Photocatalytic degradation of 4-nitrophenol in aqueous suspension by using polycrystalline TiO2 impregnated with functionalized Cu(II) -porphyrin or Cu(II)-phthalocyanine

Article abstract of DOI:10.1016/S0021-9517(03)00040-X

The photocatalytic activity of polycrystalline TiO₂ samples impregnated with functionalized Cu(II)- or metal-free porphyrins has been investigated. A probe reaction, i.e., 4-nitrophenol (4-NP) photodegradation in aqueous suspension, has been used. The results indicated that the presence of the sensitizers is beneficial for the photoactivity and suggest an important role of Cu(II). A comparison with similar samples impregnated with modified Cu(II)- and metal-free phthalocyanines showed that the presence of porphyrin is more beneficial both for the decomposition rate of 4-nitrophenol and for the disappearance of nonpurgeable organic carbon (NPOC). Experiments were carried out by using cutoff and interference filters and a cooperative mechanism involving the photoactivation of both TiO₂ and sensitizer is proposed.

Full text of DOI:10.1016/S0021-9517(03)00040-X

Journal of Catalysis 217 (2003) 334–342
www.elsevier.com/locate/jcat
Photocatalytic degradation of 4-nitrophenol in aqueous suspension
by using polycrystalline TiO2 impregnated with functionalized
Cu(II)–porphyrin or Cu(II)–phthalocyanine
Giuseppe Mele,a Roberta Del Sole,a Giuseppe Vasapollo,a Elisa García-López,b
Leonardo Palmisano,b,∗ and Mario Schiavello b
a
Dipartimento di Ingegneria dell’Innovazione, Università di Lecce, Via Arnesano, 73100 Lecce, Italy
b
Dipartimento di Ingegneria Chimica dei Processi e dei Materiali, Università di Palermo, Viale delle Scienze, 90128 Palermo, Italy
Received 16 October 2002; revised 7 January 2003; accepted 7 January 2003
Abstract
The photocatalytic activity of polycrystalline TiO samples impregnated with functionalized Cu(II)- or metal-free porphyrins has been
2
investigated. A probe reaction, i.e., 4-nitrophenol (4-NP) photodegradation in aqueous suspension, has been used. The results indicated that
the presence of the sensitizers is beneficial for the photoactivity and suggest an important role of Cu(II). A comparison with similar samples
impregnated with modified Cu(II)- and metal-free phthalocyanines showed that the presence of porphyrin is more beneficial both for the
decomposition rate of 4-nitrophenol and for the disappearance of nonpurgeable organic carbon (NPOC). Experiments were carried out by
using cutoff and interference filters and a cooperative mechanism involving the photoactivation of both TiO and sensitizer is proposed.
2
2003 Elsevier Science (USA). All rights reserved.
Keywords: Porphyrin; Cu(II)–porphyrin; Phthalocyanine; Cu(II)–phthalocyanine; TiO photocatalysts; 4-Nitrophenol photodegradation
2
1. Introduction
Few papers concern their use associated which polycrys-
talline TiO2 samples as efficient photocatalytic systems used
for the degradation of organic pollutants in water [11–14].
These combined systems showed an improved activity com-
pared to the well-known and widely studied bare TiO2 sam-
ples.
Photodegradation of organic compounds in water by
means of economically advantageous and environmental
friendly processes is a topic of growing interest and much
attention has been devoted in recent years to TiO2-based
photocatalysts for the oxidative degradation of various kinds
of organic pollutants [1–5].
A goal of both academic and industrial researches is to
obtain new catalytic systems having an enhanced activity
compared with the simply TiO2-catalyzed processes.
Phthalocyanines (Pcs) and porphyrins (Pps) have been
used as photocatalysts for a large variety of oxidation re-
actions of phenol derivatives and some mechanistic aspects
have also been investigated [6–9]. Metal–phthalocyanines
(MPcs) are efficient photosensitizers and catalysts for many
reactions [10] and, due to their high photostability and abil-
ity to absorb light, offer many advantages in comparison to
the more popular methylene blue and rose Bengal dyes.
Although phthalocyanines and metal–phthalocyanines in
oxidative media are reported to be more resistant than por-
phyrins [15], both of them present real perspectives as
(photo)-catalysts because their relative stability depends on
their molecular structure, reaction conditions, irradiation
time, temperature, light intensity, pH, and whether they are
supported or not. The chemistry of porphyrin derivatives has
played an important role especially during the past 10 years
in particular branches of new material science and many re-
searchers have on-going projects on the synthesis of vari-
ously substituted compounds aimed to obtain new “func-
tional materials.” The latter have found an application for
the construction of solar cells as light absorbents in organic
dyes displaying notable stability and unique chemical, phys-
ical, and spectroscopic properties.
*
Corresponding author.
E-mail address: palmisan@dicpm.unipa.it (L. Palmisano).
0021-9517/03/$ – see front matter
2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S0021-9517(03)00040-X

G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
335
Moreover, studies on the capability of porphyrins and
metal–porphyrins to give rise to photoinduced electron
transfer processes and to present gas sensors properties have
been also reported [16–20].
chemical shifts are reported in ppm units with respect to the
reference frequency of tetramethylsilane, Me4Si.
IR and MS spectra were performed on Perkin-Elmer
683 and Hewlett-Packard GC/Mass MSD 5971 instruments,
respectively.
In a recent paper [21] Héquet and co-workers compare
the photocatalytic activity of an hemine (which is an iron
porphyrin) with an iron sulfophthalocyanine and with a bare
TiO2 (anatase) sample affording the degradation of atrazine
by using a mercury UV lamp. TiO2 showed better degrada-
tion performance compared with the iron porphyrin and iron
porphyrin derivatives. However, all the catalyzed processes
are reported to be more efficient than the uncatalyzed ones.
The preparation of polycrystalline TiO2 samples impreg-
nated with different Cu(II)–phthalocyanines (TiO2–CuPc)
for the photocatalytic degradation of 4-nitrophenol (4-NP)
has recently been performed [22]. The presence of modified
CuPc was found to be beneficial especially for the photoac-
tivation of TiO2 (anatase), while in only few cases a slightly
enhanced photoactivity for TiO2 (rutile) was observed.
In this paper some Cu(II)–porphyrins opportunely func-
tionalized with sterically hindered alkyl groups were pre-
pared in order to obtain substrates soluble in organic solvents
and hence suitable for their impregnation onto the TiO2 sur-
face. The peripheral substitution in such molecules and the
possibility of coordinating different metals is important for
the design of functional dyes and molecular devices. More-
over, a very recent paper [23] reports that the modification of
the structure of a porphyrin dye caused a variation of the re-
combination rate between injected electrons in the TiO2 and
the anchored oxidized dye.
It is worth noting that many molecular analogies exist be-
tween phthalocyanines and porphyrins, such as a character-
istic macrocyclic structure with extended π-electron systems
and the presence of metal ions (or alternatively two hydro-
gen atoms) in the middle of the macrocycle.
Consequently the photocatalytic activity of polycristalline
TiO2 samples impregnated with a functionalized Cu(II)–
porphyrin was compared with that of some selected samples
impregnated with a functionalized Cu(II)–phthalocyanine
for a probe reaction, i.e., 4-NP photodegradation [24]. In
this paper, consequently, in order to investigate the role of
the metal some TiO2 samples impregnated with the corre-
sponding metal-free porphyrin and metal-free phthalocya-
nine were prepared and tested.
Mass spectrometry analyses were carried out by using
a LC mass spectrometer 1100 Series (Agilent) equipped
with an atmospheric pressure chemical ionization (APCI)
interface. The samples, dissolved in chloroform, were intro-
duced into the mass spectrometer injected by an autosampler
spraying a methanol solution at a flow rate of 0.5 mL/min.
A heated nebulized spray was continuously introduced into
a point corona discharge region using nitrogen to nebulize
and sheath the liquid inlet.
Ions were extracted via a heated capillary to a skimmer
lens arrangement at reduced pressure and transferred by
an octapole to the main analytical quadrupole assembly.
The instrumental conditions were as follow: drying gas
(nitrogen) 13 L/min, nebulizer pressure 60 psi, drying gas
temperature 350 ◦C, vaporizer temperature 500 ◦C, capillary
voltage 3000 V, corona current 4.0 µA, mass range 500–
2000 amu.
2.1.1. Synthesis of the [5,10,15,20-tetra(4-tert-
butylphenyl)]porphyrin (H2Pp)
4-tert-butylbenzaldehyde (1.622 g, 10 mmol), pyrrole
(0.67 g, 10 mmol), and BF3 · OEt2 (0.160 g, 1.2 mmol) were
dissolved in 200 mL of chloroform and stirred in a 250-
mL round-bottom flask at room temperature under nitrogen
atmosphere. After 2 h 1.816 g (0.8 mmol) of 2,3-dichloro-
5,6-dicyano-1,4-benzoquinone (DDQ) was added and the
solution was stirred for 2 h more.
After solvent evaporation the crude of reaction was pu-
rified by column chromatography (silica, CH2Cl2/hexane,
7/3) and product was recovered in 40% yields and char-
acterized by LC-MS, FT-IR, 1H NMR, 13C NMR, and
UV–vis analyses. The melting point of H2Pp solid was >
350 ◦C. LC-MS (APCI interface) observed (M − H+): 839
(M + 1) amu, calculated M: 838 amu. FT-IR: 3313, 2953,
2920, 2851, 1736, 1461, 1262, 1105, 964, 803, 789, 734,
1
713 cm−1. H NMR (CDCl3, 200 MHz) δ: 8.87 (s, 8H),
8.14 (d, J = 8.2 Hz, 8H), 7.75 (d, J = 8.2 Hz, 8H), 1.61
(s, 36H), −1.4 (s, br, 2H). 13C NMR (CDCl3, 200 MHz) δ:
150.4, 139.2, 134.5, 129, 128.2, 123.6, 120.1, 24.9, 31.7.
UV–vis (nm) λ = 423 (Soret band); 518, 554, 591, 648
(Q bands).
2. Experimental
2.1.2. Synthesis of the Cu(II)[(5,10,15,20-tetra(4-tert-
butylphenyl)]porphyrin (CuPp)
2.1. NMR, IR, and LC-MS measurements
An excess of CuCl2 (5.0 mmol) was added to a solution
obtained dissolving 0.840 g (1.0 mmol) of H2Pp in 200 mL
of dichloromethane. The mixture was stirred for 2 h and
monitored by TLC (thin-layer chromatoghaphy)analysis un-
til the complete disappearance of the starting material. The
crude of reaction was filtered to remove the unreacted solid
salt and further purified by column chromatography (silica,
The water used was purified by a Milli-Q/RO sys-
tem (Millipore) resulting in a resistivity, ρ, of more than
10 Mꢀ cm. Melting points were determined on an elec-
trothermal apparatus. 1H and 13C NMR spectra were
recorded on a Bruker AC-200 at room temperature and

336
G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
2.2. X-ray diffraction (XRD)
X-ray powder diffraction analysis of all of the samples
was carried out at room temperature by a Philips PW 1130
generator and PW 1050 goniometer using Ni-filtered Cu-Kα
radiation.
2.3. Specific surface area determination (BET)
The specific surface areas were measured by the single-
point BET method using a Flow Sorb 2300 apparatus
(Micromeritics International Corp.).
Fig. 1. Molecular structure of H Pc and CuPc.
2
2.4. Diffuse reflectance spectroscopy (DRS)
The spectra were obtained in air at ca. 300 K in the
wavelength range 250–800 nm using a Shimadzu UV-2401
PC spectrophotometer with BaSO4 as the reference material.
CH2Cl2/Hexane, 7/3). CuPp was recovered in 90% yields
and characterized by LC-MS, FT-IR, UV–vis analyses. The
melting point of CuPp solid was > 350 ◦C. LC-MS (APCI
interface) observed (M − H+): 900 (M + 1) amu, calculated
M: 899 amu. FT-IR: 2953, 2922, 2852, 1731, 1461, 1344,
1265, 1108, 1070, 997, 809, 802, 741, 720 cm−1. UV–vis
(nm) λ = 417 (Soret band); 540, 576 (Q bands).
2.5. Photoreactivity experiments
A Pyrex batch photoreactor of cylindrical shape contain-
ing 0.5 L of aqueous suspension was used. The photoreactor
was provided with a jacket for cooling water circulation and
ports in its upper section for the inlet and outlet of gases,
for sampling and for pH and temperature measurements.
A 125 W medium pressure Hg lamp (Helios Italquartz, Italy)
was immersed within the photoreactor and the photon flux
emitted by the lamp was Φi = 13.5 mW/cm2. It was mea-
sured by using a UVX Digital radiometer (300–400 nm)
leaned against the external wall of the photoreactor contain-
ing only pure water. O2 was bubbled into the suspensions for
ca. 0.5 h before switching on the lamp and throughout the oc-
currence of the photoreactivity experiments. The initial 4-NP
(BDH) concentration was 20 mg/L and the amount of cata-
lyst used for all the experiments was 0.8 g/L. The transmit-
ted light in the presence of this quantity, measured by using
the UVX Digital radiometer as reported before in the case
of pure water, was negligible, indicating that most of the ra-
diation was absorbed. The initial pH of the suspension was
adjusted to 4.0 by the addition of H2SO4 (Carlo Erba RPE),
and the temperature inside the reactor was held at ca. 300 K,
due to a continuous circulation of water in the jacket around
the photoreactor. The photoreactivity runs lasted 6.0 h. Sam-
ples of 5-mL volume were withdrawn from the suspensions
every 3 or 5 min during the first 30 min of irradiation and
subsequently every 30 or 60 min. The catalysts were sep-
arated from the solution by filtration through 0.45-µm cel-
lulose acetate membranes (HA, Millipore). The quantitative
determination of 4-NP was performed by measuring its ab-
sorption at 315 nm with a spectrophotometer Shimadzu UV-
2401 PC. Finally, nonpurgeable organic carbon (NPOC) de-
terminations were carried out by using a Shimadzu total or-
ganic carbon analyzer 5000-A. An additional set of selected
experiments was carried out in the setup sketched in Fig. 2
consisting of an arc lamp supply (8540 Oriel Corp., USA)
2.1.3. Synthesis of the Cu(II) tetrakis[4-(2,4-bis-(1,1-
dimethylpropyl)phenoxy)]phthalocyanine (CuPc) and
of the tetrakis[4-(2,4-bis-(1,1-dimethylpropyl)phenoxy)]-
phthalocyanine (H2Pc)
The molecular structures of H2Pc and CuPc are shown
in Fig. 1 and their preparation has been previously report-
ed [22].
2.1.4. Preparation of TiO2–CuPc, TiO2–H2Pc,
TiO2–CuPp, and TiO2–H2Pp samples
TiO2 (anatase) and TiO2 (rutile) were kindly provided
by Tioxide Huntsman. The loaded samples used as photo-
catalysts for the photoreactivity experiments were prepared
by impregnating TiO2 (Tioxide, anatase phase, specific sur-
face area 8 m2/g) with various amounts of CuPp (3.30,
5.55, 6.65, 8.35, 11.10, and 13.30 µmol/g TiO2); H2Pp (6.65
and 11.10 µmol/g TiO2); CuPc (6.65 µmol/g TiO2), H2Pc
(6.65 and 11.10 µmol/g TiO2), and TiO2 (Tioxide, rutile
phase, specific surface area 7 m2/g) with various amounts
of CuPp (3.90, 6.65 and 11.10 µmol/g TiO2); H2Pp (6.65
and 11.10 µmol/g TiO2). The sensitizers were dissolved in
10 mL of CHCl3 (or CH2Cl2) and 2 g of finely ground TiO2
was added to this solution. The mixture was stirred for 3–4 h
and the solvent was removed under vacuum.
The code used for the samples is the following: the figure
indicates the micromoles of sensitizer used for the impregna-
tion of 1 g of TiO2, A or R the anatase or rutile phases, CuPp
and CuPc the copper porphyrin and the copper phthalocya-
nine, and H2Pp and H2Pc the metal-free porphyrin and the
metal free phthalocyanine. For instance 6.65-TiO2(A)–H2Pp
represents the sample prepared by using 6.65 µmol of metal-
free porphyrin to impregnate 1 g of TiO2 (anatase).

G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
337
420-nm cutoff filter (Oriel Corp.) or a 370-nm interference
filter (Oriel Corp., bandwidth of 20 nm). The initial 4-NP
(BDH) concentration was 20 mg/L and the amount of cat-
alyst used for all the experiments was 0.8 g/L. The catalyst
also in this case was separated by filtration through 0.45-µm
cellulose acetate membranes (HA, Millipore). The duration
of the experiments was 1 h and the UV analyses of 4-NP
were carried out by using 1-mL samples withdrawn every 3
or 5 min.
In order to obtain the quantum efficiencies (reported in
Table 1) for the runs carried out in the 0.5-L photoreactor, the
ferrioxalate actinometer method was used for measuring the
photon flow inside the photoreactor [25]. The absorbance of
the actinometer sample was measured at 510 nm. The value
of the photon flow was 0.635 × 10−5 einstein/(sL).
3. Results and discussion
3.1. Synthesis of H2Pp and CuPp
Fig. 2. Setup of the runs carried out by using filters: (a) illumina-
tor–collimator, (b) position for filters, (c) photoreactor, (d) magnetic stirrer,
(e) elevator, (f) oxygen cylinder, (g) water reservoir, (h) power supply, and
(i) pump.
The synthesis of the metal-free porphyrin (5,10,15,
20-tetra(4-tert-butylphenyl)porphyrin, H2Pp, reported in
Scheme 1, was performed according to the procedure sim-
ilar to that reported in the literature [26,27] reacting 4-tert-
butylbenzaldehyde with pyrrole (1:1 molar ratio) in chloro-
form at room temperature for 2 h in the presence of 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone (DDQ). More de-
equipped with a 1000 W Hg–Xe lamp (Hanovia) and with
a system of collimating lenses. A 50-mL batch photoreactor
was used and some experiments were performed by using a
Table 1
List of the samples used together with the BET specific surface areas, the total coverage areas due to the porphyrin molecules, the initial photoreaction rates,
and the quantum efficiencies determined by using the 0.5-L batch photoreactor
a
10
ꢀ
10
Samples
BET specific
surface areas
SSA (m /g)
Calculated
r × 10
r × 10
Quantum
efficiencies
η (%)
0
0
b
2
total areas
(mol/(s L))
(mol/(s L m ))
2
2
(m )
TiO (A)
6.65-TiO (A)–H Pp
11.10-TiO (A)–H Pp
3.30-TiO (A)–CuPp
5.55-TiO (A)–CuPp
6.65-TiO (A)–CuPp
8.35-TiO (A)–CuPp
11.10-TiO (A)–CuPp
13.30-TiO (A)–CuPp
6.65-TiO (A)–CuPc
6.65-TiO (A)–H Pc
11.10-TiO (A)–H Pc
8
8
8
8
8
8
8
8
8
8
8
8
7
7
7
7
7
7
467
606
569
595
1225
1162
1087
950
854
755
454
359
16
146
189
178
190
383
363
352
315
271
236
142
122
6
0.7
0.9
0.9
0.9
1.9
1.8
1.7
1.5
1.3
1.2
0.7
0.5
0.02
0.04
0.02
0.02
0.02
0.01
2
9
15
4.5
7.5
9
11
15
18
9
2
2
2
2
2
2
2
2
2
2
2
9
15
2
2
2
2
TiO (R)
2
3.90-TiO (R)–CuPp
6.65-TiO (R)–CuPp
11.10-TiO (R)–CuPp
6.65-TiO (R)–H Pp
5
9
15
9
30
17
16
13
10.7
6.1
5.7
4.6
2.5
2
2
2
2
2
11.10-TiO (R)–H Pp
15
7
2
2
Lamp: 125 W medium pressure Hg lamp. Photocatalyst: 0.8 g/L. Initial 4-NP concentration: 20 mg/L. T = 300 K. The values reported in the table are the
average of three measurements.
a
TiO (A) and TiO (R) used for these experiments were provided by Tioxide Huntsman.
2
2
b
2
The calculated areas refer to 1 g of TiO . The surface area for a single porphyrin molecule was approximate to 2.25 nm /molecule. The porphyrins were
2
supposed flat onto the TiO surface.
2

338
G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
Scheme 1. Scheme of reaction for the synthesis of (5,10,15,20-tetra(4-tert-butylphenyl)porphyrin (H Pp).
2
Scheme 2. Scheme of reaction for the synthesis of the porphyrin metal complexes.
tails on the preparation and purification of H2Pp have been
previously reported (see Section 2).
Fig. 3 shows the diffuse reflectance spectra in air of the
bare TiO2(A) and some TiO2(A)–CuPp and TiO2(A)–H2Pp
catalysts recorded in the range 250–800 nm.
The isolated porphyrin derivative was characterized by
spectral data. UV–vis spectra showed a characteristic strong
band centered at λ = 423 nm (Soret band) and the absorp-
tions, respectively, at 518, 554, 591, and 648 nm (Q bands).
The liquid mass spectra showed the positively charged
M − H+ adduct at 839.6 (M + 1) amu, as expected by
It is worth noting that no shift of the band gap edge of
TiO2 can be observed for all the loaded samples. Neverthe-
less they reflect light less significantly than the bare support
and their absorption increases by increasing the loading.
DR spectra of TiO2(A)–CuPp and TiO2(A)–H2Pp sam-
ples present shapes, relative intensities and maximum ab-
sorption values that are similar to those observed for CuPp
and H2Pp in CHCl3 solution (see Fig. 3, inset).
1
using the APCI interface. H NMR, 13C NMR and FT-IR
spectra were also consistent with the structure of the isolated
compound.
As a consequence of this comparison, we could presume
that there is only a negligible molecule–molecule dye
aggregation onto the TiO2 particles both for the least heavily
loaded samples and for the most heavily ones, for which
some molecule–molecule interactions cannot be excluded.
TiO2(A)–CuPp samples showed a characteristic strong
absorption band centered at 417 nm (Soret band) and at 540,
and 576 nm (Q bands).
H2Pp was successively used as ligand to coordinate the
Cu(II) ion for the synthesis of the porphyrin metal complex,
CuPp, through the general protocol of metal complexation
as shown in Scheme 2 and described under Section 2. The
coordination of the metal by the metal-free porphyrin ligand
in solution was easily monitored by TLC analyses. Spectral
data for the isolated compound were consistent with the
structure of the Cu(II)–porphyrin complex.
A strong absorption at λ = 423 nm (Soret band) and
absorptions decreasing in intensity at 518, 554, 591, and
648 nm (Q bands) can be found, instead, in the spectra of
TiO2(A)–H2Pp samples.
3.2. X-ray diffraction and diffuse reflectance spectroscopy
The X-ray diffractograms of all of the loaded samples
indicated that no modification of the starting anatase or
rutile phases of the bare TiO2 supports occurred after the
impregnation treatments.
Similar DR spectra, not shown for the sake of brevity,
were observed when the sensitizers used in this work were
supported onto rutile TiO2. DR spectra of the samples
impregnated with the phthalocyanine derivatives are also not

G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
339
Fig. 3. Diffuse reflectance spectra of bare TiO and differently loaded samples obtained by impregnation of TiO with CuPp or H Pp. TiO (A) (a);
2
2
2
2
5.55-TiO (A)–CuPp (b); 6.65-TiO (A)–CuPp (c); 11.10-TiO (A)–CuPp (d); 6.65-TiO (A)–H Pp (e); 11.10-TiO (A)–H Pp (f). Inset: spectra in CHCl
2
2
2
2
2
2
2
3
of TiO (A)–H Pp (g) and TiO (A)–CuPp (h).
2
2
2
reported because they do not differ significantly from those
reported in a previous paper for similar samples [22].
ular the best samples appear to be 5.55-TiO2(A)–CuPp
and 6.65-TiO2(A)–CuPp. The 6.65-TiO2(A)–CuPc sample
is less photoactive than the corresponding 6.65-TiO2(A)–
CuPp sample, and this finding indicates that the beneficial
effect of porphyrin as sensitizer is more significant than that
showed by phthalocyanine. Moreover the metal-free por-
phyrin showed a photoactivity slightly higher than the corre-
sponding bare TiO2, indicating that the porphyrinic macro-
cycle is photocatalytically active even in the absence of the
metal. As far as the loaded TiO2(R) samples are concerned,
no significant improvement of the photoreaction rate was
observed for all of the samples compared to bare TiO2(R)
with the exception of the 3.90-TiO2(R)–CuPp. Nevertheless
the reaction rates were in all cases more than one order of
magnitude lower than those observed for the loaded TiO2(A)
samples.
As far as the quantum efficiencies values are concerned
(see Table 1), their trend is equal to that of the initial reaction
rates, and it is worth noting that they are less than the true
ones because they were calculated by taking into account the
photons impinging on the reacting system.
Nonpurgeable organic carbon concentration versus irra-
diation time is reported in Fig. 4 for some selected samples.
A complete disappearance of organic carbon can be ob-
served after ca. 4 h for the 6.65-TiO2(A)–CuPp sam-
ple. TiO2(A) appears to be the worst sample, while 6.65-
TiO2(A)–H2Pp showed a behavior similar to that of 6.65-
TiO2(A)–CuPc in the final steps.
3.3. Photoreactivity experiments
A preliminary investigation was carried out in order to
establish if the CuPp and H2Pp supported onto TiO2 were
photostable, i.e., if some decomposition or chemical modifi-
cation of the supported CuPp and H2Pp took place under the
same conditions used during the photocatalytic experiments.
NPOC determinations in the absence of 4-NP indicated no
significant release of organic degradation compounds even
after long irradiation times (5–7 h). Moreover the supported
sensitizers can be recovered quantitatively (and unchanged)
from the TiO2 surface by extraction with chlorinated sol-
vents (CHCl3 or CH2Cl2). The absence of structural modifi-
cations was confirmed by analytical and spectral data (TLC,
UV–vis, 1H NMR).
The list of the used samples is reported in Table 1 along
with the BET specific surface areas, the calculated areas due
to the porphyrin molecules, the initial reaction rates for 4-
NP disappearance, and the quantum efficiencies. A straight
line fits the data collected during the first 30 min of irradia-
tion and consequently the results are reported as zero-order
initial rates. No significant induction period for 4-NP con-
centration decrease was seen. No analyses of intermediates
was performed because it was out of the aim of this work,
and the 4-NP photodegradation was chosen only as a probe
reaction [24]. Dihydroxynitrobenzene isomers were found
initially as the intermediates and these compounds absorb
mainly in the visible range, not interfering at 315 nm [24].
It can be seen that the samples impregnated with Cu(II)–
porphyrins exhibited the highest photoactivity; in partic-
In Table 2 the results of some selected experiments
carried out in the setup shown in Fig. 2 in the absence and
in the presence of 370 nm interference and 420-nm cutoff
filters are reported with the aim of comparing the initial
reaction rates of the samples for each kind of irradiation.

340
G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
to that hypothesized for TiO2 impregnated with Fe(III)–
phthalocyanine [11].
A tentative set of reactions explaining the observed
photoreactivity behavior is reported below.
3.3.1. Role of TiO2
TiO2[Cu(II)Sens]
→
hν (<394 nm) TiO2[Cu(II)Sens]
ꢀ
ꢁ
eC−B + h+VB
.
(1)
(2)
Fig. 4. Nonpurgeable organic carbon concentration (NPOC) versus irradi-
ation time: TiO (A) ("); 6.65-TiO (A)–CuPp (2); 6.65-TiO (A)–H Pp
2
2
2
2
3.3.2. Role of electrons
(1); 6.65-TiO (A)–CuPc (Q).
2
−
3O2 + eCB → ·O2
,
−
Table 2
ꢀ1
ꢁ
1O2 ∆ + eC−B → ·O2
,
(3)
(4)
(5)
−
Initial photoreaction rates of the samples used with the setup shown in
Fig. 2, without filtered light and by using a 420-nm cutoff and a 370-nm
interference filters
H2O2 + eC−B → ·OH + OH−,
TiO2[Cu(II)Sens] + eC−B → TiO2[Cu(I)Sens].
a
8
Sample
r × 10 (mol/(s L))
0
Without filter
370-nm
420-nm
interference filter
cutoff filter
3.3.3. Role of holes
TiO (A)
6.65-TiO (A)–CuPp
6.65-TiO (A)–CuPc
6.65-TiO (A)–H Pp
9.0
14.1
11.9
10.8
9.6
2.4
3.9
3.7
2.0
1.3
Negligible
Negligible
Negligible
Negligible
Negligible
2
OH− + h+VB → ·OH,
(6)
(7)
2
2
H2O + h+VB → ·OH + H+,
H2O2 + h+VB → HO2· + H+,
HO2· + h+VB → O2 + H+,
TiO2[MSens] + h+VB → TiO2[MSens]+.
2
2
6.65-TiO (A)–H Pc
2
2
(8)
Lamp: 1000 W Hg–Xe. Photocatalyst: 0.8 g/L. Initial 4-NP concentration:
20 mg/L. T = 300 K. The values reported in the table are the average of
(9)
three measurements.
a
TiO (A) used for these experiments was provided by Tioxide Hunts-
2
(10)
man.
3.3.4. Role of sensitizer
It can be seen that a negligible photoactivity was observed
for all of the samples when the cutoff filter was used. This
indicates that the photoexcitation of TiO2 is essential for
inducing the photodegradation of 4-NP. The runs carried out
in the absence of filters confirm the results obtained by using
the other setup, i.e., both CuPp- and CuPc-loaded samples
are more photoactive than the bare TiO2 sample, but the
presence of CuPp as sensitizer was more beneficial.
The two samples impregnated with the metal-free sensi-
tizers were more active than TiO2(A) and this fact appears to
be relevant also because the experiments carried out by using
the 370-nm interference filter indicated lower photoactivi-
ties of the 6.65-TiO2(A)–H2Pp and the 6.65-TiO2(A)–H2Pc
samples in comparison to TiO2(A). Moreover, both corre-
sponding samples containing the metal were much more
photoactive than the bare TiO2, although the highest values
of reaction rate were obtained always in the absence of filter,
due to the higher light intensity impinging on the reacting
system.
intersystem crossing
hν
(isc)
1
3
[MSens] → [MSens]∗
→
[MSens]∗,
(11)
(12)
(13)
(14)
(15)
ꢀ1
ꢁ
[MSens]∗ + O2 → [MSens] + O2 ∆ ,
3
3
1
3
3
−
,
[MSens]∗ + O2 → [MSens]+ + ·O2
−
,
TiO2
[MSens]∗ → TiO2[MSens]+ + eCB
3 or 1
TiO2[MSens]+ + OH− → TiO2[MSens] + ·OH,
TiO2[MSens]+ + 4-NP → TiO2[MSens]
+oxidation products.
(16)
3.3.5. Role of some reactive intermediates in aqueous
phase
·O2− + H2O ꢀ HO2· + OH−,
(17)
(18)
(19)
(20)
(21)
(22)
·O2− + H+ ꢀ HO2 ,
·
In light of the above results it can be suggested that it is
necessary to photoexcite both the components of the system,
i.e., TiO2 and the sensitizer, in order to obtain substantial
improvement of the photodegradation rate of 4-NP.
It appears likely that the beneficial effect on the pho-
toreactivity is due to a cooperative mechanism, according
2HO· → H2O2,
2HO2· → H2O2 + O2,
H2O2 + ·O2− → HO· + OH− + O2,
2HO2· + ·O2− → HO· + OH− + 2O2.

G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
341
3.3.6. Role of metal
In Table 1 we have reported in the third column, the calcu-
lated total areas due to the porphyrin molecules, considering
the surface area occupied by a single porphyrin molecule,
supposed flat onto the TiO2 surface, approximately equal to
2.25 nm2, as reported for a similar compound [33]. Inter-
estingly, we observed the maximum of photocatalytic activ-
ity when the molecular coverage was comparable with BET
specific surface areas measured for TiO2 samples.
3
TiO2[Cu(I)Sens] + O2 → TiO2[Cu(II)Sens] + ·O2−, (23)
ꢀ
ꢁ
1
−
,
TiO2[Cu(I)Sens] + O2 1∆ → TiO2[Cu(II)Sens] + ·O2
(24)
TiO2[Cu(I)Sens] + H2O2 → TiO2[Cu(II)Sens] + HO·
+OH−.
(25)
In the system studied in this work Cu(II) could be reduced
to Cu(I) [see Eq. (5)] by electrons of the conduction band
of TiO2 where additional electrons are injected, due to
the presence of the sensitizer [11,34]. Subsequently the
reoxidation of Cu(I) to Cu(II) not only by 3O2 and 1O2(1∆)
[see Eqs. (23) and (24)] but also by hydrogen peroxide
produced in solution (25) in the presence of the sensitizer
could improve significantly the kinetics of the process due
3.3.7. Oxidation reactions in aqueous phase
ꢀ
ꢁ
1O2 1∆ + 4-NP → oxidation products,
HO2· + 4-NP → oxidation products,
·HO + 4-NP → oxidation products.
(26)
(27)
(28)
As reported in Eq. (1), the photoinduced process of
charge separation promoted by UV radiation represents the
key step of the main process, and the presence of the
sensitizer onto TiO2 does not influence appreciably this
process.
In the above set of reactions we have tried to distinguish
the roles played by conduction band electrons, valence band
holes, and metal-free or copper-coordinatedorganic sensitiz-
ers in the production of reactive intermediates effective for
the photodegradation of 4-NP in water.
1
to the higher oxidant properties of O2(1∆) and H2O2 on
respect to 3O2.
·O2 production along with Eqs.· (6)–(8) is essential
−
for inducing the formation of HO2 and ·OH radicals
responsible for oxidant attacks in photocatalytic reactions.
As far as holes are concerned, they can be delocalized
in the macrocyclic structure of the sensitizers [see Eq. (10)]
contributing to increase the lifetime of the photoproduced
pairs and/or to react with surface OH and water according to
well-known photocatalytic steps.
Analogous to that reported by Wöhrle and co-workers
[28], both metal and metal-free porphy·rin −and phthalocya-
1
nine samples can produce O2(1∆) or O2 under irradia-
It is worth noting that hole transport in p-type organic
semiconductors has been also reported [33]: when smooth
TiO2 (anatase) films coated with an amorphous thin film
of zinc tetra(4-carboxyphenyl)porphyrinmolecules acting as
sensitizer are optically excited, the porphyrin stimulates the
injection of electrons into the conduction band of TiO2 while
the remaining holes migrate toward the back electrode where
they are collected.
This phenomenon could be not important for the most
photoactive samples because the dye presumably is not
agglomerated onto the surface of TiO2 particles and it does
not form a definite solid. On the other hand, when the dye
is present as a layer of an agglomerate solid, occupying the
entire surface, no space is left for surface hydroxyl groups
and the photoactivity would be lower.
tion [see Eqs. (11)–(13)] and quantum yields of this process
are generally reported to be higher for porphyrin.
Some runs were carried out by using TiO2(A) or 6.65-
TiO2(A)–CuPp in the presence of furfuryl alcohol [29] with
1
the aim of demonstrating the presence of O2(1∆) in our
system. Unfortunately, they were unsuccessfully because
the photoreactivity decreased significantly in the presence
of furfuryl alcohol for both samples, suggesting that this
compound did not react selectively with singlet oxygen in
our heterogeneous system and/or a competition with 4-NP
on the surface sites could occur.
1
Literature reports that the route via O2(1∆) is the pre-
dominant process in both photodynamic therapy of cancer
and photooxidation of many biological and environmental
substrates [28,30].
Frackowiak and co-workers [31] report that the yields
of triplet state formation of some metal phthalocyanine via
i.s.c. (inter system crossing) producing 1O2(1∆) were higher
than for the investigated free phthalocyanine.
Finally, it should also be taken into account a possible
additional oxidation pathway in the homogeneous phase in-
volving a direct interaction between 1O2(1∆) [see Eq. (26)]
or other radical species and the substrate [see Eqs. (27)
and (28)].
Moreover it is known that aggregation and dimerization
reduce the lifetimes of the excited state of metal phthalocya-
nines, due possibly to enhanced radiationless excited state
dissipation and, therefore, quantum yields of the singlet oxy-
gen generation [32]. When this kind of sensitizer is sup-
ported and well dispersed onto TiO2 surfaces it is likely that
the above phenomena do not play a significant role. Conse-
quently, a monomolecular distribution of the dye onto the
TiO2 surface represents a favorable condition for obtaining
an improved photocatalytic activity.
4. Conclusions
TiO2 (anatase) samples impregnated with functionalized
Cu(II)–porphyrineand Cu(II)–phthalocyaninewere more ef-
ficient catalysts compared to bare TiO2 for the photodegra-
dation of 4-nitrophenol with Cu(II)–porphyrine showing
a more significant beneficial effect.

342
G. Mele et al. / Journal of Catalysis 217 (2003) 334–342
The photocatalytic activities of samples impregnated with
[7] K. Ozoemena, N. Kuznetsova, T. Nyokong, J. Mol. Catal. A 176
(2001) 29.
metal-free porphyrin and metal-free phthalocyanine were
always less significant than those of the corresponding
samples impregnated with the Cu(II) complexes.
[8] T. Matsuura, N. Yoshimura, A. Nishinaga, I. Saito, Tetrahedron 28
(1972) 4933.
[9] T. Matsuura, H. Matsushima, S. Kato, I. Saito, Tetrahedron 28 (1972)
5119.
[10] W. Spiller, H. Kliesch, D. Wöhrle, S. Hackbarth, B. Rodger, G.
Schnurpfeil, J. Porphyrins Phtalocyanines 2 (1998) 145.
[11] K.T. Ranjit, I. Willner, S. Bossman, A. Braun, J. Phys. Chem. B 102
(1998) 9397.
[12] T. Stuchinskaya, N. Kundo, L. Gogina, U. Schubert, A. Lorenz, V.
Maizlish, J. Mol. Catal. A 140 (1999) 235.
[13] V. Iliev, J. Photochem. Photobiol. A 151 (2002) 195.
[14] V. Iliev, D. Tomova, Catal. Commun. 3 (2002) 287.
[15] O.L. Kaliya, E.A. Lukyanets, G.N. Vorozhtsov, J. Porphyrins Phthalo-
cyanines 3 (1999) 592.
[16] M. Fujitsuka, O. Ito, H. Imahori, K. Yamada, Y. Sakata, Chem. Lett.
(1999) 721.
A tentative mechanism based on the formation of 1O2(1∆),
due to the presence of the sensitizers on the TiO2 surface, is
proposed. The rate of electron transfer to this very oxidant-
adsorbed species (in addition or alternative to 3O2) from the
conduction band of TiO2 and/or from Cu(I) species possi-
bly produced under irradiation is hypothesized to be mainly
responsible for the enhancement of the reaction rate of the
global photocatalytic process.
A contribution of a homogeneous oxidative process due
to the presence of 1O2(1∆) and/or other radical species
cannot be excluded.
Interestingly, we observed the maximum photocatalytic
activity when the molecular coverage was comparable with
BET specific surface areas measured for TiO2 samples.
[17] T. Nojiri, M.M. Alam, H. Konami, A. Watanabe, O. Ito, J. Phys.
Chem. 101 (1997) 7943.
[18] M. Fujitsuka, O. Ito, T. Yamashiro, Y. Aso, T. Otsubo, J. Phys.
Chem. 104 (2000) 4876.
[19] A. Tepore, A. Serra, D.P. Arnold, D. Manno, G. Micocci, A. Genga,
L. Valli, Langmuir 17 (2001) 8139.
[20] A. Maldotti, A. Molinari, R. Amadelli, Chem. Rev. 102 (2002) 3811.
[21] V. Hequet, P. Le Cloirec, G. Gonzalez, B. Meunier, Chemosphere 41
(2000) 379.
[22] G. Mele, G. Ciccarella, G. Vasapollo, E. García-López, L. Palmisano,
M. Schiavello, Appl. Catal. B 38 (2002) 309.
[23] J.N. Clifford, G. Yahiogly, L.R. Milgrom, J.R. Durrant, Chem.
Commun. (2002) 1260.
[24] V. Augugliaro, L. Palmisano, M. Schiavello, A. Sclafani, L. Marchese,
F. Martra, F. Miano, Appl. Catal. 69 (1991) 323.
[25] S.L. Murov (Ed.), Handbook of Photochemistry, Dekker, New York,
1973, p. 119.
[26] S. Shanmugathasan, C. Edwards, R.W. Boyle, Tetrahedron 56 (2000)
1025.
Acknowledgments
We thank the Ministero dell’Università e della Ricerca
Scientifica e Tecnologica (Rome) and Consorzio Interuni-
versitario Chimica per l’Ambiente (INCA) for financial
support. Prof. Tullio Caronna (Politecnico di Milano) is
gratefully acknowledged for helpful discussion and Tioxide
Huntsman for the gift of some bare TiO2 samples.
References
[1] M. Schiavello (Ed.), Photocatalysis and Environment. Trends and
Applications, Kluwer, Dordrecht, 1988.
[2] V. Augugliaro, L. Palmisano, A. Sclafani, C. Minero, E. Pelizetti,
Toxicol. Environ. Chem. 16 (1988) 89.
[3] D. Ollis, E. Pelizzetti, N. Serpone, Environ. Sci. Technol. 25 (1990)
1555.
[4] J.A.H. Meliàn, J.M. Dona Rodrìguez, A.V. Suàrez, E.T. Rendòn, C.
Valdès do Campo, J. Araña, J. Pèrez Peña, Chemosphere 41 (2000)
323.
[27] C.H. Lee, J. Lindsley, Tetrahedron 39 (2000) 11427.
[28] R. Gerdes, D. Wöhrle, W. Spiller, G. Schneider, G. Schnurpfeil, G.S.
Ekloff, J. Photochem. Photobiol. A 111 (1997) 65.
[29] K. Gollnick, A. Griesbeck, Tetrahedron 41 (1985) 2057.
[30] H. Hanabusa, H. Shirai, in: A.P.B. Lever, C.C. Leznoff (Eds.), Phtha-
locyanines: Properties and Applications, Vol. 2, VCH, New York,
1993.
[31] D. Frackowiak, A. Planner, A. Waszkowiak, A. Boguta, R.M. Ion,
K. Wiktorowicz, J. Photochem. Photobiol. A 141 (2001) 1001.
[32] A. Harriman, G. Poter, M.C. Richoux, Coord. Chem. Rev. 44 (1982)
83.
[5] V. Augugliaro, A. Di Paola, G. Loddo, G. Marcì, L. Palmisano, M.
Schiavello, Catalyst Catal. 41 (1999) 73.
[6] K. Ozoemena, N. Kuznetsova, T. Nyokong, J. Photochem. Photobiol.
A 139 (2001) 217.
[33] S. Cherian, C.C. Wamser, J. Phys. Chem. B 104 (2000) 3624.
[34] T.J. Savenije, A. Goossens, Phys. Rev. B. 11 (2001) 6411.