Efficient mineralization of hydroperfluorocarboxylic acids with persulfate in hot water

Article abstract of DOI:10.1016/j.cattod.2010.02.023

The persulfate (S₂O₈^2-)-induced decomposition of hydroperfluorocarboxylic acids (H-PFCAs), that is, HC_nF_2nCOOH (n = 4, 6, and 8), in hot water was investigated, and the results were compared with the

Full text of DOI:10.1016/j.cattod.2010.02.023

Catalysis Today 151 (2010) 131–136
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Efficient mineralization of hydroperfluorocarboxylic acids
with persulfate in hot water
∗
Hisao Hori , Misako Murayama, Naoko Inoue, Kyoko Ishida, Shuzo Kutsuna
National Institute of Advanced Industrial Science and Technology (AIST), AIST Tsukuba West, 16-1 Onogawa, Tsukuba 305-8569, Japan
a r t i c l e i n f o
a b s t r a c t
2
−
)-induced decomposition of hydroperfluorocarboxylic acids (H-PFCAs), that is,
Article history:
Available online 11 March 2010
The persulfate (S O
2
8
HCnF2nCOOH (n = 4, 6, and 8), in hot water was investigated, and the results were compared with the
results for perfluorocarboxylic acids (PFCAs). This is the first report on the use of hot water to decom-
pose H-PFCAs, which are being developed as alternative surfactants to environmentally persistent and
Keywords:
Fluorine
Surfactant
Decomposition
Perfluorooctanoic acid
Alternative
◦
bioaccumulative PFCAs. Although H-PFCAs showed almost no decomposition in hot water at 80 C in
2−
2−
the absence of S2O8 , the addition of S2O8 to the reaction solution led to efficient mineralization to
−
−
−
F
ions, with F yields [(moles of F )/(moles of fluorine in initial H-PFCAs)] of 96.7–98.2% after 6 h of
2
−
◦
treatment. The decomposition of H-PFCAs induced by S2O8 also proceeded even at 60 C, at which the
initial decomposition rates were 7.1–12.7 times those for the corresponding PFCAs. The reaction mech-
anism can be explained by nucleophilic substitution by SO4 at the carbon atom attached to the ␻-H
•
−
atom of the H-PFCAs, followed by formation of perfluorodicarboxylic acids (HOOCCn−1F2n−2COOH), which
•
−
react with SO4 to give shorter-chain perfluorodicarboxylic acids; this process eventually resulted in
−
complete mineralization to F ions.
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
confirmed that these chemicals do in fact decompose more easily
than PFCAs. Decomposition of these chemicals to F ions would
−
Perfluorocarboxylic acids (CnF2n+1COOH, PFCAs) such as perflu-
be desirable because there is a well-established waste-treatment
orooctanoic acid (C7F15COOH, PFOA) have recently received much
attention because they are ubiquitous environmental contami-
nants [1–3]. These chemicals have been used as products and as
raw materials for surfactants, surface treatment agents, and so on
because of their high surface-active effect, high thermal and chem-
ical stability, and high light transparency [2]. After it became clear
that these chemicals persist and bioaccumulate in the environ-
ment, a provisional health advisory value for PFOA was issued [4],
and efforts to eliminate these chemicals from products and facility
emissions are proceeding [5]. In parallel with the effort to eliminate
PFOA and other bioaccumulative PFCAs, there have been efforts to
develop alternatives to these chemicals.
Hydroperfluorocarboxylic acids (H-PFCAs; HCnF2nCOOH), ␻-
hydroperfluorocarboxylic acids in which a fluorine atom of the
terminal trifluoromethyl group in PFCAs has been replaced by a
hydrogen atom, are among the alternatives being developed [6].
H-PFCAs are likely to decompose more easily than corresponding
PFCAs because the former have a carbon–hydrogen bond. However,
the decomposition of H-PFCAs has not been reported; no one has
process based on reaction with Ca2+ to form the environmentally
harmless CaF , which is a raw material for hydrofluoric acid, which
2
faces an increasing global demand. Therefore, the development of
decomposition technology for these chemicals at sites where they
are emitted in large quantities is important because such tech-
nology would not only reduce the environmental impact of the
fluorochemicals but also contribute to the recycling of a fluorine
resource.
PFCAs and related chemicals are generally stable: conventional
methods for wastewater treatment, such as the use of Fenton’s
reagent (Fe2+ + H O ) and H O + UV light irradiation, are not appli-
2
2
2
2
cable, because aqueous OH radicals are only slightly reactive
toward PFCAs [7–9].
We previously reported that PFCAs such as PFOA are efficiently
−
◦
decomposed to F ions in hot water at 80 C in the presence of per-
2−
2−
sulfate (S O8 ) [10]: the thermolysis of S O produces sulfate
8
radical anions (SO4 ), which can act as strong oxidants to decom-
pose PFCAs. In this reaction system, the relatively low temperature
2
2
•
−
◦
◦
of 80 C is essential; a higher temperature, such as 150 C, is unsuit-
•
−
able because most of the SO4 is consumed by reaction with hot
water. We also reported that PFCAs are efficiently decomposed to
− •− 2−
F ions by SO4 produced by UV light irradiation of S2O8 [11].
∗ Corresponding author. Tel.: +81 29 861 8161; fax: +81 29 861 8866.
E-mail address: h-hori@aist.go.jp (H. Hori).
•
−
was applied to the decomposition of polychlori-
Recently, SO4
0
920-5861/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.02.023

1
32
H. Hori et al. / Catalysis Today 151 (2010) 131–136
•
−
−
2−
nated biphenyls [12] and chlorophenols [13,14], at which the SO
a suppressor device was used to measure the F and SO4 con-
4
2−
was obtained from reactions of S O8 or peroxymonosulfate with
centrations. The mobile phase was an aqueous solution containing
Na B O (6 mM), H BO3 (15 mM), and NaHCO3 (0.2 mM); and the
2
metal ions.
2
4
7
3
−
1
Herein we report the decomposition of typical H-PFCAs, that is,
flow rate was 0.8 mL min
.
2−
HCnF2nCOOH (n = 4, 6, 8), with S O
in hot water at low tem-
An ion-exclusion chromatograph system consisting of a guard
column (TSKgel OApak-P, 7.8-mm i.d., 1.0-cm length, Tosoh Corp.),
a separation column (TSKgel OApak-A, 7.8-mm i.d., 30-cm length,
2
8
◦
peratures (60 and 80 C). The decomposition efficiency and the
decomposition mechanisms of the H-PFCAs are compared with
those of the corresponding PFCAs.
◦
Tosoh Corp.), a pump, a column oven (40 C), and a conductivity
detector was used to determine whether H-PFCAs or PFCAs with
short (<C ) chains were formed in the aqueous phase. The mobile
phase was phthalic acid (10 mM) at a flow rate of 0.6 mL min , and
a typical sample injection volume was 5 ␮L.
The reaction substrates were quantified by HPLC with conduc-
tometric detection; the mobile phase was a mixture of methanol
4
2
. Experimental
−
1
2.1. Materials
Potassium persulfate (>99.0%) was purchased from
and aqueous NaH PO4 (20 mM, adjusted to pH 3.0 with H PO ) at
2
3
4
Wako Pure Chemical Industries (Osaka, Japan) and used as
received. 5H-Perfluoropentanoic acid (HC F COOH, >97%),
several mixing ratios (45–65 vol.% of methanol), and the separation
column was a Tosoh TSKgel Super ODS column (4.6 mm i.d., 10 cm
length × 2). When the sample injection volume was 30 ␮L and the
mobile phase was a mixture of methanol/aqueous NaH PO (55:45,
4
8
7
H-perfluoroheptanoic
acid
(HC6F12COOH,
>98%),
9H-
perfluorononanoic acid (HC8F16COOH, >98%), perfluoroazelaic
acid (HOOCC7F14COOH, >96%), and dodecafluorosuberic acid
2
4
−
1
v/v), the limits of detection (mg L ), which were calculated from a
signal-to-noise ratio of 3, were 1.54, 0.57, and 0.22 for HC8F16COOH,
HC6F12COOH, and HC F COOH, respectively.
(
HOOCC6F12COOH, >98%) were obtained from SynQuest Labo-
ratories (Alachua, FL, USA). Heptafluorobutyric acid (C F7COOH,
3
4
8
>99%), nonafluoropentanoic acid (C F COOH, >98%), undecafluo-
4 9
An ESI-mass spectrometry system (LCMS-2010 EV, Shimadzu,
Kyoto, Japan) was used to identify the intermediates in the aque-
ous phase. Analyses were carried out in negative ion mode, and
the electrospray probe voltage was 4.50 kV. Reaction samples were
delivered to the electrospray probe using acetonitrile as a mobile
rohexanoic acid (C5F11COOH, >98%), and perfluorononanoic acid
C8F17COOH, >95%) were purchased from Tokyo Kasei Kogyo
Co. (Tokyo, Japan). Tridecafluoroheptanoic acid (C6F13COOH,
96%) and PFOA (>95%) were obtained from Wako Pure Chemical
Industries.
(
>
−
1
phase at a flow rate of 0.2 mL min . LC/MS measurements were
also carried out: a separation column (TSKgel ODS-80TSQA, Tosoh
Corp.) was added to the above ESI-mass spectrometry system, and
the mobile phase was a 50:50 (v/v) mixture of methanol and aque-
2.2. Reaction procedures
A stainless steel pressure-resistant reactor (35.1 mL volume)
ous CH COONH4 (1 mM, adjusted pH 4.0 with acetic acid).
3
equipped with a thermocouple and a stainless steel screw cap
was used. The screw cap was connected to a pressure gauge for
measuring the pressure in the reactor and to a sampling port for
analyzing gas products. A gold vessel (24.6 mL, 2.8 cm i.d.) was
fitted into the reactor to prevent contamination from the reactor
material. In a typical run, an aqueous (Milli-Q) solution (10 mL) of
an H-PFCA or a PFCA (3.71–3.92 ␮mol, 371–392 ␮M) and K S O
The GC/MS system consisted of a gas chromatograph (HP5890,
Hewlett-Packard, Wilmington, DE, USA) with a Poraplot Q col-
umn (0.32-mm i.d., 25-m length; Chrompack, Bergen op Zoom,
The Netherlands), a mass spectrometer (HP 5972A), and a worksta-
tion (HP G1034CJ). The carrier gas was He, and the electron impact
source was operated at 70 eV. The analyses were conducted in full-
2
2
8
scan mode (m/z 1.2–200) to survey the products. A GC system (GC
(
50 ␮mol–0.50 mmol, 5.0–50.0 mM) was introduced into the gold
◦
3
23, GL Sciences, Tokyo, Japan) consisting of an injector (150 C), a
vessel, and the reactor was pressurized to 0.65 MPa with syn-
thetic air and sealed. The reactor was placed in an oven, and the
reactor temperature was raised to the desired reaction temper-
◦
◦
column oven (50 C), and a thermal conductivity detector (130 C),
was also used: the column was an active carbon column (60/80
mesh, 2.17-mm i.d., 2-m length), and the carrier gas was argon.
◦
ature (60 or 80 C) and then held constant for a specified time
(
e.g., 6 h), after which the reactor was quickly cooled to room
temperature using ice water. Control reactions were performed
in the absence of K S O . The gas phase was subjected to gas
3. Results and discussion
2
2
8
chromatography/mass spectrometry (GC/MS) and GC. The aqueous
phase was subjected to ion-chromatography, ion-exclusion chro-
matography, high-performance liquid chromatography (HPLC),
electrospray ionization (ESI) mass spectrometry, and LC/MS.
We also used Fenton’s reagent to attempt to decompose an
H-PFCA: an aqueous solution (22 mL; initial pH, 2.3) containing
HC6F12COOH (123 ␮mol, 5.58 mM), H O2 (1.0 M), and FeSO ·7H O
◦
3
.1. Persulfate-induced decomposition of H-PFCAs at 80 C
The time course of the reaction of aqueous HC F COOH and
8
16
2−
◦
S O
2
(130 molar excess relative to HC8F16COOH) at 80 C is
shown in Fig. 1. HC8F16COOH rapidly disappeared from the HPLC
chromatogram within 0.5 h of the treatment, and F and CO2 were
the main products in the aqueous and gas phases, respectively. After
HC8F16COOH disappeared, the amounts of F and CO2 continued
to increase, which indicates that reaction intermediates decom-
posed to F and CO2 during this period. After 6 h, the F amount
reached 61.6 ␮mol, which corresponds to a F yield [(moles of F
8
−
2
4
2
(
4.92 mM) was mixed in an oxygen atmosphere for 17 h in the dark
−
at room temperature. After the reaction, the gas and the aqueous
phases were analyzed.
−
−
−
−
2
.3. Analysis
formed)/(moles of fluorine in initial HC8F16COOH, i.e., moles of
initial HC8F16COOH × 16)] of 98.2% (Table 1, entry 1). This result
clearly indicates that the fluorine content in HC F COOH was suc-
An ion-chromatography system (IC-2001, Tosoh Corp., Tokyo,
8
16
Japan) consisting of an automatic sample injector (30-␮L injection
volume), a degasser, a pump, a guard column (TSKguard column
Super IC-A, 4.6-mm i.d., 1.0-cm length, Tosoh Corp.), a separation
column (TSKgel Super IC-Anion, 4.6-mm i.d., 15-cm length, Tosoh
cessfully mineralized.
2−
In the absence of S O8 , virtually no reaction occurred: 97.0% of
2
−
the initial HC8F16COOH remained in the aqueous phase, and the F
yield was only 0.02% after 6 h (Table 1, entry 2). These results clearly
◦
•−
Corp.), a column oven (40 C), and a conductivity detector with
indicate that SO4 acted as an oxidant to decompose HC8F16COOH.

H. Hori et al. / Catalysis Today 151 (2010) 131–136
133
◦
Fig. 2. Reaction-time dependence of the amounts of SO42− detected during the
decomposition of HC8F16COOH with S O 2−. Reaction conditions were the same
as in Fig. 1.
Fig. 1. Time course of HC8F16COOH decomposition in hot water (80 C) containing
2−
−
S2O8 : detected molar amounts of HC8F16COOH, F , and CO2. An aqueous solu-
2
8
2
−
tion (10 mL) containing HC8F16COOH (3.92 ␮mol, 392 ␮M) and S2O8 (0.50 mmol,
5
0.0 mM) was introduced in the reactor, the reactor was pressurized with synthetic
◦
air (0.65 MPa), and the reactor temperature was raised to 80 C and held constant
for 0.5–6 h.
While the reactions of HC F2nCOOH (n = 4, 6, and 8) with
n
2−
S O
(0.50 mmol; 50.0 mM) were proceeded, the pH values were
2
8
decreased from 3.3–3.4 to 1.5–1.6 after 6 h.
In accordance with the oxidative decomposition of
•
−
2−
◦
HC8F16COOH, SO4
is expected to be reduced to SO4 . Con-
3.2. Decomposition of H-PFCAs at 60 C
sistent with this expectation, SO42− accumulated in the aqueous
◦
phase, and the amount increased with reaction time (Fig. 2) while
In the above reactions, samples were heated at 80 C in the
−
2−
the decomposition of HC8F16COOH and the formation of F and
presence of a large excess of S O8 (50.0 mM; 130 molar excess
relative to the substrates). Under these conditions, the reactions
were so rapid that the H-PFCAs disappeared from the aqueous
2
2
−
−
CO2 proceeded efficiently (Fig. 1). After 6 h, the amount of SO4
2
in the aqueous phase was 0.79 mmol. The initial amount of S O8
2
was 0.50 mmol; therefore, 79% of the sulfur content in initial
phase within 0.5 h. To investigate the decomposition of H-PFCAs in
2−
2−
2−
S O
was transformed to SO4 during this period.
We carried out the decomposition reaction with the other H-
PFCAs, that is, HC F COOH and HC F COOH, at a constant reaction
the presence of S O8 in detail, we carried out the reactions under
2
8
2
◦
milder conditions (reaction temperature, 60 C; initial amount of
, 50 ␮mol; concentration, 5.0 mM). The time course of the
reaction of aqueous HC8F16COOH at 60 C is shown in Fig. 3. Under
2
−
S O
8
6
12
4
8
2
◦
◦
time of 6 h at 80 C. The results are summarized in Table 1 (entries
–6) together with the data obtained in the absence of S O
, almost all (98.0–99.7%) of the substrates remained
in the aqueous phase, and there was no formation of F and CO2
entries 4 and 6). In contrast, in the presence of S O
strates remained in the aqueous phase, and the F yields reached
2
−
3
.
these conditions, the decomposition of HC F16COOH and simulta-
2
8
8
2−
−
Without S O
neous formation of F and CO2 were observed. In the absence of
2
8
−
2−
2−
S O
, HC8F16COOH did not decompose, whereas when S O8
2
8
2
2−
(
, no sub-
was present, the amount of HC8F16COOH decreased linearly with
time, and the decrease in the substrate amount was accompanied
2
8
−
−
9
7.8% and 96.7% for HC6F12COOH and HC F COOH, respectively
by the formation of F and CO2 (Fig. 3). We used the initial rate of
4
8
2−
(
entries 3 and 5). Therefore, S O8 led to efficient mineralization
decomposition of HC8F16COOH, the slope taken from this period, as
a measure of the reactivity, and the initial rates for various H-PFCAs,
2
of these H-PFCAs.
Table 1
◦
2− a
Decomposition of H-PFCAs and PFCAs in hot water (80 C) in the presence or absence of S2O8
.
Initial S2O82 (mmol)
−
Remaining substrate (␮mol) [%]b
−
(␮mol) [yield (%)]c
F
CO2 (␮mol) [yield (%)]d
Entry
Substrate [initial amount (␮mol)]
f
f
1
2
3
4
5
6
7
8
9
HC8F16COOH [3.92]
HC8F16COOH [3.70]
HC6F12COOH [3.86]
HC6F12COOH [3.86]
HC4F8COOH [3.75]
HC4F8COOH [3.67]
C8F17COOH [3.85]
C8F17COOH [3.77]
C6F13COOH [3.74]
C6F13COOH [3.74]
C4F9COOH [3.74]
C4F9COOH [3.74]
0.50
0
0.50
0
0.50
0
0.50
0
0.50
0
0.50
0
n.d.e [0]
3.59 [97.0]
n.d. [0]
3.78 [98.0]
n.d. [0]
3.66 [99.7]
n.d. [0]
3.73 [98.9]
n.d. [0]
3.68 [98.4]
n.d. [0]
61.6 ± 0.2 [98.2]
0.01 [0.02]
45.3 [97.8]
n.d. [0]
22.3 ± 0.3 [63.2]
0.38 [1.1]
18.0 [66.6]
n.d. [0]
14.1 [75.2]
n.d. [0]
19.0 [54.8]
0.68 [2.00]
20.4 [77.9]
n.d. [0]
15.4 [82.4]
n.d. [0]
29.0 [96.7]
n.d. [0]
48.3 [73.8]
0.03 [0.05]
36.8 [75.7]
n.d. [0]
1
1
1
0
1
2
30.2 [89.7]
n.d. [0]
3.73 [99.7]
a
b
c
d
e
f
2−
◦
An aqueous solution (10 mL) of substrate with or without S2O8 was heated at 80 C under air for 6 h.
Remaining substrate (%) = [(moles of remaining substrate)/(moles of initial substrate)] × 100.
−
−
F
yield (%) = [(moles of F formed)/(moles of fluorine in initial substrate)] × 100.
CO2 yield (%) = [(moles of CO2 formed)/(moles of carbon in initial substrate)] × 100.
n.d = not detected.
Errors shown here were ascribed to replicate experiments.

1
34
H. Hori et al. / Catalysis Today 151 (2010) 131–136
◦
Fig. 4. Time course of C8F17COOH decomposition in hot water (80 C) containing
2
−
−
◦
S2O
: detected molar amounts of C8F17COOH, F , and CO . The reaction condi-
Fig. 3. Time course of HC8F16COOH decomposition in hot water (60 C) containing
8
2
2−
−
tions were the same as those described in Fig. 1, except that C8F17COOH (3.85 ␮mol,
385 ␮M) was used as the substrate.
S2O8 : detected molar amounts of HC8F16COOH, F , and CO2. An aqueous solu-
tion (10 mL) containing HC8F16COOH (3.73 ␮mol, 373 ␮M) and S2O8
2
−
(50 ␮mol,
5
.0 mM) was introduced in the reactor, the reactor was pressurized with synthetic
◦
air (0.65 MPa), and the reactor temperature was raised to 60 C and held constant
for 1–2.5 h.
−
The F yield from each of the H-PFCAs was higher than that
from the corresponding PFCAs: for example, 98.2% for HC8F16COOH
(
entry 1) compared to 73.8% for C8F17COOH (entry 7), and 97.8% for
along with other decomposition data, are summarized in Table 2,
together with other H-PFCAs. The initial decomposition rates for
HC6F12COOH (entry 3) compared to 75.7% for C6F13COOH (entry
9
faster for H-PFCAs than for the corresponding PFCAs. The difference
between the S O
obvious at 60 C and with a smaller amount of S O
time dependence of the decomposition of C F17COOH in hot water
at 60 C in the presence of S O
2
−
). This result indicates that mineralization by S O8 proceeded
2
HC8F16COOH, HC6F COOH, and HC F COOH were nearly identical
12
4 8
−1
(
1.08–1.28 ␮mol h ; entries 1–3), which indicates that the reactiv-
2−
reactivities of H-PFCAs and PFCAs was more
2
8
ities of these H-PFCAs were almost the same; that is, the reactivity
was little influenced by the carbon-chain length.
◦
2−
. The reaction-
2
8
8
◦
2−
(5.0 mM) is shown in Fig. 5,
2
8
3
.3. Comparison of H-PFCAs with PFCAs
together with that of HC8F16COOH. Whereas HC8F16COOH decom-
−
1
posed efficiently at an initial decomposition rate of 1.28 ␮mol h
(Table 2, entry 1), C8F17COOH decomposed much more slowly
We compared the S O82−-induced decomposition of H-PFCAs
2
−
1
with that of the corresponding PFCAs. Fig. 4 shows the time
course of C F COOH decomposition in hot water at 80 C with
S O
2
(Fig. 5a), at an initial decomposition rate of 0.18 ␮mol h (Table 2,
◦
entry 4), which was 14% of that of the rate for HC F16COOH.
8
17
8
2
−
−
(50.0 mM) under the same experimental conditions used
Although small amounts of F and CO2 were detected during the
8
for HC F16COOH (Fig. 1). C8F17COOH was efficiently decomposed to
reaction of C8F17COOH (Fig. 5b and c), the amounts were one order
of magnitude smaller than the amounts for HC8F16COOH over the
duration of the reaction: for example, the yields of F and CO2
at 3 h were 4.08% and 3.98% (Table 2, entry 4), whereas those for
HC8F16COOH at the shorter reaction time of 2.5 h were 31.5% and
15.5%, respectively (Table 2, entry 1). The same tendency was also
8
−
F
and CO , as previously reported for PFOA decomposition [10]. In
2
−
addition to C8F17COOH, C6F13COOH and C F COOH also completely
4
9
−
disappeared after 6 h, and F and CO2 were formed in yields of
7
1
3.8–89.7% and 54.8–82.4%, respectively (Table 1, entries 7, 9, and
2−
1). In contrast, in the absence of S O8 , almost all (98.4–99.7%)
2
of the initial amounts of the PFCAs remained in the aqueous phase,
and F and CO2 did not form meaningfully (entries 8, 10, and 12).
observed for C6F13COOH and C F COOH: the initial decomposi-
4
9
−
−1
tion rates for C6F13COOH and C F COOH were both 0.09 ␮mol h
4
9
Table 2
◦
2− a
.
Decomposition of H-PFCAs and PFCAs in hot water (60 C) in the presence of S2O8
−
Entry
Substrate (initial
amount, ␮mol)
Remaining
F
(␮mol)
CO2 (␮mol)
[yield (%)]
Short-chain PFCAs
(amount, ␮mol)
Initial decomposition
rate of substrate
(␮mol h
c
d
substrate (␮mol)
[yield (%)]
b
−1 e
)
[%]
f
5.20 [15.5]f
1
2
3
4
5
6
HC8F16COOH (3.73)
HC6F12COOH (3.80)
HC4F8COOH (3.75)
C8F17COOH (3.88)
C6F13COOH (3.71)
C4F9COOH (3.82)
0.56 [15.0]f
18.8 [31.5]
14.6 [32.0]
14.5 [48.3]
–
–
–
1.28
1.08
1.14
0.18
0.09
0.09
f
f
f
f
1.08 [28.4]
4.89 [18.4]
0.90 [24.0]f
4.99 [26.6]
f
g
g
g
g
g
g
3.38 [87.1]
3.43 [92.5]
3.46 [90.6]
2.69 [4.08]
0.66 [1.37]
0.72 [2.09]
1.39 [3.98]
PFOA (0.10), C6F13COOH (0.02)
C5F11COOH (0.08)
C3F7COOH (0.22)
g
g
g
g
0.58 [2.23]
g
g
0.94 [4.92]
a
2−
◦
An aqueous solution (10 mL) of substrate with S2O8 (50 ␮mol, 5.0 mM) was heated at 60 C under air.
Remaining substrate (%) = [(moles of remaining substrate)/(moles of initial substrate)] × 100.
b
c
d
e
f
−
−
F
yield (%) = [(moles of F formed)/(moles of fluorine in initial substrate)] × 100.
CO2 yield (%) = [(moles of CO2 formed)/(moles of carbon in initial substrate)] × 100.
The initial decomposition rate was taken from the initial period when the substrate amount decreased linearly with respect to time.
Reaction time, 2.5 h.
Reaction time, 3 h.
g

H. Hori et al. / Catalysis Today 151 (2010) 131–136
135
reaction mechanisms for the two series of compounds differ con-
siderably.
PFCAs are known to undergo little decomposition by con-
ventional techniques such as treatment with Fenton’s reagent
2+
(
Fe + H O ) [8], irradiation with UV–vis light in the presence of
2 2
H O [7], and conventional TiO2 photocatalysts [15,16], because
2
2
the reactivity of OH radicals with PFCAs in water is low. In contrast,
H-PFCAs have a carbon–hydrogen bond, which can be expected to
•
−
react with OH radicals. Also, SO4 reacts with water to form OH
−
1
radicals, with a rate constant k[H O] of 460 s [17] (Eq. (1)).
2
SO4 + H O → HSO4 + OH•
•
−
−
(1)
2
•
−
To determine whether OH radicals generated from SO4 and
water might initiate the decomposition of H-PFCAs, we exam-
ined the reactivity of an H-PFCA (HC6F13COOH) with OH radicals
produced from Fenton’s reagent and found that treatment of
HC6F12COOH (123 ␮mol in 22 mL; 5.58 mM) with Fe2+ (4.90 mM)
and H O2 (1 M) for 17 h resulted in no decrease in the amount of
2
HC6F12COOH. Therefore, OH radicals are unlikely to play a signif-
icant role at least in the initial decomposition step of H-PFCAs in
2−
the presence of S O8 , although OH radicals may play some role
2
in the transformation of reaction intermediates.
In the case of PFCAs, the mechanism of decomposition induced
•
−
by SO4 has been outlined as follows [10,11]: the oxidization of a
•
−
PFCA (CnF2n+1COOH) by SO4 causes decarboxylation that gener-
•
ates a perfluoroalkyl radical (CnF
reacts with SO4 to form CnF2n+1OSO3 . This unstable species eas-
ily undergoes hydrolysis to form CnF2n+1OH. The perfluoroalcohol
). The perfluoroalkyl radical
2n+1
•
−
−
undergoes HF elimination to form C
acid fluoride generates a PFCA shortened by one CF unit, that
F
COF. Hydrolysis of the
n−1 2n−1
2
is, C
F
COOH. The shortened PFCA is further decomposed by
n−1 2n−1
•
−
SO4 . Likewise, the decomposition of the shortened PFCAs pro-
ceeds via the formation of even shorter PFCAs.
We also expected the decomposition of H-PFCA to be initiated
by oxidative decarboxylation. However, as described above, the for-
mation of shorter-chain H-PFCAs was not observed; for example,
no HC7F14COOH was produced by decomposition of HC F16COOH
8
Fig. 5. Reaction-time dependence of the decomposition of C8F17COOH in hot water
2−
in the presence of S O8 . To determine the reaction intermedi-
2
◦
2−
(
60 C) containing S2O8 , along with data for HC8F16COOH under the same reaction
ates for H-PFCAs, we employed ESI-mass spectrometry and LC/MS.
−
conditions for comparison: (a) C8F17COOH, (b) F , and (c) CO2. The reaction condi-
◦
When the reaction of HC F COOH was carried out at 60 C in the
tions were the same as those described in Fig. 3, except that C8F17COOH (3.88 ␮mol,
8 16
presence of S2O82 (5.0 mM), the ESI-mass spectra of the reaction
−
3
88 ␮M) was used as the substrate.
aqueous phase after 1 h showed peaks at m/z 445 and 439. The peak
at m/z 445 corresponds to the anion of the initial substrate, that is,
(
Table 2, entries 5 and 6), whereas those for HC F12COOH and
−
6
[
HC8F16COO] , and the peak at m/z 439 corresponds to the anion
−1
HC F COOH were 1.08 and 1.14 ␮mol h , respectively (Table 2,
−
4
8
of perfluorodicarboxylic acid [HOOCC7F14COO] . After 2 h, another
entries 2 and 3). Thus, the initial decomposition rates for all H-
PFCAs tested here were 7.1–12.7 times as high as those for the
corresponding PFCAs.
In addition to the difference in the decomposition efficiency,
we also observed another difference in the behavior of H-PFCAs
and that of PFCAs. In the case of PFCAs, small amounts of short-
chain PFCAs, that is, PFCAs with chain lengths shorter than the
initial chain lengths, were detected in the reaction aqueous phase
peak appeared, at m/z 389. This m/z value corresponds to the anion
−
of a shorter-chain perfluorodicarboxylic acid [HOOCC6F12COO] .
These results indicate that perfluorodicarboxylic acids with var-
ious chain lengths were the reaction intermediates. We carried
out LC/MS measurements and quantified HOOCC7F14COOH in the
reaction aqueous phase. The total-ion mass chromatogram for
2−
the reaction aqueous phase (2.5 h) of HC8F16COOH with S O8
2
is shown in Fig. 6a. The peaks at 2.95 and 2.37 min correspond
to HOOCC7F14COOH and HOOCC6F12COOH, respectively, which
were identified by use of standard samples (the retention time
of HC8F16COOH is 39.8 min so that the peak is not shown here).
The peak X in Fig. 6a consist of several further shortened per-
flurodicarboxylic acids HOOCCnF2nCOOH (n = 2–5), because the
mass spectrum showed peaks which correspond to their anions
(
Table 2, entries 4–6), whereas in the case of H-PFCAs, the corre-
sponding species, that is, H-PFCAs with chain lengths shorter than
the initial chain lengths, were not observed in the ESI-mass spec-
tra or the HPLC chromatograms. This difference suggests that the
decomposition mechanism of H-PFCAs differs from that of PFCAs.
−
3
.4. Reaction mechanism
[HOOCCnF2nCOO] (n = 2–5) (Fig. 6b).
The time course of the amount of HOOCC7F14COOH is shown
in Fig. 7 (the reaction conditions were the same as those for
the reaction shown in Fig. 3). The plot clearly indicates that
HOOCC7F14COOH was formed while HC8F16COOH was decom-
posed (Fig. 3). Therefore, the initial decomposition step should be
As described above, H-PFCAs showed high reactivity at the low
◦
temperature of 60 C, whereas the reactivity of PFCAs was low
at that temperature. In addition, decomposition of PFCAs formed
shorter-chain PFCAs, whereas that of H-PFCAs did not form shorter-
chain H-PFCAs. These observations led us to suppose that the
•
−
nucleophilic substitution of SO4 at the carbon atom attached to

1
36
H. Hori et al. / Catalysis Today 151 (2010) 131–136
−
2−
detected by LC/MS. The formed HSO4 is in equilibrium with SO4
in water and was detected by ion-chromatography in the form of
2−
•−
SO4 . HOOCC7F14COOH reacts with SO4 , in a manner similar to
that described for the decomposition of PFCAs, via formation of fur-
ther shortened pefluorodicarboxylic acids, which were detected by
ESI-mass spectrometry and LC/MS.
4. Conclusion
We investigated the decomposition of H-PFCAs, that is,
2
−
HCnF2nCOOH (n = 4, 6, and 8), in the presence of S O
in hot water.
2
8
2−
The addition of S O
8
The decomposition of H-PFCAs induced by S O
to the reaction system of these H-PFCAs at
2
8
◦
−
−
0 C caused almost complete mineralization (F yields: ≈100%).
2
proceeded even
2
8
◦
at 60 C, at which temperature the initial decomposition rates were
7
.1–12.7 times those for the corresponding PFCAs. The mechanism
of H-PFCA decomposition differs from that of PFCAs: the former
•
−
is initiated by the nucleophilic substitution of SO4 at the carbon
atom attached to the ␻-H atom of the H-PFCA molecule, followed by
•
−
formation of perfluorodicarboxylic acids, which react with SO4
to give further shortened perfluorodicarboxylic acids, and the final
result was complete mineralization. Decomposition of H-PFCAs by
photochemical approaches is being investigated in our laboratory.
Fig. 6. LC/MS data for the reaction aqueous phase of HC8F16COOH with S2O82−: (a)
total-ion mass chromatogram and (b) mass spectrum for the peak X in a. The reaction
conditions were the same as those described in Fig. 3, where the reaction time was
Acknowledgements
2
.5 h.
Financial support by the Ministry of Environment is gratefully
acknowledged.
References
[
[
1] J.P. Giesy, K. Kannan, Environ. Sci. Technol. 36 (2002) 146A–152A.
2] M.M. Schultz, D.F. Barofsky, J.A. Field, Environ. Eng. Sci. 20 (2003) 487–
501.
[
[
3] M. Houde, J.W. Martin, R.J. Letcher, K.R. Solomon, D.C.G. Muir, Environ. Sci.
Technol. 40 (2006) 3463–3473.
4] US EPA, Provisional Health Advisories for PFOA and PFOS. US Environ-
mental Protection Agency, Washington, DC, USA, http://www.epa.gov/oppt/
pfoa/pubs/activities.html#advisories (last updated on 22.10.09).
5] US EPA, 2010/15 PFOA stewardship program. US Environmental Protec-
tion Agency, Washington, DC, USA, http://www.epa.gov/oppt/pfoa/pubs/
stewardship/index.html (last updated on 21.03.09).
[
[
6] N. Tsuda, C. Sawauchi, Y. Sawada, T. Shimizu, Process for producing aqueous flu-
oropolymer dispersion and aqueous fluoropolymer resin. United States Patent
Application, 2009, 2009023994.
[
[
7] H. Hori, E. Hayakawa, H. Einaga, S. Kutsuna, K. Koike, T. Ibusuki, H. Kitagawa, R.
Arakawa, Environ. Sci. Technol. 38 (2004) 6118–6124.
8] H. Moriwaki, Y. Takagi, M. Tanaka, K. Tsuruho, K. Okitsu, Y. Maeda, Environ. Sci.
Technol. 39 (2005) 3388–3392.
Fig. 7. Time course of the amount of HOOCC7F14COOH formed in the decomposition
of HC8F16COOH in hot water (60 C) containing S2O8 . The reaction conditions were
the same as those described in Fig. 3.
[9] H. Hori, A. Yamamoto, K. Koike, S. Kutsuna, M. Murayama, A. Yoshimoto, R.
Arakawa, Appl. Catal. B: Environ. 82 (2008) 58–66.
10] H. Hori, Y. Nagaoka, M. Murayama, S. Kutsuna, Environ. Sci. Technol. 42 (2008)
◦
2−
[
[
[
7438–7443.
11] H. Hori, A. Yamamoto, E. Hayakawa, S. Taniyasu, N. Yamashita, S. Kutsuna, H.
Kitagawa, R. Arakawa, Environ. Sci. Technol. 39 (2005) 2383–2388.
12] A. Rastogi, S.R. Al-Abed, D.D. Dionysiou, Appl. Catal. B: Environ. 85 (2009)
−
the ␻-H atom in the H-PFCA molecule, which releases F into the
aqueous phase (Eq. (2)).
1
71–179.
[13] G.P. Anipsitakis, D.D. Dionysiou, M.A. Gonzalez, Environ. Sci. Technol. 40 (2006)
000–1007.
HC8F16COOH + SO4• → O SOCHFC7F COOH + F−
−
(2)
3
14
1
[
[
14] A. Rastogi, S.R. Al-Abed, D.D. Dionysiou, Water. Res. 43 (2009) 684–694.
15] P. Thèron, P. Pichat, C. Guillard, C. Petrier, T. Chopin, Phys. Chem. Chem. Phys.
1 (1999) 4663–4668.
Next, the resulting species undergoes hydrolysis to form an
−
alcohol, HOCHFC7F14COOH, and HSO4 . The alcohol undergoes HF
elimination to form an aldehyde, OHCC7F14COOH, which is oxi-
dized to a perfluorodicarboxylic acid, HOOCC7F14COOH, which was
[16] R. Dillert, D. Bahnemann, H. Hidaka, Chemosphere 67 (2007) 785–792.
17] X.-Y. Yu, Z.-C. Bao, J.R. Barker, J. Phys. Chem. A 108 (2004) 295–308.
[