Application of advanced oxidation processes for removing salicylic acid from synthetic wastewaters

Article abstract of DOI:10.1016/j.cclet.2009.08.001

In this study, advanced oxidation processes (AOPs) such as anodic oxidation (AO), UV/H₂O₂ and Fenton processes (FP) were investigated for the degradation of salicylic acid (SA) in lab-scale experiments. Boron-doped diamond (BDD) film electrodes using Ta as substrates were employed for AO of SA. In the case of FP and UV/H₂O₂, most favorable experimental conditions were determined for each process and these were used for comparing with AO process. The study showed that the FP was the most effective process under acidic conditions, leading to the highest rate of SA degradation in a very short time interval. However, the results showed that Ta/BDD films had high electrocatalytic activity for complete degradation of SA; even if it employs more time for complete elimination of the SA respect to FP. Additionally, AO led to a sixfold acceleration of the oxidation rate compared with the UV/H₂O₂ process. Finally a rough comparison of the specific energy consumption shows that AO process reduced the energy consumption by at least 90% compared with the UV/H₂O₂ process.

Full text of DOI:10.1016/j.cclet.2009.08.001

Available online at www.sciencedirect.com
Chinese Chemical Letters 21 (2010) 101–104
www.elsevier.com/locate/cclet
Application of advanced oxidation processes for removing
salicylic acid from synthetic wastewaters
b
Xue Ming Chen a, , Djalma Ribeiro da Silva , Carlos A. Martınez-Huitle
*
b,c,**
´
a Environmental Engineering Department, Zhejiang University, Hangzhou 310027, China
b
´
Universidade Federal do Rio Grande do Norte, CCET – Departamento de Quımica, Lagoa Nova – CEP 59.072-970, Brazil
c University of Milan, Department of Chemistry, Celoria 2, 20133 Milan, Italy
Received 17 February 2009
Abstract
In this study, advanced oxidation processes (AOPs) such as anodic oxidation (AO), UV/H2O2 and Fenton processes (FP) were
investigated for the degradation of salicylic acid (SA) in lab-scale experiments. Boron-doped diamond (BDD) film electrodes using
Ta as substrates were employed for AO of SA. In the case of FP and UV/H2O2, most favorable experimental conditions were
determined for each process and these were used for comparing with AO process. The study showed that the FP was the most
effective process under acidic conditions, leading to the highest rate of SA degradation in a very short time interval. However, the
results showed that Ta/BDD films had high electrocatalytic activity for complete degradation of SA; even if it employs more time for
complete elimination of the SA respect to FP. Additionally, AO led to a sixfold acceleration of the oxidation rate compared with the
UV/H2O2 process. Finally a rough comparison of the specific energy consumption shows that AO process reduced the energy
consumption by at least 90% compared with the UV/H2O2 process.
# 2009 Xue Ming Chen. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
Keywords: Advanced oxidation processes; Salicylic acid (SA); Anodic oxidation; Ta/BDD; Electrocatalytic activity
Different AOPs for the elimination of organic pollutants from wastewater, such as the FP, the photo-assisted FP,
UV/Fe3+-oxalate/H2O2, photocatalysis, ozone water system, Mn2+/oxalic acid/ozone, H2O2 photolysis, O3/UV and
AO, have been developed and investigated by several research groups as alternatives for treating industrial wastewater
containing organic pollutants [1–3]. These technologies consist mainly in destroying the contaminants by chemical
oxidation, so that mineralization of the contaminants to CO2 and H2O, or at least the transformation of these
compounds into harmless products, is achieved. Oxidation process using liquid hydrogen peroxide (H2O2) in the
presence of ferrous ion (Fe2+) produces free hydroxyl radicals (ꢀOH) (Fenton process [3]). These strong oxidants can
rapidly degrade a variety of organic compounds. Whereas, other processes such as UV/H2O2, the ꢀOH are generated
through photolysis of H2O2 under ultraviolet radiation [4]. In recent years, there is growing interest in AO for toxic and
refractory organic pollutants degradation [2]. It has many advantages including high oxidation efficiency, compactness
* Corresponding author.
´
** Corresponding author at: Universidade Federal do Rio Grande do Norte, CCET – Departamento de Quımica, Lagoa Nova – CEP 59.072-970,
Brazil.
´
E-mail addresses: chenxm@zju.edu.cn (X.M. Chen), carlosmh@quimica.ufrn.br (C.A. Martınez-Huitle).
1001-8417/$ – see front matter # 2009 Xue Ming Chen. Published by Elsevier B.V. on behalf of Chinese Chemical Society. All rights reserved.
doi:10.1016/j.cclet.2009.08.001

102
X.M. Chen et al. / Chinese Chemical Letters 21 (2010) 101–104
in equipment, and simplicity in operation. This process has been effectively used to degrade several pollutants [5,6],
oxidizing them, not only to CO2 and H2O, but to biodegradable products using mediated and/or direct electrochemical
oxidation (EO). SA is a key ingredient in many skin-care products. This aromatic organic compound is very toxic and
it is widely present in pharmaceutical industrial effluents, so-called ‘‘pharmaceutical pollution’’. SA can stimulate skin
and mucous membrane, and reacts with protein, and it can bring about tinnitus, qualm, naupathia and electrolytic
turbulence. Therefore, effective removal of SA from wastewaters is very important.
1. Experimental
FP and UV/H2O2 experiments were performed in the same degradation system, applying or not radiation,
containing with a high pressure mercury lamp TQ 718 of 250 W (Heraeus Noblelight Company in Germany). The
reactor was cylindrical (25 cm  5 cm of diameter), capacity of 0.5 L, and it was made of quartz glass available for
radiation transfer. Different [H2O2]/[SA] molar ratios, [Fe2+] and pH were used to establish the most favorable
experimental conditions in this work. The electrochemical reactor (ER) was 130 mm high and 30 mm in diameter, and
was operated in batch. More details about the ER can be found elsewhere [6]. The current density (J = 200 A mÀ2) was
applied with a Tacussel model PJT24-1 (24 V–1 A) potentiostat–galvanostat at 25 8C and 0.2 mol/L Na2SO4 as
supporting electrolyte. Ta/BDD was used as anode, and a stainless steel plate was used as cathode. Samples were
analyzed for determining the concentration changes of SA and by-products by High Performance Liquid
Chromatograph (Varian Model 9050/9012 HPLC equipped with a Nucleosil C18 column (4.6 mm  250 mm); mobile
phase: phosphoric acid (solution at pH 2): acetonitrile 35:65, at a flow rate of 0.4 mL/min; the UV detector was set at
300 nm). Chromatographic retention times were compared with those of commercially available standards. Chemical
oxygen demand (COD) was measured using COD reactor and direct reading spectrophotometer. SA solutions of 1000
or 800 mg/L (Fluka reagent) were prepared using ultrapure deionized water. For each experiment, the pH was adjusted
to around 4 with NaOH. H2O2 was added as a concentrated solution (3%). Fe2+ solutions were prepared with
FeSO4Á6H2O. In the case of the FP, pH was set at the desired value before the startup, and then a given amount of iron
salt was added and mixed very well with SA before the addition of H2O2. For runs using the UV/H2O2 system, H2O2
was injected into the reactor before the beginning of each run. The temperature of the solution during the reaction was
fixed at 25 8C and maintained constant by using a water thermostat.
2. Results and discussion
To determine the most favorable concentrations of Fe2+ and H2O2 in the FP, experiments were carried out at 14
different [H2O2]/[SA] molar ratios. After that, subsequent experiments were performed at the best [H2O2]/[SA] molar
ratio for five different Fe2+ molar concentrations (0.2, 0.4, 0.6, 0.8 and 1.0 mmol/L), obtaining that at
[Fe2+] = 0.6 mmol/L and [H2O2]/[SA] = 7 were achieved the better performances. pH effect on the SA removal
efficiency for FP and UV/H2O2 was studied, experiments were carried out in the pH range 2–14, under the above-
mentioned conditions, showed that pH ꢁ4–5 were achieved high SA removal. Fig. 1A shows the elimination of SA as
well as the formation and partial disappearance of the main reaction intermediates (2,3-dihydroxy benzoic acid (2,3-
DHBA); 2,5-dihydroxy benzoic acid (2,5-DHBA); and catechol (CT)) by FP. Complete elimination of SA was
achieved after 180 min. As for the intermediates, only 2,5-DHBA and CT were almost completely eliminated; an
amount of 2,3-DHBA remained in solution increasing the time of the complete elimination (adding more Fe2+ or H2O2
for the reaction). Further intermediates (malic, maleic, and fumaric acids) [7] were oxidized to oxalic acid. Oxalic acid
is slowly transformed into CO2, but it forms Fe/oxalate complexes mainly. In the case of UV/H2O2, it occurs under
irradiation (280 nm) of the pollutant solution containing H2O2. This causes the homolytic cleavage of H2O2 [1,7].
Different [H2O2]/[SA] molar radio and pH were tested in order to determine the most favorable conditions, obtaining
that [H2O2]/[SA] = 6 and pH 4 achieved the best performances (next experiments were performed under these
experimental conditions). Then, during the UV/H2O2 oxidation of SA, the main product was 2,3-DHBA, whereas CT
and 2,5-DHBA were produced in minor amounts. The complete elimination of SA was not achieved (Fig. 1B).
Complete elimination (SA and malic, maleic and fumaric acids) was verified by analytical techniques after 23.5 h. In
the AO at Ta/BDD electrode, several experiments were realized for 200 A mÀ2 and pH ꢁ4. Fig. 2A shows that the AO
of SA is completely achieved after 8 h of electrolysis, showing that insignificant amounts of CT, 2,3 and 2,5-DBHA
were detected. Based on HPLC measurements, the rate of SA and by-products oxidation is faster due to the reactivity

X.M. Chen et al. / Chinese Chemical Letters 21 (2010) 101–104
103
Fig. 1. Elimination of SA as function of time: (A) FP and (B) UV/H2O2. pH 4 adjusted with NaOH. Evolution of the main intermediates: 2,3-DHBA;
2,5-DHBA; and CT.
of ꢀOH formed at BDD surface. Marselli et al. [8] proposed a reaction mechanism for this process that involves the
participation of ꢀOH generated at the electrode surface through the EO of water. The elimination step should occur
with the organic molecules in the proximity of the surface layer of ꢀOH and via homogeneous reaction between the
organic molecules and the ꢀOH confined within a reaction cage nearby the electrode surface. These ꢀOH impede the
evolution of higher concentration of aromatics compounds and increasing the rate of transformation of possible
intermediates to aliphatic acids, and finally CO2. Analogous results were obtained and reported by Guinea et al. [9].
Other oxidants formed at the diamond surface (H2S2O8, O3) can participate in the oxidation of the SA and
intermediates formed [2]. These reagents can oxidize organic matter leading to an increase in removal rates [2]. SA
concentration in industrial wastewaters may vary from time to time, it is necessary to investigate AO of SA under
different initial concentrations. It was found that low residual COD level could be achieved for all solutions after
oxidation. COD was reduced from 166–830 to 1–42 mg/L, with removal efficiency of over 95%, indicating that this
anode can mineralize SA effectively. In contrast, low removal SA was mineralized by other processes [10]. On the
other hand, the variation of the instantaneous current efficiency (ICE) with COD for oxidation of different
concentrations of SA was studied. Clearly, ICE decreased with COD. This behavior can be ascribed to mass transfer
limitation due to the gradual decay in pollutant concentration [11].
Fig. 2. (A) AO of SA as function of time at Ta/BDD anode. pH 4 (adjusted with NaOH), J = 200 A mÀ2 and 25 8C. Inset: Reaction of SA with the
hydroxyl radicals to form hydroxylated aromatic compounds. (B) COD removal as function of time, depending on AOPs used. Values of constants
reaction rate for SA degradation: UV/H2O2: 1.29 Â 10À3 Æ 0.00003 minÀ1
;
AO: 5.49 Â 10À3 Æ 0.00007 minÀ1
;
FP: 10.10 Â 10À3
Æ 0.00068 minÀ1
.

104
X.M. Chen et al. / Chinese Chemical Letters 21 (2010) 101–104
The residual COD variation with different values of J for AO of SA at a given charge loading and it was observed
that the residual COD increased as the J increased at a given charge loading. The residual COD was 10 mg/L at
50 A mÀ2, but increased to 190 mg/L when the J increased to 300 A mÀ2. This is consistent with that reported by
Panizza et al. [12]. The increased current beyond the limiting J is not for oxidizing pollutants but for producing O2,
resulting in a decrease in CE, and a process determined by the mass transfer rate.
For comparing all oxidation methods, SA solutions with a content of 800 ppm of COD were used. For FP and UV/
H2O2, time processes to complete the elimination of SA were substantially different; shorter times for FP; and longer
times for UV/H2O2. However, as can be seen in Fig. 2B, at FP was necessary to provide an amount of H2O2 or/and Fe2+
for removing completely the intermediates formed (after 150 min and final stages). On contrary for AO, a complete
elimination was attained, without addition of other class of chemical reagents in solution. COD removal was about
98%. Finally, data were fitted to a first-order rate model, ln(Ct/C0 = Àk0t), where C0 and Ct are the concentration of SA
at times 0 and t, respectively; k0 is a first-order rate constant (in minÀ1), and t is the time (in min). The values of k0 are
reported in the inset of Fig. 2B, in agreement with the data discussed above about the removal rates processes. Energy
required for removing 95% of 800 mg dmÀ3 of COD of SA by UV/H2O2 and AO processes was obtained. UV/H2O2 is
less economical than the AO process (13.05 kWh kgÀ1 of COD by UV/H2O2, during 23.5 h (total time used for
complete SA oxidation)); with an energy increase of 200% for removing SA from wastewaters (consumed energy
estimated from the UV radiation used which represented only a part of the total electrical energy).
The output UVenergy amount of the lamp is about 250 W. Energy consumption for AO was about 4.45 kWh kgÀ1
of COD. This energy consumption could be reduced using J values below 200 A mÀ2. It is important to remark that, FP
is more economical than other two processes; however, several conditions must be considered to obtain good
performances.
3. Conclusions
Based on the preliminary results, it is possible to ascertain that the FP and AO processes can be viable techniques for
the treatment of industrial wastewater containing SA due to its efficiency and low operational cost. FP rate is
influenced by many factors, but [SA] removal efficiencies about 96%, were achieved after 150 min and 80% of COD
removal. For UV/H2O2 oxidation, [SA] removal efficiencies were about 54% and 63% of COD removal after 700 min
of treatment. Finally, AO achieved good removal efficiencies on [SA] and COD respect to other methods, and a
complete elimination of SA was attained by AO without including other class of chemical reagents in solution,
converting this method an important alternative in the wastewater treatment. Further AO runs of SAwill be performed
in order to have kinetic information and to optimize the electrolysis parameters. Detailed data and discussions will be
reported in a separate paper in a near future.
References
[1] R. Andreozzi, V. Caprio, A. Insola, et al. Catal. Today 53 (1999) 51.
[2] C.A. Martinez-Huitle, S. Ferro, Chem. Soc. Rev. 35 (2006) 1324.
[3] E. Brillas, R. Sauleda, J. Casado, J. Electrochem. Soc. 145 (1998) 759.
[4] O. Legrini, E. Olivieros, A.M. Braun, Chem. Rev. 93 (1993) 671.
[5] X.M. Chen, G.H. Chen, F.R. Gao, et al. Environ. Sci. Technol. 37 (2003) 5021.
[6] X.M. Chen, G.H. Chen, P.L. Yue, Chem. Eng. Sci. 58 (2003) 995.
[7] J.J. Pignatello, E. Oliveros, A. MacKay, Crit. Rev. Environ. Sci. Technol. 36 (2006) 1.
[8] B. Marselli, J. Garcia-Gomez, P.A. Michaud, et al. J. Electrochem. Soc. 150 (2003) D79.
[9] E. Guinea, C. Arias, P.L. Cabot, et al. Water Res. 42 (2008) 499.
´
[10] C. Adan, J.M. Coronado, R. Bellod, et al. Appl. Catal. A: Gen. 303 (2006) 199.
[11] X.M. Chen, G.H. Chen, Sep. Purif. Technol. 48 (2006) 45.
[12] M. Panizza, P.A. Michaud, G. Cerisola, et al. J. Electroanal. Chem. 507 (2001) 206.