Role of humic acids in the TiO2-photocatalyzed degradation of tetrachloroethene in water

Article abstract of DOI:10.1016/S0043-1354(98)00368-6

The effect of humic acids on the TiO₂-mediated photocatalytic degradation of tetrachloroethene (PCE) was kinetically investigated at different pH and initial substrate concentrations. The process occurs through two parallel paths: a major oxidative route leading to mineralization and a reductive route leading to the formation of dichloroacetic acid (DCAA), also undergoing photodegradation. The rate of PCE decomposition was found to decrease in the presence of humic acids adsorbed on the semiconductor surface, while the concentration of the intermediate dichloroacetic acid increased. This is a consequence of the scavenging action of humic acids toward photoproduced surface oxidant species, which makes conduction band electrons more easily available for interface reactions. Kinetic studies on the effect of humic acids in the TiO₂-mediated photodegradation of dichloroacetic acid showed that the progressively greater accumulation of this highly toxic intermediate, observed with increasing humic acids content, was a consequence of both an increase in the rare of its production from PCE and a decrease in the rate of its oxidative photodegradation. The effect of humic acids on the TiO₂-mediated photocatalytic degradation of tetrachloroethene (PCE) was kinetically investigated at different pH and initial substrate concentrations. The process occurs through two parallel paths: a major oxidative route leading to mineralization and a reductive route leading to the formation of dichloroacetic acid (DCAA), also undergoing photodegradation. The rate of PCE decomposition was found to decrease in the presence of humic acids adsorbed on the semiconductor surface, while the concentration of the intermediate dichloroacetic acid increased. This is a consequence of the scavenging action of humic acids toward photoproduced surface oxidant species, which makes conduction band electrons more easily available for interface reactions. Kinetic studies on the effect of humic acids in the TiO₂-mediated photodegradation of dichloroacetic acid showed that the progressively greater accumulation of this highly toxic intermediate, observed with increasing humic acids content, was a consequence of both an increase in the rate of its production from PCE and a decrease in the rate of its oxidative photodegradation.

Full text of DOI:10.1016/S0043-1354(98)00368-6

Wat. Res. Vol. 33, No. 8, pp. 1827±1836, 1999
#
1999 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0043-1354/99/$ - see front matter
PII: S0043-1354(98)00368-6
ROLE OF HUMIC ACIDS IN THE TiO -PHOTOCATALYZED
2
DEGRADATION OF TETRACHLOROETHENE IN WATER
1
ELENA SELLI , DANIELA BAGLIO , LUCA MONTANARELLA and
2
2
2
GIOVANNI BIDOGLIO *
1
Dipartimento di Chimica Fisica ed Elettrochimica, Universita di Milano, Via Golgi 19, 20133 Milan,
Á
Italy and European Commission, Joint Research Centre, Environment Institute, 21020 Ispra (Va),
Italy
2
(
First received December 1997; accepted in revised form August 1998)
AbstractÐThe e€ect of humic acids on the TiO -mediated photocatalytic degradation of tetrachlor-
2
oethene (PCE) was kinetically investigated at di€erent pH and initial substrate concentrations. The pro-
cess occurs through two parallel paths: a major oxidative route leading to mineralization and a
reductive route leading to the formation of dichloroacetic acid (DCAA), also undergoing photodegra-
dation. The rate of PCE decomposition was found to decrease in the presence of humic acids adsorbed
on the semiconductor surface, while the concentration of the intermediate dichloroacetic acid increased.
This is a consequence of the scavenging action of humic acids toward photoproduced surface oxidant
species, which makes conduction band electrons more easily available for interface reactions. Kinetic
2
studies on the e€ect of humic acids in the TiO -mediated photodegradation of dichloroacetic acid
showed that the progressively greater accumulation of this highly toxic intermediate, observed with
increasing humic acids content, was a consequence of both an increase in the rate of its production
from PCE and a decrease in the rate of its oxidative photodegradation. # 1999 Elsevier Science Ltd.
All rights reserved
2
Key wordsÐTiO -mediated photodegradation, humic acids, tetrachloroethene, dichloroacetic acid
INTRODUCTION
Moreover, organic substances, such as fulvic and
humic acids, which are widespread in natural
waters, may modify the photodegradation paths in
the photocatalytic puri®cation treatments of waste
and drinking waters. Such humic substances may
also play an important role in redox reactions
photoinduced by sunlight in aquatic systems at the
surface of minerals with semiconducting properties,
such as iron oxides. Indeed, natural organic acids
have been demonstrated to be capable of sensitising
Heterogeneous photocatalysis, involving photoin-
duced redox reactions at the surface of semiconduc-
tor minerals, is a promising technique for the
treatment of waters contaminated by organic mol-
ecules (Ollis and Al-Ekabi, 1993; Pelizzetti et al.,
1
994; Ho€mann et al., 1995). Among semiconduc-
tors, titanium dioxide is the most widely used in
photocatalysis, due to its outstanding stability
(
Frank and Bard, 1977). Under irradiation with
light of suitable wavelength, charge carriers (i.e.
electrons and holes) produced at the TiO surface
photoredox
Vinodgopal and Kamat, 1992; Kamat and
Vinodgopal, 1994; Selli et al., 1996).
In the present paper, the e€ects induced by
humic acids (HA) in the TiO -mediated photode-
processes
on
semiconductors
(
2
can oxidise and/or reduce many organic and inor-
ganic substances (Serpone and Pelizzetti, 1989;
Linsebigler et al., 1995). In particular, halogenated
compounds (alkanes, alkenes and aromatic com-
pounds) were reported to undergo complete miner-
2
gradation of tetrachloroethene (PCE) were investi-
gated, to obtain insight into their possible role in
the formation and disappearance of byproducts.
Tetrachloroethene, a largely inert non-¯ammable
solvent, practically omnipresent in trace amounts in
groundwaters of industrialized countries, was
reported to give some byproducts during both its
homogeneous UV-light induced degradation
2
alization in water suspensions of TiO , with the
formation of CO and mineral acids. Earlier studies
2
focused on the complete photodegradation of pollu-
tants (Pruden and Ollis, 1983; Ollis et al., 1984),
without investigating the possible formation of in-
termediates, which might have a toxicity even (Mertens and von Sonntag, 1995; Hirvonen et al.,
higher than that of the starting substrates. 1996) and its TiO -mediated photodegradation in
2
water (Glaze et al., 1993; Kenneke et al., 1993;
Gianturco et al., 1996). In particular, Glaze et al.
(1993) proposed that the latter process proceeds
through two parallel pathways. The usually
*
Author to whom all correspondence should be addressed.
Tel.: +39-332-78-9383; fax: +39-332-78-9328; e-mail:
giovanni.bidoglio@jrc.it].
[
1827

1
828
Elena Selli et al.
mine the initial concentration of PCE. Samples (0.3 or
.5 ml) were withdrawn by means of a syringe through a
te¯on septum placed on one of the cones of the reaction
accepted oxidative route should lead to PCE miner-
alization and trichlorinated byproducts [i.e. trichlor-
oacetic acid (TCAA)], via either bound or free
0
vessel. A volume of air equal to the volume of the sample
hydroxyl radicals, formed by photogenerated holes was always injected into the reactor immediately after
trapped at the semiconductor surface through the sample collection, in order to avoid a pressure decrease. A
similar procedure was employed in the photodegradation
2
oxidation of adsorbed H O, hydroxide ions or sur-
runs of chloroacetic acids. In this case, 5-ml samples were
withdrawn from the reactor. A few runs were also carried
out in TiO suspensions purged with oxygen for several
2
face titanol groups (Turchi and Ollis, 1990). In ad-
dition, a simultaneous reductive route, involving
conduction band electrons, should produce dichlor- hours prior to irradiation. In this case, oxygen, instead of
air, was added to the suspension after taking samples at
di€erent times.
2
oacetic acid (DCAA). The TiO -mediated photode-
gradation of PCE thus appeared a particularly
suitable process for investigating how the presence
Stirred suspensions were irradiated in
Cofomegra, Italy) equipped with a 1500 W m xenon
a Solarbox
�2
(
of HA may modify the reaction pathways and the lamp having a spectral energy distribution similar to solar
fate of toxic byproducts.
radiation.
A
Einstein s
mean polychromatic radiation ¯ux of
�
4
�1 �1
was estimated by ferrioxalate actino-
10
l
metry (Hatchard and Parker, 1956).
Photodegradation runs were started by putting the reac-
tor into the Solarbox, switched on at least 20 min earlier,
in order to assure steady state irradiation conditions.
Samples for PCE analysis were put in 2-ml vials and im-
mediately sealed to avoid any loss of the analyte. Samples
employed for the analysis of chloroacetic acids were
EXPERIMENTAL
Materials
Tetrachloroethene (PCE), dichloroacetic acid (DCAA),
trichloroacetic acid (TCAA), dibromoacetic acid (DBAA),
4
-¯uorobromobenzene and methanol (purity>99%) were injected in 5-ml vials containing DBAA as internal stan-
Fluka products and were used as received. Water puri®ed dard and were purged with nitrogen for 10 min in order to
by a Milli-Q water system (Millipore) was used in the
preparation of aqueous solutions. Saturated solutions of
decrease PCE concentration, which could have interfered
in the analysis of chloroacetic acids. This procedure did
not decrease the amount of byproducts in the sample
(Glaze et al., 1993).
Kinetic runs lasted 1±2 h, depending on pH, on the in-
2
itial substrate concentration, and on TiO and HA con-
�
1
PCE, containing about 200 mg l of PCE, as determined
by gas chromatography±mass spectrometry (GC±MS)
analysis, were prepared every week by stirring this com-
pound with water for several days in the dark and stored
at 48C.
tents. The number of samples taken from the reaction
vessel during each kinetic run was calculated so that the
overall volume reduction of the irradiated suspensions was
Titanium dioxide was Degussa P25, constituted by 80%
2
�1
anatase and 20% rutile, with a surface area of 50 m g
and an average particle size of 30 nm. The TiO
sions were prepared immediately prior to use by adding
weighed amounts of powder to a 0.1 M NaClO
aqueous repeated under identical experimental conditions, in order
solution, followed by ultrasonication for 30 min. The use
of NaClO as a background electrolyte to keep ionic
suspen- limited to 10% in PCE photodegradation tests and to
15% in DCAA photodegradation tests. Some runs were
2
4
to obtain enough experimental points for kinetic analysis.
4
Analytical methods
strength constant is not expected to interfere with the
photodegradation.
The humic acid (HA) was obtained from Aldrich as a
Analytical techniques were developed for the determi-
nation of PCE and chloroacetic acids directly in the het-
erogeneous suspensions, thus avoiding any separation
procedure, which might have caused losses of the analytes.
Solid phase microextraction (SPME), coupled with GC±
MS, was used for PCE analysis. The technique is based on
the equilibration of the analyte between an aqueous phase
and an organic polymer coated onto a fused silica ®ber
�
1
sodium salt. Stock HA solutions (500 mg l ) were pre-
pared by dissolving 0.25 g of material in 2 ml of a 0.1 M
NaOH solution under nitrogen atmosphere, followed by
sonication, dilution with distilled water to 500 ml, pH
adjustment to 7 and ®ltration through a 0.45-mm pore di-
ameter Millipore membrane. Stock solutions were stored
at 48C in the dark (Selli et al., 1996).
(
Belardi and Pawliszyn, 1989), which is successively intro-
duced into the GC injector, where the analyte is thermally
desorbed, sent through the GC column and analysed by
MS.
Experimental setup and procedure
The photodegradation runs were performed in a per-
fectly sealed 100 ml Pyrex cylindrical reactor with 4 lateral
A SPME Supelco device was used, consisting of a
cones. A continuous water recirculation around it assured 10 mm long and 100 mm thick fused silica ®ber, coated
both temperature control at 20218C and ®ltering of in-
frared light. Continuous pH measurements were made by
with 95 mm thick polydimethylsiloxane. The following
optimized conditions were adopted (Nilsson et al., 1998).
means of
inserted in one of the reactor cones. Small volumes of a
.05 M NaOH aqueous solution were added by a 665
a glass electrode (Radiometer, Denmark) The ®ber was exposed to the PCE solution for 30 min,
then introduced into the GC injector and thermally des-
orbed at 2008C for 5 min. A Varian 3400 gas chromato-
0
Dosimat apparatus (Metrohm, Switzerland) through graph was used, equipped with a 75 m 0.53 mm 3 mm
another cone. Both instruments were controlled by a 682 ®lm thickness DB-624 fused-silica column (J&W
Titroprocessor (Metrohm, Switzerland), which maintained
the pH constant to 20.05 pH units, by neutralizing the
acidic species produced in the photodegradation of chlori-
nated organic compounds.
Scienti®c), with the following temperature program: 408C
�
1
for 5 min; ramp at 88C min to 2008C; 5 min at 2008C.
�
1
Helium was used as carrier gas at 25 ml min . The GC
was interfaced via a jet separator (0.3 Torr in the separa-
tor and 0.03 Torr in the analyser) and a transfer line
The reactor was ®rst almost completely ®lled with the
aqueous suspension, if necessary pre-equilibrated
TiO
2
(2208C) to
a
ITS-40 ion trap mass spectrometer
with the HA solution. The required volume of saturated (Finnigan). The ion trap was operated at 3008C in the
PCE solution was then added, the pH adjusted to the
desired value and the mixture stirred for a few minutes. A
electron impact mode, scanning from 50 to 250 m/z in
1.5 s. The GC±MS data were acquired by means of the
Saturn I software (Varian).
®rst sample of the suspension was then taken, to deter-

Humic acids in PCE photodegradation
1829
Fig. 1. Plots of ln{[PCE]/[PCE]
the presence of 1.0 g l
0
} vs irradiation time for the PCE photodegradation runs carried out in
�
1
2
of TiO , under air, at pH 7, with an impinging radiation ¯ux of
�
4
�1 �1
�5
�5
�6
1
0
Einstein s
l
. Initial PCE concentration: (.) 4.5Â 10 M, (Q) 2.3Â 10 M, (R) 8.7Â 10 M.
Open symbols refer to experimental points obtained under identical experimental conditions in oxygen
atmosphere, instead of air.
�
1
containing 1.00 g l of DCAA, TCAA or DBAA (internal
For instrument calibration, performed in the PCE con-
�
1
centration range 50±1000 mg l , solutions were prepared standard) were employed for calibration in the concen-
�
1
immediately before analysis in 2-ml vials containing water
and closed with te¯on caps, by spiking with standard PCE
methanol solutions to the desired PCE concentration. PCE
was identi®ed from its retention time and mass spectrum.
Peak areas were normalized to those of 4-bromo¯uoroben-
tration ranges of 50±500 mg l
60 mg l for TCAA. DBAA concentration was 100 mg l
in any case.
The following derivatization procedure was adopted.
5.0 ml of standard aqueous solution were acidi®ed with
for DCAA and 10±
�1
�
1
�
1
zene (139 mg l ), used as internal standard. Preliminary concentrated H SO4 (pH < 1) and mixed with 1.0 ml of
2
tests (Nilsson et al., 1998) assured that the presence of
TiO and HA did not interfere in the analysis of PCE.
dimethyl ether. After shaking, the organic phase was sep-
arated and the aqueous phase was further extracted three
2
Gas chromatography with electron capture detection times with dimethyl ether. A 2-ml volume of methanol
was added to the diethyl ether extract and the ether was
(GC±ECD) was employed for the determination of
DCAA and TCAA, after their methylation in BF
nol solutions, according to a recently reported procedure a 10% BF
3
metha- removed in a hot water bath at 65±708C. Then, 1.4 ml of
3
±methanol solution (Fluka) were added, im-
(Hargesheimer and Satchwill, 1989). Aqueous solutions mediately followed by heating at 508C for 35 min. The
2
Table 1. Experimental parameters, rate constants and DCAA percent molar yields of the TiO -mediated photodegradation of PCE
5
�1
a
�1
4
�1
b
Run No.
[PCE]
0
Â10 (M)
[HA] (mg l
)
pH
Intensity
[TiO
2
] (g l
)
k
appÂ10 (s
)
% DCAA
E1
E2
E3
E4
E5
E6
E7
E8
4.5
2.3
0.87
2.3
2.3
2.2
4.5
2.2
0.87
2.2
2.2
2.3
4.6
4.5
2.2
2.2
0
0
0
0
0
0
5
5
5
5
5
5
10
10
10
10
7
7
7
9
7
7
7
7
7
9
7
7
7
9
7
9
I
I
I
I
II
I
I
I
I
I
II
I
I
I
I
1.0
1.0
1.0
1.0
1.0
2.0
1.0
1.0
1.0
1.0
1.0
2.0
1.0
1.0
1.0
1.0
2.3720.06
3.2220.10
5.8720.16
4.0820.17
1.7420.05
4.0120.18
1.9420.08
2.6120.08
4.8020.10
2.7620.11
1.3220.02
3.3520.12
1.8320.08
2.1120.17
1.9720.03
2.3420.07
4.3
3.5
n.d.
4.3
0.9
11
16
7.1
n.d.
10
3.1
16
22
20
18
21
E9
E10
E11
E12
E13
E14
E15
E16
I
a
�2
�4
�1 �1 �2 b
l ) and (II) 1000 W m . Percent
Irradiation system operated at (I) 1500 W m (mean polychromatic radiation ¯ux = 10 Einstein s
molar amount of PCE transformed into DCAA after 30 min irradiation; n.d. = not determined.

1
830
Elena Selli et al.
Fig. 2. DCAA concentration detected during some of the PCE photodegradation runs of Table 1, in
the absence of HA. (R) E1, (Q) E2, (.) E4, (Â) E5 and (W) E6. Open symbols refer to experimental
points obtained under identical experimental conditions in oxygen atmosphere, instead of air.
mixture was then cooled and the reaction quenched by
adding 1 ml of 5% sodium sulphate aqueous solution. The
esteri®ed acids were ®nally extracted with 2 ml of cyclo-
hexane and analysed by GC±ECD. Preliminary tests on
the Langmuir±Hinshelwood rate law, r = dC/
dt = kKC/(1 + KC), with C = [PCE]. Indeed,
straight lines were obtained by plotting ln{[PCE]/
�
1
the possible interferences of suspended TiO
2
(1 g l ) and/ [PCE] } vs time (see for example Fig. 1) and the
0
�
1
or dissolved HA (5 mg l ) assured that their presence did apparent ®rst order rate constants kapp (Table 1),
not a€ect the extraction of chloroacetic acids.
5890 series II Hewlett Packard instrument was
calculated from the slope of these plots, were inver-
sely proportional to [PCE]
A
employed for GC analysis, equipped with a 15 mCi Ni
63
, the initial PCE concen-
0
electron capture detector, a 7673 injector autosampler and tration.
a
silica capillary column (SGE). It was operated with the
50 m 0.25 mm 0.25 mm ®lm thickness HT8 fused
Under the adopted experimental conditions, no
photodegradation practically occurred at pH 5,
while an increase in the reaction rate was observed
if the pH was increased (runs E4/E2 of Table 1).
following temperature program: 358C for 8 min; ramp at
�
1
5
8C min to 1458C; 10 min at 1458C. Hydrogen was used
�
1
as carrier gas at 1.8 ml min . The ECD was kept at
3
258C. The make-up gas was argon containing 5% Reaction pathways not involving the photocatalyst,
methane.
which were reported to lead to di€erent byproducts
Mertens and von Sonntag, 1995; Hirvonen et al.,
996), should require irradiation with wavelength
shorter than 300 nm and should thus be excluded in
sion, containing the amount of NaOH necessary for the the employed experimental conditions. Indeed, we
nal desired pH value, were mixed with 1.5 ml of aqueous
(
DCAA adsorption onto TiO
2
1
In order to determine the amount of DCAA adsorbed
onto titanium dioxide, 8.5 ml of a 1.0 g l TiO suspen-
2
�
1
®
veri®ed that PCE did not undergo any photodegra-
dation in the absence of TiO
As to byproducts, increasing amounts of dichlor-
solutions containing di€erent amounts of DCAA. After
stirring for 90 min in the dark, a 5-ml sample was mixed
2
.
�
1
with 0.5 ml of a solution containing 1.00 mg l of DBAA
internal standard). The suspension thus obtained was oacetic acid (DCAA) were always detected in runs
(
employed for the determination of total DCAA. The
remaining 5-ml sample was centrifuged at 2000 rpm for
carried out in di€erent experimental conditions, as
shown in Fig. 2, while the presence of trichloroace-
tic acid (TCAA), which was reported to form as
3
0 min. The surnatant was employed for the determination
of DCAA in solution, after ®ltration through a 0.45-mm
pore diameter Millipore ®lter and addition of 0.3 ml of a minor byproduct (Glaze et al., 1993; Kenneke et
al., 1993), was never revealed in our irradiated sus-
�
1
solution containing 1.00 mg l of DBAA.
�
7
pensions. However, a 5 Â10 M TCAA aqueous
suspension resulted perfectly stable under our ir-
radiation conditions. This fact, which is in agree-
ment with recent results reported by Kesselman et
RESULTS AND DISCUSSION
2
A series of preliminary systematic runs of TiO -
mediated photodegradation of PCE were carried al. (1997), excludes the possibility that TCAA could
out in the absence of HA, in order to compare kin- not be detected, being a very low concentration
etic results with those obtained in the presence of stationary state byproduct, and points to the con-
HA under identical irradiation conditions. PCE clusion that less than 0.01% of PCE was actually
photodegradation was found to occur according to transformed into TCAA.

Humic acids in PCE photodegradation
1831
2
Fig. 3. PCE and DCAA concentration pro®les during the TiO -mediated photodegradation of PCE (Ð
1
ÐÐ) in the absence of HA and in the presence of (- - -) 5 mg l and (. . .) 10 mg l of HA at (a) pH 7
�
�1
and (b) pH 9.
It is worth underlining that in the pH range of the concentration pro®les of both PCE and DCAA
the present work both DCAA and TCAA are com- during kinetic runs in di€erent experimental con-
pletely dissociated in their anionic form.
ditions are shown in Fig. 3.
The kapp values relative to runs E7±E9 of Table 1
show an increase in the rate of PCE photodegrada-
tion with decreasing initial substrate concentration,
indicating that also in the presence of HA the ex-
Photocatalytic degradation of PCE in the presence of
humic acids
PCE photodegradation runs, carried out in TiO
suspensions containing HA, followed an apparent perimental
rst order kinetics (runs E7±E16 of Table 1). The Hinshelwood rate law. The e€ect of HA was clearly
2
data
followed
the
Langmuir±
®
rate coecients kapp obtained by linearising exper- to decrease the PCE photodegradation rate: pro-
imental data of PCE concentration vs time accord- gressively lower kapp values were obtained in the
ing to such a rate law are reported in Table 1, while presence of increasing amounts of HA (see for

1
832
Elena Selli et al.
Table 2. Reactions leading to PCE photodegradation in the pre-
sence of TiO
in semiconductors from their excited state
Vinodgopal and Kamat, 1992; Kamat and
2
(
Reaction number
Vinodgopal, 1994) and of sensitising photoreduc-
tion processes at the semiconductor surfaces by
scavenging the holes photogenerated in the semi-
conductor valence band, thus stabilising the elec-
trons photopromoted in the conduction band (Selli
et al., 1996).
Initiation
hꢀ
�
cb
‡
vb
TiO2�4 e ‡ h
(1)
(2)
(3)
�
+
OHads+hvb4ÇOH
� �
O
2,ads+ecb4OÇ
2
The e€ect of their presence in the TiO -mediated
2
Oxidative Path
CCl
ÇCCl
CCl
2
2
2
=CCl
CCl OH 4 Mineralization
=CCl
=CCl
2
+ÇOH 4ÇCCl
2
CCl
2
OH
(4)
(5)
(6)
(7)
photodegradation of PCE can be discussed in the
frame of the reactions reported in Table 2. HA,
adsorbed onto titanium dioxide, act as scavengers
2
+
+
2
+hvb4(CCl =CCl Ç)
2 2
+
(
CCl
2
2
Ç) +O
2
4Mineralization
+
of valence band holes (h ), thus slowing down the
vb
predominant oxidative routes of PCE photodegra-
dation, occurring either through the hydroxyl rad-
ical attack on PCE (Table 2, reaction 4) or through
Reductive Path
�
�
CCl
(CCl
ÇCCl
2
=CCl
=CCl
±CHCl
2
+ecb4+(CCl
2
=CCl
Ç) +H 4ÇCCl
+O 4CCl OÇ
2
Ç)
(8)
(9)
(10)
-
2
2
2
±CHCl
±CHCl
2
+
2
2
2
2
2
2
the direct interaction between hvb and the substrate
Table 2, reaction 6) (Kesselman et al., 1997). The
hydrolysis
CCl2O2 ±CHCl2 �4 CHCl2 ±COOH
(11)
(12)
(13)
(
�
�
CCl
CCl
2
=CCl +OÇ
)CCl
2
2
4CCl
2
(OÇ
2
)CCl
±CHCl
2
2
�
+
reductive route, leading to DCAA as byproduct, is
accelerated by HA, by making conduction band
2
(OÇ
2
2
+H 4CCl
2
OÇ
2
�
electrons (ecb) more easily available for interface
example runs E2, E8 and E15). At pH 7 the pre- reactions. These species could either directly interact
�
1
sence of 5 mg l of HA caused a ca. 18% decrease with the substrate, yielding a radical anion (Table 2,
in rate coecient in all investigated conditions, reaction 8), which subsequently adds a proton
�
1
while in suspensions containing 10 mg l of HA the (Table 2, reaction 9) and ®nally transforms into
decrease in kapp was of the order of 30%. The e€ect DCAA in the aqueous medium (Table 2, reactions
was more pronounced at pH 9. At the same time, 10 and 11) (Glaze et al., 1993), or could interact
in all runs carried out in the presence of HA the with adsorbed molecular oxygen (Table 2, reaction
�
decrease in the rate of PCE photodegradation was 3), yielding the superoxide radical anion OÇ
, which
2
accompanied by a proportional increase in the rate attacks PCE (Table 2, reaction 12) (Schwitzgebel et
of DCAA formation (Fig. 3). This is a clear indi- al., 1995). The addition of proton follows
a
cation that this intermediate species is formed (Table 2, reaction 13), yielding the same precursor
through a reaction path di€erent from that directly of DCAA produced by reactions 8±10 in Table 2.
leading to PCE mineralization.
Several other examples are reported in the litera-
It is worth noting, ®rst of all, that the observed ture (Bahnemann et al., 1987; Sta€ord et al., 1993;
decrease in the PCE photodegradation rate could Choi and Ho€mann, 1995, 1996, 1997; Minero et al.,
not be a simple consequence of inner ®lter e€ects. 1995; Schmelling and Gray, 1995; Schmelling et al.,
Indeed, HA absorb visible light, but solutions con- 1996) on the TiO -mediated photocatalytic processes
2
�
1
taining 5 and 10 mg l of HA could not absorb of chlorinated or nitro compounds, in which, in ad-
more than 2 and 4.5% of incident radiation, re- dition to oxidative pathways, also reductive pathways
spectively. Moreover, in the conditions of the pre- are operating. In some cases, for example in the
4
sent work, HA, although carrying prevalently photocatalytic degradation of CCl , the transient free
negative charges, were totally adsorbed on the tita- radical intermediates produced by substrate reduction
nium dioxide surface (Eliet, 1996). They are elec- were also selectively trapped (Choi and Ho€mann,
tron-rich compounds, capable of injecting electrons 1996). Moreover, both conduction band electrons and
2
Fig. 4. Reaction scheme of the TiO -mediated photodegradation of PCE.

Humic acids in PCE photodegradation
1833
�
1
Table 3. Experimental parameters and rate constants of the photocatalytic degradation of DCAA in the presence of 1.0 g l of TiO
2
6
a
4
Run No.
[DCAA]
0
Â10
[HA]
pH
Intensity
k
d
Â10
A1
A2
A3
A4
A5
A6
A7
A8
A9
2.8
5.0
7.2
3.1
2.8
2.8
5.4
7.2
2.6
0
0
0
0
0
5
5
5
5
7
7
7
9
7
7
7
7
9
I
I
I
I
II
I
I
I
I
3.3420.21
3.1020.07
3.2620.21
3.3820.22
1.5120.11
2.7420.23
2.5520.17
1.6320.11
1.0520.05
a
�2
�4
�1 �1
�2
Irradiation system operated at (I) 1500 W m (mean polychromatic radiation ¯ux = 10 Einstein s
l
) and (II) 1000 W m
.
�
adsorbed O2 or OÇ2 species, besides valence band doubled in standard reaction conditions (run E8 vs
holes, have been recently demonstrated to participate run E2). Even higher increases in % DCAA were
in the photocatalytic air oxidation of organic com- induced by the presence of HA at higher substrate
pounds (Schwitzgebel et al., 1995; Fan and Yates, concentration (runs E7/E1) and at lower light inten-
1
996; Cermenati et al., 1997).
sity (runs E11/E5), while an increase in pH (runs
(runs E12/E6)
We also veri®ed that the formation of DCAA E10/E4) and in the amount of TiO
2
was not a consequence of oxygen de®ciency, which produced more or less the same e€ects both in the
is anyway rather improbable with the low substrate presence and in the absence of HA.
amounts employed in the present work and with a
In the analysis of kinetic results, however, it is
concentration of molecular oxygen in air-saturated worth reminding that also the reaction intermediate
�
4
water of 2.4 Â10 M. In fact, runs carried out by DCAA (Fig. 4) undergoes photodegradation under
saturating the reaction vessel with oxygen, instead irradiation in the presence of TiO2 (Ollis et al.,
of air, led to identical concentration vs time pro®les 1984). Thus, the greater accumulation of DCAA
of both PCE and DCAA (Figs 1 and 2).
observed in the presence of HA could either be a
The present results on HA sensitisation thus indi- consequence of an increase in the rate of DCAA
cate that DCAA, an intermediate oxidation product formation from PCE, or could be due to a decrease
of PCE, should be produced in a reaction path in the DCAA photodegradation rate, or could arise
involving conduction band electrons. The molar from the sum of both contributions. Indeed, the
percent of PCE yielding DCAA always increased in competition of HA adsorbed on TiO
the presence of larger amounts of HA (Fig. 3 and photogenerated oxidative species could induce a
2
for the
�
1
Table 1). In the presence of 5 mg l of HA, the decrease in the DCAA photodegradation rate, as in
percent amount of PCE transformed into DCAA the case of PCE photodegradation. In order to dis-
after 30 min irradiation (% DCAA) was roughly tinguish between these hypotheses, kinetic tests of
�
1
(1.0 g l ) as a function of
Fig. 5. Surface concentration of DCAA adsorbed onto Degussa P25 TiO
the DCAA concentration in the aqueous solutions.
2

1
834
Elena Selli et al.
Table 4. Rate coecients k
f
of the PCE photodegradation path
a
DCAA photodegradation were carried out both in
the presence and in the absence of HA.
leading to DCAA
5
5
�1
)
[
PCE]
0
Â10 (M)
k
f
Â10 (s
E€ect of humic acids in the photodegradation of
DCAA on TiO
�1
)
no HA
HA (5 mg l
2
4
2
.5
.3
2.8
2.9
5.6
4.2
The photodegradation of DCAA was ®rst veri®ed
to be also a heterogeneous process. Indeed, no
�
6
a
Experimental
intensity = 10 Einstein s
conditions:
l
pH = 7,
�1
amount = 1.0 g l .
light
photodegradation occurred in 2.8 Â10 M DCAA
�
4
�1 �1
, TiO
2
aqueous solutions without titanium dioxide, even in
�
1
the presence of 5 mg l of HA. This excluded HA
as photosensitizers in the homogeneous phase
(
Zepp et al., 1985; Canonica et al., 1995).
Table 3 reports the conditions and the rate par- ively.
ameters calculated from the experimental runs of
A comparison between k
from PCE and DCAA photodegradation, respect-
values obtained in the
d
TiO -mediated photodegradation of DCAA, both presence and in the absence of HA (Table 3)
2
in the absence (runs A1±A5) and in the presence of demonstrates that the rate of photodegradation of
5
�
1
mg l of HA (runs A6±A9). In the absence of DCAA decreased in the presence of HA, this e€ect
HA the disappearance of DCAA occurred accord- being more evident at pH 9 than at pH 7. This indi-
ing to a ®rst order kinetics, the rate constant values cates that the photodegradation of DCAA should
k
d
obtained from di€erent initial concentrations of prevalently occur through the usually accepted oxi-
DCAA (runs A1±A3) being coincident, within the dative path, being decelerated by the oxidant sca-
experimental error. This ®nding is in agreement vengers HA.
with the results of DCAA adsorption on TiO
2
,
The increase of DCAA accumulation observed
determined in a concentration range close to that during kinetic runs in the presence of HA (Fig. 3)
obtained in the photodegradation runs. Indeed, as was thus due, at least in part, to a decrease in the
shown in Fig. 5, at both investigated pH, the rate of disappearance of this intermediate. Also the
amount of DCAA adsorbed on the semiconductor parameters k of equation 2 could be evaluated
f
surface increased linearly with increasing DCAA from the rate d[DCAA]/dt of DCAA formation
concentration in the solution, [DCAA]sol, showing during the photocatalytic degradation of PCE, cal-
no tendency to saturation. Under these conditions culated from experimental data of Figs 2 and 3,
the Langmuir equation simpli®es to the form:
and the corresponding [PCE] and [DCAA] values at
di€erent reaction times. The values thus
obtained, reported in Table 4, show that in the
k
f
GDCAA ˆ KD‰DCAAŠsol
ꢀ1†
G
DCAA being the surface concentration of DCAA. absence of HA the PCE photodegradation path
From the slopes of the straight lines of Fig. 5 the leading to DCAA accounted for ca. 10% of the
distribution coecient
resulted equal to overall PCE disappearance (k /kapp10.1, see
8.820.6)Â10 and (4.020.2)Â10 l m at pH equation 2). In the presence of HA the rate of
and 9, respectively.
DCAA formation clearly increased (higher
A Langmuir±Hinshelwood rate law might be values, see Table 4), although the overall PCE
more appropriate in the case of DCAA photodegra- photodegradation rate decreased (lower kapp values,
dation in the presence of HA, as the k
value rela- see Table 1). This represents a ®nal con®rmation
K
D
f
�
3
�3
�2
(
7
k
f
d
tive to run A8 is smaller than kd values obtained that conduction band electrons, becoming more
from runs A6 and A7, starting from lower DCAA readily available in the presence of HA adsorbed on
initial concentrations. The lower availability of the semiconductor, are involved in the reaction
TiO surface sites for DCAA adsorption, due to the path leading to DCAA.
2
presence of totally adsorbed HA, may play an im-
portant role in this respect. In a very recent time-
resolved laser ¯ash photolysis study, Bahnemann et
CONCLUSIONS
al. (1997) evidenced that DCAA adsorption on the
TiO
appears to be a prerequisite for an ecient hole formed in the TiO
scavenging.
PCE, is produced in a reaction path involving con-
Anyhow, the following rate equation holds for duction band electrons. In the presence of HA both
2
surface prior to the band gap excitation
DCAA, an intermediate oxidation product
2
-mediated photodegradation of
the intermediate DCAA:
a decrease in DCAA photodegradation rate (lower
kd values) and an increase in the rate of DCAA
d‰DCAAŠ
ˆ
kf‰PCEŠ � kd‰DCAAŠ
ꢀ2† production from PCE (higher k
taneously contribute to a greater accumulation of
being the ®rst order (or apparent ®rst the highly toxic DCAA intermediate (Bull et al.,
order) rate coecients relative to DCAA formation 1990). This can have considerable importance both
f
values) simul-
dt
f d
k and k

Humic acids in PCE photodegradation
1835
degradation of trichloroethylene and tetrachloroethylene
in water. Environ. Sci. Technol. 27, 177±184.
Hargesheimer E. E. and Satchwill T. (1989) Gas-chroma-
tographic analysis of chlorinated acids in drinking
water. Aqua 38, 345±351.
in the heterogeneous photodegradation reactions of
PCE, occurring in natural waters at the surface of
mineral oxides with semiconducting properties and
2
when the TiO -mediated photodegradation is used
as a method for purifying natural waters containing Hatchard C. G. and Parker C. A. (1956) A new sensitive
chemical actinometer. II. Potassium ferrioxalate as a
PCE and humic substances.
standard chemical actinometer. Proc. R. Soc. London
Ser. A: 235, 518±536.
Hirvonen A., Tuhkanen T. and Kalliokoski P. (1996)
2 2
Formation of chlorinated acetic acids during UV/H O -
REFERENCES
oxidation of ground water contaminated with chlori-
nated ethylenes. Chemosphere 32, 1091±1102.
Ho€mann M. R., Martin S. T., Choi W. and Bahnemann
D. W. (1995) Environmental applications of semicon-
ductor photocatalysis. Chem. Rev. 95, 69±96.
Bahnemann D. W., Hilgendor€ M. and Memming R.
(
1997) Charge carrier dynamics at TiO
tivity of free and trapped holes. J. Phys. Chem. B 101,
265±4275.
Bahnemann D. W., Mo
2
particles: reac-
4
nig J. and Chapman R. (1987) Kamat P. V. and Vinodgopal K. (1994) Titanium dioxide
È
mediated photocatalysis using visible light: photosensiti-
zation approach. In Aquatic and Surface
Photochemistry, eds. G. R. Helz, R. G. Zepp and D. G.
Crosby, pp. 437±442. Lewis, Boca Raton, FL.
Ecient photocatalysis of the irreversible one-electron
and two-electrons reduction of halothane on platinized
colloidal titanium dioxide in aqueous suspension. J.
Phys. Chem. 97, 3782±3788.
Belardi R. P. and Pawliszyn J. B. (1989) The application Kenneke J. F., Ferry J. L. and Glaze W. H. (1993) The
of chemically modi®ed fused silica ®bers in the extrac-
tion of organics from water matrix samples and their
rapid transfer to capillary columns. Water Pollut. Res.
J. Can. 24, 179±191.
2
TiO -mediated photocatalytic degradation of chloroalk-
enes in water. In Photocatalytic Puri®cation and
Treatment of Water and Air, eds. D. F. Ollis and H. Al-
Ekabi, pp. 179±191. Elsevier, Amsterdam.
Bull R. J., Sanchez I. M., Nelson M. A., Larson J. L. and
Lansing A. J. (1990) Liver tumor induction in B6CF1
mice by dichloroacetate and trichloroacetate. Toxicology
Kesselman J. M., Weres O., Lewis N. S. and Ho€mann
M. R. (1997) Electrochemical production of hydroxyl
radical at polycrystalline Nb-doped TiO electrodes and
2
6
Canonica S., Jans U., Stemmler K. and Hoigne
Transformation kinetics of phenols in water: photosensi-
tization by dissolved natural organic material and aro-
matic ketones. Environ. Sci. Technol. 29, 1822±1831.
Cermenati L., Pichat P., Guillard C. and Albini A. (1997)
3, 341±359.
estimation of the partitioning between hydroxyl radical
and direct hole oxidation pathways. J. Phys. Chem. B
Â
J. (1995)
1
01, 2637±2643.
Linsebigler A. L., Lu G. and Yates J. T., Jr. (1995)
Photocatalysis on TiO surface: principles, mechanisms
and related results. Chem. Rev. 95, 735±758.
2
Probing the TiO
2
photocatalytic mechanisms in water Mertens R. and von Sonntag C. (1995) Photolysis
puri®cation by use of quinoline, photo-Fenton gener-
ated O HÇ radicals and superoxide dismutase. J. Phys.
Chem. B 101, 2650±2658.
(l = 254 nm) of tetrachloroethene in aqueous solutions.
J. Photochem. Photobiol. A: Chem. 85, 1±9.
Minero C., Pelizzetti E., Pichat P., Sega M. and Vincenti
M. (1995) Formation of condensation products in
advanced oxidation technologies: the photocatalytic
Choi W. and Ho€mann M. R. (1995) Photoreductive
mechanism of CCl
e€ects of electron donors. Environ. Sci. Technol. 29,
646±1654.
Choi W. and Ho€mann M. R. (1996) Kinetics and mech-
anism of CCl photoreductive degradation on TiO : the
4 2
degradation on TiO particles and
2
degradation of dichlorophenols on TiO . Environ. Sci.
Technol. 29, 2226±2234.
1
Nilsson T., Montanarella L., Baglio D., Tilio R., Bidoglio
G. and Facchetti S. (1998) Analysis of volatile organic
compounds in environmental water samples and soil gas
by solid-phase microextraction. Int. J. Environ. Anal.
Chem. 69, 217±226.
4
2
role of trichloromethyl radical and dichlorocarbene. J.
Phys. Chem. 100, 2161±2169.
Choi W. and Ho€mann M. R. (1997) Novel photocataly-
�
tic mechanisms for CHCl , CHBr
3
3
and CCl
3
CO
2
degra- Ollis D. F. and Al-Ekabi H. (eds.) (1993) Photocatalytic
dation and the fate of photogenerated trihalomethyl
radicals on TiO . Environ. Sci. Technol. 31, 89±95.
Puri®cation and Treatment of Water and Air. Elsevier,
Amsterdam.
2
Eliet V. (1996) Applications des techniques de ¯uorescence Ollis D. F., Hsiao C., Budiman L. and Lee C. (1984)
pour l'e
Â
et heteroge
tude de l'uranium dans des milieux homoge
Á
Á
nes
Heterogeneous photoassisted catalysis: conversions of
perchloroethylene, dichloroethane, chloroacetic acids,
and chlorobenzenes. J. Catal. 88, 89±96.
Pelizzetti E., Minero C. and Vicenti M. (1994)
Photocatalytic degradation of organic contaminants. In
Technologies for Environmental Cleanup: Toxic and
Hazardous Waste Management, eds. A. Avogadro and
R. C. Ragaini, pp. 101±138. Kluwer, Dordrecht, The
Netherlands.
Â
Â
nes: reactions d'hydrolyse et photoreduction
 Â
sur le bioxyde de titane. Ph.D. dissertation, Paris XI
Orsay University.
Fan J. and Yates J. T., Jr. (1996) Mechanism of photooxi-
dation of trichloroethylene on TiO
mediates by infrared spectroscopy. J. Am. Chem. Soc.
18, 4686±4692.
2
: detection of inter-
1
Frank S. N. and Bard A. J. (1977) Heterogeneous photo-
catalytic oxidation of cyanide and sul®te in aqueous sol-
utions at semiconductor powders. J. Phys. Chem. 81,
Pruden A. L. and Ollis D. F. (1983) Photoassisted hetero-
geneous catalysis: the degradation of trichloroethylene
in water. J. Catal. 82, 404±417.
1
484±1488.
Gianturco F., Vianelli L., Tatti L., Rota F., Bruzzi P.,
Rivas L., Bellobono I. R., Bianchi M. and Muntau H.
Schmelling D. C. and Gray K. A. (1995) Photocatalytic
transformation and mineralization of 2,4,6-trinitrotolu-
(1996) Pilot-plant photomineralization of dichloro-
2
ene (TNT) in TiO slurries. Water Res. 29, 2651±2662.
methane and tetrachloroethene in aqueous solution, by
photocatalytic membranes immobilizing titanium diox-
ide and photopromoters. Chemosphere 33, 1531±1542.
Glaze W. H., Kenneke J. F. and Ferry J. L. (1993)
Schmelling D. C., Gray K. A. and Kamat P. V. (1996)
Role of reduction in the photocatalytic degradation of
TNT. Environ. Sci. Technol. 30, 2547±2555.
Schwitzgebel J., Ekerdt J. G., Gerischer H. and Heller A.
(1995) Role of the oxygen molecule and of the photo-
2
Chlorinated byproducts from the TiO -mediated photo-

1
836
generated electron in TiO
reactions. J. Phys. Chem. 99, 5633±5638.
Elena Selli et al.
2
-photocatalyzed air oxidation Turchi C. S. and Ollis D. F. (1990) Photocatalytic degra-
dation of organic water contaminants: mechanisms
involving hydroxyl radical attack. J. Catal. 122, 178±
192.
Vinodgopal K. and Kamat P. V. (1992) Environmental
photochemistry on surface. Charge injection from
excited fulvic acid into semiconductor colloids. Environ.
Sci. Technol. 26, 1963±1966.
Selli E., De Giorgi A. and Bidoglio G. (1996) Humic acid-
sensitized photoreduction of Cr(VI) on ZnO particles.
Environ. Sci. Technol. 30, 598±604.
Serpone N. and Pelizzetti E. (eds.) (1989) Photocatalysis
Fundamentals and Applications. Wiley, New York.
Sta€ord U., Gray K. A., Kamat P. V. and Varma A.
(
1993) An in situ di€use re¯ectance FTIR investigation
of photocatalytic degradation of 4-chlorophenol on
TiO powder surface. Chem. Phys. Lett. 205,
5±61.
Zepp R. G., Schlotzhauer P. F. and Sink R. M. (1985)
Photosensitized transformations involving electronic
energy transfer in natural waters: role of humic sub-
stances. Environ. Sci. Technol. 19, 74±81.
2
5