Catalytic ozonation of trace nitrobenzene in water with synthetic goethite

Article abstract of DOI:10.1016/j.molcata.2007.09.030

This paper aims to reveal the effectiveness and the mechanism of catalytic ozonation with synthetic goethite (FeOOH) for the degradation of recalcitrant pollutants that cannot form surface complexes with surface metal sites of the catalyst in water. Nitrobenzene (NB) was used as the target pollutant. Results show that catalytic ozonation with FeOOH can substantially enhance NB degradation compared with ozonation alone. Scavenging test and quantification of the oxidation intermediates of NB reveal that the improved NB degradation is due to the enhanced hydroxyl radical generation. The activity of the FeOOH in water is related to its highly hydroxylated surfaces. The surface hydroxyl groups on the FeOOH in water are active sites in the catalytic ozonation. Moreover, the properties of the surface hydroxyl groups such as their quantity and charge states seem to have close relationship with the activity of the FeOOH in the catalytic ozonation of NB. We finally proposed a pathway that the uncharged surface hydroxyl groups on the FeOOH in water can induce aqueous ozone decomposition to generate hydroxyl radicals.

Full text of DOI:10.1016/j.molcata.2007.09.030

Available online at www.sciencedirect.com
Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
Catalytic ozonation of trace nitrobenzene
in water with synthetic goethite
Tao Zhang, Jun Ma∗
School of Municipal & Environmental Engineering, Harbin Institute of Technology,
Harbin 150090, China
Received 15 August 2007; received in revised form 27 September 2007; accepted 27 September 2007
Available online 10 October 2007
Abstract
This paper aims to reveal the effectiveness and the mechanism of catalytic ozonation with synthetic goethite (FeOOH) for the degradation of
recalcitrant pollutants that cannot form surface complexes with surface metal sites of the catalyst in water. Nitrobenzene (NB) was used as the
target pollutant. Results show that catalytic ozonation with FeOOH can substantially enhance NB degradation compared with ozonation alone.
Scavenging test and quantification of the oxidation intermediates of NB reveal that the improved NB degradation is due to the enhanced hydroxyl
radical generation. The activity of the FeOOH in water is related to its highly hydroxylated surfaces. The surface hydroxyl groups on the FeOOH
in water are active sites in the catalytic ozonation. Moreover, the properties of the surface hydroxyl groups such as their quantity and charge states
seem to have close relationship with the activity of the FeOOH in the catalytic ozonation of NB. We finally proposed a pathway that the uncharged
surface hydroxyl groups on the FeOOH in water can induce aqueous ozone decomposition to generate hydroxyl radicals.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Catalytic ozonation; Synthetic goethite; Nitrobenzene; Surface hydroxyl groups; Hydroxyl radical
1. Introduction
Ferric hydroxides are characterized by their extremely week
dissolution (Ksp = 3 × 10−39) in water. Moreover, the trace Fe3+
dissolved into water is nontoxic. Ferric hydroxides have been
widely used as adsorbents and ion exchangers in water [18]. Up
to date, there are only limited studies on their application as
catalysts for water treatment. Commercial goethite (␣-FeOOH)
were found to be able to promote the oxidation of some contam-
inants by H2O2 [19–21], which suggests that goethite would
be a promising catalyst for pollutant abatement. Lim et al. [22]
noticed that the commercial goethite promoted the degradation
of p-chlorobenzoic acid (p-CBA) during ozonation. Park et al.
[23] further discussed the possible mechanism of the improved
degradation of p-CBA by the commercial goethite. However,
the researchers had not managed to remove the impurities from
the commercial goethite before reactions. Alkalis, physically
adsorbed by the goethite in its aging step of preparation, usually
remain on the commercial product because of incompletely rins-
ing. For example, Park et al. [23] had mentioned that the water
pH increased from 3 to 8 when 2 g of the commercial goethite
was introduced into 1 L of acidified distilled water. Although the
reaction solutions were buffered, the alkalis on the catalyst sur-
face and at the water/catalyst interface still can react with ozone
Recalcitrant organic pollutants are relatively difficult to be
removed from water with traditional water treatment processes.
Some of these pollutants are even difficult to be efficiently
degraded by strong oxidants like ozone [1]. Advanced oxidation
processes such as O3/H2O2, O3/UV, and O3/UV/H2O2, have
been developed to improve the destruction of these pollutants
with enhanced hydroxyl radical generation. Catalytic ozonation
with heterogeneous catalysts is another alternative to improve
the degradation of the recalcitrant pollutants. The catalysts are
usually MnO2 [2–5], TiO2 [6,7], Al2O3 [8–10], noble metals
[11], and mixed metal oxides [12–17], In addition to its high
activity in catalytic ozonation under neutral pH condition of
water treatment, it is also necessary for the catalyst that its dis-
solution is very weak and the released metal ions pose no health
problem to the water.
∗
Corresponding author. Tel.: +86 451 86282292; fax: +86 451 82368074.
E-mail address: majun@hit.edu.cn (J. Ma).
1381-1169/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2007.09.030

T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
83
producing hydroxyl radicals [1], which makes it difficult to find
2.3. Analysis
out the actual effectiveness of catalytic ozonation with goethite
and the catalysis pathway. Therefore, it is crucial to remove the
adsorbed alkalis from the goethite to better reveal its activity in
catalytic ozonation of refractory organic pollutants as well as
the function of its surface sites in the catalytic ozonation.
In this study, goethite (FeOOH) particles were carefully
synthesized to remove any alkalis that might be adsorbed dur-
ing preparation. Nitrobenzene (NB) was selected as the target
organic pollutant because it reacts slowly with molecular ozone
2.3.1. Ozone, nitrobenzene, and oxidation intermediates of
NB
Aqueous ozone concentration was measured with the indigo
method [26]. Examination of NB in water followed a method
described in a previous paper [5]. The relative standard deviation
(R.S.D.) of this method was less than 5% for NB. Quantification
of the oxidation intermediates of NB was carried out on a Waters
HPLC system (717 plus Autosampler, 1525 Binary Pump, and
2478 Dual ␭ Absorbance Detector). A Symmetry C18 column
(3.9 mm × 150 mm, Waters) was used to separate and quantify
aromatic fractions of the intermediates, while an Atlantis dC18
column (4.6 mm × 250 mm, Waters) was used to separate and
quantify carboxylic fractions. For the examination of aromatic
intermediates, the detection wavelength was 280 nm, the col-
umn temperature was set at 40 ◦C, and the eluent consisted of a
70:30 (phosphate buffer solution (320 ␮L H3PO4 in 1 L Milli-
Q water) /methanol) isocratic mixture. For the examination of
carboxylic intermediates, the wavelength was set at 220 nm, the
column temperature was 30 ◦C, and the eluent was a mixture of
phosphate buffer solution (320 ␮L H3PO4 in 1 L Milli-Q water)
and acetonitrile with a volume ratio of 80:20.
(kO ,NB = 0.09 M−1
s
−1) [24], and also because it cannot form
3
surface complexes with FeOOH and has very weak adsorption
on the hydroxylated FeOOH surface. The objective is to reveal
the effectiveness of FeOOH in improving the ozonation of NB
and possible functions that surface hydroxyl groups on FeOOH
in water played in the catalytic ozonation.
2. Experimental
2.1. Catalysts
FeOOH was prepared according to a method described
by Kandori et al. [25]. Fe(OH)3 was obtained by quickly
adding 200 mL of 1 mol L−1 NaOH solution into 200 mL of
0.25 mol L−1 Fe(NO3)3 with magnetic stirring. The suspension
was aged at 60 ◦C and pH 12 for over 48 h. Then, the precipitate
was repeatedly rinsed with boiled Milli-Q water until the pH
of leachate remained constant in three consecutive rinses. The
precipitate was further dried at 70 ◦C for 16 h and grinded after-
wards. Particles with diameters ranging from 0.075 to 0.3 mm
were used in this experiment.
2.3.2. Surface characterization
BET surface area, pore volume, and average pore size of the
catalyst were obtained on a surface area and porosity analyzer
(ASAP 2020, Micromeritics). The density of surface hydroxyl
groups, expressed in mmol per unit gram, was measured accord-
ing to a method described by Tamara and Tanaka [27]. The pHpzc
(pH of point-of-zero charge) of the catalyst was measured with
a powder addition method [28].
Other oxides of copper, cobalt, iron, nickel, and manganese
were obtained by the calcination of the nitrates at 450 ◦C for 2 h.
Grinded particles with diameters ranging from 0.075 to 0.3 mm
were used.
2.3.3. Crystallographic confirmation
Thermogravimetry and differential thermal analysis (TG-
DTA) for the dried FeOOH particle was carried out on a
Rigaku TAS 100 thermal analyzer with temperature rising rate
of 10 ◦C min−1. Crystallite structures were confirmed with a
Rigaku D/MAX-rA powder diffractometer (XRD) using Cu K␣
radiations.
2.2. Experimental procedure
Experiments were carried out in batch mode with circulated
water bath. Ozone was generated from dried oxygen with a labo-
ratory ozone generator (XFZ-58I, Tsinghua Tongli). The ozone
gas was introduced into distilled water through a silica dispenser
in a cylindrical glass reactor with a diameter of 6 cm and a vol-
ume of 1000 mL. The gas flow was switched off when aqueous
ozone reached a steady concentration. Then, the NB stock solu-
tion and the catalyst were quickly added into the ozone bearing
water, and the reaction was started with magnetic stirring. The
initial aqueous ozone concentration can be controlled through
changing the electric current of the ozone generator. Phosphate
buffer of 1.0 mmol L−1 was used in most of the experiments,
which was sufficient to keep the pH constant during the reac-
tions. AcidifiedNa2SO3 solutionwasusedtoquenchtheresidual
ozone in the samples. Then, the samples were instantly filtered
with glass fiber filters, which had no influence on the NB concen-
tration. In the case of quantification of the NB intermediates, we
used a high initial NB concentration (19.7 ␮mol L−1) dissolved
in Milli-Q water.
3. Results and discussion
3.1. Effectiveness in improving NB degradation
Fig. 1 shows the degradation of NB in ozonation, cat-
alytic ozonation, and adsorption by FeOOH. It is clear that
ozonation degraded NB slowly except the initial 2 min of the
reaction. According to the references [29,30], instantaneous
ozone decomposition occurs at the initial phase of ozonation
which generates a significant amount of hydroxyl radicals.
The NB degradation at the initial phase of ozonation can thus
be attributed to the generation of hydroxyl radicals which
react quickly with NB at a secondary reaction rate constant
of 7.8 × 109 M−1 s−1 [31]. Catalytic ozonation with FeOOH
degraded 65 percent of NB within 20 min, which is about two
times higher than the NB degraded in ozonation alone. In addi-

84
T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
which occupied about 10% of the initial dose after seven times
of repeated circulation. Then, the weight loss of the FeOOH is
likely to be responsible for the slight decrease of NB degradation
in the repeated catalytic ozonation. The results indicate that the
activity of the FeOOH is relatively stable.
As mentioned in the introduction, if the alkalis adsorbed on
FeOOH during preparation were not removed, they can react
with ozone at the catalyst surface or catalyst/water interface
producing hydroxyl radicals. However, the alkalis will deplete
quickly in consecutive ozonation because of dilution as well as
their fast reaction with ozone [1]. The stable activity of FeOOH
in the repeated catalytic ozonation, therefore, confirms that the
alkalis are well removed from the FeOOH in the preparation of
the catalyst. The results also reveal that the activity of FeOOH
is due to its surface properties but not the alkalis.
Fig. 1. Degradation o−f1 NB in ozonation and catalytic ozonation.
100 Mg L−1; 1.0 mmol L−1 phosphate buffered pH 6.7; T = 18 ◦C.
[NB]0 = 0.37 ␮mol L−1
;
FeOOH
dose =
3.2. Discussion on the mechanism
[O3]0 = 0.40 0.02 mg L
;
3.2.1. Influence of hydroxyl radical scavenger
Since NB adsorption on FeOOH was very weak and the
direct ozone oxidation of NB is quite slow, the enhanced NB
degradation by the catalytic ozonation is possibly ascribed to
accelerated hydroxyl radical generation from ozone. The influ-
ences of tert-butanol (t-BuOH), a well-known hydroxyl radical
scavenger, on ozone decomposition and NB degradation for
ozonation alone and the catalytic ozonation are shown in Fig. 3a
and b. FeOOH enhanced ozone decomposition in water. In the
presence of t-BuOH, both ozone decomposition rate and NB
degradation rate significantly decreased in ozonation and cat-
alytic ozonation. Moreover, the two oxidation processes had
similar ozone decomposition rate as well as NB degradation rate
in the presence of t-BuOH. The t-BuOH is characterized by its
reaction with hydroxyl radicals [31], generating inert intermedi-
ates and consequently quenching aqueous ozone decomposition.
The results indicate that the improved NB degradation in the
catalytic ozonation is due to the enhanced ozone decomposition
which generates more hydroxyl radicals.
tion, only less than 1.6% of the NB can be adsorbed by the
FeOOH in this experiment. The results reveal that the FeOOH
is quite effective in enhancing the ozonation of NB.
The FeOOH was also compared with other typical catalysts
(i.e., CuO, Co3O4, Fe2O3, NiO, and MnO2) for their effective-
ness in catalytic ozonation of NB. Fig. 1 indicates that FeOOH
is more effective than these oxides in NB degradation.
The FeOOH was further used repeatedly in the catalytic
ozonation in order to find out whether its activity would
deplete. The FeOOH particles used in the batch experiment
were reclaimed by filtration through a glass fiber filter, dried
at 120 ◦C, and reused for different times in the catalytic ozona-
tion. NB degradation rate in the catalytic ozonation decreased
slightly as the reuse times of FeOOH increased (Fig. 2). When
the FeOOH was reused for seven times, the NB degradation
slightly decreased by about 8–11% compared with that when
the FeOOH was reused for the first time. It was noted that there
was a slight loss of the catalyst during the repeated filtration,
3.2.2. Yields of the oxidation intermediates
The intermediates of NB that can be quantified after
ozonation and FeOOH-catalyzed ozonation were 4-nitrophenol
(4-NP), 3-nitrophenol (3-NP), 2-nitrophenol (2-NP), and oxalic
acid. Fig. 4 shows the evolution of the concentrations of these
intermediates during ozonation and catalytic ozonation. It is
noted that catalytic ozonation produced the same kinds of inter-
mediates with ozonation alone. Furthermore, the yields of the
intermediates in catalytic ozonation are higher than those in
´
ozonation alone. According to Hoigne and Bader [24], the nitro-
phenols are the products that hydroxyl radicals react with NB
by means of addition reactions. Oxalic acid is formed in the
direct oxidation of the nitrophenols by molecular ozone [32].
The results indicate that the catalytic ozonation follows a similar
oxidation pathway with ozonation, i.e. hydroxyl radical oxida-
tion in combination with direct oxidation by molecular ozone.
The enhanced accumulation of the intermediates in the cat-
alytic ozonation also indicates that the FeOOH promotes NB
degradation through the enhanced hydroxyl radical generation.
Fig. 2. Influence of consecutive reutilization of FeOOH on NB degradation
in catalytic ozonation. [O3]0 = 0.40 0.02 mg L−1; [NB]0 = 0.37 ␮mol L−1
;
FeOOH dose = 200 mg L−1; 1.0 mmol L−1 phosphate buffered pH 6.7; T =
18 ◦C.

T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
85
Table 1
Major characteristics of FeOOH and hematite
␣-FeOOH
␣-Fe2O3
BET surface area (m2 g−1
)
54.78
0.250
24.5
0.502
7.0
58.12
0.497
30.1
0.380
8.8
Pore volume (cm3 g−1
)
a
Average pore diameter (nm)
Density of surface hydroxyl groups (mmol g−1
pHpzc
)
a
Cumulative volume of pores between 1.7 and 300 nm diameter.
MeOH+2 ꢀ MeOH + H+
MeOH + OH− ꢀ MeO− + H2O
(1a)
(1b)
The effect of water pH on the catalytic ozonation of NB
is depicted in Fig. 5. The catalytic ozonation substantially
enhanced NB degradation at pH close to 7. However, the
enhancement of NB degradation by the catalytic ozonation
compared with ozonation alone significantly decreased as the
solution became acidic or basic.
Andreozzi et al. [2] reported that MnO2 had no activity
in catalytic ozonation of oxalic acid when solution pH was
higher than its pHpzc. Tong et al. [34] observed similar phe-
nomena when different types of MnO2 were used to improve
the ozonation of sulfosalicylic acid. Furthermore, they both
proposed the mechanism of surface complex formation and
its subsequent oxidation by aqueous ozone. This mechanism
cannot be used to interpret the catalytic ozonation of NB
with FeOOH, because surface complexes cannot be formed
between NB molecules and the FeOOH. However, the surface
charge states determined by the pHpzc and water pH seem also
very important to the catalytic ozonation. Since the pHpzc of
FeOOH is 7.0 (Table 1), its surface hydroxyl groups would
be mostly zero charged near neutral pH and charged under
acidic or basic water pH conditions. The results indicate that
the zero-charged surfaces of FeOOH are more active than the
protonated or deprotonated in promoting hydroxyl radical gen-
eration.
Fig. 3. Influenceoft-BuOHonozonedecomposition(a)andNBdegradation(b).
[O3]0 = 0.40 0.02 mg L−1; [NB]0 = 0.37 ␮mol L−1; [t-BuOH]0 = 50 mg L−1
;
FeOOH dose = 200 mg L−1
T = 18 ◦C.
;
1.0 mmol L−1 phosphate buffered pH 6.7;
In addition, the total organic carbon (TOC) of the NB
solution was nearly not reduced after either ozonation or
catalytic ozonation (not shown here), which means that the
oxalic acid is difficult to be mineralized by molecular ozone
and hydroxyl radicals under the experimental conditions here.
However, since the intermediates are more biodegradable
than NB, the catalytic ozonation combined with a flow-up
bio-filtration process will degrade more TOC than the conven-
tional ozonation/bio-filtration process used in water treatment
plants.
3.2.4. Influence of thermal-treatment for FeOOH
Thermal-treatment for FeOOH above certain temperature
will make it transform to ferric oxide. Some changes on the
properties of the hydroxylated surface in water will subse-
quently take place. In an effort to understand the function of
the hydroxylated surface in the catalytic ozonation of NB, the
effect of thermal-pretreatment for FeOOH on its activity was
examined.
3.2.3. Influence of pH
Thermal-pretreatment at the temperatures of 70, 120, 180,
and 240 ◦C had no significant effect on NB degradation in the
catalytic ozonation (Fig. 6). However, a significant decrease of
NB degradation rate occurred after FeOOH having been pre-
treated at 300 or 500 ◦C.
Thermal behavior analysis with TG-DTA for FeOOH is
shown in Fig. 7. There was a weight loss of 5 percent with
the increment of temperature below 200 ◦C, which is caused
by the evaporation of water molecules physically adsorbed on
the FeOOH. A significant weight loss and an endothermic peak
Water pH is one of the most important factors that influ-
ence the efficiency of ozonation, as OH− can initiate aqueous
ozone decomposition to generate hydroxyl radicals. Moreover,
the water pH is also an important factor that determines the
charge properties of surface hydroxyl groups at oxide/water
interface(Eqs. (1a)and(1b))[33]. Theoxideisnearlynotsurface
charged at water pH near its pHpzc. The oxide surface becomes
protonated or deprotonated when water pH is below or above its
pHpzc
.

86
T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
Fig. 4. NB intermediates formed in ozonation and catalytic ozonation. [O3]0 = 10.0 0.02 mg L−1; [NB]0 = 19.7 ␮mol L−1; FeOOH dosage = 100 mg L−1; pH 6.3;
T = 18 ◦C.
emerged near 270 ◦C, which can be attributed to the dehydroxy-
lation of the FeOOH and the formation of a new crystalline. The
XRD diffractograms (Fig. 8) demonstrate that the FeOOH had
transformed to hematite (␣-Fe2O3) at the thermal-treatment of
300 or 500 ◦C.
As the FeOOH was transformed to ␣-Fe2O3 at the thermal-
treatment, its BET surface area, pore volume, average pore
size, and pHpzc were increased (Table 1). The density of sur-
face hydroxyl groups in water was decreased with the value
on FeOOH being 1.32 times of that on ␣-Fe2O3. The less
surface hydroxyl groups on ␣-Fe2O3 in water would provide
fewer chances for the interaction between ozone and the surface
hydroxyl groups. Moreover, the charge properties or acid/base
properties of the surface hydroxyl groups were also changed
with the increase of pHpzc from 7.0 to 8.8. At pH 6.7 of
this study, most of the surface hydroxyl groups on FeOOH
were nearly not charged. Most of them on the ␣-Fe2O3, how-
ever, would be positively charged (protonated) because of its
much higher pHpzc value than the water pH. Therefore, the
decrease of the catalytic activity as the FeOOH was trans-
formed to ␣-Fe2O3 possibly can be ascribed to two reasons:
(1) the decrease of surface hydroxyl groups’ quantity; and (2)
the increase in the proportion of protonated surface hydroxyl
groups.
Fig. 5. Influence of pH on NB degradation. [O3]0 = 0.40 0.02 mg L−1
;
Fig. 6. Influence of thermal-pretreatment for FeOOH on catalytic ozona-
[NB]0 = 0.37 ␮mol L−1; FeOOH dose = 500 mg L−1; 1.0 mmol L−1 phosphate
buffered pH 6.7; T = 18 ◦C; reaction time of 10 min.
tion of NB. [O3]0 = 0.40 0.02 mg L−1
;
[NB]0 = 0.37 ␮mol L−1
;
catalyst
dose = 100 mg L−1; 1.0 mmol L−1 phosphate buffered pH 6.7; T = 18 ◦C.

T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
87
Fig. 7. TG–DTA curves of FeOOH at a heating rate of 10 ◦C min−1
.
Fig. 9. Influence of the increase of anions from 2 mmol L−1 to 10 mmol L−1
on NB degradation in catalytic ozonation relative to that in ozona-
tion.
[O3]0 = 1.20 0.02 mg L−1
;
[NB]0 = 2.96 ␮mol L−1
,
FeOOH
3.2.5 Influence of substitution of surface hydroxyl groups by
different anions
dose = 200 mg L−1, pH 7.30, T = 16 ◦C.
Some inorganic anions can be adsorbed onto metal oxide in
water through ligand exchange with surface hydroxyl groups,
which is specific adsorption (Eq. (2)). If the specific adsorption
inhibits the catalytic ozonation of NB, it will prove in a more
direct way that the activity of FeOOH is related to its surface
hydroxyl groups.
2 mmol L−1 and 10 mmol L−1 anions, respectively. “Cc1” and
“Cc2” are final NB concentrations after catalytic ozonation in the
presence of 2 mmol L−1 and 10 mmol L−1 anions, respectively.
Results indicate that the increment of phosphate significantly
inhibits NB degradation in the catalytic ozonation. The increase
of sulfate has much less negative effect on the NB degradation.
However, the increment of nitrate has nearly no negative effect
on the NB degradation in the catalytic ozonation.
Phosphate has the highest coordination affinity for surface
Fe(III) on FeOOH among these anions [33,36]. According to
Ler and Stanforth [37], it forms inner-sphere complex with the
surface Fe(III), which leads to its maximum substitution for sur-
face hydroxyl groups on the FeOOH in water. Sulfate adsorbs
on FeOOH mainly as outer-sphere complex at neutral water
pH [38,39]. Nitrate has very weak coordination affinity for the
Fe(III), which leads to nearly no substitution of surface hydroxyl
groups of the FeOOH by this anion [33]. Therefore, the inhibi-
tion degree of the catalytic ozonation by these anions is closely
related with the extent of their substitution for surface hydroxyl
groups of the FeOOH. The results further reveal that the surface
hydroxyl groups of FeOOH in water play an important role in
the catalytic ozonation of NB.
MeOH + L− → MeL + OH−
(2)
Phosphate, sulfate, and nitrate have different coordination
affinity for surface Fe(III) of FeOOH [35]. Fig. 9 shows the inhi-
bitioneffectsoftheincrementdoseofsodiumphosphate, sodium
sulfate, andsodiumnitrateonNBdegradationofcatalyticozona-
tion, respectively. The “Reduced extent” of NB degradation was
used for the y-axis to level off differences in NB degradation
rates caused by the increase of ionic strength. It was expressed
as:
(Co1 − Cc1) − (Co2 − Cc2)
C0
“C0” is the initial NB concentration. “Co1” and “Co2” are
final NB concentrations after ozonation in the presence of
3.2.6 Pathway of FeOOH-induced hydroxyl radical
generation
In the previous paper of catalytic ozonation with manganese
hydroxide, the authors proposed that OH− in water would be
electro-statically attracted to the surface of the hydroxide when
the solution pH was above its pHpzc. The attracted OH− would
initiate radical chain decomposition of ozone at the hydrox-
ide/water interface [4]. However, this hypothesis cannot explain
the relatively low activity of some oxides with much lower pHpzc
than water pH in the catalytic ozonation. For example, MnO2
with the pHpzc of 4 [40], has extremely low activity in promoting
hydroxyl radical generation from ozone.
Metal oxides adsorb water molecules strongly in water,
because the surface cations are not saturatedly coordinated.
The adsorbed water molecules dissociate into OH- and H+,
Fig. 8. XRD for FeOOH treated at different temperatures.

88
T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
4. Conclusions
The careful preparation well removed the alkalis from the
synthetic goethite (FeOOH). Catalytic ozonation with the pre-
pared FeOOH substantially improved the degradation of NB in
water through enhanced generation of hydroxyl radicals. The
activity of the FeOOH in this case is due to its highly hydrox-
ylated surfaces in water. The surface hydroxyl groups on the
FeOOH in water play an important role in the catalytic ozona-
tion. It seems that the uncharged surface hydroxyl groups are
more active in the catalytic ozonation than the protonated or
deprotonated ones. Results indicate that the uncharged surface
hydroxyl groups of FeOOH in water can induce aqueous ozone
decomposition to generate hydroxyl radicals during the catalytic
ozonation of NB.
Fig. 10. Scheme of the proposed pathway of hydroxyl radical generation when
aqueous ozone interacts with surface hydroxyl group of FeOOH in water.
forming surface hydroxyl groups with surface cations and oxy-
gen anions, respectively [41]. The surface hydroxyl groups at
the water/oxide interface will interact with O3 and the sub-
strate in catalytic ozonation. The FeOOH has much higher
surface hydroxyl density (0.502 mmol g−1) in water than other
oxides (surface hydroxyl density = 0.1–0.3 mmol g−1) used in
this study. This property of the FeOOH seems to have con-
tributed to its high activity in the catalytic ozonation of NB.
The important function of the surface hydroxyl groups is further
confirmed when the substitution of surface hydroxyl groups by
inorganic anions inhibited the catalytic ozonation. Moreover, the
influences of water pH and FeOOH transformation on the cat-
alytic ozonation of NB indicate that both the quantity of surface
hydroxyl groups and their charge states have relationship with
the activity of the catalyst in enhancing ozone decomposition to
generate hydroxyl radicals.
Fig. 10 illustrates the possible pathway of hydroxyl radi-
cal generation induced by the surface hydroxyl groups of the
FeOOH. Since ozone has both nucleophilic site and electrophilic
site as a dipole agent [42], the ozone molecule can be combined
with the surface hydroxyl group as its H and O are electrophilic
and nucleophilic respectively (step A)−. The combined species
further decomposes into surface HO2 (step B). The surface
HO2− r•e−acts with another O3 forming •OH and O2•− (step C).
The O2 can further reacts with O3, which finally will yield
another •OH [1]. The vacant surface Fe(III) site adsorbs another
water molecule, which subsequently dissociates into surface
hydroxyl group (step D).
As far as the protonated surface hydroxyl group is con-
cerned, its O is weaker in nucleophilicity than the O of neutral
state hydroxyl group. Therefore, the protonation of the surface
hydroxyl group will disadvantage the surface binding of ozone
(step A). On the other hand, the deprotonated surface hydroxyl
group cannot provide the electrophilic H, which would also
handicap the proposed process. Then, it will explain why the
presence of FeOOH significantly improved the ozonation of NB
only nearly neutral water pH.
In addition to the quantities and the charge states of the sur-
face hydroxyl groups, their chemical properties, such as the
Brønsted acidity and coordination states with surface metal sites
of the catalyst, should be further studied to get more detailed
information on their relationship with the activity in the catalytic
ozonation.
Further researches are still needed to find out the relation-
ship between the acidity as well as the coordination sates of the
surface hydroxyl groups and their activity in inducing hydroxyl
radical generation from aqueous ozone.
Acknowledgment
The authors thank to the financial support from National
Natural Science Foundation of China (Project No.: 50578051).
References
[1] U. von Gunten, Water Res. 37 (2003) 1443–1467.
[2] R. Andreozzi, A. Insola, V. Caprio, R. Marotta, V. Tufano, Appl. Catal. A:
Gen. 138 (1996) 75–81.
[3] R. Andreozzi, A. Insola, V. Caprio, R. Marotta, V. Tufano, Water Res. 32
(1998) 1492–1496.
[4] J. Ma, N.J.D. Graham, Water Res. 34 (2000) 3822–3828.
[5] J. Ma, M. Sui, T. Zhang, C. Guan, X. Bao, Water Res. 39 (2005) 779–786.
[6] R. Gracia, S. Cortes, J. Sarasa, P. Ormad, J.L. Ovelleiro, Water Res. 34
(2000) 1525–1532.
[7] F.J. Beltra´n, F.J. Rivas, R. Montero-de-Espinosa, Appl. Catal. B: Environ.
39 (2002) 221–231.
[8] C.N. Ni, J.N. Chen, Water Sci. Technol. 43 (2001) 213–220.
[9] M. Ernst, F. Lurot, J.C. Schrotter, Appl. Catal. B: Environ. 47 (2004) 15–25.
´
[10] B. Kasprzyk-Hordern, U. Raczyk-Stanisławiak, J. Swietlik, J. Nawrocki,
Appl. Catal. B: Environ. 62 (2006) 345–358.
[11] F. Delano¨e, B. Acedo, N. Karpel Vel Leitner, B. Legube, Appl. Catal. B:
Environ. 29 (2001) 315–325.
[12] C. Cooper, R. Burch, Water Res. 33 (1999) 3695–3700.
[13] Y. Pi, M. Ernst, J.C. Schrotter, Ozone Sci. Eng. 25 (2003) 393–397.
[14] I. Udrea, C. Bradu, Ozone Sci. Eng. 25 (2003) 335–343.
[15] F.J. Beltra´n, F.J. Rivas, R. Montero-de-Espinosa, Appl. Catal. B: Environ.
47 (2004) 101–109.
[16] F.J. Beltra´n, F.J. Rivas, R. Montero-de-Espinosa, Water Res. 39 (2005)
3553–3564.
[17] F.J. Rivas, M. Carbajo, F.J. Beltra´n, B. Acedo, O. Gimeno, Appl. Catal. B:
Environ. 62 (2006) 93–103.
[18] G.L. Elizarova, G.M. Zhidomirov, V.N. Parmon, Catal. Today 58 (2000)
71–88.
[19] S.S. Lin, M.D. Gurol, Water Sci. Technol. 34 (1996) 57–64.
[20] M.D. Gurol, S.S. Lin, Water Sci. Technol.: Water Supply 1 (2001) 131–138.
[21] R. Andreozzi, V. Caprio, R. Marotta, Water Res. 36 (2002) 2761–2768.
[22] H.N. Lim, H. Choi, T.M. Hwang, J.W. Kang, WaterRes. 36(2001)219–229.
[23] J.S. Park, H. Choi, J. Cho, Water Res. 38 (2004) 2284–2291.
[24] J. Hoigne´, H. Bader, Water Res. 17 (1983) 173–183.
[25] K. Kandori, M. Fukuoka, T. Ishikawa, J. Mater. Sci. 26 (1991) 3313–3319.
[26] H. Bader, J. Hoigne´, Water Res. 15 (1981) 449–456.

T. Zhang, J. Ma / Journal of Molecular Catalysis A: Chemical 279 (2008) 82–89
89
[27] H. Tamura, A. Tanaka, J Colloid Interface Sci. 209 (1999) 225–231.
[28] G. Newcombe, R. Hayes, M. Drikas, Colloid Surf. A 78 (1993) 65–71.
[29] J. Hoigne´, J. Bader, Ozone Sci. Eng. 16 (1994) 121–134.
[30] H.S. Park, T.M. Hwang, J.W. Kang, H.C. Choi, Water Res. 35 (2001)
2607–2614.
[31] G.V. Buxton, C.L. Greenstock, W.P. Helman, W.P. Ross, J. Phys. Chem.
Refer. Data 17 (1988) 513–886.
[32] F.J. Beltra´n, J.M. Encinar, M.A. Alonso, Ind. Eng. Chem. Res. 37 (1998)
25–31.
[34] S. Tong, W. Liu, W. Leng, Q. Zhang, Chemosphere 50 (2003) 1359–1364.
[35] D.E. Yates, T.W. Healy, J Colloid Interface Sci. 52 (1999) 222–228.
[36] M.I. Tejedor-Tejedor, M.A. Anderson, Langmuir 2 (1986) 203–210.
[37] A. Ler, R. Stanforth, Environ. Sci. Technol. 37 (2003) 2694–2700.
[38] D. Peak, R.G. Ford, D.L. Sparks, J Colloid Interface Sci. 218 (1999)
289–299.
[39] H. Wijnja, C.P. Schulthess, J Colloid Interface Sci. 229 (2000) 286–297.
[40] M. Kosmulski, J Colloid Interface Sci. 253 (2002) 77–87.
[41] Y. Joseph, W. Ranke, W. Weiss, J. Phys. Chem. B 104 (2000) 3224–3236.
[42] B. Langlais, D.A. Reckhow, D.R. Brink, Ozone in Water Treatment: Appli-
cation and Engineering, Lewis Publishers, Chelsea, Michigan, 1991.
[33] W. Stumm, Chemistry of the Solid–Water Interface, John Wiley and Sons,
New York, 1992.