Aerobic biotransformation studies of two trifluoromethoxy-substituted aliphatic alcohols and a novel fluorinated C3-based building block

Article abstract of DOI:10.1016/j.jfluchem.2015.06.015

Fluorinated substances play a significant role for many industrial and consumer products, but many of these chemicals are attributed with an adverse ecological profile and persistence in the environment. Herein, three potentially more environmentally benign substitutes were assessed for aerobic biotransformation, namely 3-(trifluoromethoxy)-propan-1-ol (TFMPrOH), 6-(trifluoromethoxy)-hexan-1-ol (TFMHxOH) and 1-(2,2,3,3,4,4,4-heptafluorobutoxy)-propan-2-ol (HFBPrOH). Analytical techniques involved different HPLC-ESI-MS/MS techniques as well as determination of fluoride in order to assess the extent of mineralization. The two trifluoromethoxy-substituted alcohols showed very different results concerning mineralization. Whereas TFMPrOH only yielded approximately 15% fluoride, TFMHxOH reached nearly quantitative release of fluoride after 37 days. The latter one yielded 6-trifluoromethoxy hexanoic acid (TFMHxA) as well as trifluoromethyl carbonate (TFMC) as transient transformation products, both of which were entirely degraded. TFMC was also detected during biotransformation of TFMPrOH, however, the major transformation product was 3-trifluoromethoxy-propanoic acid (TFMPrA), which did not show any further degradation within a 47 days period. The third compound under investigation, HFBPrOH, expectedly yielded perfluorobutanoic acid as a stable biotransformation product. Several acidic biotransformation intermediates as well as heptafluorobutan-1-ol were identified and a biotransformation pathway was postulated. TFMHxOH might therefore be incorporated into fluorinated substances, for instance by ester linkages assuming that biotransformation of such substances would yield the alcohol.

Full text of DOI:10.1016/j.jfluchem.2015.06.015

Journal of Fluorine Chemistry 177 (2015) 80–89
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Aerobic biotransformation studies of two
trifluoromethoxy-substituted aliphatic alcohols
and a novel fluorinated C3-based building block
*
Tobias Fro¨mel, Thomas P. Knepper
Hochschule Fresenius, Institute for Analytical Research, Limburger Straße 2, 65510 Idstein, Germany
A R T I C L E I N F O
A B S T R A C T
Article history:
Fluorinated substances play a significant role for many industrial and consumer products, but many of
these chemicals are attributed with an adverse ecological profile and persistence in the environment.
Herein, three potentially more environmentally benign substitutes were assessed for aerobic
biotransformation, namely 3-(trifluoromethoxy)-propan-1-ol (TFMPrOH), 6-(trifluoromethoxy)-
hexan-1-ol (TFMHxOH) and 1-(2,2,3,3,4,4,4-heptafluorobutoxy)-propan-2-ol (HFBPrOH). Analytical
techniques involved different HPLC–ESI-MS/MS techniques as well as determination of fluoride in order
to assess the extent of mineralization.
Received 28 January 2015
Received in revised form 11 June 2015
Accepted 16 June 2015
Available online 22 June 2015
Keywords:
PFAS
The two trifluoromethoxy-substituted alcohols showed very different results concerning minerali-
zation. Whereas TFMPrOH only yielded approximately 15% fluoride, TFMHxOH reached nearly
quantitative release of fluoride after 37 days. The latter one yielded 6-trifluoromethoxy hexanoic acid
(TFMHxA) as well as trifluoromethyl carbonate (TFMC) as transient transformation products, both of
which were entirely degraded. TFMC was also detected during biotransformation of TFMPrOH, however,
the major transformation product was 3-trifluoromethoxy-propanoic acid (TFMPrA), which did not
show any further degradation within a 47 days period.
Biotransformation
Mass spectrometry
HPLC–MS
The third compound under investigation, HFBPrOH, expectedly yielded perfluorobutanoic acid as a
stable biotransformation product. Several acidic biotransformation intermediates as well as hepta-
fluorobutan-1-ol were identified and a biotransformation pathway was postulated. TFMHxOH might
therefore be incorporated into fluorinated substances, for instance by ester linkages assuming that
biotransformation of such substances would yield the alcohol.
ß 2015 Elsevier B.V. All rights reserved.
1. Introduction
Substitutes for the legacy chemicals containing fluorinated C8
chemistry are mainly shorter chained substances, mainly those
Perfluoroalkyl and polyfluoroalkyl substances (PFASs) have
been used for decades since they can provide properties not
achievable by non-fluorinated chemistry [1]. However, their
detrimental effects regarding toxicity, ecotoxicity and bioaccu-
mulation have become a cause for concern and have led to a
restraint of several such substances carrying C8 fluorinated alkyl
chains [2]. These concerns are reinforced by the fact that
perfluoroalkyl acids cannot be entirely degraded by microorgan-
isms due to the stability of the carbon–fluorine bond and thus are
persistent in the environment [3–6].
incorporating fluorinated C6 or C4 chains, but yet, such substances
cannot be entirely degraded due to the stable perfluoroalkyl chains
[7–9]. In search of appropriate biodegradable substitutes, a series of
substances with different fluorine-containing end groups was
studied, including 4-trifluoromethylphenoxy-, N,N-bis(trifluoro-
methyl)amino- and trifluoromethoxy (TFM)-substituted n-alkane-
1-sulfonates [10–12]. Within this framework, it was shown that
trifluoromethoxy-decane-1-sulfonates are mineralized to approxi-
mately 90% based on fluoride release, whereas the other substances
showed inferior degradation extents. The biotransformation route
for TFM-alkane-1-sulfonates was initiated by desulfonation and
formation of
v-TFM-alkyl carboxylates, which were then shortened
by -oxidation. Eventually, trifluoromethanol (TFMeOH) was
b
assumed to be formed which first decays to difluorophosgene
*
Corresponding author. Tel.: +49 6126 935264; fax: +49 6126 9352173.
and further to fluoride and carbon dioxide [13]. This proved that
E-mail address: knepper@hs-fresenius.de (T.P. Knepper).
http://dx.doi.org/10.1016/j.jfluchem.2015.06.015
0022-1139/ß 2015 Elsevier B.V. All rights reserved.

T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
81
TFM-based compounds can be degraded if the group is attached to
a biodegradable linker, for instance alkane-1-sulfonates.
triple quadrupole – linear ion trap tandem mass spectrometer (3200
Q Trap, Applied Biosystems, Foster City, CA, USA) and a Turbo
IonsprayinterfaceinESI (‘‘TurboSpray’’)mode. NitrogenusedforMS
was generated by a Membrane nitrogen generator NGM-22-LC/MS
(CMC, Eschborn, Germany) coupled to a SF 4 FF oil-free orbiting
scroll compressor (Atlas Corpo, Stockholm, Sweden). Operation and
data acquisition of the HPLC–MS system was carried out via Analyst
Software, versions 1.4 and 1.5, respectively. IonSpray Voltage was
+5.5 kV in positive mode and ꢁ4.5 kV in negative mode, respective-
ly, Curtain Gas was set to 25 psi, Nebulizer Gas was 55 psi, Turbo Gas
was 65 psi and CAD Gas was set to ‘medium’ equalling 5 on an
arbitrary scale from 1 to 12 when fragmentation was intended.
For target compound analysis, optimization of the MS
parameters was performed via syringe pump injection of a
solution of the analyte at an approximate concentration of 0.1–
In the present study, TFM-based 1-alkanols are investigated in
terms of their biodegradation behavior. Such alcohols might be
possible building blocks in TFM-based polyesters, similar to
fluorotelomer alcohols in legacy oligomeric or polymeric PFASs
[1]. These TFM-alcohols might therefore occur as biotransforma-
tion intermediates due to enzymatic or even chemical hydrolysis.
Since substances carrying only one trifluoromethyl group cannot
provide the same unique properties as those with higher fluorine
content, this study was carried out as proof of principle for
biodegradability of substances carrying the TFM group.
Moreover,
a novel fluorinated building block with a C3
perfluoroalkyl chain length was studied, namely 1-(2,2,3,3,4,4,4-
heptafluorobutoxy)-propan-2-ol (HFBPrOH). In this case, the focus
was set to examine whether perfluorobutanoic acid (PFBA) was
formed, since ecotoxicological data for this compound exist
suggesting lower concern compared with their longer-chained
analogs [14,15]. Similarly to TFMPrOH and TFMHxOH, HFBPrOH
contains an alcohol group and thus can be used to synthesize
oligomers or polymers where these structures are implemented in
the side chains.
20
m
g mLꢁ1 depending on response factors. Optimization was
either carried out automatically by the Analyst software.
For biodegradation experiments of TFMHxOH and HFBPrOH,
standards of the parent compounds were always prepared from
10% (or lower) of the theoretical concentration up to approxi-
mately 125–150% of the initial concentration. At least five
calibration points were used. Detailed HPLC–ESI-MS settings
and parameters are given in the Supplementary material.
2. Methods and materials
2.4. Biotransformation experiments
2.1. Chemicals
Biotransformation experiments were carried out under aerobic
conditions with effluent water from a municipal wastewater
treatment plant (Beuerbach, Hesse, Germany) in amber glass
bottles. Effluent water was chosen as inoculum to facilitate
detection of unknown transformation intermediates by MS, as the
more commonly used activated sludge may cause severe ion
suppression in ESI-MS. The substances under investigation were
spiked at a level of approximately 10 mg Lꢁ1. If possible, aqueous
stock solutions were used, but if water solubility was too low for
spiking purposes, methanolic stock solutions were prepared.
For all assays, sterile controls were prepared in the same way as
the active assay by addition of 10 g Lꢁ1 sodium azide. Blanks were
prepared like the active assay without addition of the active test
substance, but including methanol, if used for preparation of the
stock solution. The bottles were placed on an automatic shaker
which was set at 100 rpm during the entirebiotransformation assay.
Experiments were carried out at room temperature (between 20 8C
and 25 8C).
After addition of all substances needed for the assay, the bottles
were gently mixed for 30–45 min, after which the initial sample
was drawn. For TFM-alcohols, the bottles were then closed with a
glass plug and parafilm for four days to avoid volatilization of the
poorly water-soluble alcohols. Afterwards, the bottles were
aerated with an aquarium pump for 30 min daily. HFBPrOH
bottles were aerated daily from the beginning.
Sampling was carried out daily in the beginning of the
experiment and the sampling interval was then constantly
increased to weekly after three weeks. Samples were frozen at
ꢁ30 8C until analysis and ultrasonicated for 10 min after thawing
before samples were prepared.
All solvents and mobile phase modifiers used for HPLC, MS and
biodegradation study applications were at least of HPLC grade.
UltrapurewaterwaspreparedusingaMilliporeSimplicity185witha
Slimpak2Cartridge(Millipore,Milford, USA). Ifnotstateddifferently,
this water was used for preparation of solutions and dilution.
3-(Trifluoromethoxy)-propan-1-ol (TFMPrOH, 98% purity), 6-
(trifluoromethoxy)-hexan-1-ol (TFMHxOH, unknown purity) and
HFBPrOH (87% purity, 5% methyl isomer impurity), 2,2,3,3,4,4,4-
heptafluorobutanol (HFBOH) present in unknown fraction) were
synthesized and supplied by Merck (Darmstadt, Germany).
2.2. Fluoride determination
Fluoride was determined by anion exchange chromatography
on a Metrohm modular system consisting of a 709 IC Pump, a 752
Pump Unit, a 733 Separation Center, a 732 IC Detector and a 762 IC
Interface (all from Metrohm, Herisau, Switzerland). Separation was
carried out on a Metrosep A Supp 5 100 ꢀ 4.0 mm (Metrohm,
Herisau, Switzerland) and a Varian C18
3 mm precolumn (Varian,
Frankfurt, Germany). The eluent was aqueous Na2CO3 (1.6 mM)
and NaHCO3 (0.5 mM) at a flow rate of 0.8 mL/min.
A fluoride stock solution was prepared from analytical grade
sodium fluoride (Sigma, Seelze, Germany) at a concentration of
approximately 1 mg mLꢁ1. Standards were prepared at fluoride
concentrations of 0.1 mg Lꢁ1, 0.25 mg Lꢁ1, 0.5 mg Lꢁ1, 0.75 mg Lꢁ1
,
1 mg Lꢁ1 and 2.5 mg Lꢁ1
.
In order to determine the amount of organically bound fluorine,
one sample of TFMPrOH biotransformation assay was additionally
measured with and without UV decomposition by use of an ion-
selective electrode from Metrohm (Zofingen, Switzerland) as
explained previously [11].
Further details on biotransformation assay and preparation of
samples for analytical methods are given in the Supplementary
material.
2.3. HPLC–ESI-MS analysis
3. Results and discussion
Substance-specific analysis and structure elucidation was
carried out by HPLC–ESI-MS/MS analysis. The chromatographic
setup consisted of two Series 200 Micro Pumps, a Series 200
vacuum degasser, and a Series 200 autosampler (Perkin Elmer,
Norwalk, CT, USA). The chromatograph was coupled to a hybrid
3.1. 3-(Trifluoromethoxy)-1-propanol
Even though mass spectrometric signals were observed for
TFMPrOH, the study of biotransformation suffered from the

82
T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
incapability of measuring the initial compound by HPLC–ESI-MS.
The reason for this could be suppression of the very weakly basic
aliphatic hydroxyl group due to early elution from the HPLC and
ion suppression effects. Also GC-EI/MS was tested as a commonly
applied method for volatile analytes, but no signal of the initial
compound was observed either.
Thus, focus was set on transformation products, mainly on
acidic ones, since aliphatic alcohols are commonly transformed to
acids under aerobic conditions. Since TFM-substituted compounds
had been previously measured in our laboratory [10], negative ion
CID fragmentation of compounds bearing this functional group
was known. Fragmentation usually yielded the trifluoromethano-
late anion at m/z 85. In such cases, precursor ion scans can give
valuable information about unknown compounds. Here, a com-
pound with an m/z ratio of 157 was detected (see Supplementary
material). The molecular mass of 158 Da (m/z 157 in negative ESI
As described in the experimental section, the assay was kept
closed for the first four days to prevent volatilization of TFMPrOH
or volatile transformation products. Therefore, no data points are
available for this time period. Both TFMC and TFMPrA showed a
rapid increase in peak area. TFMC reached a maximum after seven
days and was then completely transformed showing no signal
anymore after 14 days, whereas TFMPrA reached a plateau just at
this time point. Even after 34 days, no decline was observed
suggesting stability of substance.
This is striking since so far, it was assumed that longer
v-TFM
substituted alkanoic acids were degraded by -oxidation until
b
TFMeOH is generated [10]. The stability of TFMPrA was also
verified by different fluoride measurements, as summarized in
Table 1. Besides free inorganic fluoride, fluoride was also measured
after UV decomposition of organic compounds, and after filtration
and UV decomposition. This is a very valuable method to affirm
results obtained from organic analysis, such as LC–ESI-MS, and to
determine the percentage contribution of biotransformation
pathways in case of multiple pathways [11,12].
Herein, fluoride measurement after UV decomposition yielded
4.54 mg Lꢁ1, which is ca. 17% higher than the theoretical value, as
calculated based on the initially spiked test compound. This value
can be substantiated by evaporation of the biodegradation
medium, which was kept at room temperature and aerated
regularly contributing to evaporation of water. Assumed that only
biotransformation of TFMC can lead to inorganic fluoride and
that this pathway yields 100% fluoride, it accounts for 0.75/
4.54 = 16.5%.
polarity) suggests an
v-oxidized transformation product, namely
3-trifluoromethoxypropionic acid (TFMPrA). Furthermore, a sec-
ond peak was detected at m/z 129 (data not shown), which
suggests the presence of trifluoromethyl carbonate (TFMC). Their
structure was verified by ‘enhanced product ion scans’ using the
linear ion trap to scan the product ions (see Supplementary
material). While TFMC showed neutral loss of CO and CO2 during
collision-induced dissociation, respectively, TFMPrOH merely
gives rise to the trifluoromethanolate anion by loss of acrylic
acid. It should be pointed out that no confirmation of the detected
metabolites was made by comparison with authentic standards,
thus chemical identities of transformation product remains
tentative.
The tentative biotransformation pathway is presented in Fig. 2.
The temporal evolution of the two transformation products
detected was investigated in single ion monitoring mode and is
presented in Fig. 1.
The first pathway is likely achieved by v-oxidation of the hydroxyl
group to the carboxylate group yielding TFMPrOH, probably via the
short-lived aldehyde.
Fig. 1. Temporal evolution of TFMPrOH metabolites TFMPrA and TFMC expressed as peak area by HPLC–ESI-MS versus time.

T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
83
Table 1
released as inorganic fluoride, the long-term fate and effects of the
remaining major transformation product TFMPrA are questionable
and cannot be assessed without further laborious studies.
Concentrations of inorganic fluoride and organically bound fluorine in TFMPrOH
biodegradation samples after 77 days.
b
(Fꢁ) [mg Lꢁ1
]
Fluoride
[% theoretical]
3.2. 6-(Trifluoromethoxy)-1-hexanol
Inorganic
0.75
4.50
19.3
116
After filtration and UV
decomposition
TFMHxOH biotransformation was initially investigated by the
temporal assessment of the test compound itself. In this case it was
possible by measuring the protonated molecule [M+H]+ in positive
ESI mode.
After UV decomposition
of the whole sample
4.54
117
The assay suffered from loss of TFMHxOH also in the sterilized
experiment (see Fig. 3a). This can be reasoned by a mixture of
volatilization and adsorption to glassware or particles present in the
inoculum, which is typical for biodegradation studies of fluorinated
substances [16]. Yet, the reduction in concentration of TFMHxOH
proceeds more rapidly in the active assay indicating biotransforma-
tion. Complete primary transformation was achieved after 11 days.
Similarly to the biodegradation assay of TFMPrOH, two acidic
transformation products were detected for TFMHxOH as well, in
this case 6-(trifluoromethoxy)hexanoic acid (TFMHxA) and again
TFMC.
It must be highlighted that the pathway leading to TFMC and
inorganic fluoride is highly speculative, since no intermediates
were detected. These intermediates were screened for in single ion
monitoring (SIM), but none of them were detected, probably due to
the transient nature of these intermediates. However, it can be
stated that no link between TFMPrA and the second pathway can
be established, since the HPLC–MS signal of TFMPrA remains
constant. Thus, a second pathway must exist. The detection of
TFMC must have been preceded by an insertion of an oxygen atom
into the alkyl chain. A possible but unproven explanation is given
as follows.
A likely reaction would be the in-chain oxidation of the alkyl
chain in proximity to the TFM group yielding the hemiacetal I,
which can be oxidized to the ester II. A Baeyer–Villiger-oxidation-
like reaction, that is an insertion of an oxygen atom leading to the
3-hydroxypropyl trifluoromethyl carbonate III. Ester hydrolysis
leads to TFMC, which subsequently decays to TFMeOH and thus
finally yields fluoride and carbon dioxide.
The identity of TFMHxA was verified by performing product ion
scans (see Supplementary material). Besides the common and
indicative trifluoromethanolate anion at m/z 85, a second product
ion at m/z 113 was observed, which represents an
v-unsaturated
alkenoate anion. This ion results from cleavage of neutral TFMeOH
from the precursor ion. This ion species had also been detected in
measurements of Peschka et al. when examining 10-(trifluor-
omethoxy)decane-1-sulfonate [10].
Both compounds were generated rapidly with TFMC showing a
maximum after only four days and no signal after 11 days, as
shown in Fig. 3b. TFMHxA showed its maximum concentration
after five days and – in stark contrast to TFMPrA – was completely
degraded afterwards showing complete degradation also after 11
days. The main difference between the degradation between
TFMPrOH and TFMHxOH, however, is a very high degree of
As the
v-oxidation pathway apparently proceeds much more
rapidly, the 3-(trifluoromethoxy)propoxy group does not seem to
be a good candidate for environmentally friendly substitutes of
PFASs assuming that in potential fluorosurfactants, the 3-
(trifluoromethoxy)propoxy group is cleaved off as TFMPrOH.
Whereas at least 15% of the organically bound fluorine can be
Fig. 2. Proposed biotransformation pathways of TFMPrOH; BVO, Baeyer–Villiger oxidation. The dotted rectangles indicate the transformation products identified by HPLC–
ESI-MS/MS.

84
T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
Fig. 3. (a) Temporal evolution of TFMHxOH in active and sterile biodegradation assays. The black dotted line indicates the theoretical initial TFMHxOH concentration and (b)
temporal evolution of TFMC and TFMHxA during biotransformation (please note the logarithmic scale) and percentage fluoride release.

T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
85
defluorination, which accounted for nearly 100% of the theoretical
value for TFMHxOH.
oxidized in a similar manner as TFMPrOH, leading to TFMHxA.
Unlike TFMPrA, TFMHxA is further metabolized, as suggested in
Fig. 4. In contrast to TFMPrA, beta oxidation was considered one
way to accomplish chain shortening of TFMHxA. In the latter
one, there is a longer spacer between the fluorinated functional
group and the carboxylic acid group thus facilitating attack of
enzymes.
This difference in mineralization yield between TFMPrOH and
TFMHxOH cannot be entirely explained, because no further
transformation products were detected in the course of this
experiment. Thus, the degradation pathways presented here are
speculative (see Fig. 4). It is assumed that TFMHxOH is first
Fig. 4. Proposed degradation pathways of TFMHxOH; please note that possible transitions from the pathway A to pathway B are possible by oxidation, but not depicted here.
For further information, see text. The dotted rectangles indicate the transformation products identified by HPLC–ESI-MS/MS.

86
T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
Fig. 5. (a) Temporal evolution of HFBPrOH in the active and sterilized biodegradation assay. The dotted line indicates the theoretical initial concentration and (b) temporal
evolution of biotransformation products of HFBPrOH expressed as their peak area in HPLC–MS.

T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
87
Fig. 6. Proposed degradation pathway of HFBPrOH with PFBA being the dead-end transformation product. The dotted rectangle indicates detection of the transformation
product without confirmation by MS/MS, the rectangle indicates detected and confirmed transformation products. T I and II were not detected.

88
T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
Since both TFMHxA and TFMC are degraded and no further
dissociation to an acetate ion and the neutral molecule it was
attached to, which disallows further measurement by MS. The
structure of HFBOHPrA could not be verified. In SIM mode, a
compound with an m/z of 287 was detected and its temporal
evolution recorded as shown in. However, no significant product
ions were detected in enhanced product ion mode as a result of the
low intensity. Thus, its structure was only postulated but not
confirmed. For HFBAA however, verification by MS/MS experi-
ments was carried out. The corresponding product ion spectrum
and explanation of product ions is shown in the Supplementary
material.
organic transformation product was detected, it is hypothesized
that both transformation products yield inorganic fluoride. Since
the curves for TFMHxA and TFMC are not temporally shifted, no
link between the two transformation products can be made, and
yet, TFMHxA is assumed to be transformed to TFMC in a similar
manner as TFMHxOH is metabolized itself. That is, TFMHxA could
be oxidized to an alcohol in proximity to the ether group yielding
the hemiacetal I. In turn, this can be oxidized to the ester II, which
is oxidized to the Baeyer–Villiger-like carbonate III. Ester
hydrolysis yields TFMC, which is eventually mineralized.
Simultaneously, oxidation of TFMHxOH analogous to TFMPrOH
proceeds. Even more complex, transformation products IV to VI
might be oxidized to the respective molecule on the left side. For
instance, IV might be oxidized to I by oxidation of the terminal
hydroxyl group to the carboxylic acid. Which of these reactions
take place cannot be scrutinized here because some of these
reactions are supposed to proceed very rapidly and thus the
intermediates remain undetected.
The temporal evolution of the transformation products is
illustrated in Fig. 5a. The first substances were detected after four
days, except for PFBA, which was only measured above the LOD
after 14 days, which suggests that PFBA is generated via an indirect
pathway, i.e. via non-detected transient transformation products.
It is suggested that PFBA is generated by a complex pathway
as shown in Fig. 6. Starting with a hydroxylation in
a-position (I)
and subsequent oxidation of this terminal hydroxyl group to
the aldehyde (II), the lactic acid derivative 3-(2,2,3,3,4,4,4-
heptafluorobutoxy)-2-hydroxypropanoic acid (HFBOHPrA) could
be formed. This in turn is transformed to (2,2,3,3,4,4,4-
Alternatively, TFMHxA might be shortened via
b-oxidation, as
was initially presumed during the investigation of 10-(trifluor-
omethoxy)decane-1-sulfonate by Peschka et al. [10] and also in
other studies carried out with
v
-substituted alkane-1-sulfonates,
heptafluoroboxy)acetic acid (HFBAA) by a-oxidation. b-Oxida-
which were initially transformed to the respective carboxylates
[11,12]. This would involve the two transient transformation
products VII and VIII as shown in pathway C, whereas subsequent
tion then leads to 2,2,3,3,4,4,4-heptafluorobutan-1-ol (HFBOH),
which is easily converted to PFBA, probably via the correspond-
ing aldehyde. As expected, PFBA seems to be the dead-end
transformation product of HFBPrOH, since no fluoride was
b
-oxidation of VIII would contribute to the presence of TFMeOH.
However, a -oxidation pathway would not form TFMC, indicating
b
detected. The transformation products I and II were not
that at least one other transformation pathway must be involved.
Also, VII and VIII were not detected by MS, even though they are
supposed to be analytes that can be detected in negative ESI mode
in a straightforward way being at least semi-strong acids.
The complete defluorination renders the 6-(trifluoromethox-
y)hex-1-oxy group a promising building block for environmentally
friendly substitutes of PFASs. If implemented in larger organic
molecules via an ester bridge, TFMHxOH is a likely to be formed by
chemical or enzymatic ester hydrolysis. The transformation rate is
probably highly depending on the structure of the whole molecule,
especially when it comes to enzymatic cleavage, where steric
aspects may predominate.
detected, but highly reductive compounds such as aldehydes
or vicinal diols are rarely detected in degradation studies.
4. Conclusion
Three different candidates for building blocks in novel
fluorinated surfactants were investigated with respect to their
degradability.
The two structurally related TFMPrOH and TFMHxOH showed
analogous transformation products, but a drastic difference in
mineralization yield, expressed as the molar percentage release of
fluoride. This is due to the stability of TFMPrA, the carboxylic acid
associated to TFMPrOH, which is generated in the biotransforma-
tion assay, but not further degraded. In contrast to this, the
carboxylic acid derivative of TFMHxOH, TFMHxA, is generated and
completely transformed to other transformation products. Thus,
TFMHxOH biotransformation yielded nearly 100% of the theoreti-
cal fluoride, but TFMPrOH was only defluorinated to an extent of
15%. It is assumed that the remaining 85% are accumulated in form
of TFMPrA, whose environmental behavior is unknown. Thus, from
an environmental point of view, TFMHxOH would be a suitable
substitute for longer-chained PFASs. However, it is questionable
whether the compound can compete with long-chained fluorinat-
ed compounds in terms of performance.
The reason for the dissimilar stability of TFMHxA and TFMPrA
cannot be entirely explained with the knowledge obtained. One
possible reason is a different accessibility of the methylene group
in vicinity to the TFM group in the two compounds. In TFMPrA, the
methylene group might be shielded by the polar carboxylic acid
function, so that it cannot be hydroxylated enzymatically, which in
turn could be possible in TFMHxA, where the carboxylic acid group
is separated by more methylene groups. However, these sugges-
tions remain speculative unless one or several transient transfor-
mation products are detected in other studies.
3.3. 1-(2,2,3,3,4,4,4-Heptafluorobutoxy)-propan-2-ol
HFBPrOH was found to be measurable as the acetate adduct
[M+acetate]ꢁ, similar to FTOHs [17]. In order to increase
reproducibility of the quantitative analysis, 6:2-FTOH was used
as an internal standard.
As illustrated in Fig. 5a, the initially spiked concentration of
HFBPrOH could not be verified by HPLC–MS measurement, which
is probably due to pronounced adsorption of the compound. In the
sterilized assay, the concentration remains stable around 20
mM
for 45 days, whereas in the active assay, a drop in concentration
can be observed after 8 days. Primary biodegradation then
proceeds rapidly so that almost no HFBPrOH can be detected
after 14 days.
Different transformation products were sought for by HPLC–
MS/MS in MRM mode or SIM mode with subsequent verification by
MS/MS measurements. Non-acidic potential transformation
products, i.e. those carrying only ether and hydroxyl groups,
were monitored as [M+acetate]ꢁ ! acetate. The structures of
these transformation products are shown in the Supplementary
material.
Several transformation products were observed, which were
neither detected in the sterilized nor in the control assay.
Unfortunately, for non-acidic compounds, no further verification
of the structure could be made, since acetate adducts fragment by
Interestingly, the implementation of ether bridges as in
TFMHxOH and TFMPrOH does not imply biotransformation. Such
ether bridges are also implemented into new substitutes for
traditional PFASs, such as perfluoroether carboxylic acids and

T. Fro¨mel, T.P. Knepper / Journal of Fluorine Chemistry 177 (2015) 80–89
89
perfluoroether sulfonic acids, which were also found not to be
References
biodegraded [8].
[1] R.C. Buck, J. Franklin, U. Berger, J.M. Conder, I.T. Cousins, P. de Voogt, A.A.
Jensen, K. Kannan, S.A. Mabury, S.P. van Leeuwen, Integr. Environ. Assess.
Manag. 7 (2011) 513–541.
[2] Stockholm Convention, Commussion Regulation (EU) No 757/2010 (2010),
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:223:0029:
0036:en:PDF, last access: 28-01-2015
From an environmental point of view, HFBPrOH exhibits both
advantageous and disadvantageous properties. Whereas it does
not lead to precarious and hazardous transformation products
like PFOS or PFOA, it is also not mineralized, which would
represent the ideal case. Instead, the already known substance
PFBA is generated via biotransformation. PFBA is not a novel
PFAS and has been detected in environmental samples for
several years [18–21]. PFBA has been shown to exhibit less
environmentally adverse effects, such as less pronounced
bioaccumulation and toxicity [22]. Ecotoxicity of PFBA has been
investigated and it also gives rise to dramatically reduced effects
when compared with longer chained PFCAs [14,15]. Whereas
most properties and environmental effects of short-chained
PFASs seem to be more benign as compared with their long-
chained counterparts, there is at least one drawback of the short
perfluoroalkyl chain: their mobility. It was shown that PFBA is
not retained during soil passage [23], which in turn indicates
that it may easily reach groundwater and eventually drinking
water, where it has already been detected [24,25], sometimes
[3] A. Remde, R. Debus, Chemosphere 32 (1996) 1563–1574.
[4] B.D. Key, R.D. Howell, C.S. Criddle, Environ. Sci. Technol. 32 (1998) 2283–2287.
[5] T. Fro¨mel, T.P. Knepper, Rev. Environ. Contam. Toxicol. 208 (2010) 161–177.
[6] M.P. Krafft, J.G. Riess, Chemosphere 129 (2015) 4–19.
[7] M. Scheringer, X. Trier, I.T. Cousins, P. de Voogt, T. Fletcher, Z. Wang, T.F.
Webster, Chemosphere 114 (2014) 337–339.
[8] Z. Wang, I.T. Cousins, M. Scheringer, K. Hungerbuehler, Environ. Int. 75 (2015)
172–179.
[9] J. Liu, S. Mejia Avendano, Environ. Int. 61 (2013) 98–114.
[10] M. Peschka, N. Fichtner, W. Hierse, P. Kirsch, E. Montenegro, M. Seidel, R.D.
Wilken, T.P. Knepper, Chemosphere 72 (2008) 1534–1540.
[11] M. Peschka, T. Fro¨mel, N. Fichtner, W. Hierse, M. Kleineidam, E. Montenegro,
T.P. Knepper, J. Chromatogr. A 1187 (2008) 79–86.
[12] T. Fro¨mel, M. Peschka, N. Fichtner, W. Hierse, N.V. Ignatiev, K.H. Bauer, T.P.
Knepper, Rapid Commun. Mass Spectrom. 22 (2008) 3957–3967.
[13] M.R. Zachariah, W. Tsang, P.R. Westmoreland, D.R.F. Burgess, J. Phys. Chem. 99
(1995) 12512–12519.
[14] G.H. Ding, T. Fro¨mel, E.J. van den Brandhof, R. Baerselman, W. Peijnenburg,
Environ. Toxicol. Chem. 31 (2012) 605–610.
[15] A. Hagenaars, L. Vergauwen, W. De Coen, D. Knapen, Chemosphere 82 (2011)
764–772.
[16] N. Wang, B. Szostek, R.C. Buck, P.W. Folsom, L.M. Sulecki, J.T. Gannon,
Chemosphere 75 (2009) 1089–1096.
[17] U. Berger, I. Langlois, M. Oehme, R. Kallenborn, Eur. J. Mass Spectrom. 10
(2004) 579–588.
[18] A. Moller, L. Ahrens, R. Surm, J. Westerveld, F. van der Wielen, R. Ebinghaus, P.
de Voogt, Environ. Pollut. 158 (2010) 3243–3250.
even up to the
m
g Lꢁ1 range as PFBA is as persistent in the
environment as its longer-chained homologs [6,9]. Thus, it can
be stated that this C3 perfluorinated building block represents a
compromise between satisfactory performance and acceptable
environmental and toxicological profile.
[19] J. Busch, L. Ahrens, R. Sturm, R. Ebinghaus, Environ. Pollut. 158 (2010)
1467–1471.
Acknowledgement
[20] B.F. Scott, C.A. Moody, C. Spencer, J.M. Small, D.C.G. Muir, S.A. Mabury,
Environ. Sci. Technol. 40 (2006) 6405–6410.
[21] L. Ahrens, W. Gerwinski, N. Theobald, R. Ebinghaus, Mar. Pollut. Bull. 60
(2010) 255–260.
The authors would like to thank Karl-Heinz Bauer for fluoride
measurements.
[22] K.P. Das, B.E. Grey, R.D. Zehr, C.R. Wood, J.L. Butenhoff, S.C. Chang, D.J.
Ehresman, Y.M. Tan, C. Lau, Toxicol. Sci. 105 (2008) 173–181.
[23] V. Gellrich, T. Stahl, T.P. Knepper, Chemosphere 87 (2012) 1052–1056.
[24] Y.L. Mak, S. Taniyasu, L.W.Y. Yeung, G.H. Lu, L. Jin, Y.L. Yang, P.K.S. Lam, K.
Kannan, N. Yamashita, Environ. Sci. Technol. 43 (2009) 4824–4829.
[25] M. Wilhelm, S. Bergmann, H.H. Dieter, Int. J. Hyg. Environ. Health 213 (2010)
224–232.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.jfluchem.2015.06.015.