The aqueous photolysis of TFM and related trifluoromethylphenols. An alternate source of trifluoroacetic acid in the environment

Article abstract of DOI:10.1021/es990422c

The lampricide 3-trifluoromethyl-4-nitrophenol (TFM) is added annually to the Great Lakes (approximately 50 000 kg/y), with treatment concentrations varying from 1 to 14 mg/L at source. TFM was shown to undergo photohydrolytic degradation, at 365 nm and u

Full text of DOI:10.1021/es990422c

Environ. Sci. Technol. 2000, 34, 632-637
2.5 month period. It was assumed that the trifluoromethyl
The Aqueous Photolysis of TFM and
Related Trifluoromethylphenols. An
Alternate Source of Trifluoroacetic
Acid in the Environment
substituent of the aromatic ring remained intact as no fluoride
production was observed. Under anaerobic conditions,
partial reduction of the TFM nitro group to an amine has
been observed (8). TFM is almost completely ionized in most
natural waters, so this is unlikely to be an important loss
mechanism due to the lack of adsorption on such sediments.
It has been reported (9, 10) that the trifluoromethyl group
contained within a phenolic ring is susceptible to hydrolytic
degradation when in the ortho or para positions relative to
a hydroxyl group. Conversely, when in the meta position,
the trifluoromethyl group is inert to alkaline or acidic
hydrolysis even at elevated temperatures. However, a pho-
toinduced excited state may allow the cleavage of the carbon-
fluorine bonds (11).
D A V I D A . E L L I S A N D S C O T T A . M A B U R Y*
80 St. George Street, Department of Chemistry,
University of Toronto, Toronto, Ontario, Canada M5S 3H6
The lampricide 3-trifluoromethyl-4-nitrophenol (TFM) is
added annually to the Great Lakes (approximately 50 000 kg/
y), with treatment concentrations varying from 1 to 14
mg/L at source. TFM was shown to undergo photohydrolytic
degradation, at 365 nm and under actinic radiation, to
produce trifluoroacetic acid (TFA). A mechanistic study
for the production of TFA from TFM was conducted, and
the structural parameters associated with the production of
TFA from trifluoromethylated phenols were investigated.
It was found that the yield of TFA is clearly dependent on
the nature of the trifluoromethylated phenol. The nature
of the substituents, the substitution pattern, and the pH
strongly effected the photolytic half-life of the parent
compound and the yield of TFA. The half-life of TFM at
365 nm was found to be 22 h (pH 9) and yielded 5.1% TFA,
and 91.7 h at pH 7, yielding 17.8% TFA. Converting the
nitro substituent of TFM to an amino group caused a decrease
in the half-life to 2.3 min and yielded 11% TFA. The
mechanism for the production of TFA from TFM was
deduced from the pH dependence and the effect of altering
substituents on the trifluoromethylphenol. Ultimately, the
formation of trifluoromethylquinone led to the quantitative
production of TFA.
Trifluoromethyl aromatics, which do not contain an
electron withdrawing substituent, have been shown to
undergo bacterial metabolism to produce trifluoroacetic acid
(TFA) (12). This is an unlikely mechanism for the degradation
of TFM or the subsequent production of TFA due to the
electron withdrawing nitro group.
The major source of TFA in the environment is believed
to be the degradation of the HCFC’s 123, 124, and 134a, which
occurs primarily through reaction with a hydroxy radical
followed by subsequent hydrolysis (13). From this source
alone, the projected concentrations have been calculated to
be >100 µg L-1 by the year 2010 in seasonal wetlands.
However, the concentrations which have been measured in
the environment cannot be accounted for based on current
atmospheric sources (14). These levels have been shown to
inhibit plant growth (15). TFA has also been shown to be a
biliary excretion metabolite of haloethane anesthetics in
infants younger than 5 months, but at greater than 5 months
it has been shown to be completely retained in a entero-
hepatic circulation (16).
To date there is little evidence for the degradation of TFA.
It was reported that trifluoroacetic acid was observed to
degrade in both oxic and anoxic sediments (17), although
this observation has yet to be reproduced (18).
The aqueous photolytic fate of TFM has been investigated
by Carey et al. (19, 20). In the present investigation particular
attention has been paid to certain key intermediates proposed
by Carey et al., specifically the formation of the trifluoro-
methylhydroquinone, from which it was postulated that there
were two competing loss mechanisms: the production of
the trifluoromethylquinone or a reaction which produces
the corresponding acid fluoride (2,5-dihydroxybenzoyl fluo-
ride). Furthermore, Carey et al. hypothesized that this acid
fluoride in turn leads to the formation of gentisic acid (2,5-
dihydroxybenzoic acid) and hydrogen fluoride. Investigations
within our laboratory have indicated that the fate of the
trifluoromethyl substituent was as fluoride or trifluoroacetic
acid (TFA), which had not been previously observed.
Due to the absence of naturally occurring fluorinated
materials and the large spectral windows associated with
fluorine, 19F NMR provides a very powerful method for the
direct analysis of fluorinated materials, reaction intermedi-
ates, and degradation products (21, 22). Therefore this
technique was adopted in the characterization of TFM
degradation.
Introduction
Fluorinated chemicals are rarely observed in nature (1).
However, the incorporation of fluorine in anthropogenic
products is expanding, with fluorine present in many different
types of organic materials, including agrochemicals, poly-
mers, and pharmaceuticals (2). The ultimate fate of the
fluorine contained within these molecules is of considerable
interest. From 1958 the chemical 3-trifluoromethyl-4-nitro-
phenol (TFM) has been used to control the sea lamprey
(Petromyzon marinus) in four of the Great Lakes (Superior,
Michigan, Huron, and Ontario). By 1988, more than 1 million
kg of TFM had been applied to these lakes, and usage since
has been approximately 50 000 kg/ year (3). TFM has also
been introduced in order to control tadpole infestations in
warm water ornamental fish ponds (4).
A possible loss mechanism for lampricides is through
adsorption onto bottom sediments (5). However, studies have
shown microbial degradation is not significant within this
medium (6, 7). The enzymatic biodegradation of TFM was
investigated under aerobic conditions in natural water
sediments (6). No degradation of TFM was observed over a
It was our hypothesis that the production of trifluoro-
methylquinone is of key importance to the production of
TFA. Our objectives were to investigate the role of phenolic
substitution patterns in the rate of degradation of the parent
compound, the mechanism of TFA formation, and the
influence of aromatic substituents on the yield of TFA.
* Corresponding author phone: (416)978-1780; fax: (416)978-3596;
e-mail: smabury@alchemy.chem.utoronto.ca.
9
6 3 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 4, 2000
10.1021/es990422c CCC: $19.00
2000 Am erican Chem ical Society
Published on Web 01/15/2000

Materials and Methods
Chem icals and Reagents. The chemicals 3-trifluoromethyl-
4-nitrophenol (99%), 2-nitro-4-(trifluoromethyl)phenol (99%),
2-(trifluoromethyl)phenol (97%), 3-(trifluoromethyl)phenol
(99%), and 4-(trifluoromethyl)phenol (99%) were purchased
from the Aldrich (Mississauga, Canada). Trifluoroacetic acid
(98%) was purchased from Caledon (Georgetown, Canada).
All other solvents and reagents were analytical purity grade
or better. All chemicals were used without further purification.
Synthesis. Trifluorom ethylquinone was synthesized
through the oxidation of 2-(trifluoromethyl)phenol using
Fremy’s radical in a procedure similar to that used by Zimmer
et al. (23) for the oxidation of phenols. Synthesis of 4-hydroxy-
3-trifluoromethylphenol was carried out directly using the
method of Feiring et al. (24) in which the 3-(trifluoromethyl)-
phenol was oxidized using potassium persulfate. The
4-amino-3-trifluoromethylphenol was synthesized via the
reduction of the 4-nitro-3-trifluoromethylphenol using zinc
metal in hydrochloric acid according to a standard procedure
(25). All products were verified by IR, NMR, and MS.
Photolysis Experim ents. Photolysis experiments were
carried out at 365 nm (Rayonette bulbs), which also show
minor lines at 334 and 313 nm. Photolysis was also confirmed
under actinic radiation in a solar simulator (Suntest CPS
Solar Simulator). All analytes were dissolved in buffered de-
ionized water (Borax buffer (pH 9) and potassium dihydrogen
phosphate (pH 7)) to a concentration typically in the range
of 5-10 × 10-5 M and irradiated in sealed, 100 mL, quartz
vessels. Samples were obtained at appropriate time intervals
and analyzed using the pertinent analysis equipment.
FIGURE 1. Chemical structures used in the mechanistic fate of
TFM.
TABLE 1. Molar Extinction Coefficients and Observed Half-Lives
of Trifluoromethyl Aromatics
molar extinction coefficient, Eo,
compd no.
L mol-1 cm-1, 365 nm, pH 9a
half-life, h (365 nm)
I
II
IV
V
VI
6675 (394 nm )
0.0123 (310 nm )
0.00495 (308 nm )
0.0261 (298 nm )
565 (400 nm )
22.4
0.5
0.04
0.4
∞
a
Values in parentheses correspond to m axim um absorption wave-
length in actinic spectrum .
NMR Analysis. 19F, 1H, and 13C NMR spectra were obtained
for synthesized and degradation products when appropriate,
on a Gemini 500 MHz spectrometer operating at 470.596
MHz with a broad band probe tuned to fluorine, on a Varian
VXR-S 400 MHz spectrometer operating at 376.289 MHz with
a Nalorac 4N probe tunable to H, F, H, and Si. Chemical
shifts were reported relative to CFCl3 (0.00 ppm). The internal
standard 4′-(trifluoromethoxy)acetanalide (TFMAA) was used
for quantification.
Aqueous samples (100-1500 mL) of TFA (10-6-10-5 M) were
passed through a strong anionic exchange column (SAX) at
a flow rate of 5 mL min-1. Elution was carried out using 1
mL of 2 M NaOH, or alternatively with acetone, methanol,
or DMSO containing the organic base 1,8-diazabicyclo[5.4.0]-
undec-7-ene (DBU). The eluents were then spiked with 300
mL of deuterated acetonitrile containing the internal standard
(TFMAA), and TFA was measured directly using 19F NMR
(117 ( 13% for a 500 ppb solution, and 101 ( 6% for a 120
ppb solution, n ) 3).
Chrom atographic Analysis. HPLC analysis was carried
out using either a Waters 600 pump equipped with a 486
UV/ visible tunable detector, a Waters 616 pump equipped
with a 996 photodiode array detector, or on a Perkin-Elmer
instrument equipped with a series 200-IC pump and a diode
array detector 235 C. Separation was achieved using an Alltech
Econosphere C18 5U column. Mobile phases were generally
AcCN/ buffered water mixtures. Fluoride, nitrite, nitrate, and
TFA were analyzed using a Perkin-Elmer Series 200 IC pump
equipped with a Dionex Ionpac AS14 column, and an Alltech
1000 HP conductivity suppresser and conductivity detector;
the mobile phase used was generally Borax buffered water.
Fluoride was analyzed using an Orion ISE. High-resolution
mass spectra were obtained using a Micromass 70-250S
(double focusing) spectrometer in negative EI mode. Data
were obtained at 10 000 (10% Valley) resolution. For GC-MS
analysis aqueous samples were acidified and extracted with
ethyl acetate. Derivatization, if required, was carried out using
diazomethane. Samples were run on a Perkin-Elmer GC Auto
System XL equipped with a Q-Mass 910 quadrupole mass
spectrometer, run in EI mode, with a 30 m, 0.25 mm Simplicity
5 column. The carrier gas was helium at a flow rate of 0.5
mL/ min.
Results and Discussion
Trifluoromethylated compounds used to investigate the
mechanistic pathway and structural requirements for the
production TFA from TFM are shown in Figure 1. The
trifluoromethylhydroquinone (II) was selected since it is
observed in the environment as a degradation product of
TFM (I). Carey et al. (19) postulated that the trifluorometh-
ylquinone (III) is formed from II, thus the degradation of
this species was also investigated. The aminotrifluorometh-
ylphenol (IV) and the trifluoromethylphenols (V) were used
to investigate the influence of the nitro substituent in TFM,
on the production of TFA. Finally, an isomeric form of TFM
(VI) was studied in order to determine the substitution pattern
required for the production of TFA. Molar extinction coef-
ficients for the compounds at 365 nm and their half-lives are
presented in Table 1. The rate of degradation of the parent
compound is not solely dependent upon the amount of light
absorbed at 365 nm. It is also dependent upon the quantum
yield of the molecule. For example compound IV has a lower
absorption at 365 nm but degrades more rapidly than I,
indicating that it has a higher quantum yield.
UV/vis Spectroscopy. Molar extinction coefficients and
UV/ vis spectra were obtained for the analytes using an HP-
diode array model 8452 spectrometer.
Analysis of Trifluoroacetic Acid. A new method was
developed for the analysis of TFA to overcome problems
such as unreliable liquid-liquid extractions and derivati-
zation required for currently accepted procedures (26-29).
Photolysis of TFM (I). Figure 2(a) shows the photolysis
of TFM (pH 9, 365 nm). The concentration of products
fluoride and TFA are shown as a function of time; nitrite was
also observed as a product. Trifluoromethylhydroquinone
(II) was observed as a short-lived intermediate based upon
HPLC/ UV spectral comparison with an authentic sample.
These results are in concordance with work carried out by
9
VOL. 34, NO. 4, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 6 3 3

producing a greater yield of TFA. It is hypothesized that
degradation of TFM, through an excited state of the depro-
tonated anion, occurs more readily at higher pH and does
not lead to the production of TFA. At lower pH, this type of
degradation is reduced due to an increase in concentration
of the protonated form. Given that the pKa of TFM is 6.1 (7),
structure Ib (Figure 4) would be in a 100-fold excess at pH
9 relative to pH 7. This hypothesis is further supported by
the observation that TFM absorption shifts from 297 to 394
nm as pH is changed from 7 to 9. This leads to photonu-
cleophilic substitution occurring to a greater extent at pH 7
and producing the trifluoromethylhydroquinone intermedi-
ate (II), followed by subsequent degradation to TFA. It follows
that, since the production of TFA is proposed to occur from
the intermediate II, the TFA yield would increase at lower
pH. This hypothesis also supported by the observation that
there is an enhanced production of nitrite at pH 7 relative
to pH 9. All dark control experiments showed no loss of TFM.
The photolysis of TFM was also carried out in the presence
of equimolar 4-tert-butylphenol. The rate of reaction was
observed to be identical to that in the absence of the phenol.
It is known that radicals react with aromatic rings in a manner
that superficially resembles that of a nucleophilic substitution
(30), thus kinetically resembling an Sn2 type reaction. If the
degradation of TFM had occurred through a radical reaction,
the rate would be expected to change in the presence of the
phenol due to its radical capturing ability (31).
The structural isomer of TFM, 2-nitro-4-trifluorometh-
ylphenol (Figure 1, structure VI), was photolyzed under the
same conditions. No observable degradation of the isomer
was observed over time periods up to four times greater than
the half-life of TFM, presumably due to the stabilizing
formation of an o-semiquinone.
FIGURE 2. (a) The photolysis of TFM (I) at pH 9.2. Loss of TFM and
the production of fluoride and TFA. No observable loss of TFM in
the absence of light. (b) Photolysis of TFM at pH 7. Loss of TFM and
the production of nitrite, TFA, and fluoride. No observable loss of
TFM in the absence of light.
Photolysis of Trifluorom ethylhydroquinone (II). Figure
5 shows the photolysis of trifluoromethylhydroquinone (pH
9, 365 nm). The rate of production of gentisic acid, fluoride,
and trifluoroacetic acid are indicated. The half-life of the
parent compound was observed to be 31 min. When complete
degradation of the parent molecule had occurred, the yield
of TFA was 6% and the yield of fluoride was 82%; the remain-
ing fluoride can be accounted for in residual starting material.
No degradation of the parent compound was observed in
the dark. 19F NMR of the photolyzate indicated the presence
of a short-lived trifluoromethylquinone (δ ) -59 ppm).
Photolysis/Hydrolysis of the Trifluorom ethylquinone
(III). Trifluoromethylquinone appeared to degrade rapidly
(<5 min) and quantitatively at pH 9 (observed from HPLC)
to TFA. To investigate the mechanism for the production of
the TFA from (III), the aqueous samples were extracted and
derivatized with diazomethane. From mass spectral evidence
the degradation structures shown in (Figure 6) were tenta-
tively assigned. It would appear that initial hydroxylation of
the quinone occurs to produce (1) or the 3,6-dihydroxy-
quinone isomer. This is followed by cleavage of the quinone
ring at the 3,4-bond (2). The ring is then further cleaved at
the 1,2-bond to yield dihydroxymaleic acid (3) and trifluo-
romethylglyoxal (4). Compound 4 is then presumed to
decarboxylate to produce TFA, in a manner similar to that
published for pyruvic acid (32).
Proposed Mechanism for the Production of TFA from
TFM. The proposed mechanism for the production of TFA
from TFM is outlined in Figure 7. The initial steps of this
mechanism are similar to that suggested by Carey et al. (19).
The results suggest the overall photolytic degradation of TFM
occurs through the hydrolysis of the nitro substituent via a
photonucleophilic aromatic substitution to produce a nitrite
anion; fluoride can also be lost from the excited state of the
TFM (Figure 5). The nucleophilic substitution results in the
production of a trifluoromethylhydroquinone, which would
then appear to undergo two competing reactions. One
Carey et al. (19, 20), in which aqueous field samples that had
contained TFM were analyzed using mass spectroscopy.
Evidence was presented for an initial photoinduced nucleo-
philic hydroxyl displacement of the nitro group, which led
to the tentative assignment of trifluoromethylhydroquinone
and a trifluoromethylquinone, the latter considered to have
been produced through thermal or photochemical oxidation
of the hydroquinone precursor. This postulated mechanism
is further supported by the observation that heteroaromatics
containing a trifluoromethyl substituent undergo enhanced
nucleophilic substitution (36).
In the present studies the observed pseudo-first-order
half-life for TFM was 22.4 h. Nitrite and fluoride were
produced at the time of initial irradiation, suggesting they
were produced directly from TFM. Observable TFA produc-
tion occurred only after irradiation for 10 h, suggesting that
TFA was produced from an intermediate, rather than directly
from TFM. The yield of TFA was approximately 3.9% when
76% of the parent molecule had undergone photolysis. Figure
3 shows the time-dependent 19F NMR obtained over the
course of irradiation at 365 nm. The production of TFA relative
to the internal standard TFMAA is shown. Photolysis of TFM
using actinic radiation yielded the same photodegradation
products as those observed at 365 nm. The half-life of the
TFM under these conditions was observed to be 10.7 h.
TFM photolysis was also conducted at pH 7, and identical
products were observed (Figure 2(b)). TFM was more
persistent at pH 7, with a half-life of 91.7 h. The yield of TFA
at this lower pH was 8.9%, when 50% of the parent molecule
had undergone photolysis. At a lower pH there is a shift in
the equilibrium toward the protonated form of TFM, overall
9
6 3 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 4, 2000

FIGURE 3. Time dependent 19F NMR for the production of TFA (-71.2 ppm) from TFM (-56.3 ppm). The concentration is relative to the
internal standard TFMAA (-54.3 ppm).
FIGURE 4. Mechanism for the enhanced production of TFA from
FIGURE 6. Degradation products observed from the hydrolysis
TFM by altering the pH. Shift in equilibrium at higher pH from the
production of a hydroquinone (II) toward semiquinone (Ib) which
does not lead to the production of TFA.
trifluoromethylquinone. Parent ion mass is indicated in parentheses
for the methylated product.
FIGURE 7. Proposed mechanism for the production of TFA from
TFM (I). The alternate pathways for the degradation of TFM are also
indicated.
FIGURE 5. The photolysis of trifluoromethylhydroquinone (II) and
the production of gentisic acid (VIII), fluoride, and TFA.
key intermediates that were observed. The rate of degradation
possibility is the production of an acid fluoride through the
expulsion of 2 mol of fluoride. This is then followed by loss
of hydrofluoric acid leading to the formation of the genistic
acid. The second possibility is a photochemical oxidation of
the hydroquinone to a trifluoromethylquinone. The trifluo-
romethylquinone then undergoes what would appear to be
a hydrolytic degradation to produce trifluoroacetic acid. The
mechanism of this degradation appears to be complicated
due to the numerous products observed in the GC-MS. A
mechanistic postulation is presented on the basis of certain
of TFM is strongly dependent upon the degree to which the
compound is ionized. This is evident from the observed 4-fold
increase in the half-lives of TFM from pH 9 to pH 7. As
previously indicated, the production of TFA by the photo-
nucleophilic hydroxyl substitution of the nitro group is further
supported by the increased production of TFA at lower pH
values due to the shift in equilibrium toward the production
of the hydroquinone (Figure 4).
Possible Structural Requirem ents for the Production of
TFA. Trifluoroacetic acid is produced from 3-trifluoromethyl-
9
VOL. 34, NO. 4, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 6 3 5

phenol (33). The role of this substituent on the production
of TFA was investigated; the amount of TFA produced in the
absence of this group and when it is replaced with a strong
electron donating group.
Photolysis of 4-Am ino-3-trifluorom ethylphenol (IV).
The strong nitro electron withdrawing substituent of the TFM
was replaced with a strong electron donating amino group.
Figure 8 shows the loss of the parent compound (IV) through
photolysis and the production of fluoride, 2,5-dihydroxy-
benzoic acid (VIII), and TFA. Approximately 11% of the
expected amount of TFA was produced from the parent
compound, when the parent molecule had undergone 100%
photolysis. The half-life of the parent compound was 2.3
min. Mechanistically the production of TFA from the
4-amino-3-trifluoromethylphenol appears to be complicated
as seen from the numerous peaks that arise in the HPLC
chromatogram. Further investigations toward elucidating the
mechanism will be carried out.
Photolysis of Trifluorom ethylphenols (V). The photolysis
of the three isomers o-, m-, and p-trifluoromethylphenol at
365 nm was executed. The relative rates of degradation, based
upon the amount of fluoride that is produced, are shown in
Figure 9.
As can be seen from the figure the meta isomer is stable
in the dark. Both the ortho and the para isomers are not. A
strong electrical interaction of electron donating substituents
FIGURE 8. Photolysis of 4-amino-3-trifluoromethylphenol (IV). Loss
of the parent compound, production of TFA, fluoride, and gentisic
acid (VII).
4-nitrophenol (I). An investigation toward discovering the
structural requirements within the phenol ring was carried
out. An aromatic nitro group is a very strong electron
withdrawing substituent that results in a pKa of 4.91 for a
FIGURE 9. (a-c). The photolysis of 2-, 3-, and 4-(trifluoromethyl)phenols (V) and the production of fluoride. The reaction mechanism for
the hydrolysis of each of the compounds in the dark is given, along with the relative rates.
9
6 3 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 4, 2000

Acknowledgments
This research was funded by an operating, equipment, and
strategic grant from NSERC. This work, in part, was developed
and carried out in the ANALEST facility kindly supported by
Perkin-Elmer Canada, Analytical Instruments Division.
Literature Cited
(1) Peters, S. R.; Hall, R. J. Nature 1960, 187, 573.
(2) Key, B. D.; Howell, R. D.; Criddle, C. S. Environ. Sci. Technol.
1997, 31, 2445.
(3) Hewitt, M. L.; Munkittrick, K. R.; Scott, I. M.; Carey, J. H.;
Solomon, K., R.; Servos Mark, R. Environ. Tox. Chem. 1996, 15,
894.
(4) Gabbadon, P. W.; Chapman, F., A. Progressive Fish-Culturist
1996, 58, 23.
(5) Allen, J. L., Dawson, V. K., and Johnson, D. A. Can. J. Fish. Aquat.
Sci. 1986, 43, 1515.
(6) Thingvold, D. A.; Lee, G. F. ES & T Research 1981, 15, 1335.
(7) Scott, B. F.; Carey, J. H.; Nagy, E.; Dermott, R. Water Poll. Res.
J. Canada. 1984, 19, 59.
FIGURE 10. Photolysis of 3-(trifluoromethyl)phenol and the produc-
tion of fluoride, 3-hydroxybenzoyl fluoride (IX), 3-hydroxybenzoic
acid (X), and fluoride.
(8) D. Lui, M. E. Fox. Biodegradation of TFM in a cyclone fermentor
system, in biodeterioration; Proceedings of the Fourth Inter-
national Biodeterioration; Pitman: Berlin, London, 1979.
(9) Liotta, C. L.; D. F. Smith, J.; Harry P. Hopkins, J.; Rhodes, K. A.
J. Phys. Chem. 1972, 76, 1909.
(10) Kozachuk, D. N.; Serguchev, Y. A.; Fialkov, Y. A.; Yagupol’skii,
L. M. Zh. Organicheskoi Khim. 1973, 9, 1918.
(11) Wirz, J.; Seiler, P. Helv. Chem. Acta 1972, 55, 2693.
(12) Engesser, K. H.; Cain, R. B.; Knackmuss, H. J. Arch. Microbiol.
1988, 149, 188.
(13) Scientific Assesment of Stratospheric Ozone: 1989 Global Ozone
Research and Monitoring Project; Rep. No., World Meteorological
Organization: Geneva, 1989; Vol. II.
(14) Boutonnet, C. J.; Bingham, P.; Calamari, D.; de Rooij, C.; Franklin,
J.; Kawano, T.; Libre, J.-M.; McCulloch, A.; Malinverno, G.; Odom,
J. M.; Rusch, G. M.; Smythe, K.; Sobolev, I.; Thompson, R.; Tiedje,
J. M.; Human Ecological Risk Asses. 1999, 5, 59.
(15) Tromp, T. K.; Ko, M. K. W.; Rodriguesz, J. M.; Sze, N. D. Nature.
1995, 376, 327.
was initially proposed by Roberts et al. (34). The para isomer
produces 33% fluoride immediately, after which little change
is seen over time. These results are consistent with the loss
of hydrogen fluoride through hydrolysis of the ortho and the
para isomers to produce the semiquinones, 2-(difluoro-
methylene)-2,5-cyclohexadien-1-one and 4-(difluorometh-
ylene)-2,5-cyclohexadien-1-one, respectively. The rate of
hydrolysis of the para isomer is greater than the ortho isomer
due to the linear conjugation produced as opposed to cross
conjugation. The para isomer is then presumed to undergo
polymerization as observed by Kozachuk et al. (10). The
aqueous solution became opaque over time due to the
formation of an insoluble polymer. A semiquinone is not
observed for the meta isomer and, hence, is stable in the
dark (Figure 9).
Further investigation into the mechanism for the deg-
radation of 3-(trifluoromethyl)phenol is shown in Figure 10
(pH 9, 365 nm). Photolysis is observed to initiate the loss of
2 mol of fluoride to produce the acid fluoride, 3-hydroxy-
benzoyl fluoride (IX). Assignment of this compound was
based upon an 19F NMR signal at +44 ppm, which is highly
indicative of an acid fluoride. The acid fluoride (IX) then
undergoes further hydrolysis to produce 3-hydroxybenzoic
acid (X). The half-life of the phenol was 22 min.
(16) Wark, H.; Earl, J.; Chau, D.; Overton, J. Anaesth. Intens. Care
1991, 19, 213.
(17) Visscher, P. T.; Culbertson, C. W.; Oremland, R. S. Nature 1994,
369, 729.
(18) Oremland, R. S.; Matheson, L. J.; Guidetti, J. R.; Schaefer, J. K.;
Visscher, P. T. Summary of research results on bacterial
degradation of trifluoroacetate (TFA), November, 1994-May, 1995;
Open File Report 95-OF 95-0422; USGS: Denver, CO, 1995.
(19) Carey, J. H.; Fox, M. E. J. Great Lakes Res. 1981, 7, 234.
(20) Carey, J. H.; Fox, M. E.; Schleen, L. P. J. Great Lakes Res. 1988,
14, 338.
Concentration of 500 mL of the photolyzed sample was
carried out by passing it through a SAX column. 19F NMR
and ion chromatography showed no observable TFA.
In conclusion, the photolysis of TFM under light intensities
and wavelengths which would be equivalent to those found
on a typical summers day in the Great Lakes region leads to
the production of TFA. Based on TFM concentrations of
approximately 4.6 ppm in Lake Ontario tributaries, and that
the photodegradation is the primary loss mode (19), the
average concentration of TFA produced from this source
alone would be 500 ppb. Typical environmental water sam-
ples have shown a concentration in the order of 10-100 ppt
(35), indicating that this might be an important source of
TFA in the aqueous environment surrounding Lake Ontario.
It would appear that the type and sequence of substitution
on a trifluoromethylphenol are of importance to the fate of
the fluorine; ultimately yielding fluoride or TFA. The pH also
strongly affects the production of TFA, for compounds that
contain the correct structural parameters. In addition, the
degree of electron donating/ withdrawing of the substituent
ortho to the trifluoromethyl group defines the yield of TFA.
The results also indicate that the rate of decay and the nature
of the products produced for trifluoromethyl phenols is
dependent upon the substituents of the aromatic ring and
their relative substitution.
(21) Xu, A. S. L.; Kuchel, P. W. Biochim. Biophys. Acta 1993, 1150,
35.
(22) Mabury, S. A.; Crosby, D. G. J. Agric. Food Chem. 1995, 43, 1845.
(23) Zimmer, H.; Lankin, D. C.; Horgan, S. W. Chem. Rev. 1970, 71,
229.
(24) Feiring, A. E.; Sheppard, W. A. J. Org. Chem. 1975, 40, 2543.
(25) Furniss, B. S.; Hannaford, A. J.; Rogers, V.; Smith, P. W. G.;
Tatchell, A. R. In Vogel’s Textbook of Practical Organic Chemistry;
Longman: 1978; Vol. 4.
(26) Zehavi, D.; Seiber, J. N. Anal. Chem. 1996, 68, 3450.
(27) Wujcik, C. E.; Cahill, T. M.; Seiber, J. N. Anal. Chem. 1998, 70,
4074.
(28) Frank, H.; Renschen, D.; Klein, A.; Scholl, H. J. High Res.
Chromatogr. 1995, 18, 83.
(29) Both, D. A.; Jemal, M. J. Chromatogr. 1992, 596, 85.
(30) Perkins, M. J. In Free Radicals; Wiley: New York, 1973; Vol. II.
(31) Evens, C. A. J. Am. Chem. Soc. 1975, 97, 205.
(32) Nord, A. Chem. Rev. 1940, 29, 423.
(33) Maskill In Bronsted Acids and Bases; Oxford University Press:
Stirling, 1985; Vol. I.
(34) Roberts, J. D.; Webb, R. L.; McElhill, E. A. 1950, 72, 408.
(35) Scott, B. F.; Alaee, M. Water Qual. Res. J. Can. 1998, 33, 279.
(36) Kobayashi, Y.; Kumadaki, J. Acc. Chem. Res. 1977, 11, 197.
Received for review April 14, 1999. Revised manuscript re-
ceived December 2, 1999. Accepted December 2, 1999.
ES990422C
9
VOL. 34, NO. 4, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 6 3 7