DEQ-CFW_00083481

.::::.....
Identification of Novel Perfluoroalkyl Ether Carboxylic Acids (PFECAs)
• ! Sulfonic• • Accurate Mass
i-Flight Mass Spectrometry
Mark Sttynar,*"Sonia Dagnino,'" Rebecca McMahen,' Shuang Liang" Andrew Lindstrom,'
Erik Andersen, ` Larry McMillan,` Michael Thurman," I]nma Ferrer,and Carol Ball:
National Exposure Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711,
United States
*Oak Ridge Institute for Science and Education, Oak Ridge, Tennessee 37831 United States
§National Caucus and Center on Black Aged, Inc., Durham, North Carolina 27713, United States
"Center for Environmental Mass Spectrometry, University of Colorado Boulder, Boulder, Colorado 80309, United States
1Agilent Technologies Inc., Wilmington, Delaware 19808, United States
Supporting Information
INTRODUCTION
Perfluoroalkyl and polyfluoroalkyl substances (PFASs) have
Rc:ceM,& March 10, 2015
unique physical and structural properties that make them R evis,QAugust 7, 2015
extremely resistant to chemical and thermal degradation. As a Ac1:epte h August 30, 2015
result, PFASs have been used in a wide range of consumer
a
ACS Publications C XXXX American Chemical Society A DOI: 10.1021{acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX--XXX
DEQ-CFW 00083481

Environmental Science & Technology
Step 1
E-"4Sarn;.�;ei
�.;teettiaE
Source
<achP€e
Sa;ro}e'J=
f!-potergisI
Er.ter.t�af E Source
Srurre.
E:offrsed \.rat:. <: n::rar;+;:ods
x�
yc a!--:.rt.rn ,f .faathc r. c s:ace?arr' for
co:�fircnatE_,n ;; f-,:)ss:ye
CheriicaE database searr:aiag
i.
Step 6
Step 2 Step 3
1 sary
�. Non-txgeteii
\.TR3i"P,9S t ra: r,ft cxr ni f ea'..r > u riq 1e t, ex,,e
S,.-.mP e ) fccmtml! zt aia ! a
exlfaft;LnMFP
.a ;e_Ea;zc-rent; c-,r.rpouaC.f=ex:f'sg� Piss{Cffjeraortitrcn
.............[­­­­­­_
s,. =c:int2...
Rcdsc2Ea�r of data by extr:;cted <on d" omatr
gram { EC} cf traditfcr.aE PPCs
M, estiss-,inc,o E'If Re. ncg'ative'T:ass defect
�oec%rrc,-tio� E
.u(rP) Er g dnta
�.
Figure 1. Sample workflow for TOFMS discovery.
i:)vestigat: on fif ce-e:3ting peaks latidw-ts- fermate,
acetate and CE:ners - Na-, K=)
Diaer.ostic fragrT e. dEc,rs
sowne Ir rak :aatc4,:,rg Edfstritatf ,n, s;>aGrE3
Exact mas* caifgiev tocA
Eza,t fnass off sets?Cf=,; CF-C-1 i:on,;WoFcus series
Read r:ck N13ss Defect Picas
Step S
products and industrial applications.' Between 1950 and 2000,
the eight -carbon PFASs, including perfluorooctanoic acid
(PFOA) and perfluorooctanesulfonate (PFOS), were among
the most commonly produced and used perfluorinated
compounds. PFASs have been detected in many environmental
and biological matrices across the globe for a number of years,
and this ubiquity has led to mounting concerns about exposure
and potential toxicity. As a result, many international regulatory
efforts have been enacted to reduce, substitute, or eliminate
long -chain (8 consecutive perfluorinated carbons or longer)
PFASs from products and waste streams. In the United States,
eight major fluorochemical manufacturers entered into a
voluntary stewardship agreement with the U.S. Environmental
Protection Agency (EPA) to phase out the use and production
of long -chain perfluorinated chemistries by 2015.' To take their
place, manufacturers have started using alternative chemistries
that may include both shorter -chain perfluorinated (<C8) and/
or polyfluorinated materials. For example, some of these
replacement compounds still have the traditional perfluorinated
carbon regions, but they are broken into shorter units by
insertion of ether oxygens at regular intervals. These ether
linked compounds are still technically classified as perfluori-
nated if all carbons are substituted with fluorine and
polyfluorinated if any carbon is occupied by a hydrogen rather
than a fluorine.-' The rationale is that the addition of oxygens
would likely make the replacement chemistry more labile to
degradation, and thus more favorable for usage. However, to
date the chemical manufacturers have disclosed little
information publicly concerning the chemical structures and
potential toxicities of the replacement PFASs now being
produced. This makes it extremely difficult for independent
�� Ac�rasccti
e<......
Step 4
research groups to evaluate their environmental distributions
and potential toxicities.
A rare exception is a peer -reviewed publication on 4,8-dioxa-
3H-perfluorononanoate (ADONA) (Supporting Information
Figure. SO, a polyfluorinated alternative to PFOA which is said
to have a "more favorable toxicological profile'? This single
journal article was published by its manufacturer, but standards
for this new material have not been made available (to our
knowledge) for comprehensive ecological or toxicological
evaluation. Manufacturers typically submit more detailed
information concerning compound structure, potential toxicity,
environmental fate, and projected production volumes to
regulatory authorities prior to large-scale production (e.g., the
U.S. EPA's premanufacture notice process), but this informa-
tion is held as confidential business information and cannot be
disclosed to the general research community.
One example of industrial producers manufacturing replace-
ment compounds is outlined in an executed draft consent order
from the state of West Virginia to DuPont Corporation
allowing the discharge of a new fluorinated compound into the
Ohio river, at the Washington Works facility (West Virginia).`'
This is the same facility that has been the source of historical
PFOA contamination of local surface and groundwater due to
industrial manufacture and discharge and is the center of a
court -ordered investigation by the C8 science panel to
determine if PFOA exposure in the local human population
has led to adverse health effects.'.A number of recent
epidemiological studies have indicated that exposure to
PFOA at this location has been associated with adverse
human health outcomes.`''S
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ-CFW 00083482

Environmental Science & Technology
Typical environmental occurrence investigations have for
years focused on targeted analysis for specific analytes of
interest. Highly precise and sensitive LC and GC MS/MS
methods have been developed to quantify contaminants of
concern. A more recent approach for environmental sample
analysis has been the use of high -resolution mass spectrometry
(HRMS) for these efforts. Some researchers have used
nontargeted or suspect screening effectively in demonstrating
many chemical classes in wastewater effluent without the use of
reference standards."—' These approaches have begun to rely
more heavily on the use of databases, accurate mass searches,
and HRMS molecular formula prediction capabilities for
discovery. Confirmation with authentic standards still remains
the "gold standard". Schy'lmanski et al.' t reported on five levels
of confidence in identifying small molecules using HRMS
ranging from exact mass matches, through molecular formula
prediction, tentative candidate structures, library match
searches, and finally confirmation with a reference standard.
This is the type of approach that we use in this work to identify
previously undescribed PFASs in surface water.
As it is mostly unknown what is being produced to replace
traditional perfluorinated compounds, a series of sampling
events was conducted to collect water samples with
determination of new per- and polyfluorinated substances as
a goal. In Nakayama et al.,") a sampling of surface water in the
Cape Fear River Basin in North Carolina indicated sources of
perfluorinated compounds intermittently spread throughout
the drainage basin. A follow-up investigation with more focused
sampling to determine sources was not conducted. However,
based on this early work, there was some indication that there
were PFASs sources in the Fayetteville, NC area. More recent
surface water sampling trips in this region indicated continued
elevated concentrations of traditional Perfluorinated com-
pounds, supporting the Nakayama et al.'( findings. One such
source was known to be downstream of the industrial effluent
discharge of a fluorochemical production facility. Further
investigation of these samples for identification and occurrence
of potential replacement chemistries was undertaken.
MATEMALS AND METHODS
Sample Workflow for Discovery. A diagram of the
workflow we used for novel compound discovery in surface
water samples is shown in Figure J. It is a six -step process that
starts with sequential sampling of water, followed by non -
targeted screening, visualization of detected peaks, further
investigation of suspect features, advanced TOFMS techniques,
and, last, compound confirmation with authentic standards
when available. Using this approach, it becomes possible to
detect chemicals that enter a waterway from a source (point or
nonpoint) between sampling locations. This can then be used
for such purposes as source elucidation, compound discovery,
or both. In the following sections we will be referring to Fipi e
1 and the specific step in the process to which we are referring.
Sample Collection. In the summer of 2012, water samples
were collected at locations on the Cape Fear River and its
tributaries, with some locations having previously been found
to contain measurable PFASs.' Grab samples were taken from
the bank and bridge crossings with a lab -made dip sampler or a
stainless steel Kemmerer sampler, and stored in a 1-L HDPE
bottle following the procedures described in Nakayama et al.;"
Samples (n = 9) were acquired from the main flow of the Cape
Fear River including from just south of the city of Fayetteville,
NC to the bridge crossing in the town of Tar Heel, NC.
Duplicate samples, trip spikes, and trip blanks were included in
this sampling. This sampling also included tributaries to the
Cape Fear River along the same stretch of river where
reclaimed wastewater treatment plant (wW'IP) and industrial
effluent streams occurred (Figure 1, step 1).
Sample Analysis for Traditional PFASs. Samples were
prepared and analyzed according to the methods of Nakayama
et al."' In brief, water samples were stabilized with addition of
nitric acid and stored at ambient temperature prior to analysis
(<7 days). Each water sample was measured in a graduated
cylinder, and the sampling bottle was washed with 10 mL of
methanol to solubilize analytes that may have become sorbed to
the container wall during shipping and storage. The water was
then placed back in the methanol -washed bottle with the
methanol rinsate, spiked with a mixture of internal standards,
and filtered through a glass fiber filter. A volume, typically 500
mL, of water was concentrated on a Waters WAX SPE cartridge
using a positive displacement pump. The cartridge was elated
with basic methanol (0.3% NH40H) and the eluent was
concentrated via evaporation under N2 gas to 1 mL and
prepared for analysis. A procedural blank was prepared as
described above using 500 mL of in-house DI water.
Traditional PFASs (C4—CI0 perfluorocarboxylic acids
(PFCAs); perfluorobutanesulfonate (PFBS), perfluorohexane-
sulfonate (PFHxS), and PFOS) analysis was performed using a
Waters Acquity ultra performance liquid chromatograph
interfaced with a Waters Quattro Premier XE triple quadrupole
mass spectrometer (UPLC-MS/MS) (Waters, Milford, MA,
USA).
TOFMS Investigation of Novel Fluorinated Com-
pounds. The extraction method used for traditional PFASs
analysis was also used for elucidation of novel compounds.
Elevated concentrations of traditional PFASs in samples were
used as an indication of samples that were likely closer to
sources (Ngure. S2 ). Sample analysis was performed using an
Agilent 1100 series HPLC interfaced with a 6210 series
Accurate -Mass LC-TOF system (Agilent Technologies, Palo
Alto, CA). The mass spectrometer was operated in electrospray
ionization (ESI) negative mode and any drift in the mass
accuracy of the TOF was continuously corrected by infusion of
two reference compounds (purine (m/z 119.03632) and the
acetate adduct of hexakis (1H,1H,3H-tetrafluoropropoxy)
phosphazine (m/z 980.016375)) via dual -EST sprayer.
Chromatographic separation was accomplished using an Eclipse
Plus C8 column (2.1 X 50 mm, 3.5 ym; Agilent). The method
consisted of the following conditions: 0.2 mL/min flow rate;
column at 30 °C; mobile phases: A: ammonium formate buffer
(0.4 mM) and DI water/methanol (95:5 v/v), and B:
ammonium formate (0.4 mM) and methanol/DI water (95:5
v/v); gradient: 0-15 min a linear gradient from 75:25 A/B to
15:85 A/B; with a 4 min post time for equilibration.
Identification of Suspect Features. The total ion
chromatogram (TIC) of each water sample was subjected to
the software molecular feature extraction algorithm (MFE) and
was restricted to the 100 most intense features m/z 50-1700
based on peak height. A molecular feature is defined as a single
accurate mass with a specific retention time and an integrated
area count. Compounds identified in the solvent and
procedural blank samples were used to develop a mass
exclusion list to subtract from subsequent unknown samples.
After molecular features were identified, they were exported as
a Compound Exchange File (.CEF) and imported into Mass
Profiler to compare upstream with downstream samples.
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ-CFW 00083483

Environmental Science & Technology
c}ma
9.X
03z
4 a5<
m:m
0:45,
0,4 :
Yr
Od
O 2S
0.2
c�.a=
a.ms
[2M_j_
N }mm 1:Yf 1�0 1W m 1W. 22k 2da S!9m ;r&o :#6f 32d1 344 :SE 3io on 4-aa oo 4i0 4iu Wo '!'?U miff smm wo 6.v "n m w '1iv Tv Ym xio 224E em ffi2m $4o miff m we zv w mmm M
iarz;r4s S hI vv._ AB.sc,tntis3r�nx {.: sa,5
,10, Sc>rt.:s.33.a.ss sss::n.rft..rsi..zi,p<4<sU:
4%
s
zsa.:za2
i2 ai;a
s
[2M--
mM%xes.
mix 960 32m f4ff 16.'. }Sm '..ek2 236 >.d£f 2ma 23.m :xk3 S2FF SaQ Smm :sau 4£0 420 4is'. 48m -M ,*) 52Ff s4m ,1,h 90 m:D2 64m :r22 7II3 74m 7HLt 224E m3Y". �2a? s^4m 28ir f S2m Sr+W 22ft Ct
£,awry i"k1 Yc. 5dsr �t>3"�h�ezcn (mezi
Figure 2. ESI negative spectra for two example perfluorinated ether carboxylic acids found in water: (A) C6HF1103 (retention time 5.9-6.0 min)
and (B) C6HF1106 (retention time 6.5-6.6 min). Note: in spectrum (A and B) m/z 112.9856, 119.0363, and 966.0007 are reference masses for
continuous mass calibration. Peaks not pointed out are not associated with the compound of interest.
Pairwise comparisons of sequential samples in river flow
(Figure i step 2) were used to identify those requiring further
investigation. For instance, features found in the downstream
sample that were not found in the upstream sample indicate a
source (point or nonpoint) exists between the two sampling
locations (_Eiigure 1 step 3). Further investigation of peaks with
larger area counts using the "identify compounds" feature in the
software and additional analyses of samples were used for
confirmation. Perfluorinated compounds already identified with
traditional UPLC-MS/MS analysis were eliminated from this
remaining list further narrowing the unknown compounds
requiring follow-up scrutiny.
Poly- and perfluorinated compounds tend to have a negative
mass defect due to the presence of multiple fluorine and oxygen
atoms. Negative mass defect means that the exact mass of a
compound is less than the nominal mass.For instance, the
exact mass of perfluorooctanoate (CsF1502) is 412.9664,
whereas the nominal mass is 413 Da. In this instance negative
mass defect compounds only found in the downstream sample
were further explored. An additional characteristic of samples
that are contaminated with per- and polyfluorinated com-
D DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ-CFW 00083484

Environmental Science & Technology
Table 1. Accurate Mass of Polyfluorinated Compounds and In -Source Artifacts Found in Extracted Water Samples
Monoether PFECAs
1 C,HF603
179.9846
178.9773
380.9438
358.9619
2 C4HF703
229.9813
228.9740
480.9372
458.9553
3 CSHF103
863090-89-5
279.9782
278.9709
580.9310
558.9491
4 C6HF„03
13252-13-6
undecafluoro-2-methyl-3-oxahexanoic acid
329.9750
328.9677
680.9247
658.9427
5 C7HF1303
379.9718
378.9645
780.9182
758.9363
6 C,HF1503
429.9686
428.9613
880.9118
858.9299
Polyether PFECAs
7 C71IF1307
39492-91-6
perfluoro-3,5,7,9,11-pentaoxadodecanoic
4439515
442.9442
908.8776
886,8957
acid
8 C6HF1,06
39492-90-5
perfluoro-3,5,7,9-butaoxadecanoic acid
377.9598
376.9525
776.8942
754.9123
9 C,HF90,
39492-89-2
perfluoro-3,5,7-propaoxaoctanoic acid
311.9681
310.9608
644.9108
622.9289
10 C,HF704
39492-88-1
perfluoro-3,5-dioxahexanoic acid
245.9764
244.9691
5129274
490.9455
PFESAs
11 C7HF1301S
66196-30-3"
443.9337
442.9264
12 C7H2F1406S
463.9399
462.9326
Other
Na'
22.9892
H'
1.0073
CF20
65.9917
CFz
49.9968
"Indicates the monoisotopic
mass of
the neutral species. bCAS number for Nation
copolymer.
pounds is the presence of multiple peaks that differ by exactly ±
nt/z 49.9968 and/or 65.9917, corresponding to a difference of
CFz and CF20, respectively. This happens because PFAS
synthesis is an industrial process that yields a distribution of
products that differ primarily in chain length and/or degree of
isomerization. Compounds identified as having a negative mass
defect and differing from another peak by ± na/z 49.9968 and
65.9917 are highly likely to be polyfluorinated compounds.
Selected samples were also analyzed by QTOF to aid with
confirmation of proposed structures (generously provided by
colleagues mentioned in the Acknowledgments).
TOFMS Experimental Workflow. Once an unknown
compound with negative mass defect was noted, it was isolated
from other mass spectral data in the following way: An
extracted ion chromatogram (EIC) was generated from the
total ion chromatogram (TIC) using the m/z identified. The
EIC was then used to do a background subtraction of the
spectral region both immediately before and after the detected
peak (typically ±0.1 min) (Figure 1 step 4). The center of the
detected peak was extracted for the mass spectral information,
and the background spectrum was subtracted for a clean
spectrum. This was necessary to eliminate, to the best of our
ability, competing spectral peaks that were coeluting and not
associated with the spectrum being investigated. This would
include, but is not limited to, spectral signals originating from
the reference compounds (constantly infused to maintain a lock
on mass assignments) and their fragments. Once suspect
spectral signals were isolated, a series of experiments were
conducted to induce in -source fragmentation ions to aid with
compound identification based on the work of Ferrer and
Thurman. is Specifically, a series of methods with differing
fragmentor voltages (80-190 V) were run in sequence to look
for diagnostic ions that emerged as fragmentor voltage
increased. In addition, common adducts such as Na-, NH4+
acetate, etc. were used to aid with compound identification.
Detected spectral features were subjected to molecular formula
generation using the elements (C, H, 0, N, S, P, Cl and F).19
Formulas generated were scored based on accurate mass,
isotope abundance (compound mass distributions attributable
to elemental mass distributions e.g., 1.1% i3C, 0.2% is0, etc.)
and isotope distribution (for example, slight differences in
isotope mass such as when an m+2 peak is from two 13 C or a
single 180) . Only scores >75% were considered for further
exploration. Figure 1 steps 4 and 5 outline the advanced
TOFMS techniques used to discover novel chemical species.
RESULTS
With the approach described above, sites were identified for
further investigation when a large increase in number and
magnitude (area counts) of unknown compounds was found in
downstream samples. In addition, a very large increase (greater
than 2 orders of magnitude) in concentration of historically
measured PFAAs also persisted for many river miles down-
stream of a certain point (Figure. S2) consistent with
observations in Nakayama et al."' Interestingly, the profile of
the historically measured PFASs contributing to the total
PFASs found at each location was seen to dramatically change
near this location (Figure S5). The major compounds
contributing to this increase were perfluoropentanoic acid
(PFPeA), followed by perfluoroheptanoic acid (PFHpA) and
perfluorobutanoic acid (PFBA). It was evident that there was a
significant source (total historically measured PFASs concen-
tration increased by > 100 times) of per- and polyfluorinated
compounds near this location. Plotting the samples (CFR004
upstream vs CFRO02 downstream) in MassProfiler for
visualization indicated that there were 77 features that were
unique to the downstream sample (Figure S4) with many (n =
69) having negative mass defects. In this figure (S,4) the size of
the symbol is proportional to the area counts of the molecular
feature. Traditional PFASs with known concentrations (in
Figure S2:) are included for comparison purposes. Peaks with
larger areas were investigated first under the assumption that
these concentrations would be among the highest. Additional
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX-XXX
DEQ-CFW 00083485

Environmental Science & Technology
painvise comparisons were done on all sequential samples,
however results indicated that samples downstream of CFRO03
were simply dilutions of a source input. Samples were chosen
for additional scrutiny where there was sufficient analytical
signal without saturation of the detector which compromises
mass accuracy.
Dimers and Fragments. It became evident that in multiple
instances several of the larger peaks had a number of common
characteristics: (1) negative mass defect, (2) multiple related
chromatographic peaks differing by ± m/z 49.9968 and/or
65.9917 (± a CF2 group and/or a CF20 group), and (3)
apparent noncovalent homodimers linked by either a proton or
sodium ion. For example, at a retention time of 5.9-6.0 min,
two large peaks emerged from the spectra with m/z of
658.9462 and 680.9256 (Figure ?A). After careful evaluation, it
became apparent that these were the proton -bound and
sodium -bound in -source homodimers of undecafluoro-2-
methyl-3-oxahexanoic acid (C614171103) (Table 1). These
differed by m/z 21.9819, the difference between the Na' and
H- versions of the dimer. It is important to note that the [M -
H]- peak (m/z 328.9677), which one might expect to be the
most prominent, was barely distinguishable from the back-
ground signal at this point because the in -source ionization
conditions so heavily favored the formation of these dimers.
Given the prominence of the C6HF1103 compound in this
sample, we postulated that a homologous series of related
perfluorinated ether carboxylic acids (PFECAs), differing from
the first compound by either the addition or deletion of CF2
units (m/z 49.9968) might also be present. An EIC for each of
the hypothetical masses, based on sequential addition or
deletion of CF2 units (m/z 49.9968) was extracted from the
TIC, with the resulting series of related PFECAs presented in
Figure S5. Although not specifically shown in this Figure, the
also observed the presence of the analogous sodium and proton
bound dieters for each compound identified in this
homologous series of PFECAs (C(=3-s)HF(n=s-ls)O3 (Table
1). To obtain additional evidence supporting the proposed
structure of the compounds eluting at 5.9 min, additional
experiments were conducted by altering the fragmentor voltage
in order to try to form diagnostic ions from the m/z of
658.9427 and 680.9247 dimers eluting at this time. Figure S6
shows diagnostic dimer and fragment ions resulting from in -
source CID that are consistent with the structure proposed for
the C6HF1103 compound. The major ions found are also
shown in Figure ,S 10. Extracted spectra comparable to those
shown in Rgure S6, differing only by the loss of CF2 groups,
were investigated as well (Figure S7) lending credence to the
postulated structure(s) and discussed in more detail in the
following section.
Confirmation of Structure. An online search of the
formula and proposed structure of the C6HF1103 compound
eluting at 5.9 min led to a tentative match in a "grey -literature"
citation that mentions a new -generation processing aid for the
production of high-performance fluoropolymers identified only
as C3 dimer acid/salt (CAS 13252-13-6) (Figure. S6).'With a
CAS number, it was possible to purchase an authentic standard
(Synquest Laboratories, Alachua, FL) to compare for retention
times and mass spectral ionization patterns. The authentic
standard compared very well with the tentatively identified
compound found in water samples (±0.051 min RT; ± m/z
0.0002 (0.51 ppm) (Figure 1 step 6). Additional authentic
standards for other compounds in this homologous series were
not commercially available. However, common spectral
patterns and adducts were observed differing only by repeating
CF2 units, as supporting information for compound identi-
fication of a homologous series of PFECAs (Figure. S7).
Additional Perfluorinated Ether Acids. Figure ,S4 shows
that there are a number of features that are present in the water
sample downstream of a potential input location that are not in
the upstream sample. Evident in Figure SDI- are a pronounced
series of peaks that appear as vertical lines, likely from coeluting
and related chemicals (see later discussion on in -source n-mer
formation). A number of these peaks exhibited negative mass
defects. Further investigation of the TOFMS spectra indicated
two peaks coeluting at 6.5 min (m/z 776.8942 and 754.9123)
with a mass difference of m/z 21.9819, again suggesting the
presence of in -source proton- and sodium -bound homodimers
(higure :1.B). Knowing this information, we postulated an exact
mass of m/z 376.9525 for the [M - H]- ion (Table 1). We
further postulated a homologous series based on addition and
deletion of CF2 units, but exact masses corresponding to ± m/z
49.9968 were not detected. Given reports of industrial
producers making greater use of ether oxygens to limit the
size of perfluorinated regions within a given molecule, we
postulated that a homologous series could also be based on the
repeating units of CF20, with an exact mass offset of ± (rn/z
65.9917).""' Searching the sample for m/z 376.9525 ± this
hypothetical CF20 offset did yield another homologous series
(Figure S8). In addition, as was previously the case, each of
these new compounds was also found to form sodium -bound
and proton -bound homodimers, providing further support for
the proposed structures (fable 1). Confirmation of these
compounds with authentic standards was not possible. It
should be noted that in the homologous series shown, the
chemical C3HF503 (m/z 178.9773) is a common feature
(Table 1) to the previously identified homologous series.
Taken together, the exact mass of the proposed structures, the
offset by a CF20, and the exact mass of the in -source dimers
formed, all support the occurrence of this additional
homologous series. One compound in this homologous series
(C4HF,04 [M - H] 244.9691 `fable 1) was subjected to
QTOF analysis leading to a series of accurate grass fragments
consistent with the proposed structure (figure S9). The
accurate mass fragments show the sequential breakage of the
ether oxygen bonds and the resulting fragment ions.
In -Source n-mer Formation. Multiple peaks originally
thought to be of polymeric origin (m/z 1176.8353 and
1576.7769) were determined to be sodium -bound n-mers of
m/z 376.9525 already identified (Figure S12). The retention
time of these peaks (6.5 min) indicates this is an in -source
phenomenon, as they coelute (Figure ,S4). In -source artifacts
occur in the ESI source and can include such things as the
formation of adducts (formate, Na+, H'_) NH4'), fragments,
dimers, trimers, and combinations of the above. Peaks that do
not coelute, or are chromatographically separated, are generally
considered not to be in -source artifacts. The observed spectra
(_Figure S12) include the formation of n-mers17 including 2 and
3 bound sodium cations in addition to the compounds shown
in Figure. S11. The formation of other n-mers with >1 bound
sodium was seen as well for additional PFECAs found in this
work (data not shown). Recent work by Trier et al.17 indicates
that the formation of these polyfluorinated surfactant clusters
occurs in the MS source in the gas phase and are pH and
concentration dependent." However, this phenomenon may
also be instrument -specific as sample dilution in this study did
not lead to loss of these in -source dimers being formed. Similar
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX-XXX
DEQ-CFW 00083486

Environmental Science & Technology
W: 03
A
F
r C=
hC i dr,;gr+ok r �o: rreba. s f -:,•�,�
44 92S4
Measured IfOl-H) 442.9264 l0,19 ppm rriass accuracy)
U 5
7:
r
aE
a fa ,qy gFC�?SC Ma.,'s -` 4 C'��C ` a
a's
3:
's` Measo3 e ?M Hl 4629.=27 (0-1.1 p€3rti B ass.fL?.laxy-
YA
+a: oyx
Gs;A 4s) at),: aeio- 416 431 _oft& °:, fV 4°n. svnz sons asne .w1xs c
Figure 3. Suspected perfluoro ether sulfonic acids identified in water. Oxidation species are likely products of CAS 16090-14-5, a perfluorinated ether
sulfonyl fluoride. Chromatogram shown of two resolved peaks (AB) and isotope cluster matching of measured versus exact mass. Red boxes indicate
the exact mass for the suspected formulas, compared to the accurate mass profile spectrum. Mass difference between two species is an HF (m/z
20.0063-1.41 ppm mass accuracy).
sodium- and proton -bound, as well as mobile phase, modifier
(i.e., formate) adducts have been shown by Trier et al.". for
perfluorocarboxylic acids and mono -PAPS (perfluorinated alkyl
phosphates),
Perfluoroalkyl Ether Sulfonic Acid (PFESA) Identifica-
tion. Records of groundwater monitoring at a fluorochemical
manufacturing plant with wells in close proximity to our
sampling site indicate PFOA in ugIL concentrations.21 This
document also indicated a sulfonated tetrafluoroethylene based
fluoropolymer-copolymer with the trade name Nafion was also
produced at this facility. Our data provide evidence that
materials related to this compound are also emitted into the
river at this point. figure 3 shows the chemical structures of the
suspected oxidation species and resulting transformation
products tentatively identified in our samples. An EIC of m/z
442.9264, which is the exact mass of the [M — H]- moiety of
the copolymer containing the sulfonate group (IUPAC name
1,1,2,2-tetrafluoro-2- ({ 1,1,1,2, 3, 3-hexafluoro-3- [ (trifluorovinyl) -
oxy]-2-propanyl}oxy)ethanesulfonic acid), suggests this com-
pound is present in the water. In addition, spectral evidence
supports the detection of a related compound which is
consistent with the addition of an HF ([M — H]- nt/z
462.9327) (IUPAC name 1,1,2,2-tetrafluoro-2-{[1,1,1,2,3,3-
hexafluoro-3-(1,1,2,2-tetrafluoroethoxy) prop an-2-yl]oxy}-
ethanesulfonic acid). The two suspected compounds differ by
the mass of HF (m/z 20.0062-1.41 ppm mass accuracy)
(_Figure 3). These compounds elute as two discrete chromato-
graphic peaks, likely explained by the position occupied by the
H and F additions. It is possible that these chemicals originate
from the oxidation of ethanesulfonyl fluoride, 2-[1-[difluoro-
[ (t (fluoroethenyl) oxy] methyl]-1,2,2,2-tetrafluoroethoxy] -
1,1,2,2-tetrafluoro- (CAS 16090-14-5) which may have been in
use at this industrial location ' These compounds would thus
be classified as perfluoroalkyl ether sulfonic acids (PFESA) 4
However, unlike previous perfluorinated ether carboxylic acid
compounds found in this study, in -source proton- and sodium -
bound dieters were not found for these perfluorinated ether
sulfonates. In addition, a homologous series based on CFZ or
CF20 additions was not found. We have not been able to
obtain an authentic standard for confirmation of this proposed
structure.
DISCUSSION
It has been well-known for a number of years that, due to the
toxicity and persistence of perfluorinated compounds, industrial
producers are moving toward shorter chain per- and
polyfluorinated compound homologues. However, it is not
well-known which compounds are being produced as replace-
ments for historic perfluorinated compounds (e.g., PFOS and
PFOA). Peer -reviewed literature showing the chemical
structure of manufactured per- and polyfluorinated replacement
compounds is sparse. An example of one such publication from
3M Corporation indicated that the compound ADONA (4,8-
dioxa-3H-perfluorononanoate) is being made to replace
ammonium perfluorooctanoate (APFO) as an emulsifier in
the manufacture of fluoropolymers.' The data presented in this
study and Fl gure. S 1. indicate that the manufacture and release
of per- and polyfluorinated compounds with CFZ units
interspersed with ether oxygen linkages likely is ongoing. A
more recent study suggests that polyfluoroalkyl ether
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ—CFW 00083487

Environmental Science & Technology
compounds are being produced and that the number and
spacing of ether oxygen linkages and size of the resulting per -
and polyfluorinated compound may be manufacturer -specific
including other rnalor fluorochemical producers (DuPont, 3M,
Solvay, and Asahi).i" However, there are little to no data in the
peer -reviewed literature on most of these compounds. The use
of HRMS appears to be an ideal technique to use for discovery
and occurrence of new PFASs being released into the
environment.
As illustrated in this research, identification of previously
undescribed compounds in environmental matrices can be
made difficult by the formation of a multitude of in -source
artifacts such as in -source gas phase adducts, n-mers, and the
presence of a homologous series of per- and polyfluormated
compounds. D'Agostino and Mabury identified novel fluori-
nated surfactants in aqueous film -forming foams (AFFF) and
commercial surfactant concentrates using high -resolution mass
spectrometry.``' Their results imply that there are a number of
additional fluorinated compounds being produced and emitted
into the environment that are unknown to most analysts. The
use of TOFMS or other high -resolution mass spectrometers
and the workflow we demonstrated (Figure 1) appears to be a
useful way to determine occurrence and emissions of new and
emerging fluorinated compounds for environmental occurrence
efforts.
Schymanski et al. report on identifying small molecules via
HRMS, with five levels of confidence."' As has been noted
earlier, one of the pitfalls of discovery work is the lack of
authentic standards at times. As such the highest level of
confidence for confirmed structure (Level 1) can be made to
one compound in this effort, with the remaining compounds
falling in the categories of tentative candidate(s) (Level 3) or
probable structure (Level 2).1 } An additional line of evidence
we present that Schymanski et al. appear not to is the presence
of a homologous series (CF2, CF20) as evidence for compound
identification.
The manufacture of shorter -chain perfluorinated and
polyfluorinated compounds may have an advantageous effect
on biological persistence and bioaccumulation. Two recent
publications deemed "The Helsingor Statement"21 and the
"Madrid Statement"'` explore the state of the science related to
poly- and perfluorinated alkyl substances. Shorter straight -chain
homologues of perfluorocarboxylic acids (<C6) and sulfonic
acids (<PFBS) have been shown to be cleared quickly from
mammalian species tested and thus are not expected to
bioaccumulate as readily as PFOS and PFOA.26 However, the
environmental persistence of these shorter chain perfluorinated
compounds is likely to be no different from that of the related
longer -chain acids. The conventional wisdom is the perfluori-
nated ether compounds will be more labile than the n-alkyl
perfluorinated homologues with ether oxygen linkages being
more susceptible to chemical or microbial attack. However, as
the CF2 bond in perfluorinated compounds is not at all open to
microbial attack due to its bond strength, even the slightest
degree of degradation of a per- or and polyfluorinated ether
compound would be considered "more" labile. Environmental
degradation or toxicology studies associated with the
compounds identified in this study were not found by the
authors in the peer -reviewed literature at the time of writing
this Article. Although data are scarce, the few perfluoropo-
lyethers (PFPEs) to have undergone any kind of degradation
studies show little to no biodegradation or hydrolysis."" These
data mainly come from the European Chemical Agency
(ECHA) database of registered substances and associated
information."' Past studies have demonstrated poor removal
efficiency for historical PFASs from source water to drinking
water in conventional systems (i.e., treatments that do not
include granular activated charcoal (GAC)). It is unknown if
these PFECAs and PFESAs are removed by conventional
Wth'TP processes.
Environmental Implications. We demonstrate the pres-
ence of a series of novel perfluorinated ether carboxylic and
sulfonic acid(s) found to be in natural waters using a
nontargeted workflow (Figare. 1). The compounds consist of
a homologous series of per- and polyfluorinated compounds
with repeating units of CF2 or CF20 subunits. LC-TOFMS
investigations of accurate mass dimers and n-mers (proton- or
sodium -bound) in addition to diagnostic fragment ions support
these findings. Further unidentified compounds with negative
mass defects are likely also of per- and polyfluorinated origin,
but have not yet been identified and are undergoing further
analytical scrutiny. Nontargeted screening and discovery of
novel species in samples such as reported here is an ongoing
process. Additional environmental samples from other locations
may be needed to identify additional chemistries.
Once a xenobiotic compound is identified in the environ-
ment, it falls upon the scientific community to begin
monitoring efforts to find the extent of contamination of
newly discovered species. In addition, toxicological and
degradation investigations of identified compounds can
commence once a compound has been identified. The
procurement of authentic compounds for use in these types
of investigations may be a major limiting factor in conducting
such investigations, as most of these compounds appear not to
be commercially available. More research will need to be
conducted on these perfluorinated ether carboxylic and sulfonic
acids concerning environmental occurrence, toxicology, and
degradation potential.
A,SSO IATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website. at DOI: 10.1021!=,cs.est.5b01?IS.
Additional information including tables, figures, chem-
icals structures, and spectra
AUTHOR INFORMATION
Corresponding Author
*Phone: [phone]; e-mail: stryr�ar.rr.a.tic(r>epa.ov.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
TS
We thank Agilent Technologies for their support of this effort
through a TOFMS and travel/training CRADA with the U.S.
EPA (CRADA 437-A-12) and in particular Joe Weitzel for his
support of this work. In addition, we thank Mike Hays
(USEPA), and Chris Higgins and Simon Roberts (Colorado
School of Mines) for the use of their QTOFs in confirmation of
select compounds identified. We also thank John Offenberg,
Michelle Angrish, and Mike Hays for their review of this
manuscript. The United States Environmental Protection
Agency through its Office of Research and Development
funded and managed the research described here. It has been
subjected to Agency review and approved for publication.
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ-CFW 00083488

Environmental Science & Technology
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
REFERENCES
(1) Lindstrom, A. B.; Strynar, M. J; Libelo, E. L. Polyfluorinated
Compounds: Past, Present, and Future. Environ. Sci. Technol. 2011, 4S
(19), ', 954-7961.
(2) Houde, M.; De Sflva, A. O.; Muir, D. C. G.; Letcher, K J.
Monitoring of Perfluorinated Compounds in Aquatic Biota: An
Updated Review. Environ. Sci. Technol. 2011, 4S (19), 7962-7973.
(3) U.S. EPA. 2010/1S PFOA Stewardship Program. http://,wwvc.epa.
goldoppt;'pfoa;/pubs,/stewardship / index.hti nl.
(4) Buck, K C; Franklin, J.; Berger, U; Conder, J. M.; Cousins, I. T.;
de Voogt, P.; Jensen, A. A.; Kannan, K; Mabury, S. A.; van Leeuwen,
S. P. Perfluoroalkyl and polyfluoroalkyl substances in the environment:
terminology, classification, and origins. Integr. Environ. Assess. Manage.
2011, 7 (4), 513-41.
(5) Gordon, S. C. Toxicological evaluation of ammonium 4,8-dioxa-
3H-perfluorononanoate, a new emulsifier to replace ammonium
perfluorooctanoate in fluoropolymer manufacturing. Regul. Toxicol.
Pharmacol. 2011, S9 (1), 64-80.
(6) West Virginia Department of Environmental Protection. Executed
Draft Consent Order No. 7418. lrttp:;';'w w,v.dep vrv.gov ipio/
Documents,%Duhenta4;'.Ol�irtal9�u `OVersion.pdf.
(7) Steenland, K.; Fletcher, T.; Savitz, D. A. Epidemiologic Evidence
on the Health Effects of Perfluorooctanoic Acid (PFOA). Environ.
Health Perspect. 2010, 118 (8), 1100-1108.
(8) MacNeil, J.; Steenland, N. K.; Shankar, A.; Ducatman, A. A cross -
sectional analysis of type 11 diabetes in a community with exposure to
perfluorooctanoic acid (PFOA). Environ. Res. 2009, 109 (8), 997-
1003.
(9) Steenland, K; Tinker, S.; Frisbee, S.; Ducatman, A; Vaccarino, V.
Association of Perfluorooctanoic Acid and Perfluorooctane Sulfonate
With Serum Lipids Among Adults Living Near a Chemical Plant. Am.
J. Epidemiol. 2009, 170 (10), 1268-1278.
(10) Nakayama, S.; Strynar, M.; Helfant, L.; Egeghy, P.; Ye, X.;
Lindstrom, A. B. Perfluorinated Compounds in the Cape Fear
Drainage Basin in North Carolina. Environ. Sci. Technol. 2007, 41,
5271-5276.
(11) Hollender, J.; Bourgin, M.; Fenner, K. B.; Longree, P.; Mcardell,
C. S.; Moschet, C.; Ruff, M.; Schymanski, E. L.; Singer, H. P. Exploring
the Behaviour of Emerging Contaminants in the Water Cycle using the
Capabilities of High Resolution Mass Spectrometry. Chimia. 2014, 68
(11), 793-798.
(12) Schymanski, E. L.; Singer, H. P.; Longree, P.; Loos, M.; Ruff,
M.; Strays, M. A.; Ripoll6s Vidal, C.; Hollender, J. Strategies to
Characterize Polar Organic Contamination in Wastewater: Exploring
the Capability of High Resolution Mass Spectrometry. Environ. Sci,
Technol. 2014, 48 (3), 1811-1818.
(13) Schymanski, E. L.; Singer, H. P.; Slobodnik, J.; Ipolyi, 1. M.;
Oswald, P.; Krauss, M.; Schulze, T.; Haglund, P.; Letzel, T.; Grosse, S.;
Thomaidis, N. S.; Bletsou, A.; 7wiener, C.; Ibanez, M.; Portoles, T.; de
Boer, R.; Reid, M. J.; Onghena, M.; Kunkel, U.; Schulz, W.; Guillon,
A.; Noyon, N.; Leroy, G.; Bados, P.; Bogialli, S.; Stipanicev, D.;
Rostkowski, P.; Hollender, J. Non -target screening with high -
resolution mass spectrometry: critical review using a collaborative
trial on water analysis. Anal. Bioanal. Chem. 2015, 407, 6237.
(14) Schymanski, E. L.; Jeon, J.; Gulde, R.; Fenner, K; Ruff, M.;
Singer, H. P.; Hollender, J. Identifying Small Molecules via High
Resolution Mass Spectrometry: Communicating Confidence. Environ.
Sci. Technol. 2014, 48 (4), 2097-2098.
(15) Nakayama, S.; Strynar, M. J; Helfant, L.; Egeghy, P.; Ye, X.;
Lindstrom, A. B. Perfluorinated compounds in the Cape Fear Drainage
Basin in North Carolina. Environ. Sci. Technol. 2007, 41 (15), 5271-6.
(16) Nakayama, S. F; Strynar, M. J.; Reiner, J. L.; Delinsky, A. D.;
Lindstrom, A. B. Determination of Perfluorinated Compounds in the
Upper Mississippi River Basin. Environ. Sci. Technol. 2010, 44 (11),
4103-4109.
(17) Trier, X.; Granby, K.; Christensen, J. H. 'fools to discover
anionic and nonionic polyfluorinated alkyl surfactants by liquid
chromatography electrospray ionisation mass spectrometry. J.
Chromatogr. A 2011, 1218 (40), 7094-104.
(18) Ferrer, L; Thurman, E. M. Liquid Chromatography Time -of Flight
Mass Spectrometry; Wiley: Hoboken, NJ, 2009.
(19) Kind, T; Fiehn, O. Seven Golden Rules for heuristic filtering of
molecular formulas obtained by accurate mass spectrometry. BMC
Bioinf. 2007, 8 (1), 105.
(20) Wang, Z.; Cousins, I. T.; Scheringer, M.; Hungerbiihler, K.
Fluorinated alternatives to long -chain perfluoroalkyl carboxylic acids
(PFCAs), perfluoroalkane sulfonic acids (PFSAs) and their potential
precursors. Environ. Int. 2013, 60 (0), 242-248.
(21) NCDENK Onsite PFOA Groundwater Monitoring Report 12-9-
13; p 6. hops:iiedrn.nc.giw!L?ENI-Portal;'; full text search
H V F NCD047368642 12-09-2013 GW GMR
(22) USEPA. Envirofacts. 1'sttp:i;'iaspub.epa.gov;'enviroitsca.get
chern ii co v vgi.st y ii:1=110111105,59609. 2015.
(23) D'Agostino, L. A.; Mabury, S. A. Identification of novel
fluorinated surfactants in aqueous film forming foams and commercial
surfactant concentrates. Environ. Sci. Technol. 2014, 48 (1), 121-9.
(24) Scheringer, M.; Trier, X.; Cousins, 1. T.; de Voogt, P.; Fletcher,
T.; Wang, Z.; Webster, T. F. Helsingor Statement on poly- and
perfluorinated alkyl substances (PFASs). Chemosphere 2014, 114 (0),
337-339.
(25) Blum, A.; Balan, S.A.; Scheringer, M.; Goldenman, G.; 'frier, X.;
Cousins, L; Diamond, M.; Fletcher, T.; Higgins, C.; Lindeman, A. A.;
Peaslee, G.; de Voogt, P.; Wang, 'Z.; Weber, K Madrid Statement.
Environ. Health Perspect. 2015, 123, A107.
(26) Lau, C.; Anitole, K.; Hodes, C.; Lai, D.; Pfahles-Hutchens, A.;
Seed, J. Perfluoroalkyl acids: a review of monitoring and toxicological
findings. Toxicol. Sci. 2007, 99 (2), 366-94.
(27) European Chemical Agency (ECHA). Registered Substances.
http:;';'echa.europa.erE/information-on -chemicals;'registered-
sl:rbsta:crc{rs. 2015.
DOI: 10.1021 /acs.est.5b01215
Environ. Sci. Technol. XXXX, XXX, XXX—XXX
DEQ-CFW 00083489