Degradation of phenyltrifluoromethylketone in water by separate or simultaneous use of TiO2 photocatalysis and 30 or 515 kHz ultrasound

Article abstract of DOI:10.1039/a902506e

TiO₂ photocatalysis and ultrasound are emerging technologies for the mineralization of pollutants in water. To further investigate these technologies and to assess whether advantages and synergy can be expected from their differences, phenyltrifluoromethylketone (PTMK) was selected as a test compound for pollutants generating CF₃COOH, an undesirable final product. The PTMK first-order removal rate constant k was ca. 14 times higher when the ultrasound frequency was 515 kHz instead of 30 kHz for the same energy, and ca. 2.5 times higher when a TiO₂ sample we synthesized was used instead of TiO₂ Degussa P25. On simultaneous photocatalytic and ultrasonic treatment an increase in k by a factor between 1.4 and 1.9, depending on the TiO₂ sample, was observed at 30 kHz but not at 515 kHz. On the basis of catalase enzymatic effect upon k, these observations are tentatively explained by a photocatalytic OH radical production from sonochemically formed H₂O₂, provided that the H₂O₂ residence time on TiO₂ is sufficient. PTMK ultrasonic pyrolysis was demonstrated by product analysis. The amount of CF₃COOH was ca. 8 times lower in sonicated solutions than in UV-irradiated TiO₂ suspensions, for both frequencies and both TiO₂ samples. Therefore, because of a higher k value, a high frequency ultrasonic (pre)treatment is preferable to minimize CF₃COOH formation.

Full text of DOI:10.1039/a902506e

View Article Online / Journal Homepage / Table of Contents for this issue
Degradation of phenyltriÑuoromethylketone in water by separate or
simultaneous use of TiO photocatalysis and 30 or 515 kHz
2
ultrasound
Philippe Theron,a Pierre Pichat,*a Chantal Guillard,a Christian Petrierb and Thierry Chopinc
a L aboratoire ““Photocatalyse, Catalyse et EnvironnementÏÏ, CNRS UMR ““IFoSÏÏ, Ecole
Centrale de L yon, BP 163, 69131 Ecully Cedex, France. E-mail: Pierre.Pichat=ec-lyon.fr
b L aboratoire ““Chimie Moleculaire et EnvironnementÏÏ, ESIGEC-Universite de Savoie, 73376
L e-Bourget-du-L ac Cedex, France
c Rhodia, Centre de Recherches, 93308 Aubervilliers Cedex, France
Received 29th March 1999, Accepted 19th August 1999
TiO photocatalysis and ultrasound are emerging technologies for the mineralization of pollutants in water.
2
To further investigate these technologies and to assess whether advantages and synergy can be expected from
their di†erences, phenyltriÑuoromethylketone (PTMK) was selected as a test compound for pollutants
generating CF COOH, an undesirable Ðnal product. The PTMK Ðrst-order removal rate constant k was ca.
3
14 times higher when the ultrasound frequency was 515 kHz instead of 30 kHz for the same energy, and ca. 2.5
times higher when a TiO sample we synthesized was used instead of TiO Degussa P25. On simultaneous
2
2
photocatalytic and ultrasonic treatment an increase in k by a factor between 1.4 and 1.9, depending on the
TiO sample, was observed at 30 kHz but not at 515 kHz. On the basis of catalase enzymatic e†ect upon k,
2
these observations are tentatively explained by a photocatalytic OH~ radical production from sonochemically
formed H O , provided that the H O residence time on TiO is sufficient. PTMK ultrasonic pyrolysis was
2
2
2
2
2
demonstrated by product analysis. The amount of CF COOH was ca. 8 times lower in sonicated solutions
3
than in UV-irradiated TiO suspensions, for both frequencies and both TiO samples. Therefore, because of a
higher k value, a high frequency ultrasonic (pre)treatment is preferable to minimize CF COOH formation.
2
2
3
gate the speciÐcities of these technologies and to assess
whether advantages and synergy can be expected from them.
For that, phenyltriÑuoromethylketone (PTMK, Table 1)
was chosen as the test pollutant because it contains the very
1
Introduction
Heterogeneous photocatalysis over TiO 1 and ultrasound2h7
have independently attracted attention as advanced oxidation
processes for treating water. Photoexcitation of semicon-
ducting TiO promotes valence band electrons to the conduc-
tion band, thus leaving an electron deÐciency or hole in the
valence band. Dioxygen provides a sink for electrons, forming
superoxide O ~~ and its protonated form hydroperoxide
HO ~. Holes can react with water molecules (or hydroxide
2
stable CF group and thus can lead, when degraded photo-
3
catalytically, to triÑuoroacetic acid which is not degradable by
2
photocatalysis13 and resists OH~ radicals in aqueous solution.
By contrast, ultrasonic pyrolysis of PTMK is conceivable as a
result of the relative hydrophobicity of this compound (Table
1). As many pesticides which can potentially pollute water
contain a triÑuoro group,14 the data concerning PTMK
present a broad interest. To our knowledge, PTMK degrada-
tion by photocatalysis or ultrasound has not been reported.
We determined the PTMK removal rates by sonolysis at 30
2
2
anions) to form OH~ radicals, or can also be Ðlled by an
adsorbed organic donor. On the other hand, the chemical
e†ects of ultrasound are due to the phenomenon of cavitation
which is the nucleation, growth and collapse of bubbles in a
liquid.8h11 The collapse of the bubbles induces high-energy
phenomena: high temperature and pressure, electrical dis-
charges and plasma e†ects. The consequences of these extreme
conditions are the cleavage of dioxygen molecules and water
molecules (into H atoms and OH~ radicals). From the reac-
tions of these entities (O, H, OH~) with each other and with
H O and O , HO ~ radicals and H O are formed. Therefore,
photocatalytic and ultrasonic irradiation can cause the degra-
dation of organic pollutants in water by the same species,
notably OH~ radicals, as was illustrated in a paper concerning
monochlorophenols.12 But distinct phenomena can also be
involved, e.g. direct electron transfer from the organic com-
and 515 kHz, by photocatalysis over TiO Degussa P25 and
2
Table 1 PTMK formula and physical properties15
Compound
Formula
PTMK
2
2
2
2 2
Log K
Water
a
2.15
857/4.92
ow
solubilityb
pounds to TiO in photocatalysis, and pyrolysis of the organic
compounds in ultrasonic water treatment. Furthermore, one
(mg L~1/mmol L~1)
2
Vapor pressureb/Torrc
1.22
process is basically heterogeneous since it involves a solid and
a liquid, whereas in the other process local heterogeneities are
created by the cavitation. Our purpose was to further investi-
a Octan-1-ol/water partition coefficient. b At 298 K. c 1 Torr \ 133.3
Pa.
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668
The journal is ( The Owner Societies 1999
4663

View Article Online
over a TiO sample synthesized by Rhodia (previously Rhone-
Poulenc), and by the combination of both techniques for these
2
frequencies and TiO samples. The amount of triÑuoroacetic
2
acid, a Ðnal product formed from PTMK was measured under
these various degradation conditions. Qualitative analysis of
intermediate products was carried out for PTMK degradation
by separate or simultaneous use of UV-irradiated TiO
Degussa P25 and 515 kHz-ultrasound to discriminate between
2
ultrasonic pyrolysis and other processes.
2
Experimental
2.1 Materials
PhenyltriÑuoromethylketone (99%) was obtained from
Aldrich, triÑuoroacetic acid (99.5%) from Fluka and catalase
(hydrogen peroxide: hydrogen-peroxide oxidoreductase, EC
1.11.1.6) from Boehringer Mannheim. Acetonitrile and meth-
anol (HPLC grade; Rathburn), benzoic acid (99.5%; Fluka),
lithium hydroxide (98%; Aldrich) and water (ultra-pure from
Milli-Q Plus Millipore) were used to prepare the aqueous
solutions and the eluents. All these products were employed as
supplied.
We used TiO Degussa P25 (prepared from titanium tetra-
2
chloride in a Ñame reactor, mainly anatase, nonporous, parti-
cle size: ca. 30 nm; surface area: 53 m2 g~1). Rhodia
synthesized a TiO sample from titanium sulfate. This sample
2
was calcined at 1073 K overnight. Its characteristics were:
97.5% anatase, particle size: 80 ^ 10 nm, surface area: 8 m2
g~1.
2.2 Apparatus
Two types of cylindrical jacketed glass reactors opened to air
were utilized, depending on the ultrasonic frequency (Fig. 1).
Reactor 1 (ca. 90 mL) was used for 30 kHz-ultrasonication or
photocatalytic experiments, or both simultaneously. The 30
kHz-irradiation was supplied by a Sonimasse S30 ultrasonic
processor equipped with a titanium probe (diameter 9 mm).
The bottom optical window, 3.6 cm in diameter, was screwed
onto the reactor. Reactor 2 (ca. 300 mL) was used for 515
kHz-ultrasonication or photocatalytic experiments, or both
simultaneously. The 515 kHz-ultrasonic wave was emitted
Fig. 1 Experimental devices used in this work: reactor 1 (top) and
reactor 2 (bottom).
ultrasonic waves stir the solution sufficiently and more impor-
tantly because the magnetic stirring perturbs the formation of
the cavitation bubbles. In some experiments, catalase (50
lL \ 50 mg L~1 \ 71 500 U) was added at the beginning of
the stirring period.
from
a ceramic titanateÈlead zirconate disk transducer
(Quartz et Silice P 7/62), 2.5 cm in diameter, Ðxed on a stain-
less steel plate at the bottom of the reactor. The system was
driven by a home-made high-frequency power supply. The
optical window (4.6 cm in diameter) was at the top of the
reactor. The ultrasonic power dissipated into the reactors,
estimated by a calorimetric method,16 was 16.0 ^ 0.3 W at
both frequencies.
2.4 Analyses
Before high-performance liquid chromatography (HPLC)
analysis, the aqueous samples were Ðltered through 0.45 lm
Millipore discs to remove TiO agglomerates. The HPLC
2
apparatus used for PTMK analysis comprised an isocratic
In all cases, UV irradiation was provided by a Philips HPK
125-W high-pressure mercury lamp through a 2.2 cm thick
circulating water cell (Pyrex-Ðlter: j [ 290 nm). The radiant
Ñux entering the reactor measured with a UDT 21A power-
meter radiometer was ca. 55 mW cm~2. The corresponding
pump (Thermo Separation Products (TSP) P100), a UV-
detector (TSP UV100) set at 260 nm and a TSP Data Jet Inte-
grator. A FluoÐx (Interchim) reverse-phase column (25 cm
long, 4.6 mm id) especially designed to analyze Ñuorinated
organic compounds was chosen. The mobile phase (1 mL
min~1) was a mixture of acetonitrile (30 v/v%) and water.
number of photons potentially absorbable by TiO was about
2
5 ] 1017 s~1.
CF COOH was analyzed by use of a Waters 501 isocratic
3
pump, a Waters 431 conductivity detector and an IC-PAK
2.3 Procedures
anion column (5 cm long, 4.6 mm id) with lithium benzoate
eluent (1 mmol L~1, pH 6, 1 mL min~1).
The volume of the aqueous solution of PTMK (160 lmol
L~1) introduced into reactor 1 (30 kHz-ultrasound) was 50
The intermediate products of PTMK degradations were
identiÐed by GC-MS (HP 5890 series II; HP 5971A). The
mass analyzer was operated in the electron impact (EI) mode,
and the injections were made in the splitless mode. For these
analyses, the initial PTMK concentration was unchanged (160
lmol L~1). For relatively heavy intermediate products (m/z
150È350), a Chrompack CP-SIL 5-CB column (25 m long,
0.25 mm id, Ðlm thickness 1.2 lm) was used. The irradiations
were stopped after 20% PTMK degradation. Subsequently,
mL, to which 175 mg of powder TiO was added, i.e. 3.5 g
2
L~1. In reactor 2 (515 kHz ultrasound), the quantities used
were multiplied by 4 (200 mL solution, 700 mg TiO ). The
2
degradations were carried out at 293 K and at natural pH, i.e.
6È6.5. The aerated suspension was Ðrst stirred in the dark for
15 min. This was sufficient to reach equilibrated adsorption as
was deduced from the steady-state concentrations. The mag-
netic stirring was stopped when using ultrasound because the
4664
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668

View Article Online
the solution was totally evaporated using a rotatory evapo-
rator and ca. 0.5 mL CH Cl was added. The column tem-
2
2
perature program was: 313È353 K (20 K min~1); 353È533 K
(5 K min~1, hold time: 10 min). Lighter intermediate products
were trapped by solid phase microextraction (SPME). After 20
to 25% PTMK degradation in the closed reactor, the irradia-
tions were stopped. A 75 lm Carboxen-PDMS Ðber from
Supelco was then placed for 30 min in the head space of the
stirred solution heated at 323 K. The desorption of the pro-
ducts retained on the SPME Ðber was carried out in the
GC-MS oven at 473 K. A CP-poraPLOT Q column (27.5 m
long, 0.25 mm id, Ðlm thickness 8 lm) from Chrompack was
used with the following temperature program: 310 K (hold
time: 5 min), 310È493 K (10 K min~1, hold time: 20 min).
3
Results
3.1 PTMK removal rates
Control experiments showed that the variations in PTMK
concentration c due to direct photolysis (j [ 290 nm), hydro-
lysis and evaporation of PTMK were negligible over 180 min.
The decrease in c, for c \ 160 lmol L~1, due to adsorption
Fig. 2 (a) Plots of phenyltriÑuoromethylketone concentration vs.
time for PTMK degradation in reactor 1 ([PTMK \ 160 lmol L~1,
50 mL, [TiO ] \ 3.5 g L~1) with TiO Degussa P25 or TiO Rhodia.
0
on TiO was ca. 6% for both samples at the equilibrium,
2
reached within less than 10 min, in the dark. Considering that
2
2
2
(b) Linear transforms. Degradation methods: 30 kHz-ultrasound (=);
photocatalysis with TiO Degussa P25 (+); photocatalysis with TiO
the surface occupied by one PTMK molecule is ca. 0.55 nm2,
based on the benzene molecule which is reported to occupy
2
2
Rhodia ()); 30 kHz-ultrasound and photocatalysis simultaneously
0.4 nm2,17 it was estimated that the coverage of the TiO
with TiO Degussa P25 (>); 30 kHz-ultrasound and photocatalysis
2
2
surface by PTMK was only ca. 2% for TiO Degusssa P25
simultaneously with TiO Rhodia (|).
2
and ca. 11% for TiO Rhodia. This low coverage is under-
2
2
standable since the PTMK concentration was far below the
energy alter k values. Therefore our objective was not to
compare the removal rates of the individual degradation
methods but to investigate the e†ects of various parameters.
solubility limit (Table 1).
For all degradation conditions, an apparent Ðrst-order
decay in pollutant concentration was observed (see, for
example, Fig. 2), which enables comparison of the correspond-
ing rate constants k (Tables 2 and 3). Since k values depend on
the experimental conditions, they have no absolute meaning.
The photocatalytic removal rate was faster with TiO Rhodia
2
than with TiO Degussa P25 (Fig. 2; Tables 2 and 3). The
2
sonolytic removal rate increased with increasing frequency
(Tables 2 and 3). It was not a†ected by TiO particles in the
2
The radiant Ñux and the TiO sample a†ect the values for
dark at 30 kHz (Table 2), but was decreased by ca. 20% at 515
2
photocatalysis, inter alia. For sonolysis, the frequency and the
kHz (Table 3). A synergy between photocatalysis and son-
Table 2 Apparent Ðrst-order rate constants k (103 ] min~1) for PTMK degradation in reactor 1 ([PTMK] \ 160 lmol L~1, 50 mL, [TiO ] \
2
3.5 g L~1) without and with catalase (22 mg L~1) by the methods indicated
Ultrasound
in the dark
(30 kHz)
Photocatalysis
alone
Photocatalysis and 30 kHz-
ultrasound simultaneously
Method
With TiO Degussa P25
39 ^ 3(1)
4.0 ^ 0.2(2)
4.0 ^ 0.2(4)
83 ^ 5
2
k
] k \ 43 ^ 3
(1)
(2)
With TiO Rhodia
96 ^ 6(3)
152 ^ 8
2
k
] k \ 100 ^ 6
(3)
(4)
Without TiO
È
4.0 ^ 0.2
È
2
With TiO
35 ^ 3(5)
3.7 ^ 0.2(6)
39 ^ 3
2
Degussa P25 and
k
k
] k \ 38.7 ^ 3
(5)
(6)
catalase
With TiO
2
39 ^ 3(7)
3.4 ^ 0.2(8)
43.3 ^ 3
Rhodia and
] k \ 42.4 ^ 3
(7)
(8)
catalase
Table 3 Apparent Ðrst-order rate constants k (103 ] min~1) for PTMK degradation in reactor 2 ([PTMK] \ 160 lmol L~1, 200 mL,
[TiO ] \ 3.5 g L~1) by the methods indicated
2
Ultrasound
in the dark
(515 kHz)
Photocatalysis and 515 kHz-
ultrasound simultaneously
Method
Photocatalysis alone
With TiO Degussa P25
21.5 ^ 1(1)
11.2 ^ 0.5(2)
34.5 ^ 1.5
2
k
] k \ 32.7 ^ 1.5
(1)
(2)
With TiO Rhodia
32 ^ 1.5(3)
11.3 ^ 0.5(4)
14.0 ^ 0.5
42.7 ^ 2
2
k
] k \ 43.3 ^ 2
(3)
(4)
Without TiO
È
È
2
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668
4665

View Article Online
for the reason indicated in the Discussion section. The results
are shown in Table 2.
A positive e†ect of ultrasound upon photocatalysis was
reported18 for TiO Degussa P25 and 20 kHz-ultrasound in
2
the degradations of 3-chlorobiphenyl, pentachlorophenol and
2,4-dichlorophenol. However, the degradation rates when only
ultrasound was employed were not indicated so that whether
a synergy occurred cannot be inferred. The removal rate of
4-chlorophenol by ultrasound was displayed and, curiously,
the photocatalytic and sonolytic rates did not add initially,
but apparently only after a long period, and no synergy was
observed.
Fig. 3 Plots of triÑuoroacetic acid concentration vs. time for PTMK
degradations in reactor
1
([PTMK] \ 160 lmol L~1, 50 mL,
[TiO ] \ 3.5 g L~1). Degradation methods: 30 kHz-ultrasound (=);
3.2 PTMK degradation products
2
photocatalysis with TiO Degussa P25 or TiO Rhodia (+); 30 kHz-
2
2
ultrasound and photocatalysis simultaneously with TiO Degussa
P25 or TiO Rhodia (>).
The formation of triÑuoroacetic acid (TFA) was observed,
whatever the degradation technique employed, and a nearly
constant concentration of TFA was reached over time as
shown in Fig. 3 for some cases. The concentration of TFA
formed by ultrasonic degradation of PTMK at both fre-
quencies was much smaller than that produced by photo-
2
2
olysis at 30 kHz was observed, since the k values for the
simultaneous action of photocatalysis and 30 kHz-ultrasound
were between 1.7 and 1.9 times larger for TiO Degussa P25
2
and between 1.4 and 1.6 times larger for TiO Rhodia (Table
2
catalytic degradation whatever the TiO specimen. The
2
2) than the sum of the individual photocatalytic and ultra-
simultaneous action of photocatalysis and 30 kHz-ultrasound
sonic rate constants. Within experimental accuracy, no
synergy was found for 515 kHz-ultrasound, regardless of the
slightly reduced the formation of TFA compared with photo-
catalysis alone (Figs. 3 and 4; Table 4).
The intermediate products detected (Table 5) by GC-MS
(with or without SPME) were identiÐed by use of a mass
spectra library. These experiments were carried out only for
TiO sample (Table 3). In the case of the simultaneous action
2
of photocatalysis and 30 kHz-ultrasound, catalase was added
photocatalysis with TiO Degussa P25, sonolysis at 515 kHz
2
and their combination. In all cases, hydroxylated PTMK and
a
PTMK dimer were found. Benzene, benzene mono-
substituted with F or CF , and C -alkynes were observed
3
4
only when ultrasound was used. The monohydroxy and dihy-
droxy derivatives were the main intermediate products at low
PTMK conversion.
4
Discussion
4.1 E†ect of the TiO sample on the PTMK photocatalytic
2
removal rate
Fig. 4 Plots of triÑuoroacetic acid concentration vs. PTMK conver-
sion. Conditions and symbols: see Fig. 3.
TiO Rhodia, despite its lower surface area, was almost 2.5
2
Table 4 Maximum concentration of TFA (lmol L~1) obtained and corresponding conversion percentage of PTMK degraded by the methods
indicated ([PTMK] \ 160 lmol L~1, 50 mL (reactor 1) or 200 mL (reactor 2), [TiO Degussa P25 or Rhodia] \ 3.5 g L~1)
2
Ultrasound
in the dark
(30 and 515 kHz)
Photocatalysis and ultrasound
simultaneously
(30 and 515 kHz)
Photocatalysis
alone
Parameter
Maximum concentration
of TFAa
Conversion percentage
of PTMK to TFAa
70È80
ca. 50
ca. 10
ca. 6
50È60
ca. 35
a See plateaus of TFA concentration in Fig. 3.
Table 5 Intermediate products of PTMK degradation in reactor 2 obtained by the methods indicated ([PTMK] \ 160 lmol L~1, 200 mL,
[TiO Degussa P25] \ 3.5 g L~1)
2
Photocatalysis and
515 kHz-ultrasound
simultaneously
Photocatalysis
alone
515 kHz-ultrasound
515 kHz-ultrasound
Compound
without TiO
with TiO
2
2
Monohydroxy-PTMKa
Dihydroxy-PTMKa
Trihydroxy-PTMKb
DitriÑuoroacetylbiphenyl isomer
TriÑuoromethylbenzene
Fluorobenzene
Benzene
Buta-1,3-diyne
But-3-ene-1-yne
Yes
Yes
Yes
Yes
No
No
No
No
No
Yes
Yes
Yes
Yes
Traces
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Traces
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Traces
Yes
Yes
Yes
Yes
a The three isomers. b One isomer (not identiÐed).
4666
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668

View Article Online
times more active than TiO Degussa P25 (Table 2). This
by ultrasound. Furthermore, this enzyme also remained
2
result conÐrms19 that TiO with a higher photocatalytic activ-
active, at least partially, in the presence of UV-irradiated
TiO .29 PTMK disappearance with catalase followed appar-
ently Ðrst-order kinetics, as without catalase. No synergy
2
ity than that of the commonly used TiO Degussa P25 can be
2
2
synthesized, and that the relatively high activity of the
Degussa sample is not due to an optimum ratio20 of anatase
(Table 2) between photocatalysis and sonolysis was observed
to rutile (ca. 80/20) since TiO Rhodia is 97.5% anatase.
in the presence of catalase, whatever the TiO sample. There-
2
2
Further characterizations of the samples and studies with
fore, the e†ect of catalase apparently supports the hypothesis
other pollutants are needed for a better understanding of the
parameters that determine the photocatalytic activity. That
was not the objective of this investigation.
that the synergy is due to the OH~ radicals photocatalytically
generated from the hydrogen peroxide produced by the ultra-
sonic waves.
By contrast, the absence of synergy within experimental
accuracy at 515 kHz, regardless of the TiO sample, is a priori
unexpected according to this hypothesis, since the formation
4.2 Frequency and TiO e†ects on the PTMK ultrasonic
2
2
removal rates
rate of hydrogen peroxide was reported to be more important
at high frequency.24 The working hypothesis we tentatively
suggest to explain this phenomenon is that the H O
On increasing the ultrasonic frequency from 30 to 515 kHz, k
was increased (from ca. 1 to ca. 14 ] 10~3 min~1) for the
same aqueous solution volume (200 mL). Increases in ultra-
sonic removal rates have been previously reported for CCl 16
and aromatic pollutants.21,22 The cavitation bubbles, at high
frequency, are more numerous per volume unit, are formed in
a larger fraction of the solution volume, have a reduced life-
time and correspond to a lower energy (decrease in tem-
perature and pressure).23,24 The two former phenomena may
be at the origin of the observed increase in k, whereas the two
latter phenomenaÈwhich could have had an unfavorable
e†ect on kÈdid not prevail, apparently.
In contrast to what was observed with low frequency ultra-
sound, the degradation rate of PTMK with 515 kHz-
ultrasound in the dark decreased by ca. 20% in the presence
2
2
residence time at the TiO surfaceÈand thereby the decom-
2
position into hydroxyl radicalsÈmight be reduced at 515
4
kHz, since the fraction of TiO particles submitted to the
2
ultrasound is much larger.4 Further work is clearly needed to
verify this hypothesis or to propose another one.
4.4 Information derived from the PTMK degradation
products
4.4.1 TriÑuoroacetic acid (TFA). The 6% of PTMK con-
verted into TFA (Table 4) by ultrasonic degradation at both
frequencies can be attributed to the attack of PTMK by OH~
radicals ejected from the cavitation bubbles, whereas pyrolysis
of PTMK in the cavitation bubbles presumably splits the CÈF
bonds thereby forming other products (see below). From the
practical viewpoint, using 515 kHz-ultrasound seems prefer-
able because of the higher efficiency without any change in the
amount of TFA.
of either one or the other TiO specimen. As the ultrasonic
2
power remained unchanged on adding the catalyst, we
hypothesize that the decline in k at 515 kHz might have been
caused by a decrease in the gaseous exchanges, i.e. in the con-
centration of dissolved oxygen25 and thereby in the ultrasonic
efficiency, in the presence of the TiO particles. Indeed a
Also, using TiO Rhodia instead of TiO Degussa P25 is of
2
higher ultrasonic power was needed to form the acoustic foun-
2
2
great practical interest, since the removal rate of PTMK by
photocatalysis without ultrasound can be increased by a
factor of 2.5 without producing a higher percentage of TFA.
This observation illustrates the fact that the relative impor-
tance of various photocatalytic mechanisms di†ers with the
tain.
4.3 Hypothesis concerning the origin of the synergy between
photocatalysis and sonolysis
TiO specimen. TFA was stable whatever the degradation
conditions (see, for example, Fig. 3). Like the other halogen-
Several hypotheses may explain the synergy between photo-
catalysis and sonolysis at 30 kHz.18 We checked the hypothe-
sis of a better efficiency of photocatalysis due to a change in
the physical properties of titania induced by 30 kHz-
2
ated acetic acids, TFA being miscible to water presumably
does not enter the cavitation bubbles where it could be
pyrolyzed. On the other hand, OH~ radicals in aqueous solu-
tion are very poorly reactive with respect to TFA.30,13
Clearly, to avoid TFA formation from PTMK or analogous
compounds, a sonolytic treatment or pretreatment at high fre-
quency is of interest.
ultrasound. Although a signiÐcant portion of the TiO
Degussa P25 went through the 0.45 lm Millipore discs when
2
the TiO aqueous suspension was previously sonicated for 120
2
minÈshowing that the mean diameter of the agglomerates
had diminishedÈthe photocatalytic degradation rate of
PTMK was the same whether or not TiO was sonicated in
2
pure water for 120 min before PTMK was added. A similar
4.4.2 Intermediate products. The formation of benzene, Ñu-
orobenzene, buta-1,3-diyne and but-3-ene-1-yne by sonication
of PTMK at 515 kHz proves the existence of pyrolysis in the
cavitation bubbles or at their surface (Table 5). For instance,
buta-1,3-diyne and 3-but-3-ene-1-yne are known intermediate
products of the high-temperature thermalÈphotolytic oxida-
tion of monochlorobenzene.31 Formation of Ñuorobenzene
observation was reported for 2,4-dichlorophenol.18 The
hypothesis of a positive inÑuence of titania on the ultrasonic
disappearance rate of PTMK in the dark was also ruled out
because this rate was identical with and without TiO , irre-
2
spective of the TiO specimen (Table 2). The hypothesis we
2
propose assumes the formation of additional OHÕ radicals by
photocatalysis from the hydrogen peroxide produced by ultra-
sound. In sonochemistry, hydrogen peroxide results from the
recombination of either hydroxyl radicals or hydroperoxyl
radicals, and it accumulates in the solution over time.26,27 On
the other hand, in UV-irradiated TiO solutions, H O is
and of triÑuoromethylbenzene traces means that F~ and CF~
3
radicals were formed by sonochemistry. Thus, unlike mono-
chlorophenols,12 PTMK, as, presumably, other relatively
hydrophobic compounds, allows one to reveal the di†erences
between the ultrasonic and the photocatalytic treatment of
water (see the Introduction).
2
2 2
rapidly degraded to OH~ radicals at the TiO surface.28 To
verify the validity of this hypothesis concerning H O , cata-
2
Formation of aromatic hydroxylated products is common
2
2
2
lase, which catalyzes the dismutation of H O into H O and
in TiO photocatalysis in the aqueous phase.1a Since the
2
2
2
O , was added to the suspensions. Although a small decrease
PTMK structure is thus maintained, this may explain the
2
in the enzymatic activity was found when an aqueous solution
of catalase was ultrasonicated for 3 hours, the enzyme
remained sufficiently active to degrade all the H O formed
marked increase in TFA concentration when PTMK had
almost completely disappeared (Fig. 4). Formation of a coup-
ling product containing two PTMK moieties is consistent
2
2
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668
4667

View Article Online
Weinheim, 1997, vol. 4, pp. 2111È2122; (c) A. Mills and S. Le
with the presence of a PTMK radical and shows that micro-
movements in a high-frequency sonicated PTMK solution do
not prevent coupling reactions.
In short, the identiÐcation of the PTMK intermediate pro-
ducts shows that PTMK pyrolysis occurs in the cavitation
bubbles. And the absence of additional intermediate products
when photocatalysis and ultrasound (at both frequencies) were
simultaneously used indicates that the degradation pathways
pertaining to each technique are not qualitatively changed
when the techniques are combined.
Hunte, J. Photochem. Photobiol. A, 1997, 108, 1.
A. Kotronarou, G. Mills and M. R. Ho†mann, J. Phys. Chem.,
1991, 95, 3630.
N. Serpone and P. Colarusso, Res. Chem. Intermed., 1994, 20, 635.
C. Petrier, M.-F. Lamy, A. Francony, A. Benahcene, B. David, V.
Renaudin and N. Gondrexon, J. Phys. Chem., 1994, 98, 10514.
I. Hua and M. R. Ho†mann, Environ. Sci. T echnol., 1997, 31,
2237.
J. N. Jensen, Hazard. Ind. W astes, 1996, 28, 265.
N. Serpone, R. Terzian, P. Colarusso, C. Minero and E. Peliz-
zetti, Res. Chem. Intermed., 1992, 18, 183.
H. G. Flynn, in Physical Acoustics, ed. W. P. Mason, Academic
Press, New York, 1964, vol. 1, pp. 58È172.
A. J. Walton and G. T. Reynolds, Adv. Phys., 1984, 33, 595.
2
3
4
5
6
7
8
9
5
Conclusion
10 A. A. Atchley and L. A. Crum, in UltrasoundÈIts Chemical,
Physical and Biological E†ects, ed. K. S. Suslick, VCH, New
York, 1988, ch. 1, p. 1.
Synthesis of the TiO Rhodia sample conÐrms that samples
2
more active than TiO Degussa P25 can be prepared19 and
2
demonstrates that an optimum rutile/anatase ratio20 is not
11 I. Hua, R. H. Hoechemer and M. R. Ho†mann, J. Phys. Chem.,
necessary to obtain more active TiO . Also, the e†ect of cata-
1995, 99, 2335.
2
12 N. Serpone, R. Terzian, H. Hidaka and E. Pelizzetti, J. Phys.
Chem., 1994, 98, 2634.
13 D. Mas, H. Delprat and P. Pichat, in Photoelectrochemistry, ed.
K. Rajeshwar, L. M. Peter, A. Fujishima, D. Meissner and M.
Tomkiewich, The Electrochemistry Society Inc., Pennington NJ,
1997, PV 97-20, p. 289.
14 Merck Index, Merck, Rahway, NJ, 12th edn., 1996.
15 Handbook of Physical Properties of Organic Chemicals, ed. P. H.
Howard and W. M. Meylan, CRC Lewis, Boca Raton, FL, 1997.
16 A. Francony and C. Petrier, Ultrasonics, 1996, 3, 77.
17 S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and Poro-
sity, Academic Press, London, 1967, p. 80.
18 A. J. Johnston and P. Hocking, Am. Chem. Soc., Symp. Ser., 1993,
518, 106.
19 P. Pichat, C. Guillard, C. Maillard, L. Amalric and J.-C.
DÏOliveira, in Photocatalytic PuriÐcation and T reatment of W ater
and Air, ed. D. F. Ollis, H. Al-Ekabi, Elsevier, Amsterdam, 1993,
p. 207.
lase apparently suggests that the role of photocatalytically
formed H O in degrading organic pollutants can vary
2
2
largely, depending on the TiO sample.
2
The very favorable inÑuence of increasing the ultrasonic fre-
quency upon organic pollutant removal, previously reported
for aromatic pollutants21,22 and CCl ,16 is also observed for
4
PTMK, at least for an increase from 30 to 515 kHz without a
change in the delivered energy. Our study clearly shows that
the phenomena occurring in the cavitation bubbles can be of
interest in water treatment to split groups, such as CF , that
3
withstand oxidation technologies. Finally, for the treatment of
wastewater that may contain Ðne solid particles, our results
indicate that the degradation efficiency of ultrasound at 30
kHz will presumably not be modiÐed by these particles and at
515 kHz will undergo only a modest decrease.
Simultaneous use of TiO photocatalysis and ultrasound
20 K. Tanaka, M. F. V. Capule and T. Hisanaga, Chem. Phys. L ett.,
2
does not appear of interest with respect to the successive use
1991, 187, 1, 2, 73.
of these technologies, since a synergy between both technol-
ogies was observed only at the lower ultrasound frequency
(which is less e†ective than the higher frequency). Further-
more, the advantage of using ultrasound to minimize the for-
mation of triÑuoroacetic acid, an undesirable and stable
21 B. David, M. Lhote, V. Faure and P. Boule, W ater Res., 1998, 32,
8, 2451.
22 C. Petrier, B. David and S. Laguian, Chemosphere, 1996, 32, 9,
1709.
23 R. M. G. Boucher, Br. Chem. Eng., 1970, 14, 263.
24 C. Petrier and A. Francony, W ater Sci. T echnol., 1997, 35, 175.
25 N. Gondrexon, V. Renaudin, P. Boldo, Y. Gonthier, A. Bernis
and C. Petrier, Chem. Eng. J., 1997, 66, 21.
26 P. K. Bhattacharyya and R. Veeraghavan, Int. J. Chem. Kinet.,
1977, 9, 629.
27 C. Petrier, M. Micolle, G. Merlin, J.-L. Luche and G. Reverdy,
Environ. Sci. T echnol., 1992, 26, 1639.
product, from CF -containing pollutants is very much miti-
3
gated when both technologies are employed concurrently.
Acknowledgements
The authors are grateful to Region Rhone-Alpes (France) for a
PhD scholarship to P.T. and partial Ðnancial support
(““Programme EmergenceÏÏ).
28 B. Jenny and P. Pichat, L angmuir, 1991, 7, 947.
29 L. Amalric, C. Guillard and P. Pichat, Res. Chem. Intermed.,
1994, 20, 579.
30 K.-D. Asmus, R. Fliount, H. Hungerbuehler and O. Makogen, in
AFEAS (Alternative Fluorocarbons Environmental Acceptability
Study) W orkshop on ““Decomposition of T riÑuoroacetic Acid in the
EnvironmentÏÏ, February 1994, Washington, DC, SPA-AFEAS,
Washington, DC.
References
1
See, for example: (a) D. Bahnemann, J. Cunningham, M. A. Fox,
E. Pelizzetti, P. Pichat and N. Serpone, in Aquatic and Surface
Photochemistry, ed. G. R. Helz, R. G. Zepp and D. G. Crosby,
Lewis, Boca Raton, FL, 1994, ch. 21, pp. 261È316 and refs.
therein; (b) P. Pichat, in Handbook of Heterogeneous Photo-
catalysis, ed. G. Ertl, H. Knozinger and J. Weitkamp, VCH,
31 J. L. Graham, J. M. Berman and B. Dellinger, J. Photochem. Pho-
tobiol. A, 1993, 71, 65.
Paper 9/02506E
4668
Phys. Chem. Chem. Phys., 1999, 1, 4663È4668