PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact

Wat Res 37, 2003, 551 560 .pdf

Original filename: Wat Res 37, 2003, 551-560.pdf
Title: PII: S0043-1354(02)00311-1

This PDF 1.3 document has been generated by / Acrobat Distiller 4.05 for Windows, and has been sent on pdf-archive.com on 01/11/2015 at 18:01, from IP address 71.17.x.x. The current document download page has been viewed 600 times.
File size: 148 KB (10 pages).
Privacy: public file

Download original PDF file

Document preview

Water Research 37 (2003) 551–560

Anaerobic microbial and photochemical degradation
of 4,40-dibromodiphenyl ether
Sierra Raynea, Michael G. Ikonomoub, MacMurray D. Whalec,*

Department of Chemistry, P.O. Box 3065, University of Victoria, Victoria, BC, Canada V8W 3V6
Marine Environment and Habitat Science Division, Pacific Region, Institute of Ocean Sciences, Fisheries and Oceans Canada,
9860 West Saanich Road, Sidney, BC, Canada V8L 4B2
Department of Mechanical Engineering, P.O. Box 3055, University of Victoria, Victoria, BC, Canada V8W 3P6

Received 1 January 2002; received in revised form 1 July 2002; accepted 3 July 2002

The anaerobic microbial and photochemical degradation pathways of 4,40 -dibromodiphenyl ether (BDE15) were
examined. BDE15 was reductively debrominated within a fixed-film plug-flow biological reactor at hydraulic retention
times of 3.4 and 6.8 h, leading to exclusive production of 4-bromodiphenyl ether (BDE3) and diphenyl ether (DE). A
suite of potential BDE15 metabolites arising from reductive debromination, hydroxylation, and methoxylation of the
aromatic C–Br and C–H bonds were not observed. Following initial debromination of BDE15, degradation of BDE3 to
DE readily occurs, suggesting the rate-limiting step for anaerobic BDE15 degradation is conversion of BDE15 to
BDE3. The photochemical degradation of BDE15 was also examined in organic (CH3CN and CH3OH) and aqueous
(H2O:CH3CN; 1:1 v/v) solvent systems at 300 nm. Only photochemically induced reductive debromination was found
to occur via homolytic C–Br bond cleavage, with no evidence of C–O bond cleavage or products arising from
heterolytic bond cleavage.
r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: 4,40 -Dibromodiphenyl ether; Polybrominated diphenyl ethers; Anaerobic biodegradation; Photodegradation

1. Introduction
Polybrominated diphenyl ethers (PBDEs) are additive
flame retardants in a wide range of materials including
plastics, textiles, and foams at concentrations up to 30%
by weight [1]. Recent evidence suggests concentrations
of these compounds are increasing rapidly in biota from
both populated and remote regions [2–4]. High concentrations of PBDEs have also been observed in
freshwater and marine sediments [5–7], air [8], and in
sewage sludge and the resulting biosolids [9,10]. While
current concentrations of PBDEs in the environment are
*Corresponding author. National Bank, Financial Bankers
Hall, 855-2nd Street SW, Suite 2000, Calgary, AB, Canada,
T2P 4J8. Tel.: +1-403-290-5626; fax: +1-403-269-7099.
E-mail address: MacMurray.Whale@nbfinancial.com
(M.D. Whale).

generally 10–500 times lower than PCBs and DDT [2,3],
equivalent concentrations of PBDEs, PCBs, and DDT
have been recently reported in some aquatic organisms
and sewage sludge [11]. This suggests that, in certain
regions, accumulation of PBDEs in the environment
may equal or exceed that of other halogenated aromatic
At present, three major commercial PBDE formulations are produced. These include the fully brominated
deca-BDE mixtures (>90% deca-BDE with B10%
nona- and octa-BDE congeners), and several octa-BDE
(hepta- and octa-BDE congeners), and penta-BDE
(tetra- and penta-BDE congeners) mixtures [12,13].
Prefixes used to describe these technical mixtures only
indicate the average degree of bromination, and a wide
range of congeners are observed in these products [13].
Of these products, the deca-BDE product is the major
PBDE formulation in use, accounting for B75% of the

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 2 ) 0 0 3 1 1 - 1


S. Rayne et al. / Water Research 37 (2003) 551–560

1999 worldwide PBDE production of 70,000 tonnes
[14,15]. In contrast, the major congeners typically found
in environmental compartments are those with four and
five bromine substituents (e.g., 2,20 ,4,40 -BDE47,
2,20 ,4,40 ,5-BDE99, and 2,20 ,4,40 ,6-BDE100) [16,2,3].
Hence, sources of the dominant tetra- and penta-BDE
congeners in aquatic environments are the subject of
interest. The predominance of these homologue groups
in the environment could result from (1) previous tetraBDE production in the late 1970s and 1980s followed by
continuous cycling in aquatic systems over the past two
decades, (2) present day commercial penta- and octaBDE production, or (3) biotic or abiotic debromination
of the major PBDE product in current use, the fully
brominated deca-BDE [13]. At present, it is difficult to
assess which of these potential processes is dominant
because little is known regarding the environmental fate
of PBDEs, especially on a congener specific basis. It has
been suggested that PBDEs will be more labile towards
biotic and abiotic environmental degradation than PCBs
and other chlorinated aromatics as the aryl–bromine
bond is weaker than the corresponding aryl–chlorine
bond [15], although this remains to be tested.
As a tool to help elucidate the environmental fate of
PBDEs, microbiological reactors have demonstrated
utility for evaluating the effectiveness of pollutant
degradation technologies and for directly degrading
toxic organic substances [17,18]. Several types of
bioreactors are able to achieve elevated biomass levels
and are in use on laboratory, pilot, and full scales,
including trickling filters, rotating biological contactors,
activated sludge reactors, fluidized bed reactors, and
biomass recycle reactors [18,19]. Of these, fixed biomass
microbiological reactors are perhaps the simplest and
most cost-effective to design, colonize, and operate. In
addition, fixed biomass reactors are likely more representative of the nature of microbial activity taking place
in sediments, where most hydrophobic contaminants
such as PBDEs generally reside. Furthermore, previous
studies have successfully demonstrated the utility of
using a fixed-film plug-flow biological reactor (FFPFBR) design to simulate biodegradation in natural and
engineered environments [20].
Within FF-PFBR systems, mixed assemblages of
microbes have been shown to dehalogenate aromatic
organic pollutants under primarily anaerobic conditions
at mg/L to mg/L concentrations [21,17,22,23]. Removal
of halogen atoms from these compounds makes them
more susceptible to complete mineralization [24,22]. For
example, microbial degradation of highly chlorinated
PCBs, phenols, and benzenes occurs only via reductive
dehalogenation [21,25,23]. As aerobic degradation takes
place at higher rates than under anaerobic conditions
[24,22], multistage treatment processes using sequential
anaerobic and aerobic phases [26] may be the optimum
means of degrading PBDEs in engineered environments.

In natural systems, anaerobic degradation may remove
the halogen substituents, thereby decreasing the hydrophobicity of a compound and allowing it to migrate
back into aerobic aqueous environments for complete
mineralization. In the case of both the mono- and dihalogenated (with F, Cl, and Br substituents) and the
parent diphenyl ether systems, previous work has shown
these compounds to be degraded by bacteria and fungi
under aerobic conditions. The resulting metabolites
typically underwent hydroxylation or methoxylation
(i.e. removal of an H atom followed by addition of an
OH or OCH3 group, respectively) prior to, or in
conjunction, with aromatic ring cleavage [27–34]. However, as noted above, halogen substituents on aromatic
compounds are generally not removed under aerobic
conditions. Thus, aerobic degradation of halogenated
aromatics such as PBDEs may result in the formation of
toxic halogenated aliphatic metabolites following initial
ring cleavage. Hence, anaerobic pretreatment of waste
streams containing halogenated aromatics such as
PBDEs may be preferred.
Because of the difficulty in synthesizing specific PBDE
congeners in large quantities, most of these compounds
are not readily available in sufficient quantities for large
laboratory- or pilot-scale studies. Only 4,40 -dibromodiphenyl ether (BDE15) and 4-bromodiphenyl ether
(BDE3) can be purchased at the low-cost gram scale
from commercial suppliers. Using these available
compounds, the objectives of this study were to
investigate the photochemical (abiotic) and anaerobic
microbial (biotic) degradation of BDE15 to assist in
guiding future wastewater treatment process developments and to better understand the environmental fate
of the more prevalent higher-brominated PBDEs.

2. Methods and materials
2.1. Bioreactor design, characterization, colonization,
and operation
Two individual microbiological reactors were constructed of 200 mm diameter PVC pipe each 1 m in
length (Fig. 1) filled completely with 1–2 cm nominal
diameter gravel. Pore volume tests showed each reactor
had an equivalent free space of 7.1 L. The columns were
supplied continuously with feed water in an upflow,
once-through mode at 34 mL/min using a peristaltic
pump (Masterflex, Cole-Parmer, Vernon Hills, IL) and
1/400 Tygons tubing (Cole-Parmer, Vernon Hills, IL).
Colonization water for the bioreactors was obtained
from a nearby wetland on the campus of the University
of Victoria in Victoria, BC, Canada (481250 N,
1231210 W). The site is hydraulically downgradient from
a historical munitions dump and was expected to have
microorganisms capable of degrading both natural and

S. Rayne et al. / Water Research 37 (2003) 551–560


2.2. Measurement of major chemical and physical

Fig. 1. Fixed-film plug-flow biological reactor design.

anthropogenic compounds. Water samples from the
wetland were pumped through the bioreactor in a
continuous recycle mode for a period of 1 week, with
fresh feed water replaced daily. After the 1-week initial
colonization stage, the feed stream was switched to a
simulated grey water [18] having the following components: 2.55 mM NaCl, 1.5 mM NaHCO3, 0.2 mM
MgSO4, 4 mM CaCl2, and 12 mM FeCl3. Macronutrients
(C, N, and P) were added as 200 mg/L CH3OH C,
10 mg/L KNO3 N, and 2.5 mg/L K2HPO4 P in accordance with the C:N:P ratio of 100:5:1 typically observed
in municipal wastewaters [35,19]. The simulated wastewater was fed to the reactor in a once-through mode
continuously for 8 months prior to trials for BDE15
biodegradation in order to ensure a steady-state microbial population.
To determine the mean hydraulic residence time
(HRT) in the reactors, 1 mL pulses of a concentrated
fluorescent rhodamine WTC dye (Cole-Parmer Brand,
Cat. no. U-00298-16, Cole-Parmer, Vernon Hills, IL)
were injected into the bioreactor influent. Movement of
the dye tracer through the bioreactors was determined
by measuring the absorbance of the effluent at 555 nm
(lmax of dye) on an UV–visible spectrophotometer
(Varian Cary 50). Operating the bioreactors either
individually, or in parallel (see Fig. 1), resulted in
hydraulic residence times (HRTs) of 3.4 and 6.8 h,
respectively. The HRTs determined using the dye tracer
were in good agreement with those calculated using the
free space determined by pore volume tests (3.5 and
7.0 h, respectively). To achieve a HRT of 3.4 h, one
reactor was isolated from the flow using valves located
immediately before and after the reactor, while the
reactor of interest received the full system flow (34 mL/
min). For a HRT of 6.8 h, the reactors were linked in
parallel where equal volumes of influent (17 mL/min)
pass through each bioreactor.

Concentrations of chemical oxygen demand (COD),

nitrate (NO
3 ), nitrite (NO2 ), ammonia (NH3), total
dissolved nitrogen (TDN), soluble reactive phosphorus
4 ), and total dissolved phosphorus (TDP) in
both the influent and effluent were monitored on a
monthly basis during the 8 month colonization period
prior to when BDE15 was added to the feed water.
During the 8-week period where BDE15 biodegradation
was present in the feed water, these nutrients were
monitored on a weekly basis. Measurement of pH and
temperature followed the same temporal patterns.
Samples were analyzed for COD in accordance with
standard method 5220D [35]. Filtered samples were
diluted 10-fold to avoid chloride or other interferences
and purged with N2 for 5 min prior to analysis to strip
any H2S or NH3 from solution. Prepared Hach COD
test vials (Loveland, CO; range 0–150 mg/L) and an
UV–visible spectrophotometer (Varian Cary 50; Mississauga, ON, Canada) were used for the analysis.
Ammonia was determined using the Nessler standard
methods 4500-NH3-B and C [36], while nitrite, nitrate,
and TDN were analyzed by standard methods

2 -B, 4500-NO3 -E, and 4500-Norg, respectively
[35]. SRP and TDP were determined using standard
method 4500-P-E with TDP undergoing the persulfate
digestion method (4500-P-B) prior to 4500-P-E [35].
2.3. Bioreactor sample extraction and cleanup
All organic solvents used were pesticide residue
analysis grade. Internal, recovery, and calibration
standards containing 12C and 13C labelled chloro- and
bromo-diphenyl ethers were purchased from Cambridge
Isotope Laboratories (Andover, MA). Bioreactor influent and effluent samples (0.500 L) were collected in
glassware washed with standard laboratory detergent
(Alconox; White Plains, NY) and rinsed with successive
washes of tap water, distilled water, toluene, hexane,
dichloromethane, and acetone, in that order. Glassware
was then baked overnight at 1051C to eliminate solvent
residues. Samples were spiked with 1 ng 13C-3,30 ,4,40 tetrachlorodiphenyl ether, 2 ng 13C-2,3,30 ,4,40 ,5-hexachlorodiphenyl ether, and 3 ng 13C-2,20 ,3,30 ,4,40 ,5,50 octachlorodiphenyl ether as internal standards and then
transferred to a separatory funnel where 100 mL of
saturated NaCl solution in distilled water was added.
The resulting solutions were then acidified to pH 2 using
6 N HCl to allow extraction of any acidic, phenolic, and
hydroxylic metabolites. Extractions (3 100 mL) were
performed using CH2Cl2, after which the extracts were
dried using anhydrous MgSO4, filtered through a
0.45 mm filter, and rotoevaporated down to a volume
of B10 mL. Extracts were then further concentrated to


S. Rayne et al. / Water Research 37 (2003) 551–560

less than 100 mL on a rotoevaporator, transferred to an
amber microvial using toluene, and then reduced under
N2 and gentle heating to a final volume less than 25 mL.
A 13C-labelled method performance standard (1 ng 13C3,30 ,4,40 -tetrabromodiphenyl ether) was added prior to
capping the microvial for analysis.
2.4. HRGC–HRMS analysis of bioreactor extracts
PBDE extracts were analyzed by HRGC–HRMS
using a VG-Autospec high resolution mass spectrometer
(Micromass, Manchester, UK) equipped with a HewlettPackard Model 5890 Series II gas chromatograph. The
GC was operated in the splitless injection mode, and the
splitless injector purge valve was activated 2 min after
sample injection. The volume injected was 1 mL of
sample plus 0.5 mL of air. A standard 30 m DB-5 column
(0.25 mm I.D. 0.25 mm film thickness) from J&W
Scientific (Folsom, CA) was used with UHP-He as the
carrier gas at a constant head pressure of 25 psi to
maintain a linear velocity of 35 cm/s. The temperature
program used under constant pressure was as follows:
hold at 1001C for 1 min; 21C/min to 1401C; 41C/min to
2201C; 81C/min to 3301C; and hold 1.2 min. The splitless
injector port, direct GC–MS interface, and the MS ion
source were maintained at 3001C, 2701C, and 3101C,
For all analyses, the MS was operated under positive
EI conditions with the filament in the trap stabilization
mode at 600 mA, an electron energy of 35 eV, and
perfluorokerosene used as the calibrant. The instrument
operated at 10,000 resolution and data were acquired in
the selected ion monitoring (SIM) mode. For each
analyte, the two most abundant isotopic peaks were
monitored for each molecular ion cluster. Identities of
the 32 PBDE congeners from mono- through hexabrominated, other than BDEs 15 and 3, analyzed in this
study are available elsewhere [2]. Compounds were
identified only when the HRGC–HRMS data satisfied
all of the following criteria: (1) two isotopes of the
analyte were detected by their exact masses with the
HRMS operating at 10,000 resolution or higher during
the entire chromatographic run; (2) the retention time of
the analyte peak was within 3 s to the predicted time
obtained from analysis of authentic compounds in the
calibration standards (where available); (3) the maxima
for both characteristic isotopic peaks of an analyte
coincided within 2 s; (4) the observed isotope ratio of the
two ions monitored per analyte were within 15% of the
theoretical isotopic ratio; and (5) the signal-to-noise
ratio resulting from the peak response of the two
corresponding ions was X3 for proper quantification of
the analyte. Concentrations of identified compounds
and their method detection limits (MDLs) were calculated by the internal standard isotope-dilution method
using mean relative response factors (RRFs) determined

from calibration standard runs made before and after
each batch of samples was analyzed.
Possible PBDE metabolites involving the replacement
of one or both Br atoms by a methoxy or hydroxy
function, or other metabolites where either one or two
aromatic C–H bonds were converted to an aryl-OH or
aryl-OCH3 moiety, were also investigated by gas
chromatography (GC) and HRGC–HRMS. A quantitative diphenyl ether (DE) standard was also prepared
for analysis. SIM analyses of the expected two most
abundant isotopic peaks of the molecular ion cluster for
each potential metabolite shown in Fig. 2 were performed at 10,000 resolution on all extracts. Compounds
would have been identified only if they met the QA/QC
criteria noted above. As discussed in the text, none of
these potential metabolites were observed by HRGC–
HRMS. GC analyses of these extracts were also
performed on a Hewlett Packard 5890 Series II gas
chromatograph with a flame ionization detector, containing a 15 m DB-5 (0.25 mm I.D. 0.1 mm film
thickness) column using the instrumental conditions
discussed below in the Photochemical studies subsection.
No peaks other than those of BDE15, BDE3, or DE
eluted after the toluene solvent.
Non-microbially exposed blanks were performed with
HRTs of 3.4 and 6.8 h to ensure no abiotic transformation of BDE15 was occurring in the solvent matrix.
These blanks were conducted on the bioreactor prior to
its microbial colonization. Complete recovery (95–
105%) of BDE15 was obtained from these blanks.
Extracts were also taken of the bioreactor effluent prior
to seeding the influent with BDE15 to ensure the reactor
materials or feed chemicals did not elute analytes of
interest and thereby constitute interferences. No analytes of interest were detected in any of these blanks.
2.5. Photochemical studies
Proton nuclear magnetic resonance (1H-NMR) spectra were determined on a Bruker WM 300 instrument
(Milton, ON, Canada). Preparative photolyses were
performed in 100 mL quartz tubes in a Rayonet RPR100 photochemical reactor (Southern New England
Ultraviolet Company, Brandford, CT) equipped with
16–300 nm lamps. Organic solvents were commercially
available ACS grades and were used as received, with
the exception that CH3CN was dried over CaH2 via
reflux. Authentic standards were used as received from
Aldrich (St. Louis, MO) and confirmed to be >99%
pure by GC.
Preparative photolyses were carried out with 50 mg
samples of BDE15 first dissolved in 100 mL of 100%
CH3OH, 100% CH3CN, or H2O:CH3CN (1:1 v/v) and
subsequently transferred to a quartz photolysis tube.
The respective solutions were then irradiated at 300 nm
with continuous cooling by a coldfinger to maintain a

S. Rayne et al. / Water Research 37 (2003) 551–560















Fig. 2. Anaerobic microbial degradation pathways of BDE15.

temperature of B181C. Solutions were also purged by a
stream of Ar for approximately 15 min before and
continuously during irradiation to remove O2 from the
reaction chamber. Photolysis times ranged from 5 to
60 min depending on the desired product conversion.
Following the photolysis period, 50 mg of hexabromobenzene was added to the photolysis tube as an internal
standard. Aqueous samples were then transferred to a
separatory funnel, 100 mL of a saturated aqueous NaCl
solution was added, and the solution was acidified to pH
2 to ensure any phenolic or acidic photoproducts were
captured by extraction. The acidified aqueous solutions
were then extracted with 3 100 mL washes of CH2Cl2
and concentrated to B10 mL on a rotoevaporator.
Direct rotoevaporation of the organic solvent to a final
volume of B10 mL was used when samples were
irradiated in 100% CH3OH or 100% CH3CN. A 1 mL
aliquot of this reduced solution was then injected
directly into a Hewlett-Packard 5890 Series II GC with
a flame ionization detector and containing a 15 m DB-5
(0.25 mm inside diameter 0.1 mm film thickness) column from J&W Scientific (Folsom, CA). The following
temperature program was used: 801C for 2 min following injection; 101C/min to 3001C; and held at 3001C for
20 min.
Structural confirmation of photoproducts was obtained by comparison to the GC retention time and 1HNMR of authentic standards. To allow photoproduct
identification by 1H-NMR after preparative photolysis
and extraction, solvents were removed completely by
rotoevaporation and subsequent placement under va-

cuum with gentle heating. The remaining compounds
were then re-dissolved in CDCl3 and subjected to 1HNMR analysis. The absence of photoproducts other
than BDE3 and DE in the crude 1H-NMR spectra
(clearly identified by comparison to published Aldrich
spectra, and spectra obtained from authentic standards)
negated the requirement for thin-layer chromatography
(TLC) or silica-gel column chromatography to effect
photoproduct separation and identification.
Further photolyses were performed by direct irradiation of samples in NMR tubes. A 10 mg quantity of
BDE15 was dissolved in B2 mL of 1:1 D2O–CD3CN.
This solution was then purged gently with Ar for 10 min,
transferred to an NMR tube, and subsequently placed in
the Rayonet RPR-100 photochemical reactor. Samples
were irradiated at 300 nm while an internal fan in the
reactor cooled the sample during the photolyses.
Photolysis times ranged from 5 to 15 min depending
on the conversion desired. 1H-NMR spectra were taken
directly after photolysis without further treatment. No
unknown peaks suggesting the presence of compounds
other than BDE15, BDE3, and DE, either before or
after photolysis, were observed by either GC or 1HNMR analysis. In addition, the maintenance of a full
mass balance (10072%) throughout all photolyses also
suggests that no other products were formed. To ensure
no ground state reactions were taking place, thermal
blanks were performed by subjecting the BDE15
solution to all experimental procedures except irradiation. No conversion of the starting material (o0.01%)
was observed during these blanks.


S. Rayne et al. / Water Research 37 (2003) 551–560

3. Results and discussion
Anaerobic microbial and photochemical degradation
studies show that BDE15 only undergoes reductive
debromination with replacement of a Br atom by a H
atom. These results suggest that in anaerobic anthropogenic or natural environments, or upon exposure to
UV light in organic and aqueous solvent systems,
BDE15 will sequentially debrominate to the parent
diphenyl ether (DE). Using BDE15 as a model for
higher-brominated analogs such as the more environmentally prevalent tetra- through hexa-BDEs, it appears
as though environmental debromination may be a
significant transformation pathway. This may help
explain why PBDE congener profiles in some environmental compartments differ significantly from the
commercial mixtures in present use [2,3,6,13].
Under anaerobic microbial conditions in the FFPFBR (Fig. 1) at both 3.4 h and 6.8 h HRTs and an
influent concentration of 7.92 10 7 M (0.2 mg/L),
BDE15 underwent sequential reductive debromination
to BDE3 and DE (Table 1). Potential microbial
products resulting from either hydroxylation (replacing
a Br or H atom with an OH group) or methoxylation
(replacing a Br or H atom with an OCH3 group) were
not observed (Fig. 2). The lack of these products
suggests these metabolic pathways are not operating
with respect to BDE15, and that under anaerobic
conditions, this compound only debrominates to the
parent DE. The parent DE system can then serve as a
carbon source for further aerobic degradation by
bacteria and fungi as shown elsewhere [27–30,32–34].
Furthermore, a suite of 32 PBDEs from mono- through
hexa-brominated, other than BDEs 15 and 3, were
analyzed for to determine if either microbial activities
could transfer Br atoms from one position on the
aromatic rings to another, or whether the reactor
materials were leaching PBDEs that could undergo

debromination and interfere with analyses for BDEs 15
and 3 and the parent DE. Furthermore, since this work
was performed using potable schedule 200 PVC pipe as
would be installed in water distribution systems, the
absence of PBDE leaching is an important finding in its
own right regarding public health. The lack of a
complete mass balance (90% and 40% accounted for
after 3.4 and 6.8 h, respectively), and the relative
shortage of BDE3 or DE in the effluent, suggests the
rate of debromination increases in moving from
BDE15-BDE3-DE, and that the rate-limiting step
is reduction of BDE15-BDE3, after which it readily
debrominates to DE and then undergoes rapid anaerobic transformation to as yet unknown bioproducts.
Prior to performing reproducible biodegradation
experiments, the bioreactor must achieve a sufficiently
stable microbial consortium to be considered at steady
state. FF-PFBRs, with a continuous, once-through
exposure to aqueous samples, provide a unidirectional
supply of carbon, nutrients, and energy to heterotrophic
organisms attached to an inert matrix. It is assumed the
most biologically labile and energetically advantageous
molecules are be initially metabolized. Recalcitrant
compounds are metabolized over a longer residence
time (=distance) in the reactor. Because of this longitudinal gradient of diminishing substrate quantity and
quality in the reactor, multiple niches select for a wide
assemblage of organisms capable of substrate metabolization [37]. Adsorption of highly hydrophobic halogenated aromatic contaminants similar to BDE15, such as
chlorobenzenes, has been shown to account for o1% by
mass of the influent pollutant mass [38]. Similar
adsorption losses are expected for BDE15 and its
metabolites, thus allowing a quantitative analysis of
anaerobic biodegradation products in the FF-PFBR
effluent. At steady state, values of the various process
parameters will approach a constant value with respect
to time. Among the process variables measured in the

Table 1
Concentrations of BDE15 and its anaerobic microbial biodegradation products (in moles/L795% confidence limits for triplicate
trials) in bioreactor effluent after 3.4 and 6.8 h HRTs. Molecular formulas are those of potential metabolites corresponding to the
structures shown in Fig. 2
3.4 h HRT

6.8 h HRT

6.9770.64 10 (88%)
1.6070.12 10 8 (2.0%)

3.1270.38 10 7 (39%)
2.4074.70 10 9 (0.3%)

Percent of mass balance of BDE15 in influent is shown in parentheses. ND=below method detection limit at a signal-to-noise ratio of 3.

S. Rayne et al. / Water Research 37 (2003) 551–560

first-order decay as expected for unimolecular photochemical processes, to BDE3 and ultimately to the
parent DE (Fig. 3). Similar results were found when the
solvent was changed to 100% CH3OH. Both solvent
systems provided good first-order BDE15 decay fits
(R2 ¼ 0:993 and 0.997 for CH3CN and CH3OH,
respectively) with an increase in the rate constant from
1.98 10 2 to 3.10 10 2 min 1 in moving from
CH3CN to CH3OH. This is consistent with the proposed
photochemical mechanism involving homolytic cleavage
of the aryl–Br bond (paths a and c in Fig. 4), resulting in
formation of an aryl radical which subsequently
abstracts a H-atom from the organic solvent. Hence,
because CH3OH is a better H-donor than CH3CN (i.e.,
a weaker C–H bond strength on the methyl function in
CH3OH), the aryl radical resulting after absorption of a
photon and C–Br bond dissociation can more readily
abstract a H-atom from the solvent to complete the
reductive debromination process. If the aryl–Br bond
was cleaving heterolytically to produce bromide (Br )
and an aryl carbocation, we would expect to find the
hydroxylated (in H2O/CH3CN) or methoxylated (in
CH3OH) products resulting from such photonucleophilic substitution [40]. As these products were not observed
by GC or 1H-NMR, homolytic cleavage is likely the sole
operative pathway for BDE15 aryl–Br bond cleavage.

Concentration (mM)








1.50 0


















% Mass Balance

Concentration (mM)



% Mass Balance

present study were pH, temperature, and concentrations

of the major nutrients (COD, NO
3 , NO2 , NH3, TDN,
PO4 , and TDP). During the 8-month colonization
period and the subsequent 8-week period of BDE15
biodegradation trials, pH and temperature were always
constant in both the influent and effluent at 7.55–7.60
and 21.570.51C, respectively. The constant pH is
indicative of the high buffering capacity of the influent
feed (1.5 mM NaHCO3), which maintains the pH in the
range optimum for microbial biodegradation (pH=7–8)
[19]. However, for the first 2 months of colonization,
nutrients (N and P) were retained within the reactors as
evidenced by the inability to complete a mass balance
for these compounds (B20–80% accounted for in the
effluent). During the period from two to 8 months after
colonization, nitrogen and phosphorus mass balances
were maintained in the influent and effluent, as was
observed during the subsequent 8-week period where
BDE15 investigated. During the 8-week period of trials,
approximately 6972% and 5571% of the influent
COD of 800 mg/L remained after 3.4 and 6.8 h HRTs,
respectively, demonstrating a significant degree of
mineralization of the CH3OH substrate. At these
influent levels of COD, anaerobic conditions were
established quickly in the reactors because the influent
DO concentration was always o5 mg/L.
Concentrations of reduced and oxidized forms of
dissolved nitrogen were also monitored during this 8week period. Provided anaerobic conditions were prevalent within the reactors, the influent NH+
4 feed should
appear in the effluent also in this reduced form, and not

as NO
3 or NO2 . Only NH3 was detected in the effluent

at 9171% of the influent feed, with no NO
3 or NO2
detected. Measurement of effluent TDN accounted for
9875% of the influent nitrogen mass, suggesting that
loss of nitrogen to N2 was minimal and that the
microbial population was at steady state, such that
there was no net retention of nitrogen in the reactor
biomass. Additionally, 9872% of the influent TDP and
6271% of the influent SRP was recovered in the
effluent, suggesting conversion of the simple PO3
influent feed to more complex forms of dissolved
phosphorus and no net retention of phosphorus in the
reactor biomass. Phosphorus was also the limiting
nutrient in the synthetic waste stream, consistent with
conditions observed in many natural and engineered
aquatic environments [19,39]. These nutrient results, in
addition to the COD data, indicate that the reactors
were at steady state and that there was no net
accumulation of biomass within the reactors over the
period when the BDE15 biodegradation studies were
The photochemistry of BDE15 was also examined in
organic and aqueous solvent systems at an initial
concentration of 500 mg/L (1.52 mM). In 100%
CH3CN, BDE15 undergoes strict debromination, via



Time (min)
Fig. 3. Time resolved photolyses of BDE15 at 300 nm in 100%
CH3CN and 100% CH3OH.

S. Rayne et al. / Water Research 37 (2003) 551–560

















Fig. 4. Photodegradation pathways of BDE15 in aqueous and organic solvent systems.

Previous work with the parent diphenyl ether and its
substituted analogs has shown the existence of a photoFries type rearrangement following C–O bond cleavage
[41,42], which would lead to 4-bromo-40 -hydroxybiphenyl (BHBP). One would also expect another photo-Fries
product analogous to BHBP, except that the hydroxy
group is in the ortho, and not para position (as in
BHBP), relative to the biphenyl linkage. Unfortunately,
an analytical standard was not available for this orthophoto-Fries product, although the absence of unassigned peaks in the gas chromatogram and 1H-NMR
spectra suggests it was not present. 4,40 -dibromobiphenyl (DBB), bromobenzene (BB), and 4-bromophenol
(BP) would also result from C–O bond cleavage; DBB
being the result of coupling between the two radical
intermediates leading to BB. The lack of these products
suggests path b was not operative.
Duplicate trials were also performed in H2O:CH3CN
(1:1 v/v) to model the aqueous photochemical fate of
BDE15 and see if it differed from that in organic
solvents. The solubility of BDE15 is sufficiently low to
prevent dissolution in higher concentrations of water
(e.g. 2:1 v/v H2O:CH3CN) for the photochemical
studies, although such higher concentrations of water
are desired as they more accurately represent environmental conditions. After 30 min irradiation, the photoproduct distribution of BDE15 in H2O:CH3CN (1:1 v/v)

was as follows: 7372% BDE15, 2571% BDE3, and
1.370.2% DE. Thus, reductive debromination is much
slower in the aqueous system compared to 100%
CH3CN and CH3OH, where only 51% and 41%,
respectively, of the starting material remained after
30 min. This is consistent with the lack of H-atom
donating ability by H2O as the O–H bond strengths are
too great in water to allow H abstraction. When
combined with the relatively poor H-donating properties
of the CH3CN co-solvent, irradiation in H2O:CH3CN
(1:1 v/v) produced the slowest rate constant among the
three solvent systems. As shown in Fig. 4, no products
other than strict reductive debromination were observed
in any of the solvent systems. This lack of photochemical C–O bond cleavage in BDEs 15 and 3, while taking
place when Cl or F atoms are present on the diphenyl
ether system [41,42], suggests that the aryl–halogen bond
strength (B350–390 kJ/mol; [40]) in the chlorinated and
fluorinated diphenyl ethers is of the same order as the C–
O bond. In contrast, for PBDEs the strength of the aryl–
Br bond (B335 kJ/mol; [40]) must be significantly lower
than that of the C–O bond such that no rearrangement
or cleavage products are observed. Based on these
findings, the C–O bond strength in PBDEs likely resides
between B335–390 kJ/mol.
Anaerobic microbial and photolytic debromination of
PBDEs has both advantages and disadvantages in

S. Rayne et al. / Water Research 37 (2003) 551–560

environmental terms. While little is known regarding the
congener specific toxicity of PBDEs, octanol–water
partition coefficients for this group of compounds
decrease with decreasing bromination [43]. Hence,
debromination through natural or engineered anaerobic
or photochemical processes may reduce the accumulation potential of PBDEs in biota and sediments. PBDE
debromination could also be utilized in wastewater
treatment plant processes using sequential anaerobic–
aerobic biological treatment (e.g. A2O, Bardenpho), as
has recently been demonstrated on PCBs [44,45]). These
coupled anaerobic–aerobic processes may allow complete mineralization of PBDEs in natural and engineered

4. Conclusions
A model polybrominated diphenyl ether, 4,40 -dibromodiphenyl ether (BDE15), has been shown to undergo
exclusive and sequential reductive debromination under
anaerobic microbial conditions in a plug-flow bioreactor, and via photolysis in organic and aqueous solvents,
resulting in formation of the parent diphenyl ether
system. No evidence of other biological or photochemical degradation products resulting from aryl-oxygen
bond cleavage or alternative pathways were observed.
These findings may help elucidate the environmental fate
of PBDEs and provide for advances in treatment
processes that will reduce fluxes of PBDEs into the

SR extends special thanks to Peter Wan for guidance,
financial assistance, and expertise on the photochemical
studies. MGI thanks the Department of Fisheries and
Oceans Canada (DFO) and the Toxic Substances
Research Initiative (TSRI) for funding. SR and MGI
thank Maike Fischer for the HRGC–HRMS analyses.

[1] World Health Organization. Environmental health criteria
162: brominated diphenyl ethers. Geneva: AB Firmatrych,
[2] Ikonomou MG, Rayne S, Addison RF. Exponential
increases of the brominated flame retardants, polybrominated diphenyl ethers, in the Canadian arctic from 1981 to
2000. Environ Sci Technol 2002;36(9):1886–92.
[3] Ikonomou MG, Rayne S, Fischer M, Fernandez MP,
Cretney W. Occurrence and congener profiles of polybrominated diphenyl ethers (PBDEs) in environmental
samples from coastal British Columbia, Canada. Chemosphere 2002;46(5):649–63.


[4] Meironyte D, Noren K, Bergman A. Analysis of polybrominated diphenyl ethers in Swedish human milk. A
time-related trend study, 1972–1997. J Toxicol Environ
Health, A 1999;58:329–41.
[5] Allchin CR, Law RJ, Morris S. Polybrominated diphenylethers in sediments and biota downstream of potential
sources in the UK. Environ Pollut 1999;105:197–207.
[6] Christensen JH, Platz J. Screening of polybrominated
diphenyl ethers in blue mussels, marine and freshwater
sediments in Denmark. J Environ Monitoring 2001;
[7] de Boer J, van der Zande TE, Pieters H, Ariese F, Schipper
CA, van Brummelen T, Vethaak AD. Organic contaminants and trace metals in flounder liver and sediment from
the Amsterdam and Rotterdam harbours and off the
Dutch coast. J Environ Monitoring 2001;3(4):386–93.
[8] Strandberg B, Dodder NG, Basu I, Hites RA. Concentrations and spatial variations of polybrominated diphenyl
ethers and other organohalogen compounds in Great
Lakes air. Environ Sci Technol 2001;35(6):1078–83.
[9] Hagenmaier H, She J, Benz T, Dawidowsky N, Dusterhoft
L, Lindig C. Analysis of sewage sludge for polyhalogenated dibenzo-p-dioxins, dibenzofurans, and diphenylethers. Chemosphere 1992;25(7–10):1457–62.
[10] Hale RC, La Guardia MJ, Harvey EP, Gaylor MO,
Mainor TM, Duff WH. Persistent pollutants in landapplied sludges. Nature 2001;412:140–1.
[11] Asplund L, Athanasiadou M, Sjodin A, Bergman A,
Borjeson H. Organohalogen substances in muscle, egg and
blood from healthy Baltic salmon (Salmo salar) and Baltic
salmon that produced offspring with the M74 syndrome.
Ambio 1999;28:67–76.
[12] Buser HR. Polybrominated dibenzofurans and dibenzopara-dioxins-thermal-reaction products of polybrominated
diphenyl ether flame retardants. Environ Sci Technol 1986;
[13] Rayne S, Ikonomou MG. Reconstructing source PBDE
congener patterns from semipermeable membrane devices
in the Fraser River, British Columbia, Canada: comparison to commercial mixtures. Environ Toxicol Chem, 2002,
in press.
[14] Arias PA. Brominated flame-retardants—an overview. The
Second International Workshop on Brominated FlameRetardants May 14–16, Stockholm, Sweden, 2001. p. 3–4.
[15] Hooper K, McDonald TA. The PBDEs: an emerging
environmental challenge and another reason for breastmilk monitoring programs. Environ Health Perspect 2000;
[16] de Boer J. Polybrominated biphenyls and diphenylethers.
In: Paasivirta J, editor. The handbook of environmental
chemistry. Part K: New types of persistent halogenated
compounds, vol. 3. Berlin: Springer, 2000.
[17] Bouwer EJ, McCarty PL. Removal of trace chlorinated
organic compounds by activated carbon and fixed-film
bacteria. Environ Sci Technol 1982;16(12):836–43.
[18] Konopka A, Zakharova T, Oliver L, Camp D, Turco RF.
Biodegradation of organic wastes containing surfactants in
a biomass recycle reactor. Appl Environ Microbiol 1996;
[19] Viessman J, Warren, Hammer MJ. Water supply and
pollution control, 5th ed. New York: Harper Collins, 1993.

Related documents

wat res 37 2003 551 560
environ int 32 2006 575 585
environ toxicol chem 21 2002 2292 2300
environ sci technol 43 2009 7992 7993
science news ringed seals
j environ sci health a 45 2010 1322 1346

Related keywords