Separation and Purification Technology 34 (2004) 35–42

Sonocatalytic oxidation processes for the removal of contaminants

containing aromatic rings from aqueous effluents
Maria Papadaki a,∗ , Richard J. Emery a , Mohd A. Abu-Hassan b ,
Alex Dı́az-Bustos a , Ian S. Metcalfe b , Dionissios Mantzavinos c
aDepartment of Chemical Engineering, University of Leeds, Clarendon Road, Leeds LS2 9JT, UK
b Department of Chemical Engineering, UMIST, PO Box 88, Manchester M60 1QD, UK
c Department of Environmental Engineering, Technical University of Crete, Polytechneioupolis, GR-73100 Chania, Greece

Abstract

The effect of low frequency (20 kHz) ultrasonic irradiation on the removal of sodium dodecylbenzene sulfonate (SDBS),
phenol, 2-chlorophenol and 3,4-dichlorophenol from aqueous solutions has been investigated. Sonolytic degradation of SDBS
(at an initial concentration of 1 g/l) results in the formation of lower molecular weight compounds and is accompanied by
low total oxidation rates. In parallel, water sonolysis results in the formation of hydrogen peroxide. Several heterogeneous
catalysts (three noble metals and a metal oxide) were tested with respect to their effect on SDBS sonolysis. Of these, a CuO·ZnO
supported on alumina catalyst appears to enhance both SDBS fragmentation and total oxidation rates as well as hydrogen peroxide
formation. Phenolic compounds (at an initial concentration of 0.1 g/l) are only partially removed by ultrasonic irradiation with
2-chlorophenol being more susceptible to degradation than phenol and 3,4-dichlorophenol. However, the presence of Fe2+ ions
at concentrations as low as about 10−3 g/l generally increases the rate of the uncatalysed sonolytic degradation. This is attributed
to iron being capable of readily decomposing hydrogen peroxide in a Fenton-like process to form reactive hydroxyl radicals as
well as being an effective oxidation catalyst.
© 2003 Elsevier B.V. All rights reserved.

Keywords: Catalysts; Fenton oxidation; Phenols; Surfactants; Ultrasound; Wastewaters

1. Introduction treatment systems. Given that biological processes

have little effect in successfully treating biorefractory
The second half of the last century has witnessed wastewaters, several thermochemical and advanced
a rapidly deteriorating water environment as the out- oxidation processes have been developed for either
come of enormous use of complex organic com- the partial or complete destruction of hazardous and
pounds, the spent parts of which were discharged toxic organic pollutants.
into conventionally designed and operated wastewater Only very recently, has considerable interest been
shown in the application of an innovative treatment
∗ Corresponding author. Tel.: +44-113-343-2420; for hazardous chemical destruction based on the use
fax: +44-113-343-2405. of ultrasound. The chemical effects of ultrasound
E-mail address: [Link]@[Link] (M. Papadaki). derive from acoustic cavitation i.e. the formation,

1383-5866/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S1383-5866(03)00172-2
36 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42

growth and implosive collapse of cavitation bub- 2. Experimental and analytical

bles in a liquid. Extreme temperatures of several
thousand degrees and pressures of several hundred Sodium dodecylbenzene sulfonate (SDBS), phenol,
atmospheres are developed locally within the bub- 2-chlorophenol (CP) and 3,4-dichlorophenol (DCP)
bles during their collapse with these bubbles serving were supplied by Fluka and used without further pu-
as hot spot microreactors in an otherwise cold liq- rification. Aqueous solutions containing either 1 g/l
uid. Destruction of chemicals is usually achieved of SDBS or 0.1 g/l of the phenolic compound were
through a combination of pyrolytic reactions oc- prepared and loaded into a glass vessel. A horn-type
curring inside or near the bubble and hydroxyl sonifier (Ultrason 250 manufactured by Labplant, UK)
radical-mediated reactions occurring in the liquid capable of operating either continuously or in a pulse
bulk [1–3]. mode at a fixed frequency of 20 kHz and a variable
In previous studies, the beneficial effect of ultra- electric output power up to 250 W was used for son-
sonic irradiation on the removal of several target ication experiments. In all cases, sonication was per-
compounds from model aqueous solutions has been formed continuously at an output power of 125 W.
demonstrated. Such compounds include, amongst Sonication of 1 l SDBS solutions was carried out
others, phenol, chlorophenols, nitrophenols, polychlo- with or without catalysts. Several heterogeneous cat-
rinated biphenyls, pesticides, polycyclic aromatic hy- alysts were used in catalytic experiments as follows:
drocarbons and surfactants [4–10]. Typical treatment (a) noble metals: platinum (Pt), palladium (Pd) and
conditions include dilute aqueous solutions of initial ruthenium (Ru) at a noble metal concentration of 5%
concentrations less than about 10−3 M (such concen- w/w supported on alumina, and (b) a CuO·ZnO sup-
trations are typical of refractory compounds found in ported on alumina. In all cases, 0.05 g/l of catalyst
natural waters or in already treated effluents), low to was used in the form of slurry. All experiments with
medium ultrasound frequencies from 20 to 500 kHz SDBS were performed at ambient conditions without
and irradiation times from several minutes to several temperature control.
hours. These studies are comprehensively analysed in 0.05 l solutions containing phenolic compounds
a recent review paper [3]. were subject to various treatments as follows: (a)
However, it is notable that relatively few studies sonication alone (US), (b) sonication in the presence
report the use of ultrasonic irradiation in conjunction of Fe2+ ions (US + Fe2+ ), (c) oxidation by the Fen-
with an effective oxidation catalyst which could syner- ton reagent (Fe2+ + H2 O2 ), and (d) a combination
gistically enhance degradation rates and consequently of US and the Fenton reagent (US + Fenton). In all
allow milder treatment conditions to be employed, cases where iron ions were used as a homogeneous
thus reducing capital and operating costs. Ultrasonic catalyst, the appropriate amount of FeSO4 ·7H2 O was
irradiation in the presence of Cu2+ ions was em- added to the solution to give a metal concentration of
ployed as an effective pretreatment step coupled with 0.81 × 10−3 g/l; in those experiments where hydrogen
subsequent wet air oxidation to treat effluents from peroxide was added to the solution, its initial concen-
the petrochemical [11] and textile [12] industries. In tration was 0.35 g/l. All experiments with phenolic
further studies [8], the beneficial synergy between compounds were carried out at a constant liquid phase
Fe2+ ions and ultrasound to enhance 3-chlorophenol temperature of 35 ◦ C in a temperature-controlled wa-
degradation rates in aqueous solutions has been ter bath. Samples periodically drawn from the vessel
demonstrated. were analysed as follows.
The scope of this work is to study the effect of High performance liquid chromatography (HPLC)
low frequency ultrasound on the removal of various was used to follow concentration profiles of the orig-
organic pollutants, such as surfactants and chlori- inal organic substrates. Different analytical methods
nated phenols, typically found in wastewaters pro- were developed to analyse SDBS and phenols as fol-
duced by chemical processing. In particular, the lows: (a) SDBS was analysed on a 250 × 4.6 mm
synergy between ultrasonic irradiation and several ThermoHypersil C18 column using a gradient mobile
heterogeneous and homogeneous catalysts has been phase comprising a 70% acetonitrile solution (buffered
investigated. with 0.15 M sodium perchlorate) and pure water (also
 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42 37

buffered with 0.15 M sodium perchlorate) at a flowrate

of 1 ml/min and ambient temperature. The injection
volume was 20 ␮l and detection was through a UV
detector set at 210 nm, (b) phenols were analysed on
a 150 × 3 mm Waters RP8 SymmetryShield column
using 40:60 acetonitrile:water as an isocratic mobile
phase at 1 ml/min and ambient temperature. The in-
jection volume was 20 ␮l and detection was through
a UV detector set at 276 nm.
The liquid phase total organic carbon content (TOC)
was measured by catalytic combustion/non-dispersive
infra red (NDIR) gas analysis with a Shimadzu
5050 TOC analyser. Total carbon was measured first
followed by inorganic carbon and TOC was deter-
mined by subtracting inorganic carbon from total Fig. 1. TOC-time profiles during the sonication of SDBS solu-
carbon. The relative standard deviation of three sep- tions. --, uncatalysed; -䊏-, CuO·ZnO/Al2 O3 ; -䉱-, Pt/Al2 O3 ;
-䊐-, Pd/Al2 O3 ; -䊉-, Ru/Al2 O3 .
arate measurements was always less than 2%. TOC
proved to be useful in assessing the extent of to-
tal oxidation of organics to carbon dioxide that had
occurred. various catalysts. Of the various catalysts tested,
Chemical oxygen demand (COD) was determined CuO·ZnO/Al2 O3 appears to be the most effective in
by the dichromate method. The appropriate amount terms of TOC removal. In all cases though, the final
of sample was introduced into commercially available (i.e. after 120 min) TOC removal was never greater
digestion solution (Hach Europe, Belgium) containing than about 20%. The pH of the solution dropped from
potassium dichromate, sulfuric acid and mercuric sul- its initial value of 8 to 6.5–7 within 10–20 min of son-
phate. The mixture was then incubated for 120 min at ication and then remained almost constant throughout
150 ◦ C in a COD reactor (Model 45600-Hach Com- the course of reaction up to 120 min. These results
pany, USA) and the COD concentration was measured imply that SDBS sonolysis is accompanied by the
colorimetrically using a DR/2010 spectrophotometer formation of acidic compounds, and that their concen-
(Hach Company, USA). The average value of six sep- tration remains quite stable. At the conditions under
arate measurements was taken, and the deviation of consideration, these acidic compounds appear to be
them never exceeded 1.5% for the range of COD con- refractory to total oxidation; however, interestingly,
centrations measured. AD analysis showed that sonified solutions had de-
The concentration of hydrogen peroxide generated creased levels of detergency. For instance, a 30% AD
during sonication experiments was determined spec- decrease was recorded after 180 min of uncatalysed
trophotometrically at 410 nm using the titanium com- sonication. Fig. 2 shows the molecular weight (MW)
plexing method according to the procedures described distribution of original SDBS solution as well as of
in detail elsewhere [13]. In some cases, SDBS samples solutions obtained after 120 min of sonication with
were also analysed with respect to active detergency and without CuO·ZnO/Al2 O3 , where the UV detector
(AD) content using the crystal violet method [14]. response is plotted against the elution time from the
chromatographic column. All peaks of both sonication
runs have been magnified by a factor of 2.4 for clar-
3. Results and discussion ity. As can be seen, peaks originally present at elution
times greater than about 8.6 min which correspond to
3.1. Degradation of SDBS higher mass fractions of SDBS, do not appear in the
treated solutions, thus implying that partial fragmen-
Fig. 1 shows the normalised TOC–time profiles tation of SDBS has occurred. (Similar results were
during the sonication of SDBS with and without also obtained for those experiments performed with
38 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42

Fig. 2. The MW distribution of original and sonified SDBS solutions.

noble metal catalysts). It has to be pointed out that To test this, 0.15 l of pure water was subject to un-
separation of peak clusters to narrower MW ranges catalysed sonication and hydrogen peroxide concen-
was not possible with the present equipment and tech- tration was recorded as a function of the sonication
niques; consequently, quantification based on peak time. As can be seen in Fig. 3, hydrogen peroxide con-
areas was found to be inaccurate. However, for the centration appears to increase linearly with sonication
CuO·ZnO/Al2 O3 -catalysed run most of the remaining time (a correlation coefficient r2 = 0.996 is obtained
peaks are clearly smaller than those of the original for the linear regression of the concentration against
solution as well as of the corresponding uncatalysed time). These results are in agreement with those re-
run, and this is also consistent with the TOC profiles ported in previous studies [15,16], where the concen-
shown in Fig. 1. tration of hydrogen peroxide generated during water
Aqueous phase sonolysis is likely to result in the sonication (carried out at various experimental condi-
formation of hydrogen peroxide which may be formed tions such as frequencies between 20 and 800 kHz,
through the recombination of hydroxyl radicals at the water saturated with various gases (i.e. air, oxygen,
gas–liquid interface and/or in the solution bulk. More- nitrogen, argon)) was found to increase linearly with
over, if the solution is saturated with air or oxygen, increasing sonication time.
peroxyl (O2 H) and more hydroxyl radicals are formed Fig. 3 also shows hydrogen peroxide concentration–
in the bubble and the recombination of the former at time profiles during the sonication of SDBS with and
the interface and/or in the solution bulk results in the without catalysts. As can be seen, the concentration
formation of additional hydrogen peroxide [1,15,16]. of hydrogen peroxide reaches a maximum within
 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42 39

haviour of SDBS against hydroxyl radicals. From the

results presented so far, it could be hypothesised that
interactions between ultrasound and the metal cata-
lyst (i.e. CuO·ZnO/Al2 O3 ) result in increased rates of
formation of hydroxyl radicals that partly recombine
to hydrogen peroxide. Moreover, radicals are capable
of oxidising the organic substrate to lower organics
and eventually to carbon dioxide and water, a process
promoted by the catalytic activity of copper, which
is known [17] to be an effective oxidation catalyst.
It could be argued that the beneficial effects arising
from increased hydroxyl radical formation are com-
promised by the fact that radicals are wasted through
recombination reactions. This is so since hydrogen
Fig. 3. H2 O2 concentration–time profiles during the sonication
of SDBS solutions or pure water. --, uncatalysed, SDBS; -䊏-,
peroxide in its molecular form is a rather ineffective
CuO·ZnO/Al2 O3 , SDBS; -䉱-, Pt/Al2 O3 , SDBS; -䊐-, Pd/Al2 O3 , oxidant for the aqueous phase oxidation of organics.
SDBS; -䊉-, Ru/Al2 O3 , SDBS; -䊊-, uncatalysed, H2 O (secondary However, copper is known [18] to be an effective cat-
axis). alyst for hydrogen peroxide decomposition to several
reactive species (hydroxyl radicals mainly) and this
is consistent with the concentration profiles shown in
the first 30 min of sonication with CuO·ZnO/Al2 O3 , Fig. 3.
Pt/Al2 O3 and Ru/Al2 O3 after which it decreases
presumably due to its decomposition. Conversely, 3.2. Degradation of phenol, CP and DCP
hydrogen peroxide concentration remains relatively
low throughout the experiments with Pd/Al2 O3 and Figs. 4–6 show the normalised concentration–time
without catalyst, which implies that it is not formed profiles of phenol, CP and DCP, respectively, during
at appreciable concentrations or, if it is formed, it various treatments. As can be seen, all phenols are only
readily decomposes. In a recent study [6], the mecha- partially removed by US treatment alone with about
nisms of uncatalysed sonolytic (at 200 kHz) degrada- 10, 15 and 20% removals being recorded after 180
tion of various surfactants including SDBS amongst min for phenol, DCP and CP, respectively. Given that
others, have been investigated. It was found that (a) sonolytic degradation of phenol and chlorophenols
SDBS mainly decomposes at the gas–liquid interface
through both hydroxyl radical oxidation and pyrolytic
reactions (for instance, methane and carbon monoxide
were identified as the primary pyrolysis by-products)
rather than in the solution bulk. This was attributed
to the fact that SDBS, although a water soluble and
non-volatile molecule, tends to accumulate at the
gas–liquid interface. Interestingly, it was also found
that the concentration of hydrogen peroxide formed
during SDBS sonication experiments was substan-
tially lower than that formed during water sonication
(i.e. in the absence of substrate). Moreover, hydrogen
peroxide formation decreased with increasing initial
substrate concentration and was nearly completely
suppressed at SDBS initial concentration of 10−3 M
(this is about a third of the initial concentration em- Fig. 4. Concentration–time profiles of phenol during various treat-
ployed in this study), thus implying the scavenging be- ments. -䊉-, US; -䊊- US + Fe2+ ; -- US + Fenton; -䉱-, Fenton.
40 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42

Fig. 5. Concentration–time profiles of CP during various treat- Fig. 7. COD–time profiles during the sonication of phenolic com-
ments. -䊉-, US; -䊊- US + Fe2+ ; --, US + Fenton; -䉱-, Fenton. pounds. -䊉-, phenol; -䊊-, CP; --, DCP.

is known to occur predominantly through reactions dation of the organic substrate. As can be seen, with
involving hydroxyl radicals [5,8,16], and (b) water the exception of DCP, iron (at a concentration as low
sonolysis generates hydroxyl radicals partial recom- as 0.81 × 10−3 g/l) substantially increases the rate of
bination of which would result in the formation of substrate removal. This was more pronounced in the
hydrogen peroxide, it was decided to investigate the case of CP, where a 55% removal was recorded after
effect of Fe2+ ions on the degradation of phenols. The 180 min of sonication. These results are in line with
rational behind this is associated with the fact that those reported by Nagata et al. [8] who studied the
iron (a) is an effective oxidation catalyst for several sonolysis of 3-chlorophenol (initial concentration of
classes of organics capable of degrading the organic 10−4 M) at 200 kHz and found that degradation rates
radicals formed via reactions of free radicals with the increased by as much as 2.4-times in the presence of
organic substrate, and (b) readily decomposes hydro- 56 × 10−3 g/l of Fe2+ .
gen peroxide to regenerate more hydroxyl radicals It has to be pointed out that, in this series of ex-
(classic Fenton chemistry), thus promoting the oxi- periments, hydrogen peroxide concentration was not
determined analytically. However, an indication of
its presence in the reaction mixture came through
measuring the COD content of various samples taken
during a series of screening experiments with phenol
and several chlorophenols (including CP and DCP
amongst others). Fig. 7 shows normalised COD–time
profiles obtained during the sonication of phenol, CP
and DCP at an initial substrate concentration of 0.2
g/l. In contrast to what would have been expected,
COD values are scattered showing no clear decreas-
ing trend; interestingly, COD values, in several cases,
exceed that of the original solution, implying the pres-
ence in the reaction mixture of compounds that inter-
fere with the COD test. It is well documented in the
literature [19,20] that hydrogen peroxide has a posi-
tive effect on the COD content of organic-containing
Fig. 6. Concentration–time profiles of DCP during various treat- samples which is due to the reduction of potas-
ments. -䊉-, US; -䊊-, US + Fe2+ ; --, US + Fenton; -䉱-, Fenton. sium dichromate in an acidified solution to Cr3+
 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42 41

as follows: of hydroxyl radicals and hydrogen peroxide both

of which are formed during water sonolysis or/and
K2 Cr 2 O7 + 3H2 O2 + 4H2 SO4 pyrolytic reactions occurring in the vicinity of the
→ K2 SO4 + Cr2 (SO4 )3 + 7H2 O + 3O2 bubble. A stable and inexpensive material capable
of promoting the oxidation of organic substrate and
It has to be pointed out that increased COD values effectively utilising hydrogen peroxide would make a
are not likely to be due to chloride interference as this promising catalyst for the sonocatalytic degradation of
was prevented by the addition of mercuric sulphate organic pollutants. In this respect, a copper-containing
in the digestion solution. HPLC analysis also showed heterogeneous catalyst and homogeneous iron were
that the degradation of phenolic compounds is accom- capable of promoting the sonolytic degradation of
panied by the formation of several reaction interme- SDBS and phenols, respectively.
diates; however, the identity of such compounds was Current research focuses on the effect of various
not checked for. treatment conditions such as ultrasound frequency,
Phenols were also treated by the Fenton reagent output power and mode of operation (i.e. continuous
with and without ultrasonic irradiation. As can be vs. pulse mode sonication), volume and concentration
seen, all phenols are readily oxidised by the Fen- of the liquid phase and type of catalyst on the chem-
ton process with about 85% substrate removal be- ical and biological properties of the waste stream.
ing recorded after 130, 40 and 30 min for phenol, Such fundamental studies are needed to thoroughly
DCP and CP, respectively. However, coupling the Fen- understand and rationally develop the proposed
ton reagent with ultrasonic irradiation had an unex- process.
pectedly detrimental effect on the efficiency of the
Fenton process. It can be hypothesised that iron ions
and the ultrasound compete for hydrogen peroxide Acknowledgements
decomposition through different routes which could
possibly involve (a) catalytic decomposition to hy- The financial support of this work from Glaxo-
droxyl radicals, and (b) thermal decomposition (which SmithKline (UK), EPSRC and the Royal Society is
may occur in the vicinity of the cavitation bubble greatly acknowledged.
where extreme local conditions exist) to water and
oxygen rather than to reactive radical species. If this
is the case, hydrogen peroxide is partly consumed References
through wasteful reactions and this would explain the [1] N.H. Ince, G. Tezcanli, R.K. Belen, I.G. Apikyan, Appl.
decreased efficiency of the combined Fenton and ultra- Catal. B—Environ. 29 (2001) 167.
sound treatment. However, further investigations are [2] L.H. Thompson, L.K. Doraiswamy, Ind. Eng. Chem. Res. 38
needed to understand the mechanisms involved in the (1999) 1215.
process. [3] Y.G. Adewuyi, Ind. Eng. Chem. Res. 40 (2001) 4681.
[4] H. Okuno, B. Yim, Y. Mizukoshi, Y. Nagata, Y. Maeda,
Ultrason. Sonochem. 7 (2000) 261.
[5] S. Goskonda, W.J. Catallo, T. Junk, Waste Manage. 22 (2002)
4. Conclusions 351.
[6] B. Yim, H. Okuno, Y. Nagata, R. Nishimura, Y. Maeda,
Ultrason. Sonochem. 9 (2002) 209.
The sonolytic degradation of an anionic surfac-
[7] C. Wu, X. Liu, D. Wei, J. Fan, L. Wang, Water Res. 35
tant and three phenolic compounds with and without (2001) 3927.
several catalysts has been investigated. Sonolysis of [8] Y. Nagata, M. Nakagawa, H. Okuno, Y. Mizukoshi, B. Yim,
organic compounds in aqueous phase occurs through Y. Maeda, Ultrason. Sonochem. 7 (2000) 115.
complex reaction mechanisms and pathways de- [9] E. Taylor Jr, B.B. Cook, M.A. Tarr, Ultrason. Sonochem. 6
(1999) 175.
pending on the physicochemical properties of the
[10] H.M. Cheung, A. Bhatnagar, G. Jansen, Environ. Sci. Technol.
substrate as well as the operating conditions under 25 (1991) 1510.
consideration. Such mechanisms are believed to in- [11] M.N. Ingale, V.V. Mahajani, J. Chem. Technol. Biotechnol.
clude free radical reactions involving the participation 64 (1995) 80.
42 M. Papadaki et al. / Separation and Purification Technology 34 (2004) 35–42

[12] A.D. Dhale, V.V. Mahajani, Ind. Eng. Chem. Res. 38 (1999) [17] D. Mantzavinos, R. Hellenbrand, A.G. Livingston, I.S.
2058. Metcalfe, Water Sci. Technol. 35 (4) (1997) 119.
[13] H. Pobiner, Anal. Chem. 33 (1961) 1423. [18] J.F. Perez-Benito, Monatshefte für Chemie 132 (2001)
[14] C.E. Hedrick, B.A. Berger, Anal. Chem. 38 (1966) 791. 1477.
[15] C.A. Wakeford, R. Blackburn, P.D. Lickiss, Ultrason. [19] Y.W. Kang, M.J. Cho, K.Y. Hwang, Water Res. 33 (1999)
Sonochem. 6 (1999) 141. 1247.
[16] C. Pétrier, A. Francony, Ultrason. Sonochem. 4 (1997) 295. [20] I. Talinli, G.K. Anderson, Water Res. 26 (1992) 107.