Wat. Res. Vol. 32, No. 11, pp.

3223±3234, 1998
# 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(98)00118-3 0043-1354/98 $19.00 + 0.00

PHOTODEGRADATION KINETICS OF 4-NITROPHENOL IN

TiO2 SUSPENSION
DINGWANG CHEN and AJAY K. RAY**
M

Department of Chemical Engineering, National University of Singapore, 10 Kent Ridge Crescent,

Singapore 119260, Singapore

(First received October 1997; accepted February 1998)

AbstractÐA new two-phase swirl-¯ow monolithic type reactor was designed to study the kinetics of
heterogeneous photocatalytic processes on semiconductor catalysts. Most of the earlier work investi-
gated the e€ects of various parameters on the initial degradation rate (or the formation of CO2) of the
pollutants. In the present study, photomineralization of 4-nitrophenol (4-NP) sensitized by Degussa
P25 TiO2 in aqueous solution has been investigated as a function of the following di€erent experimental
parameters: initial concentration of pollutant (CS0), light intensity (Ia), partial pressure of oxygen (pO2 ),
catalyst concentration, pH, chloride ion and temperature. The kinetic expression obtained can be
described by the following formula rs=k0 exp (ÿE/RT)IabKO2 pO2 =…1 ‡ KO2 pO2 †KsCs/(1 + KsCs0). The
apparent overall activation energy for 4-NP obtained was 7.42 kJ molÿ1 by ®tting the above equation
with experimental data and the values for b, KO2 and Ks were 0.84, 9.98 atmÿ1 and 0.075 ppmÿ1 re-
spectively. The kinetic equation can predict the concentration of 4-NP during the entire degradation
process at pH value of 4.3 and TiO2 concentration of 2 g lÿ1, and is therefore, more meaningful. It can
be used in the development of large-scale photocatalytic reactor and optimization of experimental con-
ditions for any particular studies. Furthermore, the photodegradation of 4-NP on two di€erent cata-
lysts (P25 and UV100) was compared and the photonic eciency of the process has also been
investigated. # 1998 Elsevier Science Ltd. All rights reserved

Key wordsÐsemiconductor photocatalysis, advanced oxidation, water treatment, titanium dioxide,

4-nitrophenol, slurry reactor, kinetic reactor, photocatalytic degradation

INTRODUCTION The most suitable semiconductor photocatalyst for

Water puri®cation by semiconductor photocatalysis sensitizing above reaction is TiO2. Although there
has been attracting an immense research interest in are many di€erent sources of TiO2, Degussa P25
TiO2 has e€ectively become a research standard due
recent years. A growing interest in the puri®cation
to its reasonably well-de®ned nature and consider-
of water by semiconductor photocatalysis, especially
able photocatalytic activity. The adopted wave-
in the removal of toxic organic pollutants, is evi-
length of UV light for TiO2 is about 365 nm in
denced by the ever-increasing number of di€erent
order to avoid the direct photolysis of organic com-
studies in this area (Ollis et al., 1989, 1991,
pounds. Organic pollutants in above equation
Matthews, 1988; Mills and Morris, 1993; Inel and
include alkanes, haloalkanes, aliphatic alcohols, car-
Okte, 1996). The semiconductor photocatalytic pro-
boxylic acids, alkenes, polymers, surfactants, herbi-
cess is based on aqueous phase hydroxyl radical
cides, pesticides, and dyes as well as aromatics and
chemistry and couples low energy UV-A light with
haloaromatics which are commonly present in
semiconductors acting as photocatalyst. It has been industrial wastewater.
proved that this process can completely destroy It has been demonstrated (D'Oliveira et al., 1990;
many toxic organic compounds in wastewater into Al-Sayyed et al., 1991; Inel and Okte, 1996) that
carbon dioxide (Ollis et al., 1989; D'Oliveira et al., various parameters have e€ects on the degradation
1990; Okamoto et al., 1985). The overall process rate of compounds. A few degradation kinetic ex-
can be summarized by the following reaction pressions of di€erent compounds have been
equation: reported in literature (Lu et al., 1993; Chen and
Organic Pollutants ‡ O2 Chou, 1993). Unfortunately, almost all these studies
investigated the e€ects of various parameters on the
Semiconductor‡
UVÿA light;
initial degradation rate rather than the degradation
ÿ4 CO2 ‡ H2 O ‡ mineral acid …1† rate during the whole photomineralization process.
The degradation kinetic rate expressions have so far
focused on the initial disappearance rate of organic
*Author to whom all correspondence should be addressed. compounds (Inel and Okte, 1996) or the initial for-
3223
3224 Dingwang Chen and Ajay K. Ray

mation rate of CO2 (Matthews, 1988, Mills and Photoreactor and light source
Morris, 1993). Meanwhile, initial rate data are The batch photoreactor consists of two circular glass
tedious to obtain and prone to variation, this will plates that are placed between soft padding housed within
reduce the reliability of the results. For the design stainless steel and aluminum casings. The TiO2 aqueous
solution, which was circulated by a peristaltic pump, was
of large-scale reactor, these kinetic expressions are introduced tangentially between the two glass plates, and
not always meaningful, and therefore, not useful. exited from the center of the top plate. The tangential
In this study 4-nitrophenol (4-NP) was chosen as introduction of liquid created a swirl-¯ow, thereby making
model compound. It is one of the 114 organic pol- the liquid solution well mixed. The lamp (Philips BPR
125 W high-pressure mercury vapour) was placed about
lutants listed by EPA in U.S.A. and its maximum 10 cm underneath the bottom glass plate on a holder that
allowed concentration is 20 ppb. So far only a few could be moved to create a di€erent angle of incidence of
papers (Augugliaro et al., 1991, 1993; Palmisano et light on the bottom glass plate. Primary radiation of this
al., 1989; Dieckmann et al., 1992; Dieckmann and lamp is emitted at 365 nm and the incident light intensity
is 213 W mÿ2. The lamp and reactor were placed inside a
Gray, 1996) have reported the photodegradation of
wooden box painted black so that no stray light can enter
4-NP. In this work, e€ects of initial pollutant con- the reactor. The lamp was constantly cooled by com-
centration, light intensity, partial pressure of oxy- pressed air to keep the temperature down and protect the
gen, TiO2 catalyst concentration, pH value of lamp from overheating. Provision was made for placement
suspension, temperature and chloride ion on the of several metal screens of di€erent mesh size between the
lamp and bottom glass plate to obtain variation in light
degradation rates of 4-NP have been studied. A intensity.
kinetic expression, which can be used in the devel-
opment of large-scale photocatalytic reactor and Experimental procedures
optimization of experimental conditions, has been The reaction solution was made by adding 145 ml of
obtained. Furthermore, the photodegradation of 4- Milli-Q water at natural pH containing the target concen-
NP on two di€erent catalysts (P25 and UV100) was tration of 4-NP into the beaker with a water jacket. Then
0.29 g (0.2%) powder photocatalyst was dispersed into
compared and the photonic eciency of the process above solution and in some cases, hydrochloric acid or
has also been investigated. nitric acid or sodium hydroxide was used to adjust the sol-
ution to the desired pH. The mixture was magnetically
stirred and circulated in the system by peristaltic pump for
about 20 min. About 5 ml sample was taken to determine
EXPERIMENTAL DETAILS its initial concentration prior to illumination. The tem-
perature of the photodegradation system was adjusted by
Materials a water bath in which heating or cooling water was recir-
Two di€erent brand TiO2 catalysts, Degussa P25 pro- culated through the jacket of the beaker.
vided by Degussa Company and Hombikat UV 100 pro-
Analyses
vided by Sachtleben Chemie GmbH (Germany), were used
throughout this work. The physiochemical properties of In order to remove particulate before analyses, all water
these two titanium dioxide catalysts were listed in Table 1. samples were ®ltered by Millex-HA ®lter (Millipore,
Hydrochloric acid (37%) and nitric acid (65% by weight) 0.45 mm). A SHIMADZU TOC-5000A analyzer with an
were obtained from Merck chemicals, 4-nitrophenol (4- ASI-5000 autosampler analyzed the total organic carbon
NP; 98 + %) from BDH chemicals, sodium hydroxide (TOC) in samples. Analysis of 4-NP was carried out by
(98 + %) from Baker chemicals. All chemicals were used using a Shimadzu UV 3101 PC spectrophotometer.
as received. Water used to make up solutions in this work
was Milli-Q water.
REACTION MECHANISM AND RATE EXPRESSIONS

It has been well demonstrated that when TiO2 is

Table 1. Physicochemical properties of the two di€erent brand of illuminated by light (l < 380 nm) in the presence of
titanium dioxide
water containing dissolved oxygen and organic
Degussa P-25 TiO2 Hombikat UV100 compounds, photodegradation of the organic com-
Property TiO2
pounds will occur. Many organic compounds are
Physical state White powder White powder known to be degraded to CO2 under these con-
Composition (%) 180% anatase, 120% 100% anatase
rutile
ditions. The primary step in photodegradation is
Density (g cmÿ3) 3.8 3.9 certainly the generation of electrons and holes
BET surface area 155 >250 within the TiO2 particle.
(m2 gÿ1)
Average primary 130 <10 h, l<380 nm
particle size (nm) TiO2 ÿÿÿÿÿ ÿ TiO2 …eÿ ‡ h‡ †
ÿ4 …2†
pH in aqueous 3±4 16
solution
Contents
The majority of these recombine with the liberation
SiO2 (%) <0.2 ÿ of heat. However, in the presence of dissolved oxy-
Al2O3 (%) <0.3 ÿ gen and an electron donor, OH
radical is formed
Fe2O3 (%) <0.01 ÿ
TiO2 (%) >97 >99 by the reaction between the valence band holes
HCl (%) <0.3 ÿ (h+) and the TiO2 surface active OH
group or
Porosity nonporous porous (mesopores, H2O.
about 5.6 nm in
diameter) h‡ ‡ OHÿ ad ÿ4OHad


…3†
 Photodegradation of 4-NP in TiO2 suspension 3225

on the TiO2 surface inhibits the recombination

h‡ ‡ H2 Oad ÿ4OH
ad ‡ H‡ …4†
process by trapping the electrons. In addition,
The photogenerated conduction band electron (eÿ) the generation of hole±electron pairs is a€ected
is trapped by the dissolved oxygen to form a super- by the light intensity. Therefore, equation 7 can
oxide ion Oÿ2. be rewritten as
eÿ ‡ O2 ÿ4O
ÿ
2 …5† rs ˆ kP I ba yO2 ys , …8†
4-NP will be attacked by hydroxyl radicals formed where
in above equations and generates organic radicals
KO2 PO2
or some other intermediates. NOÿ 2 was found and yO2 ˆ …9†
1 ‡ KO2 PO2
oxidized to NOÿ 3 (Augugliaro et al., 1991).
Eventually all the parent compounds and intermedi-
Ks Cs
ates will be oxidized into CO2 with an overall reac- ys ˆ X …10†
tion stoichiometry: 1 ‡ Ks Cs ‡ Ki Ci
i
C6 H4 OHNO2 ‡ 7O2 ÿ46CO2 ‡ 2H2 O ‡ HNO3
yO2 and ys, are the fractions of the sites covered
…6†
by oxygen and 4-NP. Ia is light intensity, sub-
In the photomineralization of 4-NP sensitized by scripts s and i are 4-NP and intermediates, KO2 ,
TiO2, it is postulated that oxygen, 4-NP and its in- Ks, and Ki are equilibrium adsorption constants
termediates are adsorbed at di€erent sites on the for O2, s and i respectively. Augugliaro et al.
surface of TiO2 (Augugliaro et al., 1993) and the (1991) made the following assumption:
rate-determining step is the reaction between the X
primary oxidizing species OH
, and the adsorbed 4- Ks Cs ‡ Ki Ci ˆ Ks Cs0 , …11†
i
NP. In terms of Langmuir±Hinshelwood kinetics
the second-order decomposition rate of 4-NP may where Cs0 is the initial concentration of 4-NP.
be expressed as That is, the adsorption constants for the de-
composition intermediates were same as that of
rs ˆ k 0 yOH
ys , …7†
the parent compound. Substituting equations 9±
The number of electron±hole pairs that escape 11 into equation 8 and rewriting the apparent
from recombination signi®cantly in¯uences con- rate constant as a function of temperature, more
centration of hydroxyl radical. Oxygen adsorbed complete photodegradation kinetics is obtained

Fig. 1. 4-NP, intermediates and pH pro®les during degradation. [CS0=20 ppm; pH = normal; TiO2
dosage = 2 g lÿ1; ¯owrate = 250 ml minÿ1 and O2 saturated].
3226 Dingwang Chen and Ajay K. Ray

with the following expression. NP, a series of UV spectra were recorded in the
  200±500 nm range with the increasing illumination
dns E KO2 pO2
rs ˆ ÿ ˆ k0 exp ÿ Ib time. Typical spectra are shown in Fig. 2. The
S dt RT a …1 ‡ KO2 pO2 †
characteristic absorption of 4-NP almost disap-
Ks Cs peared after 95 min of irradiation. No peak for in-
 …12† termediates was detected due to the low
1 ‡ Ks Cs0
concentration of intermediates.
Here, the photodegradation rate (g mÿ2 sÿ1) is
de®ned as the mole (ns) reduction of pollutant
per irradiated reactor window area (S). E€ect of mass of TiO2
The variation in degradation rate constant of 4-
RESULTS AND DISCUSSIONS
NP against the concentration of TiO2 was deter-
mined (Fig. 3) over the [TiO2] range 0.2±4 g lÿ1. As
As shown in Fig. 1, a very small decrease in the expected, the rate constant was found to increase
concentration of 4-NP was observed when only with the increasing concentration of TiO2 and
UV-A is present or when only TiO2 is present. approaching a limiting value at high concentration.
However, 4-NP could be degraded quickly by illu- The limiting value (2 g lÿ1 for 4-NP) mainly results
minated TiO2. The pH value of the solutions from following two factors: (a) aggregation of TiO2
decreased as the reaction proceeded due to the for- particles at high concentrations, causing a decrease
mation of HNO3 or HNO2. Two aromatic inter- in the number of surface active sites, and (b)
mediates (4-nitrocatechol and hydroquinone) were increase in opacity and light scattering of TiO2 par-
identi®ed during the degradation of 4-NP ticles at high concentration leading to a decrease in
(Dieckmann and Gray, 1996). In this work, some the passage of irradiation through the sample.
intermediates during the reaction were also found Similar results have been reported in previous stu-
(Fig. 1) and the total maximum concentration of dies (Al-Sayyed et al., 1991; Mills and Morris,
these intermediates was less than 8% of the initial 1993; Terzian and Serpone, 1995) and the amount
concentration of 4-NP. However, in this study no of catalyst at the beginning of the plateau was
attempt was made to identify these intermediates. between 0.5 and 2 g lÿ1 for di€erent photosystem.
As described in the experimental section, an UV When the experimental data was ®tted with a
absorption experiment of the reaction solution was Langmuir-type expression: kp=kp0 Kapp[TiO2]/
carried out after the ®ltration of photocatalyst. In (1 + Kapp[TiO2]), values of kp0 and Kapp were
order to monitor the gradual disappearance of 4- obtained from the linear transform of the

Fig. 2. Typical UV spectrum pro®les during degradation of 4-NP on P25 catalyst.

 Photodegradation of 4-NP in TiO2 suspension 3227

Fig. 3. E€ect of TiO2 dosage on the degradation rate. [CS0=20 ppm; pH = natural; T = 303 K; ¯owra-
te = 250 ml minÿ1 and O2 saturated].

expression as kp0=3.33  10ÿ5 m sÿ1 and Kapp= is necessary to investigate its dependence. The e€ect
2.48 l gÿ1. of initial concentration on the photocatalytic
degradation of 4-NP is shown in Fig. 4. Di€erent
E€ect of initial concentration concentration pro®les were observed during the
As the e€ect of pollutant concentration is of im- degradation for di€erent initial concentrations.
portance in any process of water treatment, it Furthermore, degradation rates at same concen-

Fig. 4. E€ect of initial concentration on the degradation rate. [pH = 4.3; TiO2 dosage = 2 g lÿ1;
T = 303 K; ¯owrate = 250 ml minÿ1 and O2 saturated].
3228 Dingwang Chen and Ajay K. Ray

tration with di€erent initial concentrations were E€ect of light intensity

di€erent. This is quite di€erent from ordinary ther- Previous studies (D'Oliveira et al., 1990; Ollis et
mal reaction. But, all the concentration pro®les al., 1991; Al-Sayyed et al., 1991) on photodegrada-
could be correlated by the following exponential tion rate vs illumination intensity indicated that the
function with good agreement. reaction rate increases with the square root of
Cs ˆ Cs0 exp…ÿks t† …13† intensity at high intensity level when mass transfer
limit is not reached. At suciently low levels of illu-
The apparent rate constant ks (l minÿ1) in above mination (catalyst dependent), on the other hand,
equation decreased with the increasing of initial degradation rate is of ®rst-order in intensity.
concentration of 4-NP when other parameters are Increased illumination results in an increase in volu-
kept unchanged. Therefore, the degradation rate metric reaction rate, until the mass transfer limit is
was pseudo-®rst-order with respect to the concen- encountered. The transition points between these
tration within the experimental range. regimes, however, will vary with the photosystem.
In recent years the Langmuir±Hinshelwood rate In the degradation of 3-CP in TiO2 aqueous suspen-
expression has been used successfully for hetero- sions, the initial degradation rate was reported to
geneous photocatalytic degradation to describe the be proportional to radiant ¯ux for values smaller
relationship between initial degradation rate and in- than 20 mW cmÿ2 while above this value it was pro-
itial concentration (Al-Sayyed et al., 1991; Lu et al., portional to its square root (D'Oliveira et al., 1990).
1993). A linear expression can be conveniently The initial degradation rate of 4-CP against radiant
obtained by plotting reciprocal initial rate against ¯ux was also investigated in the range of 2±
reciprocal initial concentration 50 mW cmÿ2 (Al-Sayyed et al., 1991), where similar
1 1 1 1 rate shift from ®rst-order to halforder in intensity
ˆ ‡ …14† was observed. Other investigations on the photode-
rs0 k kKs Cs0
gradation of phenol (Okamoto et al., 1985), isopro-
A reasonable ®t of above equation was obtained panol, pesticides, xylenols and malonic acid (Inel
with values of 0.072 ppmÿ1, 9.55  10ÿ4 g mÿ2 sÿ1 and Okte, 1996) also reported similar results.
for Ks and k respectively. This indicates that the However, no report on 4-NP was found in litera-
degradation of 4-NP occurred mainly on the surface ture. Figure 5 shows the result of the e€ect of light
of TiO2. However, Turchi and Ollis (1990) investi- intensity on the degradation rate constant of 4-NP
gated the photodegradation of organic pollutants in in this study. Following equation correlated the
illuminated TiO2 slurries and found it plausible that e€ect
the reaction occurs not only on the TiO2 surface
but also in solution. kAI ba …15†

Fig. 5. E€ect of light intensity on the degradation rate. [CS0=20 ppm; pH = 4.5; TiO2 dosage =2 g lÿ1;
T = 303 K; ¯owrate = 250 ml minÿ1 and O2 saturated].
 Photodegradation of 4-NP in TiO2 suspension 3229

Fig. 6. E€ect of temperature on the degradation rate. [CS0=20 ppm; pH = 4.3; TiO2 dosage = 2 g lÿ1;
¯owrate = 250 ml minÿ1 and O2 saturated].

with the b value 0.84. The value is quite close to perature of 4-NP is shown for the temperature
reported value of 0.74 in the photodegradation of range from 288 to 323 K in Fig. 6. It was observed
4-CP (Mills and Morris, 1993). It indicated that the that the rate constants increased with increasing
degradation rate was neither catalyst dependent nor temperature. The Arrhenius plot yielded a straight
mass transfer dominated. Several studies line from which the overall apparent activation
(D'Oliveira et al., 1990; Ollis et al., 1991; Al-Sayyed energy obtained was 7.42 kJ molÿ1 (Fig. 6).
et al., 1991; Inel and Okte, 1996) attributed the Ia1.0 Obviously, this value was smaller than that of
to Ia0.5 rate transition to the recombination of ordinary thermal reactions and the photocatalytic
photogenerated electron±hole pairs at high light reaction was less temperature dependent. Although
intensity. This is obviously detrimental to the no activation energy of 4-NP photodegradation was
photocatalytic process as the quantum eciency reported in literature, the value is comparable with
decreases. Recombination process may be slowed those listed in Table 2. The value is also quite close
possibly by the addition of better electron accep- to that for a hydroxyl radical reaction (Matthews,
tors, such as H2O2, Cu+ and Ag+ (Ollis et al., 1987), suggesting that the photodegradation of 4-
1991). NP was governed by hydroxyl radical reaction.
However, for TiO2 photocatalysis, irradiation is the
E€ect of temperature primary source of electron±hole pair generation at
The overall process of semiconductor photocata- ambient temperature as the band gap energy is too
lysis is usually not very temperature sensitive. The high to be overcome by thermal activation.
dependence of degradation rate constant on tem- Therefore, the increase in rate constant was most

Table 2. Comparison of activation energy, E, obtained in this study with those in literature
Reactant TiO2 catalyst E (kJmolÿ1) T (K) Investigator(s)

4-CP deposited 20.6 283±333 Hofstadler et al., 1994

4-CP suspension 16 288±328 Mills and Morris, 1993
4-CP suspension 5.5 283±318 Al-Sayyed et al., 1991
Methyl orange suspension 18 300±318 Chen and Chou, 1993
Dichlorvos deposited 28.4 293±313 Lu et al., 1993
Malonic acid suspension 9.99 294±324 Inel and Okte, 1996
Phenol suspension 10 293±323 Okamoto et al., 1985
Salicylic acid deposited 11 298±318 Matthews, 1987
Xylenols suspension 8.8 279±333 Terzian and Serpone, 1995
4-NP suspension 7.42 288±323 this work
3230 Dingwang Chen and Ajay K. Ray

likely due to the increasing collision frequency of TiO2 is at pH between 5.6 and 6.4 (Lu et al., 1993;
molecules in solution that increases with increasing Terzian and Serpone, 1995). Hence, at more acidic
temperature. pH values, the catalyst surface is positively charged,
while at pH values above 5.6, it is negatively
E€ect of dissolved oxygen concentration charged
Photomineralization of organic compounds will TiIV ÿ OH ‡ H ‡ ÿ4TiIV ÿ OH ‡ …17†
2
cease unless sucient oxygen is present in the sol-
ution. The partial pressure of oxygen was adjusted TiIV ÿ OH ‡ OH ÿ ÿ4TiIV ÿ Oÿ ‡ H2 O …18†
by mixing the oxygen stream with a nitrogen stream
while the total ¯owrate of gas was maintained at a Therefore, pH value will have signi®cant e€ect on
constant value. the adsorption±desorption properties at the cata-
Figure 7 shows that the degradation rate constant lyst's surface.
of 4-NP increased with increasing oxygen partial The e€ect of pH on the degradation of 4-NP is
pressure and it has reached about 70% of its maxi- shown in Fig. 8. The result indicates that pH value
mum value when the partial pressure of oxygen was has a signi®cant e€ect and at both low and high pH
0.2 atm. This indicates that in commercial appli- values photodegradation rates are quite slow. The
cation it is possible to replace pure oxygen with air. best pH value for the degradation is near the pzc
The e€ect of oxygen partial pressure on the degra- point of TiO2. This probably explains that the e€ect
dation of 4-NP could be well described by noncom- of pH on degradation rate is perhaps due to its
petitive Langmuir kinetic equation as follows. e€ect on TiO2 particle itself This includes the
kO2 pO2 charge separation on the particles. The high OHÿ
kp A …16† ion content of the system enhances the e±h separ-
1 ‡ kO2 pO2
ation. However, the photogenerated CO2 will be
Straight line was obtained with good correlation trapped in the solution, and bicarbonate, carbonate
coecient by plotting pO2 /k against pO2 . The main are formed in alkaline system. Both bicarbonate
in¯uence of the dissolved oxygen is to enhance the and carbonate are ecient scavengers of hydroxyl
separation of photogenerated electron±hole pairs, radicals due to their very high rate constants with
hereby increasing hydroxyl radical concentration. the hydroxyl radicals (k = 3.9  108 Mÿ1 sÿ1 for car-
bonate and k = 8.5  106 Mÿ1 sÿ1 for bicarbonate)
E€ect of pH value and chloride ion (Buxton et al., 1988). On the other hand, according
The pH value is an important parameter in to equation 3 low OHÿ in acidic system hinders the
photodegradation that takes place on the surface of formation of hydroxyl radicals and subsequently
photocatalyst. The point of zero charge (pzc) for reduces the degradation rate. Maximum degra-

Fig. 7. E€ect of dissolved oxygen on the degradation rate. [CS0=20 ppm; pH = 4.5; TiO2
dosage = 2 g lÿ1; T = 303 K; ¯owrate = 250 ml minÿ1; O2 saturated].
 Photodegradation of 4-NP in TiO2 suspension 3231

Fig. 8. E€ect of pH and chloride ion on the degradation rate. [CS0=20 ppm; TiO2 dosage = 2.0 g lÿ1;
T = 303 K; ¯owrate = 250 ml minÿ1 and O2 saturated].

dation rate of 4-NP at pH 4.5 was reported tory agreement.

(Palmisano et al., 1989), but the degradation rate  
increased a little bit with the increasing pH value dns ÿ3 7415:3
rs ˆ ÿ ˆ 1:87  10 exp ÿ
beyond 10. Augugliaro et al. (1993) investigated the S dt RT
e€ect of pH in the range of 3±11 on the degra-
9:98pO2 0:075Cs
dation of 4-NP in TiO2 suspension and found that  I 0:84
a

…19†
the degradation rate decreased slightly with the 1 ‡ 9:98pO2 1 ‡ 0:075Cs0
increasing pH value. In order to investigate the
chloride e€ect on the degradation, HCl and HNO3 Above equation indicates that equilibrium adsorp-
were used to adjust the pH value of solution. The tion constants for oxygen (KO2 ) and 4-NP (Ks) on
result in Fig. 8 shows that chloride had substantial TiO2 surface are 9.98 atmÿ1 and 0.075 ppmÿ1 re-
inhibitory e€ect on the degradation of 4-NP, par- spectively. Literature reported (Augugliaro et al.,
ticularly at low pH value. One possible explanation 1991) values of KO2 and Ks for 4-NP-TiO2 sus-
is that chloride ion can competitively adsorb on the pension are 1.82 atmÿ1 and 0.064 ppmÿ1 respect-
surface of TiO2 with big adsorption constant. ively.
Similar result was reported elsewhere (Augugliaro
Comparison of the two di€erent TiO2 catalysts
et al., 1991).
Although Degussa P25 TiO2 catalyst has been
extensively used in most of the photocatalytic
Degradation kinetics of 4-NP processes, recently it was reported that a novel
In literature, no perfect degradation kinetics has TiO2 photocatalyst (Hombikat UV100) has a
been reported that could predict the concentration good performance in the photodegradation of
pro®les during photoreaction. A perfect degradation some model compounds. The physiochemical data
kinetic expression is very useful in the design of of these two titanium dioxide catalysts were listed
large-scale photoreactor. In this study, a series of in Table 1. In the present study, the photominera-
experiment has been carried out for 4-NP at natural lization of 4-NP has been carried out under
pH value of 4.3 and constant TiO2 concentration of otherwise identical conditions using both P25 and
2 g lÿ1. While other parameters are varied as fol- UV100 photocatalyst. The disappearance of 4-NP
lows: T = 288±323 K, I = 1.5±18 mW cmÿ2, on both catalysts is illustrated in Fig. 9. The
CS0=10±120 ppm, pO2 =0.05±1.0 atm. The follow- photodegradation is via apparent ®rst order
ing kinetic expression was obtained with satisfac- kinetics: kp=1.54  10ÿ3 m sÿ1 and kp=6.03 
3232 Dingwang Chen and Ajay K. Ray

Fig. 9. Comparison of degradation rate for 4-NP on Hombikat UV100 and Degussa P25 catalysts.
[pH = natural; TiO2 dosage = 2 g lÿ1; T = 303 K; ¯owrate = 250 ml minÿ1 and O2 saturated].

10ÿ4 m sÿ1 for Degussa P25 and Hombikat UV100 Eciency of the photocatalytic process
respectively. Obviously, P25 is more active than
UV100 in the photomineralization of 4-NP under So far, many photocatalytic processes have been
given experimental conditions. investigated in laboratory scale. It is obviously

Fig. 10. Average photonic eciency for the degradation of 4-NP on P25 and UV100 catalysts.
[CS0=20 ppm; pH = natural; TiO2 dosage = 2 g lÿ1; T = 303 K; ¯owrate = 250 ml minÿ1 and O2 satu-
rated].
 Photodegradation of 4-NP in TiO2 suspension 3233

necessary to compare the photocatalytic eciencies behavior. The in¯uence of pH was remarkable for
of di€erent processes. In homogeneous photochem- 4-NP and photodegradation was not suitable at
istry, quantum yield has been extensively used as it both low and high pH values. Chloride ion had a
provides a means of comparing two or more photo- negative e€ect on the degradation. The overall
chemical events. By contrast, a similar description apparent activation energy was 7.42 kJ molÿ1,
in heterogeneous photochemistry is further compli- implying that the degradation of 4-NP was less tem-
cated by many fundamental unknown quantities. perature sensitive. Equilibrium adsorption constants
In this work, we use the relative photonic e- for oxygen and 4-NP on TiO2 P25 obtained in this
ciency to describe the eciency of the photocataly- work were 9.98 atmÿ1 and 0.075 ppmÿ1 respectively.
tic process. The relative photonic eciency is Although P25 is more active for the photodegra-
de®ned as the number of molecules converted dation of 4-NP than UV100, at the current state of
relative to the total number of photons incident on development the degradation eciency on both two
the reactor window, i.e. catalysts is very low. Therefore prior to commercia-
lizing this process, a signi®cant progress in the
rate of reaction…mol sÿ1 †
j…t† ˆ improvement of the photo-eciency must be made
rate of photon incident on window…mol sÿ1 † by modifying currently used photocatalysts or
…20† developing a novel photocatalyst.
For our system (batch reactor):
rs
j…t† ˆ …21† REFERENCES
f
Al-Sayyed G., D'Oliveira J. C. and Pichat P. (1991)
where f is light intensity (mol photon mÿ2 sÿ1). Semiconductor-sensitized photodegradation of 4-cholo-
When the reaction is ®rst order, i.e.: phenol in water. J. Photochem. Photobiol. A Chem. 58,
99±114.
C…t† ˆ C…0† exp…ÿkTOC t† …22† Augugliaro V., Lopez-Munoz M. J., Palmisano L. and
Soria J. (1993) In¯uence of pH on the degradation kin-
The relative photonic eciency is etics of nitrophenol isomers in a heterogeneous photoca-
talytic system. Appl. Catal. A Gen. 101, 7±13.
VR C…0† ‰1 ÿ exp…ÿkTOC t†Š Augugliaro V., Palmisano L., Schiavello M., Sclafani A.,
 …t† ˆ
j , …23†
fSt Marchese L., Martra G. and Miano F. (1991)
Photocatalytic degradation of nitrophenols in aqueous
where VR is the volume of the reaction solution and titanium dioxide dispersion. Appl. Catal. 69, 323±340.
 …t† is the average value within reaction time t. In
j Buxton G. V., Greenstock C. L., Helman W. P. and Ross
above equations kTOC is the degradation rate con- A. B. (1988) Critical review of rate constants for reac-
tions of hydrated electrons, hydrogen atoms and hy-
stant measured in terms of the TOC reduction. C is
droxyl radicals in aqueous solution. J. Phys. Chem. Ref.
the concentration of carbon (mol lÿ1) and we Data 17, 513.
assume that one photon required per carbon atom. Chen L. C. and Chou T. C. (1993) Kinetics of photodeco-
Typical values of the relative photonic eciency are lorization of methyl orange using titanium dioxide as
indicated in Fig. 10 for the degradation of 4-NP on catalyst. Ind. Eng. Chem. Res. 32, 1520±1527.
Dieckmann M. S. and Gray K. A. (1996) A comparison
both P25 and UV100 photocatalysts. The reasons of the degradation of 4-nitrophenol via direct and sensi-
for the very low relative photonic eciencies are tized photocatalysis in TiO2 slurries. Water Res. 30,
mainly due to the recombination of photogenerated 1169±1183.
electron±hole pairs and the light absorption by the Dieckmann M. S., Gray K. A. and Kamat P. V. (1992)
Photocatalyzed degradation of adsorbed nitrophenolic
reaction solution. equation 23 can also be used to
compounds on semiconductor surfaces. Water Sci.
compare the degradation of di€erent compounds on Technol. 25, 277±279.
the photocatalyst. D'Oliveira J. C., Al-Sayyed G. and Pichat P. (1990)
Photodegradation of 2- and 3-chlorophenol in TiO2
aqueous suspensions. Environ. Sci. Technol. 24, 990±
996.
CONCLUSIONS
Hofstadler K., Bauer R., Novalic S. and Heisler G. (1994)
New reactor design for photocatalytic wastewater treat-
In this study, a meaningful perfect kinetic ex- ment with TiO2 immobilized on fused-silica glass ®bers:
pression for the degradation of 4-NP was obtained. photomineralization of 4-chlorophenol. Environ. Sci.
equation 19 can predict the concentration of 4-NP Technol. 28, 670±674.
during the photodegradation in addition to the in- Inel Y. and Okte A. (1996) Photocatalytic degradation of
itial degradation rate, and is therefore, more reliable malonic acid in aqueous suspensions of TiO2: an initial
kinetic investigation of CO2 photogeneration. J.
because it was obtained under a variety of di€erent Photochem. Photobiol. A Chem. 96, 175±180.
experimental conditions. The degradation rate of 4- Lu M. C., Roam G. D., Chen J. N. and Huang C.
NP was found to be pseudo ®rst-order with respect P. (1993) Factors a€ecting the photocatalytic degra-
to its concentration within the experimental concen- dation of dichlorvos over titanium dioxide supported on
glass. J. Photochem. Photobiol. A Chem. 76, 103±110.
tration range. The apparent rate constant decreased Matthews R. W. (1987) Photooxidation of organic impuri-
with the increasing initial concentration. It indicates ties in water using thin ®lms of titanium dioxide. J.
that the photodegradation of 4-NP has a saturate Phys. Chem. 91, 3328±3333.
3234 Dingwang Chen and Ajay K. Ray

Matthews R. W. (1988) Kinetics of photocatalytic oxi- Ollis D. F., Pelizzetti E. and Serpone N. (1991)
dation of organic solutes over titanium dioxide. J. Destruction of water contaminants. Environ. Sci.
Catal. 111, 264±272. Technol. 25, 1523±1529.
Mills A. and Morris S. (1993) Photomineralization of 4- Palmisano L., Augugliaro V., Schiavello M. and Sclafani
chlorophenol sensitized by titanium dioxide: A study of A. (1989) In¯uence of acid±base properties on photoca-
the initial kinetics of carbon dioxide photogeneration. J. talytic and photochemical processes. J. Mol. Catal. 56,
Photochem. Photobiol. A Chem. 71, 75±83. 284±295.
Okamoto K., Yamamoto Y., Tanaka H. and Itaya Terzian R. and Serpone N. (1995) Heterogeneous photoca-
A. (1985) Kinetics of heterogeneous photocatalytic de- talyzed oxidation of creosote components: mineraliz-
composition of phenol over anatase TiO2 powder. Bull. ation of xylenols by illuminated TiO2 in oxygenated
Chem. Soc. Jpn. 58, 2023±2028. aqueous media. J. Photochem. Photobiol. A. Chem. 89,
Ollis D. F., Pelizzetti E. and Serpone N. (1989) 163±175.
Heterogeneous photocatalysis in the environment: appli- Turchi C. S. and Ollis D. F. (1990) Photocatalytic degra-
cation to water puri®cation. In Photocatalysis: dation of organic water contaminants: mechanisms invol-
Fundamentals and Applications, ed. N. Serpone and E. ving hydroxyl radical attack. J. Catal. 122, 178±192.
Pelizzetti. Wiley Interscience, New York, pp. 603±637.