Water detoxification: Photocatalytic decomposition of phenol on Au/TiO2

Article abstract of DOI:10.1007/s007060170044

Photodeposition of gold on TiO₂ was conducted in situ in the photoreactor used for the reaction of phenol photodestruction. The Au/TiO₂ system shows better photocatalytic properties in the destruction of phenol contaminants in water than pure TiO₂. The gold cocatalyst prevents also the decrease of the rate of the photocatalytic reaction at higher phenol concentration observed for pure TiO₂. Hydroquinone, p-benzoquinone, and catechol are main intermediates of the photoreaction, which finally leads to total phenol mineralization.

Full text of DOI:10.1007/s007060170044

Monatshefte f uÈ r Chemie 132, 1037±1045 (2001)
Water Detoxi®cation: Photocatalytic
Decomposition of Phenol on Au/TiO2
Anna Dobosz and Andrzej Sobczy n skiÃ
Faculty of Commodity Science, Pozna n University of Economics, PL-60967 Pozna n , Poland
Summary. Photodeposition of gold on TiO2 was conducted in situ in the photoreactor used for the
reaction of phenol photodestruction. The Au/TiO2 system shows better photocatalytic properties in
the destruction of phenol contaminants in water than pure TiO2. The gold cocatalyst prevents also the
decrease of the rate of the photocatalytic reaction at higher phenol concentration observed for pure
TiO2. Hydroquinone, p-benzoquinone, and catechol are main intermediates of the photoreaction,
which ®nally leads to total phenol mineralization.
Keywords. Gold; Titania; Phenol; Photodegradation.
Introduction
Photocatalytic water puri®cation on semiconductors is one of the newer AOP
(
[
Advanced Oxidation Processes) techniques, developed during the last two decades
1±9]. The process, which takes place on a surface of irradiated semiconductors,
leads to water puri®cation from organic substances. It is similar to those described
earlier for semiconductor catalyzed hydrogen photogeneration from water in the
presence of sacri®cial agents [10±12]. In both cases, photogenerated holes react
Á
with hydroxyl groups from water giving hydroxyl radicals. The generated OH
radicals react further with organic matter dissolved in water and cause its oxidation,
‡
®nally to CO and H O. Photogenerated electrons in deaerated water reduce H O
2 2 3
causing hydrogen evolution. In aerated water, the electrons reduce dissolved
oxygen.
Hydrogen evolution is not catalyzed well by the titanium dioxide surface.
Therefore, it was necessary to modify the surface by introducing a cocatalyst
‡
improving H O reduction, recombination of the reduced hydrogen atoms to H ,
3
2
and gaseous product evolution. Besides platinum [13±17], also other cocatalysts like
palladium [18], ruthenium [19], copper [20], nickel [21, 22], ruthenium dioxide
[23], molybdenum, and tungsten sul®des [24, 25] have been applied. From the
electrochemical point of view, islands of the noble metals or other substances on the
TiO surface act as microelectrodes (microcathodes) on which reduction processes
2
occur [26].
à Corresponding author. E-mail: biuropromocji@novci1.poznan.pl

1038
A. Dobosz and A. Sobczy n ski
Besides, the bare titania surface catalyzes the reduction of oxygen dissolved
in water. Therefore, photooxidation (mineralization) of organic water pollutants
proceeds relatively ef®ciently on pure TiO particles [1±9, 12]. However, it has been
2
shown [27] that introduction of a Pt cocatalyst enhancees the activity of titania in
photoassisted oxidation of hydrocarbons dissolved in water. Based on theoretical
considerations, Gerischer and Heller [28] suggested that the quantum ef®ciency of
oxidation of organic compounds in water on illuminated n-TiO was limited by
2
the O reduction rate. According to these authors, the semiconductor-like n-TiO
2
2
does not possess a suf®cient density of shallow, near-surface electron traps to assist
in the O reduction process. It causes, in turn, accumulation of electrons on the
2
semiconductor and acceleration of the rate of radiationless electron-hole recom-
bination, a process which leads to a loss of the semiconductor photoactivity. To
improve the ef®ciency of the photoreactions, they have suggested application of
group VIII metals as cocatalysts. The authors' predictions have been proven
0
experimentally [29]. It has been found that incorporation of Pd on titania particles
promotes the electron transfer to O dissolved in water, thus eliminating negative
2
charge on TiO and increasing the quantum ef®ciency of the photoassisted oxidation
2
of 2,2-dichloropropionate. An increased rate of photodegradation of organic
0
compounds like 1,4-dichlorobenzene and salicylic acid upon incorporation of Ag ,
0
0
Au , and Pd onto the surface of TiO particles has also been observed [30±32].
2
Recently, papers by Chen et al. [33±35] concerning the photocatalytic properties of
0
0
Pt - and Pd -loaded titania particles have discussed kinetic processes and possible
photocatalytic mechanisms which take place on the metallized surface of TiO2.
In this paper, the photoactivity of Au/TiO in phenol destruction is reported. The
2
reagent was prepared by in situ metal photodeposition in the reaction slurry.
Results and Discussion
As mentioned above, metallic gold was photodeposited on the surface of TiO2
in situ in the reaction vessel under a neutral atmosphere. For that purpose, the slurry
of TiO in an aqueous phenol solution to which an appropriate amount of HAuCl Á
2
4
3
H O had been previously added was purged with argon in order to remove
2
dissolved oxygen. Next, the mixture was illuminated with a 180 W medium pressure
Hg lamp for 30 min. It had been checked before that this illumination time was
suf®cient for the total reduction of metals like Pt [16], Ru [19], or Cu [20] on the
surface of TiO . During the metal reduction the slurry was mixed by an argon
2
stream applied at the bottom and ¯owing through a draft tube placed in the centre of
the photoreactor. The photocatalytic reaction was conducted in a non-continuous
mode (batch reactor).
The catalytic effect of Au on the photoactivity of TiO has been studied
2
previously by Gao et al. [30] in the reactions of 1,4-dichlorobenzene (DCB) and
salicylic acid photodecomposition. The novelty of the present studies lies in the
preparation of the Au/TiO system. Here the metal was photodeposited on TiO
2
2
in situ, i.e. in the reaction environment. The method has been described previously
for the preparation of such photocatalysts as Pt/TiO , Ru/TiO , and Cu/TiO used
2
2
2
in systems for hydrogen photogeneration from water in the presence of methanol
as a sacri®cial agent [16, 19, 20]. The method allows to omit steps like ®ltration/

Photocatalytic Decomposition of Phenol on Au/TiO2
1039
Fig. 1. TEM micrographs of pure TiO2 (A) and TiO2 doped with 0.05 (B), 1.0 (C), and 2.0 (D)
weight percent of gold
centrifugation and drying which can cause some changes of the photocatalytic
activity of Me/TiO . This way, photooxidation of phenol proceeds on a freshly
prepared active surface.
2
The metal photoreduction and deposition on the surface of titania was monitored
by transmission electron microscopy (TEM). The TEM images of the samples
containing 0.05, 1.0, and 2.0% (w/w) Au, and, for comparison, of pure TiO , are
2
shown in Fig. 1. The black dots of photodeposited Au islands can well be seen. It
follows from the images that the higher metal content favours more dense and
bigger gold islands on the titania surface. However, if we turn back to SEM
investigations on Ru/TiO [19] it can be predicted that also the naked part of the
2
surface may contain some highly dispersed metal. After photodeposition of Au and
30 min of photoreaction under argon, the slurry was aerated in the dark and
illuminated again. Therefore, the real phenol photooxidation on Au/TiO in the
presence of air started after 1 hour (Fig. 2).
2
The rate of phenol photooxidation was monitored by the 4-aminoantipyrine
method [4]. The spectrophotometric method gives as a result a so-called phenolic
index, i.e. the sum of phenolic compounds present in the solution. However, for

1040
A. Dobosz and A. Sobczy n ski
Fig. 2. Changes of phenolic index in the photoreactor during illumination of pure TiO2 and titanium
dioxide with various amounts of gold
simpli®cation, the term phenol instead of phenolic index is used in the paper. Real
phenol consumption and the difference between phenol and phenolic index changes
during the photoreaction will be shown below. The changes of phenol concentration
(changes of the phenolic index) during illumination of the slurry of pure titanium
dioxide and Au-loaded TiO are shown in Fig. 2. It can be seen that the photo-
2
reaction rate increases with increasing of Au content in the range of 0±0.5% (w/w).
For bigger Au contents (0.75, 1.0, 1.5, and 2% (w/w)), the curves of phenol
disappearance cover partly that for 0.05% (w/w) Au and are not shown in Fig. 2.
Some conclusions can be drawn from the data of Fig. 2. Firstly, illumination of
the slurry under a neutral atmosphere causes ± besides metal reduction ± also phenol
photooxidation. Its consumption increases with rising amount of metal. Secondly,
the metal admixture shows a catalytic (electrocatalytic) effect on the photoreaction
in the presence of oxygen: the rate of phenol destruction increases with increasing
amount of metal in the range from 0.02 to 0.5% (w/w) Au. Further increase of the
amount of the photodeposited gold does not in¯uence considerably the photo-
catalytic activity.
st
Phenol photodecomposition in the presence of titania obeys 1 order kinetics,
and usually the Langmuir-Hinshelwood expression is used in order to describe the
reaction rate [36±38]. Thus, the comparison between the reaction rates has to be
carried out at the same substrate (phenol) concentration, i.e. at the beginning of the
process.
When studying the photocatalytic activity of titania in DCB degradation, Gao
et al. [30] observed an increase of the activity by 30% caused by deposition of 1
at.% Au on the surface of TiO after 30 min. In the present investigation the increase
2
of the activity amounts to 120% after deposition of 0.5% (w/w) Au on the titania
surface (optimum metal loading). The signi®cant difference between the results of
Gao et al. and those reported here can be accounted for by both different substrates
decomposed (DCB and phenol) and different ways of Au/TiO preparation. The
2

Photocatalytic Decomposition of Phenol on Au/TiO2
1041
Fig. 3. HPLC and TOC analysis of the products of the photocatalytic reaction on 1% (w/w) Au on
TiO2; for comparison, a phenolic index curve is also depicted
observed small decrease of the photoreaction rate at a higher metal loading (> 0.5%
(w/w)) can be caused by the so-called screening effect. This effect has been
previously described for Pt/TiO , Ru/TiO , and Cu/TiO systems used for
2
2
2
photocatalytic hydrogen production from water (the deposited metal makes part
of the surface less accessible for photons) [19, 20].
Figure 3 shows the results of HPLC and total organic content (TOC) studies of
phenol photooxidation on Au/TiO (1% (w/w)); the curve of the phenolic index is
2
also included. One can see from the ®gure that the decrease of phenol concentration
during the process of Au photodeposition under neutral atmosphere (15±45 min of
the process) is much faster than the decrease of the phenolic index. At the same
time, hydroquinone, benzoquinone, and catechol appear in detectable amounts.
Dark saturation of the reaction mixture with air (45±60 min of the process) does not
cause noticeable changes in the concentration of substrate and intermediates.
Illumination of the reaction mixture in the presence of dissolved air causes fast
phenol photooxidation. Slower changes of the phenolic index show the existence of
phenolic-type reaction intermediates. Hydroquinone, p-benzoquinone, and catechol
are the main intermediate compounds. Small concentrations of p-benzoquinone in
relation to that of hydroquinone have been observed before in studies of photo-
oxidation of these compounds on pure titania [37]. Oxidation of catechol also leads
to o-benzoquinone. However, the absence of o-benzoquinone in the reaction
mixture is probably a result of its instability and, hence, fast decomposition.
The concentration of the intermediates rises with time at the beginning of
illumination; in parallel, that of phenol decreases. Photooxidation of all compounds
proceeds via a reaction with hydroxyl radicals photogenerated on the surface of
titania. It has been con®rmed before that the photooxidation of phenol, catechol,
st
resorcinol, and hydroquinone/benzoquinone proceeds according to 1 order
kinetics; the corresponding results have been partly published [37, 39]. Therefore,
the rate of the photooxidation of the mentioned compounds depends on their con-

1042
A. Dobosz and A. Sobczy n ski
Fig. 4. Degree of phenol photodecomposition on pure TiO2 (a) and TiO2 doped with 1.0% (w/w) of
Au at different substrate concentrations: 0:2 Á 10 (A), 0:6 Á 10 (B), 1:0 Á 10� 3 (C), and 1:6 Á 10� 3
�
3
� 3
�
3
(
D) mol Á dm
centration in the reaction mixture. After about 90 minutes of photoreaction, the
concentrations of the intermediates remain nearly constant. This means that the
rates of their formation and destruction are equal (see ¯at parts of the curves of
catechol, hydroquinone, and benzoquinone). Finally, all organic compounds are
fully oxidized to CO and H O, although their mineralization is rather slow (see
2
2
TOC curve in Fig. 3).
Figures 4a and 4b show the amounts of phenol which reacted during illu-
mination of the slurries of pure TiO (4a) and 1% (w/w) Au on TiO (4b) with
different initial phenol concentration. The steps of gold photoreduction and dark
2
2
periods are omitted in the ®gures. On pure TiO , the rate of the photoreaction
2
�
4
� 4
increases with increasing phenol concentration in the range of 2 Á 10 ±0:6 Á 10
�
3
� 3
� 3
� 3
mol Á dm and is strongly retarded at 1:0 Á 10 and 1:6 Á 10 molÁ dm . There
can be at least three reasons for the observed phenomenon. Firstly, higher substrate
concentration can favour the zero-order reaction kinetics (e.g. when phenol

Photocatalytic Decomposition of Phenol on Au/TiO2
1043
adsorption is not concentration dependent). Secondly, the fast formation of phenol
radicals (in the reaction of phenol with photogenerated hydroxyl radicals) can result
in their combining into bi- or polyphenols. The compounds are not water soluble
and are expected to adsorb strongly on the surface of TiO as it has been observed
2
before [36]. The third reason lies in the rate of oxygen photoreduction. Limitations
of O reduction on pure TiO have been considered and experimentally proved by
Gerischer, Heller, and Wang [28, 29] as was mentioned in the Introduction.
2
2
No photoreaction rate set-back was observed for 1% (w/w) Au on TiO over the
2
�
4
� 3
� 3
whole phenol concentration range from 1 Á 10 to 2 Á 10 mol Á dm . Here the
in¯uence of the Au additive is distinct: strong enhancement of the photoreaction
rate at lower substrate concentration (in comparison with pure TiO ) and lack of
2
retardation of the rate at higher concentrations. Probably, the gold microelectrodes
on the TiO surface cause a strong increase of the rate of oxygen reduction in the
2
reduction route of the photoreaction on the semiconductor as predicted [28, 29]. It is
also possible that the faster generation of hydroxyl radicals in the oxidation route of
the photoreaction prevents the production of polyphenols.
Conclusions
Introduction of metallic Au on a TiO surface improves the photocatalytic pro-
2
perties of TiO . The in situ preparation of the cocatalyst results in an increase of the
2
rate of phenol destruction on illuminated TiO of 120%. The Au cocatalyst prevents
2
also the decrease of the rate of the photocatalytic reaction at higher phenol con-
centrations which has been observed for pure TiO . Hydroquinone, p-benzoquinone,
2
and catechol are the main intermediates of the process. The existence of higher
hydroxylated products con®rms a radical mechanism of the photoprocesses on TiO2
[27, 38]. The photocatalytic phenol destruction leads ®nally to its total mineral-
ization to CO and H O.
2
2
Experimental
Photocatalytic experiments
The photocatalytic activities of various catalyst samples were evaluated by degradation of phenol
(
99‡%, Aldrich) in an aqueous environment. A stock solution of HAuCl4 Á 3H2O (Sigma) containing
3
3
5
.002 mg Au/cm was prepared. Various amounts of the stock solution were added to 100cm of the
3
�
4
reaction mixture containing 2 Á 10 mol=dm of phenol and 0.02 g of TiO2 anatase (99.9%, Aldrich).
This reaction mixture was sonicated for 3 min and poured in the photoreactor. The phenol solution
3
was added to achieve a ®nal volume of 165cm . In the case of different initial concentrations of
3
phenol, the same procedure was utilized. To get 1% (w/w) Au per TiO2, 40 mm of the stock solution
of Au were added to 0.02 g TiO2 suspended in the phenol solution.
The photoreactor consisted of an external glass tube with a conical bottom equipped with a septum
and a draft tube placed in the centre of the external one. The slurry in the reactor was mixed by a gas
stream (argon or air) applied at the bottom using a needle and ¯owing up through the draft tube. The
dimensions of the reactor were as follows: inner diameter of the external tube, 40 mm; h, 160 mm;
3
inner diameter of the draft tube, 10 mm; h, 120 mm; total volume of the reactor, ca. 190cm . The
reactor was illuminated from a side wall with a 180 W medium pressure Hg lamp. The Pyrex walls of
the reactor determined a lower limit of entering light (about 300nm cutoff ®lter). The photoreactor

1044
A. Dobosz and A. Sobczy n ski
ꢀ
was cooled by a stream of air; the reaction temperature was maintained at 28Æ2 C. For a detailed
reactor description, see Refs. [36, 37].
At ®rst, the reaction mixture was saturated with Ar for 15 min. Then, the mixture was illuminated
for 30 min in presence of Ar in order to photodeposit metallic Au on the TiO2 surface. Afterwards, the
reaction mixture was saturated with air for 15 min and illuminated for 3 h.
3
Subsequently, the apparatus was saturated with air. During the further photoreaction, 1 cm of the
slurry was withdrawn every 30 min using a syringe. After ®ltration over Millipore Millex GV3 units
the samples were analyzed by absorptiometric determination of the phenol index (concentration of
phenolic type compounds). The 4-aminoantipyrine method was used according to the procedure
described in Ref. [4]. HPLC (Waters) analysis was also used to identify intermediates of the
photodegradation of phenol. All reactants were of p.a. purity. In all experiments Millipore 18 Mohm
water was used.
The photocatalyst
2
� 1
The titania (TiO2 anatase, 99.9%, Aldrich) possessed a BET speci®c surface area of 9:85 m Á g and
showed a typical anatase spectrum with an adsorption onset at about 400 nm [23]. The particle size of
the TiO2 Aldrich powder is in the range of 0.10±0.35 mm as was deduced from SEM micrographs.
SEM and TEM investigations
Scanning electron micrographs were taken on a SEM 515 Philips apparatus using an acceleration
voltage of 20 kV. TEM analysis was performed using a Jeol JEM-1200 EX electron microscope. An
acceleration voltage of 80 kV was applied.
References
[
[
[
[
[
[
[
1] Hsiao H-Y, Lee CL, Ollis DF (1983) J Catal 82: 418
2] Pruden AL, Ollis DF (1983) J Catal 82: 404
3] Nguyen T, Ollis DF (1984) J Phys Chem 88: 3386
4] Matthew RW (1984) J Chem Soc Faraday Trans 1 80: 457
5] Matthew RW (1986) Water Res 20: 569
6] Okamoto K, Yamamoto Y, Tanaka H, Tanaka M, Itaya A (1985) Bull Chem Soc Jpn 58: 2015
7] Barbeni M, Pramauro E, Pelizzetti E, Borgarello E, Gr aÈ tzel M, Serpone N (1984) Nouv J Chem 8:
547
[
8] Hidaka H, Kubota H, Gr aÈ tzel M, Serpone N, Pelizzetti E (1985) Nouv J Chem 9: 67
9] Al-Ekabi H, Serpone N (1988) J Phys Chem 92: 5726
10] Bard AJ (1982) J Phys Chem 86: 172
[
[
[
[
[
[
[
[
11] Bard AJ (1988) Ber Bunsenges Phys Chem 92: 1187
12] Fox MA, Dulay MT (1993) Chem Rev 93: 341
13] Sato S, White JM (1980) J Amer Chem Soc 102: 7206
14] Chen B-H, White JM (1982) J Phys Chem 86: 3534
15] Kiwi J, Gr aÈ tzel M (1984) J Phys Chem 88: 1302
16] Sobczy n ski A, Bard AJ, Campion A, Fox MA, Mallouk TE, Webber SE, White JM (1987) J Phys
Chem 91: 3319
[
[
[
[
[
17] Nakamtsu H, Kawai T, Koreeda A, Kawai S (1985) J Chem Soc Faraday Trans 1 82: 527
18] Oosawa J, Chen J (1987) J Chem Soc Faraday Trans 1 80: 507
19] Sobczy n ski A, Jakubowska T, Zieli n ski S (1989) Monatsh Chem 120: 101
20] Sobczy n ski A (1992) Monatsh Chem 123: 211
21] Sakata T, Kawai T (1981) Chem Phys Letters 80: 341

Photocatalytic Decomposition of Phenol on Au/TiO2
1045
[
[
[
22] Prahov L, Disdier J, Herrmann JM, Pichat P (1984) Int J Hydrogen Energy 9: 397
23] Sayama K, Arakawa H (1994) J Photochem Photobiol A 77: 243
24] Sobczy n ski A, Yildiz A, Bard AJ, Campion A, Fox MA, Mallouk TE, Webber SE, White JM
1989) J Phys Chem 93: 401
(
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
25] Sobczy n ski A (1991) J Catal 131: 156
26] Bard AJ (1979) J Photochem 10: 50
27] Izumi I, Dunn WW, Wilbourn KO, Fan FRF, Bard AJ (1980) J Phys Chem 84: 3207
28] Gerischer H, Heller A (1992) J Electrochem Soc 139: 113
29] Wang C-M, Heller A, Gerischer H (1992) J Amer Chem Soc 114: 5230
30] Gao Y-M, Lee W, Trehan R, Kershaw R, Dwight K, Wold A (1991) Mater Res Bull 26: 1247
31] Papp J, Shen H-S, Kershaw R, Dwight K, Wold A (1993) Chem Mater 5: 284
32] Lee W, Shen H-S, Dwight K, Wold A (1993) J Solid State Chem 106: 288
33] Chen J, Ollis DF, Rulkens WH, Bruning H (1999) Water Res 33: 661
34] Chen J, Ollis DF, Rulkens WH, Bruning H (1999) Water Res 33: 669
35] Chen J, Ollis DF, Rulkens WH, Bruning H (1999) Water Res 33: 1173
36] Sobczy n ski A, Gimenez J, Cervera-March S (1997) Monatsh Chem 128: 1109
37] Sobczy n ski A, Duczmal è, Dobosz A (1999) Monatsh Chem 130: 377
38] Turchi CS, Ollis DF (1990) J Catal 122: 178
39] Duczmal è, Sobczy n ski A (1999) React Kinet Catal Lett 66: 289
Received February 5, 2001. Accepted March 19, 2001