Catalytic decomposition of the reactive dye Uniblue A on hematite. Modeling of the reactive surface

Article abstract of DOI:10.1016/S0043-1354(00)00295-5

Experimental results from the adsorption and subsequent catalytic combustion of the reactive dye Uniblue A on hematite indicate that this iron oxide can be used as an affordable catalyst for environmental purposes. Uniblue A was adsorbed on hematite and the products of the catalytic oxidation in O₂ atmosphere were analyzed by thermal programmed gas-chromatography/mass spectrometry (STDS-GC-MS) analysis. The catalytic combustion of Uniblue A in the presence of hematite led to about 40% conversion of the dye C-content into CO₂ at T=275°C. The activation energy (E(a)) for the desorption of CO₂ and other polyaromatic hydrocarbons (PAHs) from the hematite surface was determined to be 23.4kcalmol^-1. Identification of the species of Uniblue A in solution and those existing on the hematite surface was carried out in the framework of the generalized two-layer diffuse model. The modeling of the amount of dye absorbed on hematite is in good agreement with the experimental data. Copyright (C) 2001 Elsevier Science Ltd. Experimental results from the adsorption and subsequent catalytic combustion of the reactive dye Uniblue A on hematite indicate that this iron oxide can be used as an affordable catalyst for environmental purposes. Uniblue A was adsorbed on hematite and the products of the catalytic oxidation in O₂ atmosphere were analyzed by thermal programmed gas-chromatography/mass spectrometry (STDS-GC-MS) analysis. The catalytic combustion of Uniblue A in the presence of hematite led to about 40% conversion of the dye C-content into CO₂ at T = 275 °C. The activation energy (E_a) for the desorption of CO₂ and other polyaromatic hydrocarbons (PAHs) from the hematite surface was determined to be 23.4 kcal mol^-1. Identification of the species of Uniblue A in solution and those existing on the hematite surface was carried out in the framework of the generalized two-layer diffuse model. The modeling of the amount of dye absorbed on hematite is in good agreement with the experimental data.

Full text of DOI:10.1016/S0043-1354(00)00295-5

Wat. Res. Vol. 35, No. 3, pp. 750–760, 2001
# 2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(00)00295-5
0043-1354/01/$ - see front matter
CATALYTIC DECOMPOSITION OF THE REACTIVE DYE
UNIBLUE A ON HEMATITE. MODELING OF THE REACTIVE
SURFACE
F. HERRERA1, A. LOPEZ2, G. MASCOLO2, P. ALBERS and J. KIWI1*3M
1 Institute of Physical Chemistry II, EPFL, 1015 Lausanne, Switzerland; 2 CNR–IRSA Department
of Water Chemistry and Technology, Via F de Blasio 5, Bari 70123, Italy and 3 Degussa AG, Wolfgang,
D-63403 Hanau 1, Germany
(First received 30 September 1999; accepted in revised form 25 May 2000)
Abstract}Experimental results from the adsorption and subsequent catalytic combustion of the reactive
dye Uniblue A on hematite indicate that this iron oxide can be used as an affordable catalyst for
environmental purposes. Uniblue A was adsorbed on hematite and the products of the catalytic oxidation
in O2 atmosphere were analyzed by thermal programmed gas-chromatography/mass spectrometry (STDS-
GC-MS) analysis. The catalytic combustion of Uniblue A in the presence of hematite led to about 40%
conversion of the dye C-content into CO2 at T¼ 2758C. The activation energy (Ea) for the desorption of
CO2 and other polyaromatic hydrocarbons (PAHs) from the hematite surface was determined to be
23.4 kcal mol 1. Identification of the species of Uniblue A in solution and those existing on the hematite
surface was carried out in the framework of the generalized two-layer diffuse model. The modeling of the
amount of dye absorbed on hematite is in good agreement with the experimental data. # 2001 Elsevier
Science Ltd. All rights reserved
Key words}Uniblue A, catalytic combustion, adsorption, hematite, by-products, modeling
INTRODUCTION
tion. Other oxides like geothite, alumina, silica and
titania were also screened for the low-temperature
The treatment of textile dyeing waters by conven-
catalytic combustion of Uniblue A. Hematite
tional methods such as flocculation, reverse osmosis,
activated carbon adsorption, ozonization and ad-
vanced oxidation technologies (AOTs) is a common
practice but has drawbacks due to the increasing
number of refractory materials found in wastewater
under aerobic conditions (Zollinger, 1987). During
effluents and the difficulties in the complete removal
of color (Pulgarin and Kiwi, 1996; Roques, 1996).
varying amounts of dyes and their metabolites are
The cost of ozonization which is widely used still
needs to be reduced for competitiveness.
This study addresses an alternative technology for
of colors. They link through chemical bonding to
the removal of non-biodegradable reactive dyes like
Uniblue A and Remazol Brilliant Blue R (RB19)
the action of atmospheric gases and oxidants, and
from textile-mill effluents (Hoechst Portuguesa,
1997). The objective of this research is to explore
make these dyes suitable dyestuffs, make them
catalytic combustion after dye absorption on a
suitable substrate as a treatment for the destruction
Few studies have been devoted in recent years to the
of the reactive dye Uniblue A. This study reports on
(a-Fe2O3) was found to be the most suitable material
combining significant dye adsorption from solution,
acceptable kinetic efficiency and stability.
Textile dyes in general are non-biodegradable
the biological treatment of textile waste waters
sorbed in bioflocs (Ganesh et al., 1994). Reactive
dyes are produced in increasing quantities in a variety
fibers. These dyes are resistant to light degradation,
some acids and bases. But the same properties that
difficult to degrade (decolorize) in water bodies.
removal of this rather recent generation of reactive
(a) the adsorption process and modeling of the
dyes like Uniblue A and Remazol Brilliant Blue R
reactive dye species at the catalyst surface, (b) the
identification of the gas products generated during
the catalytic oxidation and finally (c) the physical
characterization of hematite by way of gas adsorp-
(Pasti and Crawford, 1991; Lonergan et al., 1995;
Ince et al., 1997; Herrera et al., 1999).
Scheme 1 shows the structure and hydrolysis of
Remazol Brilliant Blue (RB19) and the way it is
subsequently linked to fibers like cellulose. Scheme 1
shows how RB19 reacts and partly hydrolyses
forming a covalent bond with textile fiber/cellulose
*Author to whom all correspondence should be addressed.
Tel.: +41-21-693-3621; fax: +41-21-693-3621; e-mail:
john.kiwi@epfl.ch
750

Catalytic combustion of reactive dye Uniblue A
751
Scheme 1.
through the vinyl-sulfonate anchor reactive group.
This process involves the addition of a nucleophilic
–OH group of the fiber to the vinyl-sulfonate through
a Michael-type 1,4-nucleophilic addition mechanism.
The fixation rate between the reactive dye and the
fiber is usually below 90% (Camp and Sturrock,
1990).
MATERIALS AND METHODS
Materials
Uniblue
MW=506.5, lmax ¼ 594 nm with e594 ¼ 5570 M 1 cm
used as received. High surface area a-Fe2O3 (157.6 m2g
A salt was Aldrich 29,640-9 p.a. grade,
1
Fig. 1. Adsorption isotherm of Uniblue A on hematite. The
inset shows the linearization of the initial part of the
Langmuir isotherm. For other details see text.
.
Acid and bases used to adjust the pH were Fluka p.a. and
1
)
was prepared according to known procedures (Cornell and
Schwertmann, 1980). Transmission electron microscopy
(TEM) showed crystals with a narrow size distribution of
20–30 nm.
same molar absorption characteristics for the next 24 h.
Independent tests proved that equilibrium was reached
within this time period. The suspension containing the
hematite and the dye was centrifuged at 14,000 RPM for
15 min. The supernatant containing the non-adsorbed
Uniblue A was analyzed spectrophotometrically to deter-
mine the concentration of Uniblue A in solution at
l ¼ 594 nm. The results are reported in Fig. 1.
Adsorption isotherm
To a solution containing a-Fe2O3, 400 mg l 1 aliquots of
Uniblue A (1.69 mM) were added keeping the pH at 5.5.
After the addition of each aliquot of Uniblue A and
equilibration during 5–10 min, the solution presented the

752
F. Herrera et al.
Determination of the pKa values of Uniblue A and a-Fe2O3
species in the solution were computed by this method. The
generalized two-layer diffuse model (diffuse/GTL) has
been chosen to account for the electrostatic interaction
between dye species at the iron-oxide surface. It takes into
account the specific surface area, the iron-oxide surface
pKa values, the concentration of oxide surface sites and
the ionic strength of the solution. The numerical calcula-
tions for some of the parameters used in the Chem EQL
matrix were carried out with a MatLab 5.2 (Math Works
Inc.). The results from these calculations are shown in
Figs 2 and 3.
The pKa values for Uniblue A and a-Fe2O3 have been
determined with a Tritrino DMS 716 titrator of Metrohm
AG Herisau, Switzerland. The titration was carried out with
a solution of NaOH 0.01 N. The pKavalues of Uniblue A
were: for the sulfonic group ꢀ 1, for the primary amine
group=3.97 and for the secondary amino group=6.37. The
two pKa values of a-Fe2O3 were found to be 6.7 and 10.4.
Catalytic combustion of Uniblue A on hematite
The dye Uniblue A (1.69 mM) was fully adsorbed on
hematite (0.4 g l 1) at pH 5.5. This is equivalent to a loading
of 2137 mg of Uniblue A per gram hematite. The dye during
the adsorption process separates from the solution. After
drying, the samples were weighed and put in a cylindrical
oven. The samples were introduced into the oven in the
form of a tablet. The temperature in the oven and the O2 gas
flow (15 ml min 1) were stabilized prior to each experiment.
Subsequently, the samples were reacted for 30 s at each of
the temperatures shown in Fig. 5 and the CO2 collected in
the vessel provided for this purpose with NaOH. Subse-
quently, the carbonate was titrated with HCl.
Adsorption as a function of pH
A known amount of hematite (1.5 g l 1) was added to
Uniblue A (0.197 mM) and the pH was adjusted by addition
of HCl under continuous stirring. The temperature was
maintained at 228C in a thermostated bath to avoid losses
due to evaporation. After a 15 min equilibration period,
aliquots were removed from the suspension and centrifuged
at 14,000 RPM for 15 min. The concentration of Uniblue A
in the supernatant was determined spectrophotometrically
and subtracted from the initial concentration to determine
the adsorbed amount.
Concentration of hematite active surface sites
The surface site density was determined from fluoride
adsorption. To carry out the measurements a hematite
1
solution containing 2 g l
was used and the range of
5
fluoride concentrations was varied from 1 Â 10
to
concentration in the supernatant
1 Â 10 3 M. The F
solution was determined with the help of a Dionex Ion-
Liquid Chromatograph (ICL) provided with an AG14
column. The maximum extent of F was found by fitting
the experimental data to the linearized Langmuir equation.
BET area and hematite surface parameters
Surface area, pore size diameter, and porosity were
determined with a BET Carlo Erba Sorptomatic-1900 unit.
The N2 desorption isotherm was calculated using the
cylindrical pore model of BJH-model (Barret et al., 1951).
Hematite presented a surface area of 157.7 m2 g
1
.
Modeling
The ionization of Uniblue A as a function of pH was
calculated using the pKa values found experimentally via the
ChemEQL program (version 2.0. for Macintosh written by
B. Muller, Kastanienbaum, Switzerland). The equilibria
Fig. 3. (a) Uniblue A species existing in solution and
hematite surface species as a function of pH. (b) Adsorption
of Uniblue A (1.69 mM) on hematite (0.4 g/L) as a function
of solution pH. Experimental and calculated values. For
other details see text.
Fig. 2. Distribution of the Uniblue A in solution vs. pH.

Catalytic combustion of reactive dye Uniblue A
753
System for thermal degradation studies coupled gas-
chromatography and mass-spectrometry (STDS-GC-MS)
measurements
SIL 8CB with i.d.=0.32 mm (60 m length) and 0.25 mm film
thickness. The initial temperature of the column was 608C
1
and then increased at 158C min
measurements.
up to 3208C during the
The system has been recently described (Mascolo et al.,
1997) and allows the identification of combusted organic
species in air. The instrumentation consists of four
integrated components: (a) controlling and data acquisition
unit, (b) thermal reaction compartment where the air was let
in at a flow rate of 3.0 ml min 1, (c) cryogenic trapping gas-
chromatograph (GC) and finally (d) a quadrupole mass
spectrometer. The GC-column used was a Chromopack CP-
In a typical experiment, 2 mg of Uniblue A (by itself or
adsorbed on hematite) was placed into a quartz capillary
tube (1.5 mm i.d., 15 mm length from Vitro Dynamics,
Rochaway, NJ, USA). The quartz tube was inserted into the
probe for 5 min periods at the temperatures specified in
Tables 1 and 2. The organic compounds formed during the
incineration step were transferred and subsequently kept at
Table 1. By products identification during Uniblue combustion by STDS
No.
Compound
Retention time
2008C
3008C
3758C
4508C
6008C
1
2
SO2
CH3CHO
1–2
6.3
X
X
X
X
X
X
3
4
10
X
X
CH3CN
10.4
10.6
X
X
5
6
X
X
X
X
X
10.8
X
X
X
7
8
HCl
11
X
X
CH3COOH
11.3
9
12.5
15.8
16.4
17.7
X
X
X
X
X
X
X
X
10
11
12
X
X
X
X
X
X
13
14
18.7
29.3
X
X

754
No.
F. Herrera et al.
Table 2. By-products identification during Uniblue on hematite combustion by STDS
Compound
Retention time
2008C
3008C
3758C
4508C
6008C
1
2
SO2
CH3CHO
1–2
6.3
X
X
X
X
X
X
X
X
3
10
X
X
X
4
5
CH3CN
10.4
10.6
X
X
X
X
X
X
X
X
X
6
10.8
X
X
7
8
HCl
CH3COOH
11
11.3
X
X
9
12.5
14.1
15.8
16.4
17.7
18.7
X
X
X
X
X
X
X
10
11
12
13
14
X
X
X
X
X
X
X
X
15
29.3
X
X
X
X
X
2508C to analyze the organic compound without degrada-
tion. These compounds were then transferred into
the cryogenic trapping GC-chamber where they were
RESULTS AND DISCUSSION
Adsorption isotherm experiments
focused on
a capillary GC column, separated by a
Figure 1 shows the adsorption isotherm of Uniblue
A on hematite in a plot of the number of adsorbed
temperature gradient and identified by mass spectrometry
(MS).

Catalytic combustion of reactive dye Uniblue A
755
moles (Nads
concentration of dye used. Details of the determina- and forms the species L with pKa3 ¼ 6:37 from the
tion have been reported (see Materials and methods amino group. The species HL has relevant
section). The adsorption has been carried out at pH concentration at pH values between 2 and 6.
)
as
a
function of the equilibrium pKa2ꢀ 3:97. HL in Fig. 2 is present as a zwitterion
a
5.5. This pH has been chosen after preliminary
The positively charged species H3L2+ is shifted to
experiments which revealed that: (a) at pH55 more acid pH values than H2L+ and these species
dissolution of the iron surface takes place leading coexist in solution with the neutral HL species.
to iron-complexes. The absorption of these com-
The distribution of the surface species of hematite
plexes interferes with the optical absorption of as a function of pH has already been reported
Uniblue A. It was also observed that at pH>5.5 (Bandara et al., 1999). Neutral surface sites, denoted
the adsorption of Uniblue A on hematite begins to by >FeOH, predominate in the pH region between
decrease with increasing pH (Fig. 3(a)). Figure 1 6.5 and 11. The >FeOH+2 sites increase from a
shows monolayer adsorption of Uniblue A on negligible fraction at pH 8 to 50% at pH 6 and to
hematite. The inset of Fig. 1 shows the linearization 100% at pH 4. The >FeO fraction begins to be
of the Langmuir isotherm at concentrations of meaningful at pH>8 reaching 100% at pH 14. The
Uniblue A below 0.2 mM. The equilibrium constant generalized two-layer model (diffuse/GTL) is used to
for the dye adsorption at the iron-oxide surface can model the hematite surface species (Stumm and
be estimated from
Morgan, 1996). To calculate the surface species
distribution of hematite the values considered were
Nads
ka½UniblueŠ
1
¼
ð1Þ pKa 6.7 and 10.4, BET area of 150 m2 g
and
Nsites kd þ ka½UniblueŠ
1
4.7 Â 10 3 mol g
of sites gram of a-Fe2O3.
where Nads is the number of adsorbed molecules of
Uniblue A on a-Fe2O3, Nsites is the number of a-
Fe2O3 sites, ka is the adsorption constant and kd is
the desorption constant for Uniblue A molecules on
a-Fe2O3. Linearizing equation (1)
Modeling the interaction of Uniblue A and hematite
Figure 3(a) presents the modeling of the speciation
of Uniblue A when interacting with hematite as a
function of pH. The surface complexation taking
place between Uniblue A and the hematite surface
sites is analyzed in terms of equations (3)–(11). The
criteria used to establish these equations are: (a)
sorption is only possible at specific surface coordina-
tion sites; (b) sorption reactions can be described by
mass law equations; (c) the surface complex modeling
½UniblueŠ Nsites ½UniblueŠ
¼
þ
ð2Þ
Nads
K
Nsites
where the equilibrium constant K ¼ ka=kd has a
1
value of K¼ 1:73 Â 106 L mol
Uniblue A and hematite speciation in solution
.
Scheme 2 shows the Uniblue A structure as a does not consider the charges of the dye species in
function of solution pH in the absence of hematite. In solution and (d) the effect of surface charge on
Fig. 3 the H3L2+ is the Uniblue A species in the acid sorption is taken into account by applying
a
range protonating the two amino groups. The species correction factor from the electric double-layer
H2L+ originates from H3L2+ by losing one proton theory to the mass law constants for surface reactions
from the sulfonic group at a pKa around 1. HL (Diffuse/GTL) and finally (e) an incoming ligand like
originates from the deprotonation of H2L+ at Uniblue A will coordinate to one or two Fe sites
Scheme 2.

756
F. Herrera et al.
forming a one or two Fe-centered surface complex. hematite proceeds with four possible complex stoi-
Fig. 3(a) involves the species >FeOH+2 , >FeOH, chiometries (a) two complexes with a 1 : 1 ratio of
>FeL complex, L and HL. The residual species are Fe : Uniblue A ligand and (b) two further complexes
present at a very low concentration (50.05 mM) and with a 2 : 1 ratio of Fe and Uniblue. Surface
the experimental points overlap on a straight line at complexes are denoted as > FeHL+, >Fe2HL2+
,
the lower end in Fig. 3(a).
The matrix used for the ChemEQL calculations
takes into account only the hematite charges
>FeO( 1), >FeOH (0), >FeOH+2 (+1) and of Catalytic combustion of Uniblue A on hematite
>FeL and >Fe2L+.
the complexes >FeHL+(+1), >Fe2HL2+(2),
The amount of CO2 for each sample (Fig. 4) was
normalized to the weight of each of the tablets. Below
2008C (473 K), the quantity of gas collected was
extremely small. At temperatures between 275 and
4008C in Fig. 3 about 40% of the initial C-content of
the Uniblue A adsorbed on hematite was detected as
CO2. In addition, about 6% of the initial C-content
(ashes) was found by elementary analysis on the
hematite surface. But about 50% of the total organic
carbon was transformed into smaller molecules
which desorb from the hematite surface. These
incomplete combustion gaseous by-products are
>FeL(0) and >Fe2L+(1). The matrix is established
with the known pKa values for Uniblue A and
hematite, as stated above. The Keq values for the
complexes >FeHL+, Fe2HL2+
,
>FeL and
>Fe2L+ were found from the mass equations and
the sites balance taking into account: (a) the amount
of L (Uniblue A in solution) from spectral observa-
tions and (b) the surface active sites before and after
Uniblue A adsorption on the oxide surface. Solving
the four simultaneous equations (8)–(11) by MatLab
5.2 allowed to obtain the values: 8.66, 17, 5, 2.6
and 7.05 for Keq of the complexes >FeHL+,
>Fe2HL2+, >FeL and >Fe2L+.
reported in the section below. The units in the inset
1
of Fig. 4 are s
because the reaction was observed
The reaction schemes 3–11 modeled in Fig. 3(a)
follows the criteria outlined above.
to be pseudo-first order in excess O2 atmosphere.
Uniblue A was combusted in the absence of
hematite to assess the influence of hematite on the
combustion of Uniblue A. The control points (Ctrl)
in Fig. 4 show the combustion in the absence of
hematite. In this case the amounts detected were 2%
CO2 at 2758C (548 K) and 27% CO2 at 3508C
(623 K). The amounts of CO2 are far below the
amounts of CO2 shown in Fig. 4 for the combustion
of Uniblue A at the same temperatures. The C-ashes
in this case were about 8% and the by-products of
the order of 65% (3508C). The latter result shows the
drastic enhancement of the amount of Uniblue A
undergoing complete oxidation due to hematite
during catalytic combustion. A much higher com-
plete oxidation percentage is observed in terms of
CO2 as compared to the case when the combustion
was carried out in the absence of hematite.
H3L2þ Ð H2Lþ þ Hþ
H2Lþ Ð HL þ Hþ
ð3Þ
ð4Þ
ð5Þ
ð6Þ
ð7Þ
HLþ Ð L þ Hþ
> FeOHþ2 Ð > Fe2OH þ Hþ
> FeOH Ð > Fe2O þ Hþ
> FeOHþ2 þ H2Lþ Ð > FeHLþ þ H2O þ Hþ ð8Þ
2 > FeOHþ2 þ H2Lþ Ð > Fe2HL2þ þ 2H2O þ Hþ
ð9Þ
> FeOHþ2 þ HL Ð > FeL þ H2O þ Hþ
ð10Þ
2 > FeOHþ2 þ HL Ð > Fe2Lþ þ 2H2O þ Hþ ð11Þ
Fig. 3(b) can be understood in terms of reactions
(3)–(11). This latter set of equations was found to
match the experimental points obtained for the
Uniblue absorption as a function of pH (Fig. 3(b)).
The combination of reactions shown by equations (3)
and (11) fits the experimental data and seem to be the
reactions responsible for the observed absorption.
Fig. 3b shows that Uniblue A adsorption begins at
pH 8 and increases to pH 5.5 where it levels off. At
pH>8 no adsorption is observed since the adsorbent
and adsorbate exist in their negative form: L and
>FeO . The >FeOH+2 interacts electrostatically
with the Uniblue A negative SO23 -group (L , Fig. 2)
at pH58 leading to the adsorption of the Uniblue A
on hematite. The adsorption of Uniblue A on
Fig. 4. Generation of CO2 in air during the catalytic
combustion of Uniblue A adsorbed on hematite as a
function of temperature. The control points (Ctrl) show
the combustion in the absence of hematite. The inset shows
the plot used for the determination of the activation
energy (Ea).

Catalytic combustion of reactive dye Uniblue A
757
System for thermal degradation studies coupled
gas-chromatography and mass-spectrometry
(STDS-GC-MS) measurements
in Table 1. This by-product in the presence of
hematite is formed at temperatures as low as
2008C (Table 2). In the absence of hematite it
appears only at 6008C (Table 1). This latter effect
Figures
chromatograms of the thermal products obtained the iron oxide during the combustion of Uniblue.
for the Uniblue A alone and Uniblue A adsorbed In addition Table shows the formation of
on hematite as function of three different cholorobenzene (by-product 10). This by-product
5 and 6 show the STDS-GC-MS is indicative of the favorable catalytic effect of
2
a
temperatures. Tables 1 and 2 report the gaseous is not found in the absence of hematite and can
by-products found and report the structure of be ascribed to residual chloride content of iron
these compounds at 200, 300, 375, 450 and oxide (Cornell and Schwertmann, 1980). During
6008C. By-product 10 in Table 2 is not observed combustion processes chlorides in the presence
Fig. 5. STDS-GC-MS chromatograms of Uniblue A at 200, 300 and 4508C. For further details identifying
the peaks see Table 1.

758
F. Herrera et al.
Fig. 6. STDS-GC-MS chromatograms of Uniblue A adsorbed on hematite at 200, 300 and 4508C. For
further details identifying the peaks see Table 2.
Table 3. Surface properties of hematite determined by physisorption measurements before and after catalytic combustion cycles with
Uniblue A at 2738C
Hematite alone
Hematite+Uniblue
Cycle 1
Cycle 2
1
BET surface area (m2 g
)
157.7
0.0046
13.37
171.0
70.8
131.9
0.0043
8.17
171.5
70.5
101.9
0.0024
1.58
165.2
65.4
85.1
0.0020
1.20
170.1
62.8
Micropore volume (cm3 g
1
)
Micropore area (m2 g
1
)
Average pore diameter (A)
Porosity (%)
of iron give rise to the formation of HCl which compounds like
formaldehyde,
acetaldehyde
is transformed into molecular chlorine leading to and glyoxal and glyoxal-derivatives has been
the formation of chlorinated organic compounds. detected besides the compounds reported in
In addition, the formation of additional organic Tables 1 and 2.

Catalytic combustion of reactive dye Uniblue A
759
Fig. 7. (a) Pore volume as a function of pore diameter of hematite loaded with Uniblue A after the first
catalytic combustion at 2758C. (b) Pore volume as a function of pore diameter of hematite loaded with
Uniblue A after the second catalytic combustion at 2758C.
The compounds evolved in the catalytic combus- hyde, acetone, acetonitrile and acetic acid of Tables 1
tion as shown in Table 2 are irritating if inhaled in and (Merck Index, 1989). Open-air catalytic
2
large amounts leading to respiratory disorders. The combustion of Uniblue A will not be hindered by
former consideration applies to: benzene, tiophene, the former consideration. This is due to the relative
phenol, aniline, naphthalene, chlorbenzene, acetalde- amounts involved in open-air combustions.

760
F. Herrera et al.
Physical characterization of the catalyst by N2
physisorption
(OFES Contract No. 96.350 Bern) is gratefully acknowl-
edged.
The specific surface area, the pore average size
diameter and the porosity of hematite with adsorbed
Uniblue A before and after two catalytic combustion
cycles of Uniblue A at 2738C do not show any
meaningful variation. This is shown in Table 3. For
hematite alone the values found are shown in the first
column. The pore average size diameter is seen to
vary between 171.0 and 170.5 A for hematite and
hematite with Uniblue A, respectively. No change
occurs in the pore size after the first and second
catalytic cycles as can be seen from Figs 7a and b,
the system being hematite with Uniblue A. The
parameters in Table 3 were calculated as reported
recently by our group (Bacsa and Kiwi, 1998). The
hematite density of 5.82 g cm 3 was used to calculate
the sample porosity.
REFERENCES
Bacsa R. and Kiwi J. (1998) Effect of the rutile phase on the
photocatalytic properties of nanocrystalline titania during
the degradation of p-coumaric acid. Appl. Catal. B. 16,
19–29.
Barret E., Joyner L. and Halenda P. (1951) The calculation
of pore size distribution from adsorption data. J. Amer.
Chem. Soc. 73, 373–381.
Bandara J., Mielczarski A. J. and Kiwi J. (1999) Molecular
mechanism of surface recognition. Azo-dyes degradation
on Fe, Ti, and Al-oxides through metal sulfonate
complexes. Langmuir 15, 7670–7679.
Camp S. and Sturrock P. (1990) Covalent reactive dye-fiber
linkage for vinyl sulfone remazols. Water Res. 24,
1275–1281.
Cornell R. and Schwertmann U. (1980) Preparation of large
surface area iron-oxides. Colloid Polym. Sci. 258,
1171–1225.
Ganesh R., Boardman G. and Nicholson D. (1994) Fate of
azo-dyes in sludges. Water Res. 28, 1367–1376.
Herrera F., Kiwi J., Lopez A. and Nadtochenko V. (1999)
Photochemical decoloration of remazol brilliant blue and
uniblue A in the presence of Fe3+ and H2O2. Environ. Sci.
Technol. 33, 3145–3152.
CONCLUSION
Adsorption and catalytic combustion of the
reactive dye Uniblue A has been studied on hematite. Hoechst Portuguesa, S. A., 1997. Aqua Ambiente, Aparta-
do 6, P-2726. Martins Cedex, Portugal.
Ince N., Stefan M. and Bolton J. (1997) UV/H2O2
Degradation and toxicity reduction of textile azo dyes:
remazol black-B, a case study. J. Adv. Oxid. Technol. 2,
Complete combustion of Uniblue A (as reflected
by the amount of CO2 evolved) was observed to be
the preferential process for Uniblue A adsorbed
on hematite. Incomplete oxidation products, on the
other hand, predominate when Uniblue A was
combusted in the absence of hematite. At the same
temperatures the formation of gaseous by-products
in the latter case was observed to be higher than in
the presence of hematite. Modeling of the surface
442–448.
Lonergan T., Panow A., Jones C. and Mainwaring D.
(1995) Physiological and degradative behavior of white-
rot fungus in a 200 liter packed bed bioreactor.
Fermentation Technol. 5, 107–117.
Mascolo G., Spinosa L., Lolito V., Mininni G. and Bagnolo
G. (1997) Lab-scale evaluations on formation of products
of incomplete combustion in hazardous waste incinera-
tion: influence of process variables. Water Sci. Technol.
36, 219–227.
Merck Index, 1989. Merck and Co. Rahway, NJ, USA.
Pasti M. and Crawford D. (1991) Relationships between
the abilities of streptomycetes to decolorize anthron-type
dye and degrade lignocellulose. Can. J. Microbiol. 37,
902–907.
Pulgarin C. and Kiwi J. (1996) Overview on photocataly-
tical and electrocatalytical treatment of industrial non-
biodegradable pollutants and pesticides. Chimia 50,
50–55.
Roques H. (1996) Chemical Water Treatment Principles and
Applications. VCH, Weinheim, Germany.
speciation and the Uniblue
A charged-ions in
solution allows for a quantitative prediction of the
amount of dye adsorbed as a function of pH. In the
main, electrostatic effects between the oxide surface
and the Uniblue A species on the hematite surface
determine the extent of adsorption as a function of
pH. The activation energy was determined for the
catalytic combustion of the adsorbed Uniblue A on
hematite. Physical characterization of the catalyst by
N2 physisorption was carried out and gives insight
into the nature of the surface responsible for the
catalytic combustion.
Stumm W. and Morgan J. (1996) Aquatic Chemistry. Wiley
Interscience, New York.
Acknowledgements}The financial support of the European Zollinger H. (1987) Properties and Applications of Organic
Communities Environmental Program ENV-CT-95-0064 Dyes and Pigments. VCH, Pub, New York.