclosan is 7.9 (8), and thus the anionic phenolate form
(described hereafter as phenolate-triclosan) predominates
pH 7.9, the neutral phenolic form is the primary species.
Both the photodecomposition (8, 13, 22) and the MnO2
surface-catalyzed oxidation (24) of triclosan are extremely
sensitive to triclosan speciation.
Some source waters used for drinking water supply have
been found to contain triclosan (25, 26). In general, con-
plants rarely analyze for PPCPs and when they do, concen-
no U.S. federal regulations that necessitate periodic moni-
toring for the presence of PPCPs in drinking water, and the
FDA only requires testing for a particular PPCP if the
1 µg/L and 100 µg/kg, respectively (27). Recent cases have,
however, reported the presence of PPCPs, in general (25, 28,
A complete understanding of the fate of PPCPs within water
treatment plants and in treated drinking waters is therefore
necessary to address the impacts of these micropollutants
on the safety of drinking water.
In general, PPCPs are not effectively removed by coagu-
lation-flocculation and thus for many treatment plants the
is well established. Ozone, in particular, is effective at
oxidizing many PPCPs, although the toxicological effects of
the products formed from these reactions are not well
understood (27, 31, 32). Free chlorine is not as strong of an
PPCPs. A recent study has shown that highly chlorinated
analogues are produced when free chlorine interacts with
and ecological ramifications of these chlorinated analogues
is a subject of concern.
the reactions between free chlorine and compounds con-
reactions of chlorine and triclosan. Numerous studies have
examined the kinetics and products of the reactions that
occur between free chlorine and phenolic compounds (34-
40). When phenols undergo bimolecular electrophilic sub-
stitution (SE2), the -OH/-O-substituent of the phenol ring
activates its ortho- and para- positions toward electrophilic
attack by oxidants such as chlorine (40, 41). Thus, when free
chlorine reacts with phenol, 2-chlorophenol and 4-chlo-
rophenol are the initial products. These species are further
chlorinated to form 2,6-dichlorophenol or 2,4-dichlorophe-
nol, both of which are further chlorinated to produce 2,4,6-
trichlorophenol (38, 40). In the presence of excess free
to produce the intermediate 2,6-dichloro-p-benzoquinone.
of chlorinated carboxylic acids (34, 35) and trihalomethanes
Although two prior studies (42, 43) examined the chlo-
rination of triclosan, the applicability of these studies to
drinking water treatment conditions is unknown. In these
previous studies, the reactions between triclosan and free
chlorine resulted in production of two monochlorinated
triclosan intermediates, a dichlorinated intermediate, 2,4-
dichlorophenol, and 2,3,4-trichlorophenol. Unfortunately,
one of these studies did not provide quantitative product
measurements (42) and the other was performed using
triclosan impregnated fabric as the source material (43).
Furthermore, neither study examined the kinetics of the
reactions and the prospect that triclosan could act as a
precursor to chloroform. At this time, there have been no
comprehensive investigations examining the reactions be-
was to characterize the kinetics and products of triclosan-
free chlorine reactions under conditions typical of drinking
Materials and Methods
Reagent grade water was purified by deionization and
it in a 10% nitric acid water bath and then in a concentrated
chlorine bath. Triclosan was purchased from Aldrich (>98%
purity) and was used without further purification. Stock
triclosan solutions were prepared by dissolving 100 mg
triclosan in 50 mL of reagent grade methanol. Stocks of free
dibromopropane were purchased from Chem Service. The
pH measurements were obtained with a Fisher Scientific
model 60 pH meter coupled with a Thermo-Orion Ross
PerpHect Combination Electrode.
performed using reagent grade water containing 2 mM
sodium bicarbonate pH buffer. Sodium hydroxide, hydro-
pH. Kinetic experiments were conducted in 40-mL screw-
top amber vials containing 25 mL of free-chlorine solution
triclosan concentrations ranged from 2.5 to 27.6 µM (0.72-
8.0 mg/L). Chlorine concentrations were determined using
the DPD photometric method (44).
stock into a reaction vial using a Cheney Adaptor equipped
syringe. The final methanol concentration in the reaction
vials never exceeded 0.2% and was therefore below the level
where cosolvent effects occur (45). Control experiments
indicate that this concentration of methanol does not exert
which free-chlorine decay was monitored, the free chlorine
aliquots of N,N-dimethyl-p-phenylenediamine (DPD) indi-
cator (4.19 mM) and 1.5 mL of phosphate buffer (0.507 M
PO43-). The vessel contents were mixed and the indicator
color was allowed to develop for 1 minute. Absorbance
readings at 515 nm were then compared to a standard curve
to determine the free-chlorine concentration. The rate
constant for the DPD-free-chlorine reaction (1.4-1.7 s-1at
the phosphate buffer pH of 6.2; ref 46) is considerably larger
than those determined for the chlorination of triclosan.
and free chlorine. Overall reaction progress was determined
by measuring the free-chlorine concentration as a function
of time. Each measurement was obtained in triplicate.
Triclosan and Daughter-Product Analysis. Samples for
quantification of triclosan and its nonvolatile daughter
products were quenched with a 3× molar excess of sodium
sulfite. This quenching agent was unreactive toward both
triclosan and its daughter products. The quenched samples
were adjusted to pH 2 with 0.1 M HCl and solid phase
extracted with 3M Empore High Performance SDBS Extrac-
tion Cartridges. Prior to use, each cartridge was rinsed with
1 mL acetone and dried under vacuum. Conditioning of the
VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93177
methanol and 1.0 mL reagent grade water; care was taken
to avoid drying out the solid phase during pretreatment. An
aliquot of 20 mL was drawn through the cartridge at a rate
of 5 mL/min and samples were eluted with 1 mL acetone.
Following solid-phase extraction, triclosan and daughter
phenolic compounds were derivatized with pentafluoroben-
zyl bromide (PFBBr; ref 47). Aliquots of 100 µL of 5% PFBBr
in acetone and 100 µL 10% aqueous potassium carbonate
were spiked into the acetone eluates. The sample vials were
reaction period was determined to be sufficient for the
complete derivatization of triclosan. After cooling, the
at that time, 1.0 mL methylene chloride and 5 µL internal
standard (1040 mg/L 1,3,5-tribromobenzene) were injected
into the sample vials.
derivatized phenolic daughter products was performed on
an Agilent 6890/5973 system containing a DB-5ms GC-
column (Agilent Technologies, 30 m × 0.25 mm, film
thickness ) 0.25 µm). Helium served as the carrier gas with
a column flow rate of 1.3 mL/min. After being held at 70 °C
for 1.5 min, the temperature was ramped to 160 °C at 20 °C/
min, followed by a second ramp at 8 °C/min to 280 °C. The
temperature was held at 280 °C for 1 min prior to oven cool
down. Pulsed splitless injection was employed with a pulse
pressure of 206.8 kPa (1.1 min) and a 1.0-min purge time
delay. An aliquot of 1 µL was injected, and the samples were
run in full scan mode (range m/z ) 80-550). Derivatized
triclosan was identified on the basis of its elution time of
a fragment ion at m/z 252 (M+-Cl-PFB). Derivatized
chlorophenols were identified on the basis of their elution
times and major ions were determined using purchased
were performed to monitor chloroform formation under
headspace-free conditions in 40-mL amber screw-top vials.
quenched with sodium sulfite. Chloroform and bromodi-
chloromethane were quantified using either a liquid-liquid
extraction (LLE) procedure according to U.S. EPA Method
according to U.S. EPA Method 502.2.
Liquid-liquid pentane extraction was used for THM
of 2 mL of pentane to surfactant-containing soap samples
often resulted in formation of an emulsion of pentane and
water, these samples were centrifuged at 2700 rpm and 2 °C
for 30 min to separate this emulsion. This step made it
possible to isolate 1 mL of the organic pentane layer.
Following this step, the chloroform and bromodichloro-
methane concentrations in the pentane extracts were
quantified using an Agilent 5890 gas chromatograph. The
ID × 30 m) and an electron capture detector (ECD). The
instrument was set according to the following parameters:
The initial oven temperature was 30 °C and was held for 9
min. The temperature was then increased to 40 °C at a rate
response of the ECD detector to chloroform and bromodi-
chloromethane was calibrated using serial dilutions of
were purged in a Tekmar 2016 Purge and Trap Autosampler
attached to a Tekmar 3000 Purge and Trap Concentrator
equipped with a Supelco VOCARB 300 Purge Trap K. A
Tremetrics 9001 gas chromatograph with a Tracer 1000 Hall
DB-624 column (30 m × 0.53 mm, film thickness ) 3 µm)
and the carrier gas was nitrogen. Prior to analysis, 5-mL
samples were spiked with 10 µL of 10 mg/L 1,2-dibromopro-
pane internal standard. Samples were purged for 7 min and
then were baked from the trap at 250 °C for 10 min. The GC
temperature program involved an initial temperature of 45
°C held for 3 min, followed by a temperature ramp to 200 °C
Packard Series II Integrator.
Results and Discussion
Experiments show that triclosan and free chlorine readily
react and that the kinetics are a function of the solution pH
was negligible, and when free chlorine was absent, triclosan
the reactions between triclosan and free chlorine, experi-
for free chlorine loss is appropriate. Pseudo-first-order rate
constants (kobs; s-1) were determined at several pH values
using the method of initial rates. The linear portion of the
kobs (percent decays for each experiment are tabulated in
Supporting Information Table S1 along with the corre-
sponding regression coefficients). The kobsvalues were then
decrease as pH increases above pH 8. This pH effect can be
rationalized on the basis of the pH dependent speciation of
both free chlorine and triclosan:
On the basis of this reaction mechanism, the loss of
triclosan and free chlorine can be described as follows:
where [FC]Tand [triclosan]Trepresent the total concentra-
tions of free chlorine (i.e., [HOCl] + [OCl-]) and triclosan
([triclosan]T) [triclosan] + [phenolate-triclosan]), respec-
shown in eq 4 for both free chlorine and triclosan are
appropriate (Supporting Information Figures S1-S4). For
an excess triclosan concentration, eq 4 can be simplified:
where kobs) kArO-[phenolate-triclosan].
Collected kobsdata for pH values between 6 and 11 was
evaluated using a Marquardt-Levenberg least-squares mini-
of 5.40 ((1.82) × 103M-1s-1for kArO- was determined. As
shown in Figure 2, this single parameter provides a good fit
to the collected data. The rate constant (kArO-) for triclosan
is given in Table 1 along with literature values for the
HOCl 79 8
triclosan 79 8
Ka,triclosanphenolate-triclosan + H+
phenolate-triclosan + HOCl9 8
31789ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005
the kArO-rate constant for HOCl and phenolate-triclosan is
within the range of values obtained for other substituted
When developing the model for this system, other
potential pH dependent reactions were considered in ad-
dition to reactions 1-4; on the basis of alternate model fits,
however, the reactions of OCl-with triclosan, OCl-with
phenolate-triclosan, and HOCl with triclosan were deemed
insignificant. For phenols, the reactivity of OCl-is typically
negligible in comparison to HOCl (37-39). Furthermore,
phenolic compounds, such as triclosan, are generally more
reactive upon deprotonation. This effect occurs because O-
is better at activating the aromatic ring toward substitution
reactions than OH (48). A reaction between HOCl and the
neutral form of substituted phenols has been included in
yet excluded in others (48). Under our conditions, inclusion
of such a term was unnecessary. Some studies on the
chlorination of phenols have noted the presence of an acid-
low pH values, and these studies have included such terms
in their reaction models (37, 38). Although our data suggest
that there could be a slight catalytic effect of H+below pH
the kinetic characterization to become difficult and we thus
refrain from including such a reaction in the model.
FIGURE 1. Experimental results and model predictions for triclosan and free-chlorine loss as a function of pH. (A) pH 4, (B) pH 7, (C) pH
10. Reaction conditions: [triclosan]0) 5.05 µM; [free chlorine]0) 14.2 µM; [NaHCO3] ) 2 mM.
FIGURE 2. Observed pseudo-first-order rate constants (kobs) vs pH
conditions: [free chlorine]0 ) 2.33-3.23 µM; [triclosan]0 )
27.5 µM; [NaHCO3] ) 2 mM.
VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93179
cause an increase in the apparent reaction rate at pH values
shifts toward elemental chlorine (Khydrolysis ) 4.0 × 10-4;
than HOCl, and its presence could result in faster reaction
rates. At the outset of these experiments, HCl was used for
pH adjustment below pH 7. To determine if reaction rates
was conducted at pH 4 and pH 5 where the pH was adjusted
with H2SO4. As shown in Figure 2, the measured kobsvalues
are considerably lower when H2SO4is used to lower the pH
effects of chloride addition.
Because the reaction kinetics are complicated at low pH
values by the presence of Cl-and the formation of Cl2, only
kobs values obtained above pH 6 were employed in the
development of the model. When H2SO4was used to adjust
5 are close to those predicted by the model. The continued
H2SO4for pH adjustment and the model is presumed to be
a result of the ∼4.5 µM chloride present in the solutions
because of the chloride content of the stock sodium hy-
pochlorite. Were no chloride present in these experiments,
the reaction rate constants would more closely align with
the model prediction.
Kinetic Model Validation. To validate the performance
under excess free-chlorine conditions at pH values of 4, 7,
free chlorine and triclosan with model predictions for both
species. As shown, the model predictions for both triclosan
and free chlorine at pH 4 and pH 10 correlate reasonably
well with the experimental data. At pH 7, however, the
decay even though triclosan removal is well predicted.
Underprediction of free-chlorine loss by the model at this
pH value is a result of the chlorine demand exerted by the
Product Identification and Kinetic Evaluation. A GC
chromatogram illustrating the intermediates and products
free chlorine is shown in Figure 3. Two monochlorinated
triclosan intermediate (denoted C) were identified on the
basis of mass spectral analysis. Intermediates A and B had
identical mass spectra with a molecular ion at m/z 504 and
fragment ions at m/z 323 (M+-PFB) and 286 (M+-PFB-Cl).
ions at m/z 357 (M+-PFB) and m/z 322 (M+-PFB-Cl).
Chlorine isotope analysis of the mass spectra for the
that intermediates A and B were triclosan plus one chlorine
and that intermediate C was triclosan plus two chlorines
(Supporting Information Figure S5 and Table S2). On the
basis of the mass spectrum and a review of the literature
(42), we identify products A and B to be the isomers 5,6-
dichloro-2-(2,4-dichlorophenoxy)phenol and 4,5-dichloro-
2-(2,4-dichlorophenoxy)phenol and the dichlorinated in-
The reaction solutions were monitored for several di- and
trichlorophenol daughter products, but only 2,4-dichloro-
phenols were detected when triclosan reacted under excess
free-chlorine conditions. 2,4-Dichlorophenol was identified
on the basis of its elution time and its molecular ion at m/z
phenol was identified using its elution time as well as the
molecular ion atm/z 376 and fragment ion atm/z 197 (M+-
and thus they have not been quantified at this time. The
formation and decay of these intermediates, however, is a
B forming and then decaying. Within the time scale of these
measurements, however, the concentration of intermediate
C continually rises. Over longer time scales, intermediate C
eventually decays (Supporting Information Figure S6). The
delayed rate at which intermediate C is degraded indicates
and intermediates A and B.
Unlike the kinetics of triclosan decay and the formation
was greatest at circumneutral pH, 2,4-dichlorophenol ac-
cumulated fastest at pH 4 (Figure 4D). This behavior is
consistent with the known reactivity of 2,4-dichlorophenol
at acidic pH values. At a pH of 4, only 0.014% of 2,4-
dichlorophenol exists in the reactive phenolate ion form,
and thus once 2,4-dichlorophenol is formed it is relatively
is believed to occur via oxidative cleavage of the carbon-
ring in triclosan. Production of 2,4,6-trichlorophenol is
delayed relative to 2,4-dichlorophenol formation and is pH
TABLE 1. Rate Constants for the Electrophilic Substitution of Triclosan and a Series of Chlorinated Phenols
kArO-(M-1s-1) half-life (t1/2)b
2.19 ((0.08) × 104
2.17 ((0.33) × 103
3.16 ((0.22) × 103
2.42 ((0.08) × 103
3.03 ((0.09) × 102
1.94 ((0.08) × 102
1.28 ((0.07) × 101
5.40 ((1.82) × 103
6 × 102
3 × 102
4 × 101
estimated using LFER
estimated using LFER
estimated using LFER
aΣσ+o,m,pvalues calculated using literature values (52) for σ+mand σ+p. σ+owas estimated using (53): σ+o) 0.66 σ+p.bEstimated by assuming
pseudo-first-order conditions with a free-chlorine excess. pH 7; [free chlorine]0) 14.1 µM.cEstimated by assuming that the σ+pvalue for -OC6H5
() -0.5; ref 54) is the same as that for -OC6H3Cl2.dReference 37.eReference 38 as cited by ref 37.
H++ Cl-+ HOCl T Cl2+ H2O (6)
31809ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005
of 2,4-dichlorophenol toward free chlorine confirmed previ-
ously reported results (38, 40) that indicated 2,4,6-trichlo-
rophenol forms via chlorination of 2,4-dichlorophenol
(Supporting Information Figure S7).
no additional dichlorophenols or trichlorophenols were
detected in the reaction solutions. On the basis of the
structures of intermediates A, B, and C and the observed
formation of 2,4-dichlorophenol, it is plausible that 2,3-
dichlorophenol, 3,4-dichlorophenol, and 2,3,4-trichlorophe-
nol could form in this system. Since none of these chloro-
chlorophenol intermediates D and E; (iii) mass spectrum of PFB-derivatized monochlorinated triclosan intermediate (intermediate A or
B); (iv) mass spectrum of PFB-derivatized dichlorinated triclosan intermediate (intermediate C).
FIGURE4. FormationanddecayofA: 5,6-dichloro-2-(2,4-dichlorophenoxy)phenol,B: 4,5-dichloro-2-(2,4-dichlorophenoxy)phenol,C: 4,5,6-
trichloro-2-(2,4-dichlorophenoxy)phenol,D: 2,4-dichlorophenol,E: 2,4,6-trichlorophenolforpH4,7,and10.Reactionconditions: [triclosan]0
) 5.05 µM; [free chlorine]0) 14.2 µM; [NaHCO3] ) 2 mM. Note that the y-axes vary considerably.
VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93181
react too quickly to be detected. To address this second
possibility, a set of experiments was conducted to examine
2,3,4-dichlorophenol with free chlorine under pseudo-first-
order conditions with a 10× chlorophenol molar excess at
pH 7. Under these conditions, kobs values for 3,4-dichlo-
rophenol ()2.04 × 10-3s-1) and 2,3,4-trichlorophenol
()1.75 × 10-3s-1) are similar to the measured kobsvalue for
2,4-dichlorophenol ()1.83 × 10-3s-1) at this pH. This
observation suggests that if these species were formed as
intermediates that they should be detectable like 2,4-
dichlorophenol. Although 2,3-dichlorophenol reacts faster
(kobs) 9.70 × 10-3s-1) than the other three chlorophenols,
the reaction rate is still 1.5× slower than that of triclosan,
and thus this species should also be detectable if it were
formed. These observations suggest that cleavage of the
phenol ring of triclosan occurs at or near the same time as
the breaking of the ether linkage between the phenol ring
and the formation of 2,4-dichlorophenol.
Onodera et al. (42) reported the formation of 2,3,4-
trichlorophenol and 2,4-dichlorophenol when triclosan
reacted with free chlorine. These authors, however, did not
detect 2,4,6-trichlorophenol. On the basis of their observa-
from the 2,4-dichlorophenoxy ring of triclosan and 2,3,4-
trichlorophenol from the 4,5,6-trichlorophenol ring. Unfor-
tunately, it is difficult to compare the results obtained in the
present study with those reported by Onodera et al. because
their paper does not fully describe their experimental
trichlorophenol have identical mass spectra and similar
retention times, it is possible that 2,4,6-trichlorophenol was
mistakenly identified as 2,3,4-trichlorophenol. This hypoth-
esis seems likely given that Onodera et al. did not detect
2,4,6-trichlorophenol, a well-known product of the chlorina-
tion of 2,4-dichlorophenol (38, 40).
Linear Free-Energy Relationship. A linear free-energy
relationship (LFER) correlating the rate of electrophilic
substitution of chlorinated phenols by hypochlorous acid
σ-) have previously been used to correlate phenol reactivity
(37, 49-51), electrophilic substitution by HOCl is most
(σ+; ref 52). This scale accounts for through-resonance
between the chloro ring substituents and the electron-
each compound in Table 1, the sum Σσ+o,m,pwas obtained
using literature values for σ+mand σ+p(52). Values for σ+o
were estimated on the basis of the following (53): σ+o )
0.66σ+p. For triclosan, the σ+ovalue was determined using
and kArO-for HOCl. The resulting linear regression for the
The negative sign for the Hammett slope (F) reflects an
increased deactivating effect as additional chlorine substit-
uents are added to the phenol ring. By using this LFER and
(Table 1), kArO-,i values were determined for 5,6-dichloro-
2-(2,4-dichlorophenoxy)phenol (kArO- ) 600 M-1s-1), 4,5-
and 4,5,6-trichloro-2-(2,4-dichlorophenoxy)phenol (kArO-)
40 M-1s-1). As shown, the predicted reactivity of the
chlorinated triclosan intermediates decreases with the ad-
dition of chlorine substituents. This result is consistent with
the reactivity trends depicted in Figure 4.
and its daughter products were estimated for pH 7 under
excess free-chlorine conditions (Table 1). Under these
both 4,5,6-trichloro-2-(2,4-dichlorophenoxy)phenol and 2,4,6-
trichlorophenol are relatively unreactive (t1/2> 20 min), and
thus these compounds would be expected to form and be
fairly stable over the course of a given experiment, a result
55). To assess the potential for triclosan to act similarly, we
conducted experiments employing a 10× excess of free
of pH. Figure 6 shows that chloroform is readily produced
when free chlorine reacts with triclosan at pH 5, 6, 8, and 9.
Chloroform yields were greatest at circumneutral pH and
lower under acidic or basic conditions. This pH trend is
consistent with prior studies of chloroform formation when
the pH sensitive speciation of the reactants. After 48 h, 1.53
triclosan concentration of 2.5 µM (data not shown). This
molar yield of chloroform ()0.612 µM chloroform/µM
triclosan) is considerably larger than the yields observed for
other chlorophenol species chlorinated under similar pH
conditions (37) and suggests that the ortho-OC6H3Cl2 and
meta-Cl functional groups in triclosan strongly activate this
compound toward chloroform production. Prior studies
have shown that chloroform production is enhanced when
the phenol ring is chlorinated in the meta-position (37,
55); however, this is the first work to suggest that the
ortho-OC6H3Cl2group enhances production of chloroform.
This finding suggests that the ether linkages prevalent in
humic materials (58) could affect the formation of several
disinfection byproducts (DBPs) observed when humic-
containing waters are chlorinated.
log kArO- ) -(10.7 ( 2.2)Σσ+
o,m,p+ 4.43 ((0.35)
R2) 0.96, n ) 8 (7)
in boxes refer to the compounds listed in Table 1.
31829ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005
Chloroform production in the free chlorine-triclosan
reactions could result from cleavage of the phenol ring in
triclosan or could result from cleavage of the 2,4-dichloro-
in this system, an experiment examining chloroform pro-
duction when 2,4-dichlorophenol reacts with free chlorine
was performed (Supporting Information Figure S8). Under
similar reaction conditions to those used in the triclosan
2,4-dichlorophenol and free chlorine react are an order of
magnitude lower than the concentrations of chloroform
formed during the triclosan/free chlorine reactions. This
finding suggests that the majority of CHCl3produced in this
system originates from oxidation and ring cleavage of the
phenol moiety of triclosan and not from reactions involving
the 2,4-dichlorophenol produced via ether cleavage.
Hypothesized Reaction Pathway. On the basis of the
kinetic experiments presented herein (Figures 4 and S6), it
is possible to develop a reaction pathway for this system. In
this pathway, triclosan can either undergo ether cleavage
resulting in production of 2,4-dichlorophenol and products
not detected via PFBBr derivatization-GC/MS analysis (e.g.,
quinones) or it can be chlorinated to produce one of two
chlorophenoxyphenol intermediates: 5,6-dichloro-2-(2,4-
dichlorophenoxy)phenol or 4,5-dichloro-2-(2,4-dichlorophe-
either are further chlorinated to produce 4,5,6-trichloro-2-
to produce 2,4-dichlorophenol and products not detected
via GC/MS. The 2,4-dichlorophenol produced during these
interactions can either be further chlorinated to produce
2,4,6-trichlorophenol or can undergo ring cleavage as
documented elsewhere (37, 55).
The authors’ contend that triclosan, 5,6-dichloro-2-(2,4-
noxy)phenol, and 4,5,6-trichloro-2-(2,4-dichlorophenoxy)-
phenol all serve as precursors to 2,4-dichlorophenol. This
contention is supported by the fact that 2,4-dichlorophenol
can be detected under conditions when no chlorinated
triclosan intermediates are present (Figure 4, pH 10) and
2,4-dichlorophenol production continues to occur after
triclosan, 5,6-dichloro-2-(2,4-dichlorophenoxy)phenol, and
4,5-dichloro-2-(2,4-dichlorophenoxy)phenol are no longer
present (Figure S6). The former observation indicates that
triclosan can undergo ether cleavage to produce 2,4-
trichloro-2-(2,4-dichlorophenoxy)phenol can do the same.
It therefore is reasonable to conclude that the dichlorinated
triclosan intermediates (5,6-dichloro-2-(2,4-dichlorophe-
noxy)phenol and 4,5-dichloro-2-(2,4-dichlorophenoxy)phe-
nol) also undergo ether cleavage.
Reaction pathways in addition to ring chlorination and
ether cleavage may exist and may result in intermediates or
products that could not be detected and identified with the
procedures employed in this project. However, as shown in
Figure S6, the molar yield for the two chlorophenols at 120
min is ∼45% and a substantial amount of 4,5,6-trichloro-
2-(2,4-dichlorophenoxy)phenol is still present. These ob-
servations suggest that the pathway described here is a
significant reaction pathway, if not the most important
Engineering and Health Significance of Results. The
reactions between triclosan and free chlorine are rapid at
reaction conditions, when only 2.5 µM (750 µg/L) triclosan
reacts with an excess of free chlorine, significant quantities
triclosan measured in source waters suggest that the im-
portance of these reactions toward DBP formation within
drinking water treatment plants and distribution systems is
“Antimicrobial” cleaning agents currently on the market
the potential for chloroform formation during household
use of triclosan-containing antibacterial products, an experi-
ment was conducted in which two dish soaps of brand X,
one formulation containing triclosan (quantified as 1.4 mg
water at a concentration of 0.25 g/L. Trihalomethane
formation in these samples was then monitored with time
(Figure 7). Under these test conditions, the measured
chloroform level was 15 µg/L after 5 min and 49 µg/L within
120 min. Over this same time frame, the chloroform levels
for the nontriclosan formulation were near the detection
FIGURE 6. Chloroform formation as a function of solution pH.
Reaction conditions: [triclosan]0 ) 2.5 µM; [free chlorine]0 )
25 µM; [NaHCO3] ) 2 mM.
FIGURE 7. Formation of trihalomethanes when dish soap brand X
) 84.9 µM; [dish soap] ) 0.25 g/L; pH ) 7; triclosan concentration
in antibacterial soap was measured at 1.4 mg/g. At 24 h, the molar
concentration of CHCl3is 581 nM and the CHCl2Br molar concentra-
tion is 44 nM for the triclosan dishsoap samples.
VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93183
these test conditions, the potential exists for substantial
chloroform production to occur via daily household use of
inhalational exposures to triclosan-mediated production of
and free chlorine result in the production of 2,4-dichloro-
phenol, 2,4,6-trichlorophenol, and several chlorinated tri-
closan intermediates. Although the health implications of
the chlorinated triclosan intermediates are currently un-
known, 2,4-dichlorophenol is a known source of taste and
odor problems (40). In the reactions where 2,4-dichlorophe-
pH 7 experiment) were typically 1 to 2 orders of magnitude
higher than the average threshold odor concentration of 2
µg/L (∼12 nM; ref 40). These yields resulted from an initial
triclosan concentration of 5.05 µM (1.46 mg/L). For soap
concentrations similar to those employed in the chloroform
formation experiment and for a solution pH of 7, one could
This study also has application to the wastewater treat-
ment industry. Free chlorine is often employed during
wastewater treatment to disinfect the final effluent. The
kinetic results obtained herein would appear to suggest that
triclosan should be rapidly removed via this chlorination
process. However, when wastewater effluent is chlorinated,
the high levels of ammonia present effectively cause forma-
tion of inorganic and organic chloramines. In general,
chloramines are weaker oxidants than free chlorine and it is
The lower reactivity of chloramines could explain why
Research is currently underway to characterize the reaction
rates and the products formed when triclosan reacts with
chloramines. The results of these studies will help further
characterize triclosan’s fate during wastewater effluent
The insightful and detailed comments of three anonymous
reviewers to help clarify and strengthen this paper are
gratefully acknowledged. This work was supported by a
National Science Foundation Graduate Student Fellowship
to K.L.R. and through a research grant from the American
Water Works Association Research Foundation (AwwaRF).
We thank Jody Smiley and Julie Petruska for their help with
the analytical methods and instrumentation.
Supporting Information Available
Tables of chlorine decay percentages, chlorine isotope MS
analysis, and molecular weights for PFBBr functionalized
triclosan and its daughter products; figures illustrating
reaction orders with respect to free chlorine and triclosan;
figure depicting chlorine isotope MS spectra; figures il-
lustrating triclosan and free-chlorine decay and subsequent
product formation; a figure illustrating chloroform produc-
material is available free of charge via the Internet at http://
(1) Jones, R. D.; Jampani, H. B.; Newman, J. L.; Lee, A. S. Triclosan:
A review of effectiveness and safety in health care settings. Am.
J. Infect. Control 2000, 28, 184-196.
(2) Meincke, B. E.; Kranz, R. G.; Lynch, D. L. Effect of Irgasan on
1980, 1980, 133-147.
(3) McMurry, L. M.; Oethinger, M.; Levy, S. B. Overexpression of
marA, soxS or acrAB produces resistance to triclosan in
Escherichia coli. FEMS Microbiol. Lett. 1998, 166, 305-309.
(4) McMurry, L. M.; McDermott, P. F.; Levy, S. B. Genetic evidence
that InhA of Mycobacterium smegmatic is a target for triclosan.
Antimicrob. Agents Chemother. 1999, 43, 711-713.
(5) Slayden, R. A.; Lee, R. E.; Clifton E. Barry, R. Isoniazid affects
multiple components of the type II fatty acid synthase system
of Mycobacterium tuberculosis. Mol. Microbiol. 2000, 38, 514-
(6) McDonnell, G. D. P. Action and targets of triclosan. ASM News
1998, 64, 670-671.
(7) Schweizer, H. P. Triclosan: a widely used biocide and its link
to antibiotics. FEMS Microbiol. Lett. 2001, 202, 1-7.
(8) Lindstrom, A.; Buerge, I. J.; Poiger, T.; Bergqvist, P.-A.; Muller,
M. D.; Buser, H.-R. Occurrence and environmental behavior of
the bactericide triclosan and Its methyl derivative in surface
(9) McAvoy, D. C.; Schatowitz, B.; Jacob, M.; Hauk, A.; Eckhoff, W.
S. Measurement of triclosan in wastewater treatment systems.
Environ. Toxicol. Chem. 2002, 21, 1323-1329.
and personal care products in sewage treatment works. J.
Environ. Monit. 2003, 5, 823-830.
(11) Bester, K. Triclosan in a sewage treatment process-balances
and monitoring data. Water Res. 2003, 37, 3891-3896.
(12) Sabaliunas, D.; Webb, S. F.; Hauk, A.; Jacob, M.; Eckhoff, W. S.
Res. 2003, 37, 3145-3154.
(13) Singer,H.;Muller,S.;Tixier,C.;Pillonel,L.Triclosan: Occurrence
and fate of a widely used biocide in the aquatic environment:
Field measurements in wastewater treatment Plants, Surface
Waters, and Lake Sediments. Environ. Sci. Technol. 2002, 36,
(14) Federle, T. W.; Kaiser, S. K.; Nuck, B. A. Fate and effects of
triclosan in activated sludge. Environ. Toxicol. Chem. 2002, 21,
wastewater. Their transport into sediments. Environ. Sci.
Technol. 1980, 14, 1382-1390.
(16) Aguera, A.; Fernandez, A. R.; Piedra, L.; Mezcua, M.; Gomez, M.
J. Evaluation of triclosan and biphenylol in marine sediments
and urban wastewaters by pressurized liquid extraction and
solid phase extraction followed by gas chromatography mass
spectrometry and liquid chromatography mass spectrometry.
Anal. Chim. Acta 2003, 480, 193-205.
(17) Boyd, G. R.; Palmeri, J. M.; Zhang, S.; Grimm, D. A. Pharma-
ceuticals and personal care products (PPCPs) and endocrine
disrupting chemicals (EDCs) in stormwater canals and Bayou
St. John in New Orleans, Louisiana, USA. Sci. Total Environ.
2004, 333, 137-148.
(18) Kolpin, D. W.; Furlong, E. T.; Meyer, M. T.; Thurman, E. M.;
Zaugg, S. D.; Barber, L. B.; Buxton, H. T. Pharmaceuticals,
streams, 1999-2000: A national reconnaissance. Environ. Sci.
Technol. 2002, 36, 1202-1211.
(19) McBain, A. J.; Bartolo, R. G.; Catrenich, C. E.; Charbonneau, D.;
Ledder, R. G.; Price, B. B.; Gilbert, P. Exposure of sink drain
microcosms to triclosan: Population dynamics and antimi-
(20) Hay, A. G.; Dees, P. M.; Sayler, G. S. Growth of a bacterial
(21) Latch, D. E.; Packer, J. L.; Arnold, W. A.; McNeill, K. Photo-
chemical conversion of triclosan to 2,8-dichlorodibenzo-p-
dioxin in aqueous solution. J. Photochem. Photobiol., A: Chem.
2003, 158, 63-66.
(22) Tixier, C.; Singer, H. P.; Canonica, S.; Muller, S. R. Phototrans-
process for this widely used biocide-Laboratory studies, field
measurements, and modeling. Environ. Sci. Technol. 2002, 36,
(23) Morrall, D.; McAvoy, D.; Schatowitz, B.; Inauen, J.; Jacob, M.;
water (Cibolo Creek, TX). Chemosphere 2004, 54, 653-660.
(24) Zhang, H.; Huang, C.-H. Oxidative transformation of triclosan
and chlorophene by manganese oxides. Environ. Sci. Technol.
2003, 37, 2421-2430.
(25) Stackelberg, P. E.; Furlong, E. T.; Meyer, M. T.; Zaugg, S. D.;
31849ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 9, 2005
compounds and other organic wastewater contaminants in a Download full-text
2004, 329, 99-113.
(26) Boyd, G. R.; Reemtsma, H.; Grimm, D. A.; Mitra, S. Pharma-
ceuticals and personal care products (PPCPs) in surface and
Environ. 2003, 311, 135-149.
(27) Snyder, S. A.; Westerhoff, P.; Yoon, Y.; Sedlak, D. L. Pharma-
ceuticals, personal care products, and endocrine disruptors in
water: Implications for the water industry. Environ. Eng. Sci.
2003, 20, 449-469.
(28) Heberer, T.; Fuhrmann, B. Occurrence of Pharmaceutical
DC, 2001; Vol. 791.
M. T. Presence of Pharmaceuticals in Wastewater Effluent and
1999. In Proceedings of the 2001 Georgia Water Resources
Conference; Hatcher, K. J., Ed.; The University of Georgia:
(30) Henderson, A. K.; Moll, D. M.; Frick, E. A.; Zaugg, S. Presence
of Wastewater Tracers and Endocrine Disrupting Chemicals in
Metropolitan Atlanta, 1999. In Proceedings of the 2nd Interna-
Chemicals in Water, National Ground Water Association:
Minneapolis, MN, 2001.
processes. J. Environ. Eng. (N.Y.) 2002, 128, 253-260.
(32) Ternes, T. A.; Meisenheimer, M.; McDowell, D.; Sacher, F.;
Brauch, H.-J.; Haist-Gulde, B.; Preuss, G.; Wilme, U.; Zulei-
Seibert, N. Removal of pharmaceuticals during drinking water
treatment. Environ. Sci. Technol. 2002, 36, 3855-3863.
activities. Environ. Sci. Technol. 2003, 37, 5665-5670.
(34) Onodera, S.; Yamada, K.; Yamaji, Y.; Ishikura, S.; Suzuki, S.
Chemical changes of organic compounds in chlorinated water
(Predioxins) during chlorination of methylphenols in dilute
aqueous solution. J. Chromatogr. 1986, 354, 293-303.
of organic compounds in chlorinated water VI. Gas chromato-
graphic and mass spectrometric studies of the reactions of
phenylphenols with hypochlorite in dilute aqueous solution. J.
Chromatogr. 1983, 265, 201-213.
(36) Onodera, S.; Tabata, M.; Suzuki, S.; Ishikura, S. Gas chromato-
graphic identification and determination of chlorinated qui-
nones formed during chlorination of dihydric phenols with
Hypochlorite in Dilute Aqueous Solution. J. Chromatogr. 1980,
(37) Gallard, H.; von Gunten, U. Chlorination of phenols: Kinetics
and formation of chloroform. Environ. Sci. Technol. 2002, 36,
(38) Lee, G. F.; Morris, J. C. Kinetics of chlorination of phenol-
chlorophenolic tastes and odors. Int. J. Air Water Pollut. 1962,
(39) Soper, F. G.; Smith, G. F. The halogenation of phenols. J. Chem.
Soc. 1926, 6, 1582-1591.
(40) Burttschell, R. H.; Rosen, A. A.; Middleton, F. M.; Ettinger, M.
B. Chlorine derivatives of phenol causing taste and odor. J.s
Am. Water Works Assoc. 1959, 51, 205-214.
(41) Gallard, H.; von Gunten, U. Chlorination of natural organic
matter: kinetics of chlorination and of THM formation. Water
Res. 2002, 36, 65-74.
compounds in chlorinated water XIII. Gas chromatographic-
mass spectrometric studies of the reactions of Irgasan DP 300
[5-chloro-2-(2,4-dichlorophenoxy)phenol] with chlorine in di-
lute aqueous solution. J. Chromatogr. 1987, 392, 267-275.
(43) Kanetoshi, A.; Ogawa, H.; Katsura, E.; Kaneshima, H. Chlorina-
tion of Irgasan DP300 and formation of dioxins from its
chlorinated derivatives. J. Chromatogr. 1987, 389, 139-153.
Water Works Association, American Public Health Association,
Water Pollution Control Federation: Washington, DC, 1989.
(45) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M.
Environmental Organic Chemistry, 2nd ed.; John Wiley & Sons:
Hoboken, NJ, 2003.
(46) Jensen, J. N.; Johnson, J. D. Interferences by monochloramine
and organic chloramines in free available chlorine methods. 2.
of phenols as pentafluorobenzyl derivatives by gas chroma-
Analyst 2001, 126, 835-839.
(48) Rebenne, L. M.; Gonzalez, A. C.; Olson, T. M. Aqueous
chlorination kinetics and mechanism of substituted dihydroxy-
benzenes. Environ. Sci. Technol. 1996, 30, 2235-2242.
(49) Canonica, S.; Tratnyek, P. G. Quantitative structure-activity
relationships for oxidation reactions of organic chemicals in
water. Environ. Toxicol. Chem. 2003, 22, 1743-1754.
(50) Hansch, C.; Leo, A. Exploring QSAR: Fundamentals and
Applications in Chemistry and Biology; American Chemical
Society: Washington, DC, 1995.
Res. 2003, 37, 2883-2892.
in aqueous solution. J. Phys. Chem. B 1999, 103, 6653-6656.
(53) Jonsson, M.; Lind, J.; Eriksen, T. E.; Merenyi, G. O-H bond
strengths and one-electron reduction potentials of multisub-
(54) Hansch, C.; Leo, A.; Taft, W. A survey of hammett substituent
of halogenated phenols or halogenated anilines with sodium-
hypochlorite. Chemosphere 1982, 11, 81-87.
(56) de Laat, J.; N., M.; Dore, M. Chlorination of organic com-
pounds: Chlorine demand and reactivity in relationship to the
trihalomethane formation. Water Res. 1982, 16, 1437-1450.
formation from the halogenation of dihydroxyaromatic model
compounds for humic-acid. Environ. Sci. Technol. 1983, 17,
(58) Leenheer, J. A.; Wershaw, R. L.; Reddy, M. M. Strong Acid,
Carboxyl-Group Structures in Fulvic Acid from the Suwannee
Receivedforreview July9,2004. Revisedmanuscriptreceived
January 31, 2005. Accepted February 14, 2005.
VOL. 39, NO. 9, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93185