Degradation of fluorotelomer alcohols: A likely atmospheric source of perfluorinated carboxylic acids

Article abstract of DOI:10.1021/es049860w

Human and animal tissues collected in urban and remote global locations contain persistent and bioaccumulative perfluorinated carboxylic acids (PFCAs). The source of PFCAs was previously unknown. Here we present smog chamber studies that indicate fluorotelomer alcohols (FTOHs) can degrade in the atmosphere to yield a homologous series of PFCAs. Atmospheric degradation of FTOHs is likely to contribute to the widespread dissemination of PFCAs. After their bioaccumulation potential is accounted for, the pattern of PFCAs yielded from FTOHs could account for the distinct contamination profile of PFCAs observed in arctic animals. Furthermore, polar bear liver was shown to contain predominately linear isomers (>99%) of perfluorononanoic acid (PFNA), while both branched and linear isomers were observed for perfluorooctanoic acid, strongly suggesting a sole input of PFNA from "telomer"-based products. The significance of the gas-phase peroxy radical cross reactions that produce PFCAs has not been recognized previously. Such reactions are expected to occur during the atmospheric degradation of all polyfluorinated materials, necessitating a reexamination of the environmental fate and impact of this important class of industrial chemicals.

Full text of DOI:10.1021/es049860w

Environ. Sci. Technol. 2004, 38, 3316-3321
with carbon chain lengths of between 9 and 13 are more
Degradation of Fluorotelomer
Alcohols: A Likely Atmospheric
Source of Perfluorinated Carboxylic
Acids
bioaccumulative than PFOA (13). PFCAs with chain lengths
of 9-11 are present in wildlife at higher concentrations than
PFOA, while chain lengths of 12 and 13 are comparable in
concentration to PFOA (3).
The source of most PFCAs observed in the environment
is unclear (1). There are no known natural sources of long-
chain PFCAs. Water-soluble PFCA salts are used in the
processing of fluoropolymers, and as such may be expected
to enter local aquatic environments directly (14). However,
it is difficult to explain how involatile PFCA salts would be
transported to remote regions because they are removed
from the atmosphere via wet and dry deposition on a time
scale of a few days (15). The simplest explanation for their
ubiquity in biota in remote regions is that some unknown
precursor is emitted to the atmosphere and ultimately
degrades to yield PFCAs.
D A V I D A . E L L I S , †
J O N A T H A N W . M A R T I N , †
A M I L A O . D E S I L V A , †
S C O T T A . M A B U R Y, * , †
M I C H A E L D . H U R L E Y, ‡
M A D S P . S U L B A E K A N D E R S E N , ‡ , § A N D
‡
T I M O T H Y J . W A L L I N G T O N
Department of Chemistry, University of Toronto,
80 St. George Street, Toronto, Ontario, Canada M5S 3H6,
Ford Motor Company, SRL 3083, P.O. Box 2053,
Dearborn, Michigan, 48121-2053, and
Department of Chemistry, University of Copenhagen,
Universitetsparken 5, DK-2100 Copenhagen, Denmark
Fluorotelomer alcohols (FTOHs) are plausible candidates
for the “unknown” PFCA precursors. FTOHs have the generic
formula F(CF2)nCH2CH2OH, where n is an even number, and
are named according to the relative number of fluorinated
to hydrogenated carbons: e.g., F(CF2)8CH2CH2OH is 8:2
FTOH. FTOHs are volatile, appear to be ubiquitous in the
North American atmosphere (17-135 pg m-3 (16, 17)), have
an atmospheric lifetime (20 days) sufficient for widespread
hemispheric distribution (18), and have a long perfluoroalkyl
moiety. The worldwide production is ∼5 × 106 kg yr-1, with
40% produced in North America (18).
Human and animal tissues collected in urban and remote
global locations contain persistent and bioaccumulative
perfluorinated carboxylic acids (PFCAs). The source of PFCAs
was previously unknown. Here we present smog chamber
studies that indicate fluorotelomer alcohols (FTOHs)
can degrade in the atmosphere to yield a homologous
series of PFCAs. Atmospheric degradation of FTOHs is likely
to contribute to the widespread dissemination of PFCAs.
After their bioaccumulation potential is accounted for, the
pattern of PFCAs yielded from FTOHs could account for
the distinct contamination profile of PFCAs observed in arctic
animals. Furthermore, polar bear liver was shown to
contain predominately linear isomers (>99%) of perfluo-
rononanoic acid (PFNA), while both branched and linear
isomers were observed for perfluorooctanoic acid, strongly
suggesting a sole input of PFNA from “telomer”-based
products. The significance of the gas-phase peroxy radical
cross reactions that produce PFCAs has not been
recognized previously. Such reactions are expected to
occur during the atmospheric degradation of all polyfluorinated
materials, necessitating a reexamination of the environmental
fate and impact of this important class of industrial
chemicals.
While it is known that FTOHs are metabolically trans-
formed to PFOA in rats (19) and through microbial metabo-
lism (20), the atmospheric degradation mechanism of FTOHs
is unknown. Smog chamber experiments were performed to
investigate the possibility that FTOHs may degrade to yield
PFCAs in the atmosphere. Oxidation of FTOHs in the
atmosphere is initiated by reaction with OH radicals (18). In
the present work we employ Cl atoms to initiate the oxidation
of FTOHs. It is believed that the majority (>90%) of reactions
of both OH radicals and Cl atoms with FTOHs proceed via
hydrogen atom abstraction from the -CH2- group bearing
the alcohol functionality (18, 20). Hence, Cl atoms provide
a reasonable surrogate for OH radicals in a study of the
products of atmospheric oxidation of FTOHs.
Experimental Section
Chem icals. All reagents were of purity 97% or greater and
were used without further purification. The 4:2, 6:2, and 8:2
telomer alcohols, 2,2,3,3,4,4,5,5,5-nonafluoro-1-pentanol,
2,2,3,3,4,4,5,5,6,6,7,7,7-tridecafluoro-1-heptanol, and 2,2,3,3,-
4,4,5,5,6,6,7,7,8,8,9,9,9-heptadecafluoro-1-nonanol and per-
fluorohexanoic acid were purchased from Oakwood Research
Chemicals (West Columbia, SC). Heptadecafluorononanoic
acid, heptafluorobutyric acid, nonadecafluorodecanoic acid,
pentadecafluorooctanoic acid, perfluorododecanoic acid,
pentafluoropropionic acid, and tridecafluoroheptanoic acid
were purchased from Aldrich Chemical Co. (Milwaukee, WI).
Trifluoroacetic acid was purchased from Caledon Labora-
tories Ltd. (Georgetown, ON, Canada) and nonafluoropen-
tanoic acid from Fluka (Oakville, ON, Canada).
Introduction
The U.S. Environmental Protection Agency (EPA) has stated
its concern over the environmental effects of long-chain
PFCAs such as perfluorooctanoic acid (PFOA, C7F15COOH)
(1). PFOA is detectable in the blood of most humans and
animals worldwide (2-7), which is problematic because it
is only slowly eliminated in mammals (8), is potentially toxic
(9-11), has no known metabolic or environmental degrada-
tion pathway, and is potentially carcinogenic (12). PFCAs
Interm ediate Synthesis. Polyfluorinated acids and al-
dehydes were synthesized using literature procedures and
authenticated by use of IR, NMR, and mass spectrometry.
Fluorotelomer acids CxF2x+1CH2C(O)OH (FTACs) were syn-
thesized using the method of Le´veˆque et al. (22). Perfluori-
nated aldehydes CxF2x+1C(O)H (PFALs) were synthesized by
the method of Miller et al. (23). Fluorotelomer aldehydes
* Corresponding author phone: (416) 978-1780; fax: (416) 978-
3596; e-mail: smabury@chem.utoronto.ca.
† University of Toronto.
‡ Ford Motor Co.
§
University of Copenhagen.
9
3 3 1 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004
10.1021/es049860w CCC: $27.50
2004 Am erican Chem ical Society
Published on Web 05/12/2004

CxF2x+1CH2C(O)H (FTALs) were synthesized by the method
of Achilefu et al. (24).
from subsistence hunters in the Ittoqqortoormiit/ Scoresby
Sound area of Greenland between 1999 and 2001. The livers
were removed from the animals post mortem and stored
separately in polyethylene plastic bags. Liver samples were
kept at outdoor temperatures (-30 to -5 °C) until storage
in a freezer (-20 to -10 °C). Tissue samples were shipped
at -20 °C from Greenland and stored in a freezer (-20 to
-10 °C) until analysis. Approximately 1-2 g (wet) of each
liver was used for analyses. Extraction of PFCAs from samples
was performed in a manner described elsewhere (3) using
an ion pairing agent. The dried extracts were reconstituted
in 50 mL of distilled water for derivatization with 2,4-
difluoroaniline. The derivatization method employed has
been reported elsewhere (28, 29). Structural isomers of PFOA
and PFNA in the volatile, derivatized sample extracts were
resolved by GC/ MS and determined by selected ion moni-
toring of the molecular ion. For comparison, an authentic
standard of PFOA provided by 3M Co. which had been
produced by an electrochemical fluorination method was
analyzed. A paper describing details of the analytical
methodology and comprehensive findings of structural
isomers of PFCAs in polar bears is in preparation (30).
General Sm og Cham ber Methods. Experiments were
performed in 750 Torr of air at 296 K in a 140-L Pyrex reactor
interfaced to a Mattson Sirus 100 FTIR spectrometer described
in detail elsewhere (25). The reactor was surrounded by 22
fluorescent blacklamps (GE F15T8-BL) described elsewhere
(26). The oxidation of 4:2 FTOH, 6:2 FTOH, or 8:2 FTOH was
initiated by reaction with Cl atoms generated by photolysis
of molecular chlorine. Atmospheric oxidation of FTOHs is
initiated by reaction with OH radicals. In the present work,
chlorine atom initiation was used because Cl atoms are easier
to generate and they react with FTOHs in the same manner
as OH radicals. Initial reaction mixtures consisted of 5-7
mTorr of 4:2 FTOH, 6:2 FTOH, or 8:2 FTOH and 80-160
mTorr of Cl2 in 750 Torr of dry (<1% relative humidity) air
diluent at 296 K. Reaction mixtures were subjected to 0.5-15
min of UV irradiation, leading to consumption of FTOH in
the range from 66% to >98%. FTOH loss was monitored using
IR spectroscopy (18). The formation of COF2 and CF3OH
products was monitored via IR absorption spectroscopy. The
formation of other carbon-containing products was moni-
tored as described below. Photolysis of intermediate products
was not observed under these conditions and is reported
elsewhere (27).
Results and Discussion
Atm ospheric Chem istry. The in situ analysis of FTOH
photooxidation products by FTIR indicated the presence of
small amounts (<10% yield) of one, or more, gas-phase PFCA.
Unfortunately, due to the similarity of the IR spectra of PFCAs,
and their low yields in the chamber (<10%), FTIR spectros-
copy was not well suited to identification or quantification
of these species. Offline analysis of the trapped smog chamber
products by LC/ MS-MS indicated that each FTOH yielded
a PFCA having an intact perfluorinated chain (e.g., 8:2 FTOH
yielded PFNA). Interestingly, LC/ MS-MS analysis also
indicated production of a homologous series of shorter chain
PFCAs. For example, Cl atom initiated oxidation of 8:2 FTOH
yielded the entire suite of PFCAs ranging from trifluoroacetic
acid (TFA) to PFOA (in addition to perfluorononanoic acid).
In addition to PFCAs, 19F NMR analysis indicated the presence
of perfluorinated acid fluorides (i.e., RCFOs; Figure 1). Acid
fluorides will be hydrolyzed by electrospray conditions typical
of LC/ MS-MS to give the corresponding carboxylic acids;
thus, it remains ambiguous whether the homologous series
of perfluorinated acids (e.g., TFA through PFOA) are in fact
formed from the hydrolysis of acid fluorides. This discussion,
however, is purely semantic because in the troposphere, the
sole atmospheric fate of acid fluorides is hydrolysis to the
corresponding PFCA. For this reason, further discussion will
simply refer to PFCAs.
As indicated in Table 1 a wide variety of products were
identified following the Cl atom oxidation of the 4:2, 6:2, and
8:2 FTOHs. When techniques as sensitive as GC/ MS and LC/
MS-MS are used, it is important to consider possible artifacts
caused by contamination of the samples. There are two
reasons for believing that the results given in Table 1 reflect
the true oxidation products of the FTOH rather than problems
associated with contamination. First, samples of FTOH/ Cl2/
air mixtures taken prior to UV irradiation did not reveal any
evidence of the products listed in Table 1. Second, as seen
from Table 1, in experiments using 4:2 FTOH no products
with carbon length greater than C6 were observed and in
experiments using 6:2 FTOH no products with carbon length
greater than C8 were observed. The chamber was exposed to
8:2 FTOH and 6:2 FTOH prior to experiments involving 4:2
FTOH, and to 8:2 FTOH prior to experiments involving 6:2
FTOH. The absence of longer chain products from experi-
ments using shorter chain precursors indicates the absence
of cross contamination.
Interm ediate Collection. Smog chamber samples were
collected in duplicate at several time points through stainless
steel tubes using precalibrated portable pumps that sampled
5 L of air at a rate of 1 L min-1 through 200-mg Amberlite
XAD-2 cartridges (Supelco, Bellefonte, PA). XAD cartridges
were eluted with methanol or ethyl acetate and analyzed by
gas chromatography/ mass spectrometry (GC/ MS), liquid
chromatography/ tandem mass spectrometry (LC/ MS-MS),
and 19F NMR. A second plug of XAD-2 resin downstream of
the sample media was analyzed separately in the same
manner. No analytes were observed from this sample, which
suggested that trapping on the XAD-2 resin was quantitative.
Interm ediate Analysis by 19F NMR. NMR was performed
on ethyl acetate extracts, following addition of 300 µL of
deuterated acetone. All spectra were obtained at 25 °C on a
Varian Unity 500 three-channel spectrometer, operating at
470.297 MHz, equipped with a 5-mm Nalorac 19F proton
decoupling probe. An OPH (0, 0°; 1, 90°; 2, 180°; 3, 270°)
pulse phase was performed. Free induction decays (FIDs)
were zero filled. Chemical shifts were recorded relative to
the peak for CFCl3 (0.000 ppm).
Interm ediate Analysis by GC/MS. Analysis of target
analytes was performed by GC/ MS (Hewlett-Packard 5973
mass-selective detector) using chemical ionization in positive
(PCI) and negative (NCI) modes with methane as the reagent
gas (PCI, 1 mL min-1; NCI, 2 mL min-1). Gas chromatographic
separation was performed on a 30m DB-Wax column (0.25
mm i.d., 250-m film thickness, J&W Scientific, Folsom, CA)
using helium as the carrier gas. Pulsed splitless injections (1
L) were performed at an initial pressure of 40 psi and 200 °C,
returning to 10 psi at 1 min, and followed by an injector
purge. The initial oven temperature was 60 °C for 1 min,
ramped at 3 °C min-1 to 75 °C, and subsequently at 20 °C
min-1 to 240 °C, followed by a 1-min hold. All mass spectral
data were acquired in full scan mode.
Interm ediate Analysis by LC/MS-MS. Liquid chroma-
tography was performed on a Genesis C8 column using an
isocratic elution program, consisting of 70% methanol/ 30%
water (0.005 M ammonium acetate). Mass spectral data were
obtained in full scan and multiple reaction monitoring (MRM)
modes using a triple-quadrupole mass spectrometer equipped
with an electrospray source operating in negative mode
(Micro Quattro, Micromass, Manchester, U.K.).
Structural Isom er Determ ination of PFCAs in Polar
Bears. Liver samples from seven polar bears were obtained
The overall yield of PFCAs from 8:2 FTOH, at an
intermediate time point in the degradation, was ∼5%, while
9
VOL. 38, NO. 12, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 3 1 7

FIGURE 1. OH-initiated oxidation pathways for telomer alcohols, where n ) 2, 4, or 6, leading to formation of PFCAs. R and R′ are nonspecific
alkyl species in the smog chamber, similar species of which would play an identical role. x is an integer less than n. Significant previously
unreported reactions are indicated with large arrows.
the yield of individual PFCA homologues was proportional
to the chain length. For example, both PFNA and PFOA were
produced at carbon yields of ∼1.5%, while the yield of each
shorter PFCA was less than 0.5%. The overall yield of PFCAs
is expected to be greater than 5% because many intermediates
were still observed in these samples, a portion of which would
then yield additional PFCAs upon further oxidation. Within
experimental uncertainties, we were able to account for 100%
of the reacted FTOH in terms of the listed products for the
8:2 FTOH experiments (see Table 1).
The data in Table 1 show that the chlorine atom initiated
oxidation of FTOHs leads to the formation of small, but
significant, yields of PFCAs. This raises the following ques-
tion: What is the mechanism by which PFCAs are formed
in the chamber, and to what degree does the oxidation of
FTOHs in the smog chamber simulate that in the atmosphere?
The proposed mechanism given in Figure 1 is the simplest
explanation for their production. As noted in the Introduction,
the products observed following Cl atom initiated oxidation
of FTOHs in the smog chamber should be the same as those
formed following OH radical initiated oxidation of FTOHs in
the atmosphere. Hence, the mechanism shown in Figure 1
deduced from the results of Cl atom initiated experiments
should provide an accurate picture of the atmospheric
degradation mechanism of FTOHs in low NOx environments.
The initial step involves the reaction of OH• with FTOHs
to give CxF2x+1CH2CH•OH in 100% yield (18), whose sole
atmospheric fate is reaction with O2 to give the corresponding
FTAL CxF2x+1CH2CHO.
The FTAL then reacts with OH• to give acyl radicals which
subsequently add O2 to give an acyl peroxy radical, CxF2x+1
-
CH2C(O)OO•. In the atmosphere, this will then react with
NO, NO2, HO2, or other peroxy radicals. In the present
experiments (conducted in the absence of NOx) the fate of
CxF2x+1CH2C(O)OO• will be reaction with HO2• or other peroxy
radicals. By analogy with the observed behavior of CH3-
C(O)OO• (30) and C2F5C(O)OO• (32), reaction with HO2 will
yield the acid CxF2x+1CH2C(O)OH in approximately 25% yield.
Self-reaction of CxF2x+1CH2C(O)OO• will give CxF2x+1
-
CH2C(O)O•, which will eliminate CO2, add O2 to give a peroxy
radical, undergo reaction with other peroxy radicals to give
an alkoxy radical, and finally react with O2 to yield the
perfluorinated aldehyde CxF2x+1CHO. As shown in Figure 1,
the perfluorinated aldehyde is expected to undergo analogous
reactions to the FTAL, leading to formation of the perflu-
orinated acid CxF2x+1C(O)OH or the perfluorinated alkoxy
radical CxF2x+1O•. The atmospheric fate of CxF2x+1O• is
elimination of COF2 and, in the presence of O2, formation of
the Cx-1F2x-1OO• peroxy radical. While the dominant fate of
C
x-1F2x-1OO• is expected to be reaction with other peroxy
radicals to form alkoxy radicals, one channel of the reaction
with R-H-containing peroxy radicals will lead to formation
of perfluorinated alcohols (Figure 2, eqs a and b) (33, 34).
9
3 3 1 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004

(34). Reaction of C2F5C(O)O2 radicals with NO2 produces the
thermally unstable peroxynitrate C2F5C(O)O2NO2, which will
decompose back to reactants or possibly hydrolyze to give
C2F5C(O)OH and HNO3 (34). Competition with NO for the
available peroxy radicals will decrease the overall PFCA yield,
while competition with NO2 may decrease the PFCA yield
depending on the fate of the peroxynitrates. Given the
abundance of NO, it is unlikely that the chemistry outlined
in Figure 1 will play a role in typical urban air masses.
However, in remote areas (e.g., Arctic environments and over
the oceans) the concentrations of HO2 and other peroxy
radicals are comparable to those of NO and NO2 (35). Given
the broad similarity in rate constants for reaction of acyl
peroxy and peroxy radicals with NO, NO2, HO2, and other
peroxy radicals (31), it is likely that reactions such as those
outlined in Figure 1 will play an important role in the
atmospheric oxidation of FTOHs in remote locations.
In the above discussion we make the implicit assumption
that atmospheric loss of PFALs is dominated by reaction
with OH radicals (see Figure 1). However, at least two other
loss mechanisms need to be considered, photolysis and
reaction with water.
It is well established that PFALs absorb strongly in the
region 300-350 nm where there is significant solar flux
(36-39). Unfortunately, there are no photodissociation
quantum yield data available, and it is not possible at this
time to assess the importance of photolysis as an atmospheric
loss mechanism for PFALs. By analogy to the case for CH3-
CHO, it is expected that photolysis of PFALs will proceed via
C-C bond scission, giving perfluorinated alkyl and HCO
radicals as products. Photolysis will preclude formation of
PFCAs via the acyl peroxy + HO2 reaction route, but will
contribute to PFCA formation via the CxF2x+1O2 + RO2 route
(see Figure 1). Further research is needed to establish the
rate and products of atmospheric photolysis of PFALs.
PFALs are supplied commercially as hydrates (CxF2x+1CH-
(OH)2). In view of the stability of the hydrates it seems
reasonable to speculate that reaction with water to form
hydrates may be a significant atmospheric fate of PFALs.
Very little is known about the formation or fate of such
hydrates, and it is not possible at this time to assess their
environmental significance. It can be argued that the
formation of hydrates will preclude formation of PFCAs via
either the acyl peroxy + HO2 or the CxF2x+1O2 + RO2 route
and hence reduce PFCA formation from FTOH degradation.
Alternatively, it can be argued that formation of hydrates
opens up an efficient route to PFCA formation (reaction with
OH radicals and subsequent reaction of the diol radical with
O2) that will increase PFCA formation from FTOH degrada-
tion. Further research is needed to clarify the importance of
hydrate formation in the atmospheric chemistry of PFALs.
Long-range transport and degradation of FTOHs could
explain the presence of long-chain PFCAs in arctic animals,
and may also explain the reported contamination profile.
Martin et al. (3) observed consistent PFCA concentration
profiles in polar bears and other arctic animals whereby, in
general, there was a decreasing trend in PFCA concentrations
with increasing chain length. This trend can be explained by
the fact that the atmospheric concentration of FTOHs
decreases with increasing chain length (16) and explains the
same overall decreasing trend of liver PFCA concentrations
with increasing chain length. Despite the overall decreasing
trend with perfluoroalkyl chain length, Martin et al. observed
that odd-chain length PFCAs exceeded the concentration of
the shorter even-chain-length PFCAs in the wildlife samples.
We hypothesize that FTOH degradation and subsequent
bioaccumulation can explain this observation. Using the
degradation of 8:2 FTOH as an example, PFOA and PFNA are
produced in equal carbon yields, but given that PFNA is more
bioaccumulative than PFOA (13), animals that are exposed
TABLE 1. Products Observed in the Cl Atom Oxidation of
FTOHs in Aira
4:2
6:2
8:2 C mass balance,
e
productf
FTOH FTOH FTOH 8:2 FTOH (%)
Perfluorinated Acids
PFNA
PFOA
x
x
x
x
x
x
x
x
1.6
1.5
0.32
0.24
0.10
<0.1
<0.1
<0.1
perfluoroheptanoic acid
perfluorohexanoic acid
perfluoropentanoic acid
perfluorobutanoic acid
perfluoropropionic acid
trifluoroacetic acid
x
x
x
x
x
x
x
x
x
x
Fluorotelomer Acids
8:2 FTCA (C8F17CH2COOH)
6:2 FTCA (C6F13CH2COOH)
4:2 FTCA (C4F9CH2COOH)
x
x
x
x
x
26
6
x
x
Telomer Aldehydes
8:2 FTAL (C8F17CH2CHO)
6:2 FTAL (C6F13CH2CHO)
4:2 FTAL (C4F9CH2CHO)
x
x
Perfluoroaldehydes
b
C8 PFAL (C8F17CHO)
C6 PFAL (C6F13CHO)
C4 PFAL (C4F9CHO)
21d
3.9
6
b
x
b
x
Odd Telomer Alcoholse
8:1 FTOH
6:1 FTOH
4:1 FTOH
x
x
Unreacted FTOH
8:2
Single Carbon Products
COF2
CF3OH
CO2
x
x
x
x
x
x
x
x
x
22
2.6
<21c
total
91-112
a
In the absence of NO. “x” indicates the analyte was detected in
b
the corresponding experim ent. Identification based on FTIR spectra
relative to those of authentic standards. CO2 m ay be form ed from
c
unwanted reactions on cham ber walls; the observed yield is the upper
d
lim it. Quantification determ ined by the proportional TIC response in
e
GC/EI/MS. For further spectroscopic evidence concerning the 4:2 FTOH
please see Hurley et al. (27). f The possibility that observed products
were artifacts caused by sam ple contam ination was considered.
However, sam ples of FTOH/Cl2/air m ixtures taken prior to UV irradiation
did not reveal any evidence of the products listed in the table.
The fate of Cx-1F2x-1OH will be heterogeneous elimination
of HF to give the acyl fluoride Cx-2F2x-3C(O)F (32), which
will hydrolyze in the atmosphere to give the PFCA
C
x-2F2x-3C(O)OH. The formation of the odd-chain-length
fluorotelomer alcohols 8:1 FTOH, 6:1 FTOH, and 4:1 FTOH
(Table 1) is explained by self-reaction of CxF2x+1CH2OO•, which
will proceed via both radical and molecular channels, with
the odd fluorotelomer alcohol being formed in the molecular
channel (Figure 2, eqs c and d).
Environm ental Im plications of Atm ospheric Chem istry.
It is reasonable to assume that reactions similar to those
given in Figure 1 will occur in the atmosphere and, hence,
that emission and atmospheric degradation of FTOHs will
result in the global dissemination of PFCAs. It should be
noted that the present experiments were performed in the
absence of NOx, while in the real atmosphere NOx is present
(particularly in polluted urban air). The proposed mecha-
nisms by which PFCAs are formed (Figure 1) involve reaction
of peroxy radicals with either HO2 or other peroxy radicals.
In the presence of excess NOx the fate of peroxy radicals will
be reaction with either NO or NO2. The reaction of C2F5C(O)O2
radicals with NO proceeds via an “unzipping cycle” to
produce COF2, and does not lead directly to PFCA formation
9
VOL. 38, NO. 12, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 3 1 9

FIGURE 2. Mechanism leading to the production of perfluorinated aldehydes, eqs a and b, and to the odd-chain-length fluorotelomer
alcohols, eqs c and d.
FIGURE 3. Gas chromatogram with mass spectroscopic analysis showing an extracted and derivatized polar bear liver. The inset shows
a selective ion comparison between the isomers of an authentic PFOA standard produced electrochemically and PFOA detected in a polar
bear liver extract with the largest concentration of isomers observed for each of the extracts. No branched isomers are observed for PFNA.
The values in parentheses indicate the percent of acid present in the straight chain form.
to these two degradation products in the environment will
carry a higher load of PFNA than PFOA, with the assumption
that the only source of PFOA and PFNA would be from the
atmospheric degradation of the 8:2 FTOH.
ratio of 5:1, suggesting PFOA input from both sources. PFNA
analysis revealed no branched isomers. The hypothesis that
PFNA present in arctic animals is attributable to FTOHs (3)
is strongly supported by the absence of branched isomers
for PFNA; furthermore, the odd-chain-length/ even-chain-
length profile observed by Martin et al. (3) was clearly visible.
Thus, atmospheric degradation of FTOHs offers a quali-
tative explanation for the observed contamination of arctic
animals with PFOA and PFNA, which, for example, were
detected in polar bears between 10 and 110 ng g-1 (liver wet
mass), respectively. From the data in Table 1, it seems
reasonable to estimate a carbon PFCA yield from the
atmospheric oxidation of FTOHs in the range 1-10%. Using
the estimated flux of FTOHs into the Northern Hemisphere
Polyfluorinated materials are commercially produced by
two major methods, telomerization (the process used to
produce FTOHs) and electrochemical fluorination (14).
Telomerization leads predominately to straight polyfluori-
nated chains, while electrochemical fluorination produces
both linear and branched isomers. The livers of seven polar
bears were extracted and analyzed for their PFCA content
(Figure 3). Linear and branched PFOA was found to be
present, with the ratio of linear to branched isomers being
23:1. An authentic standard of electrochemical PFOA had a
9
3 3 2 0 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 12, 2004

(100-1000 t yr-1) (18), it follows that the atmospheric
oxidation of FTOHs gives rise to a PFCA flux in the range
1-100 t yr-1. Assuming an even distribution across the
Northern Hemisphere, this equates to the delivery of ap-
proximately 0.1-10 t yr-1 to the Arctic, which is comparable
to the annual Arctic loading of persistent organochlorine
pesticides, including hexachlorobenzene (1.8 t yr-1) (40).
Although bioaccumulation factors, spatial variation, and
accumulated burdens are also important, it is notable that
hexachlorobenzene is detectable in polar bears at a con-
centration similar to that of PFCAs (183 ng g-1 (mean wet
mass of adipose)) (3).
(14) Kissa, E. In Fluorinated Surfactants, Synthesis, Properties,
Applications; Schick, M. J., Fowkes, F. M., Eds.; Marcel Dekker:
New York, 1994.
(15) Hurley, M. D.; Sulbaek Andersen, M. P.; Wallington, T. J.; Ellis,
D. A.; Martin, J. W.; Mabury, S. A. J. Phys. Chem. A 2004, 108,
615-620.
(16) Martin, J. W.; Muir, D. C. G.; Moody, C. A.; Ellis, D. A.; Kwan,
W. C.; Solomon, K. R.; Mabury, S. A. Anal. Chem. 2002, 74, 584-
590.
(17) Stock, N. L.; Lau, F. K.; Ellis, D. A.; Martin, J. W.; Muir, D. C. G.;
Mabury, S. A. Environ. Sci. Technol. 2004, 38, 991-996.
(18) Ellis, D. A.; Martin, J. W.; Mabury, S. A.; Hurley, M. D.; Sulbaek
Andersen, M. P.; Wallington, T. J. Environ. Sci. Technol. 2003,
37, 3816-3820.
(19) Hagen, D. F.; Belisle, J.; Johnson, J. D.; Venkateswarlu, P. Anal.
Biochem. 1981, 118, 336-343.
(20) Dinglasan, M. J.; Ye, Y.; Edwards, E.; Mabury, S. A. Environ. Sci.
Technol., in press.
(21) Hurley, M. D.; Wallington, T. J.; Sulbaek Andersen, M. P.; Ellis,
D. A.; Martin, J. W.; Mabury, S. A. J. Phys. Chem. A 2004, 108,
1973-1979.
(22) Le´veˆque, L.; Le Blanc, M.; Pastor, R. Tetrahedron Lett. 1998, 39,
8857-8860.
(23) Miller, A. O.; Peters, D.; Zur, C.; Frank, M.; Miethchen, R. J.
Fluorine Chem. 1997, 82, 33-38.
(24) Achilefu, S.; Mansuy, L.; Selve, C.; Thiebaut, S. J. Fluorine Chem.
1995, 70, 19-26.
(25) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399-409.
(26) Taniguchi, N.; Wallington, T. J.; Hurley, M. D.; Guschin, A. G.;
Molina, L. T.; Molina, M. J. J. Phys. Chem. A 2003, 107, 2674-
2679.
(27) Hurley, M. D.; Ball, J. C.; Wallington, T. J.; Sulbaek Andersen,
M. P.; Ellis, D. A.; Martin, J. W.; Mabury, S. A. J. Phys. Chem. A
2004, in press.
(28) Scott, B. F.; Alaee, M. Water Qual. Res. J. Can. 1998, 33, 279-
293.
(29) Ellis, D. A.; Martin, J. W.; Muir, D. C. G.; Mabury, S. A. Analyst
2003, 128, 756-764.
(30) De Silva, A. O.; Mabury, S. A. Environ. Sci. Technol., manuscript
in preparation.
(31) Tyndall, G. S.; Cox, R. A.; Granier, C.; Lesclaux, R.; Moortgat, G.
K.; Pilling, M. J.; Ravishankara, A. R.; Wallington, T. J. Geophys.
Res.-Atmos. 2001, 106 (D11), 12157-12182.
(32) Sulbaek Andersen, M. P.; Hurley, M. D.; Wallington, T. J.; Ball,
J. C.; Martin, J. W.; Ellis, D. A.; Mabury, S. A. Chem. Phys. Lett.
2003, 381, 14-21.
(33) Wallington, T. J.; Schneider, W. F.; Worsnop, D. R.; Nielsen, O.
J.; Sehested, J.; Debruyn, W. J.; Shorter, J. A. Environ. Sci. Technol.
1994, 28, 320A-326A.
FTOH degradation is likely an important source of PFCA
pollution in remote areas. The importance of the peroxy
radical cross reactions, postulated to explain the formation
of PFCAs, has not been recognized previously. Such cross
reactions are expected to occur to some degree during the
atmospheric degradation of all polyfluorinated materials,
necessitating a reexamination of the environmental impact
of this class of important industrial chemicals.
Acknowledgments
This work was supported by a Strategic Project Grant from
the National Sciences and Research Council of Canada
(NSERC) and Ole John Nielsen (Copenhagen University) for
helpful discussion. M.P.S.A. thanks the Danish Research
Council (FUR) for a research stipend. The assistance of
substance hunters and of their local hunting and trapping
associations at all sample locations is greatly appreciated.
Thanks are due to Christian D. Sonne-Hansen, Denmark
Institute of Environmental Research, for his assistance in
the collection of wildlife samples. Derek C. G. Muir and
Christine Spencer are gratefully acknowledged for GC/ MS
assistance. Equal contribution of D.A.E. and J.W.M. toward
the research presented in this paper is noted.
Note Added in Proof
Sellevåg et al. (41) have reported an upper limit of 0.02 for
the photodissociation yield of CF3CHO suggesting, but not
proving, that photolysis of long chain PFALs does not preclude
formation of PFCAs via the acyl peroxy + HO2 reaction route.
(34) Sulbaek Andersen, M. P.; Hurley, M. D.; Wallington, T. J.; Ball,
J. C.; Martin, J. W.; Ellis, D. A.; Mabury, S. A.; Nielsen, O. J. Chem.
Phys. Lett. 2003, 379, 28-36.
(35) Yang, J.; Honrath, R. E.; Peterson, M. C.; Dibb, J. E.; Sumner, A.
L.; Shepson, P. B.; Frey, M.; Jacobi, H. W.; Swanson, A.; Blake,
N. Atmos. Environ. 2002, 36, 2523-2534.
(36) Meller, R.; Boglu, D.; Moortgat, G. K. STEP-HALOCSIDE/AFEAS
Workshop, March 23-25, 1993, Dublin, Ireland; Chemistry
Department, University College Dublin: Dublin, Ireland; 1993;
pp 130-138.
(37) Borkowski, R. P.; Ausloss, P. J. Am. Chem. Soc. 1962, 84, 4044-
4048.
(38) Lucazeau, G.; Sandorfy, C. J. Mol. Spectrosc. 1970, 35, 214.
(39) Rattigan, O. V.; Wild, O.; Cox, R. A. J. Photochem. Photobiol., A
1998, 112, 1-7.
(40) Cotham, W. E., Jr.; Bidleman, T. F. Chemosphere 1991, 22, 165-
188.
Literature Cited
(1) Hogue, C. Chem. Eng. News 2003, 81, 24-24.
(2) Masunaga, S.; Kannan, K.; Doi, R.; Nakanishi, J.; Giesy, J. P.
Dioxin 2002. Organohalogen Compd. 2002, 59, 888-891.
(3) Martin, J. W.; Smithwick, M. M.; Braune, B. M.; Hoekstra, P. F.;
Muir, D. C. G.; Mabury, S. A. Environ. Sci. Technol. 2004, 38,
373-380.
(4) Kannan, K.; Choib, J.-W.; Isekic, N.; Senthilkumar, K.; Kim, D.
H.; Masunaga, S.; Giesy, J. P. Chemosphere 2002, 49, 225-231.
(5) Hansen, K. J.; Clemen, L. A.; Ellefson, M. E.; Johnson, H. O.
Environ. Sci. Technol. 2001, 35, 766-770.
(6) Olsen, G. W.; Burris, J. M.; Burlew, M. M.; Mandel, J. H. J. Occup.
Environ. Med. 2003, 45, 260-270.
(7) Gilliland, F. D.; Mandel, J. S. Am. J. Ind. Med. 1996, 29, 560-568.
(8) Kudo, N.; Kawashima, Y. J. Toxicol. Sci. 2003, 28, 49-57.
(9) Berthiaume, J.; Wallace, K. B. Toxicol. Lett. 2002, 129, 23-32.
(10) Upham, B. L.; Deocampo, N. D.; Wurl, B.; Trosko, J. E. Int. J.
Cancer 1998, 78, 491-495.
(41) Sellevåg, S. R.; Kelly, T.; Sidebottom, H.; Nielsen, C. J. Phys.
Chem. Chem. Phys. 2004, 6, 1243.
(11) U.S. Environmental Protection Agency, Office of Pollution
Prevention and Toxics Risk Assessment Division, April 10, 2003.
(12) Biegel, L. B.; Hurtt, M. E.; Frame, S. R.; O’Connor, J. C.; Cook,
J. C. Toxicol. Sci. 2001, 60, 44-55.
(13) Martin, J. W.; Mabury, S. A.; Solomon, K. R.; Muir, D. C. G.
Environ. Toxicol. Chem. 2003, 22, 189-195.
Received for review January 27, 2004. Revised manuscript
received March 27, 2004. Accepted March 31, 2004.
ES049860W
9
VOL. 38, NO. 12, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 3 3 2 1