Decomposition of 2-naphthalenesulfonate in electroplating solution by ozonation with UV radiation

(1)

Decomposition of 2-naphthalenesulfonate in electroplating

solution by ozonation with UV radiation

Yi-Hung Chen

a

, Ching-Yuan Chang

a,∗

, Shih-Fong Huang

a

, Neng-Chou Shang

a

, Chun-Yu Chiu

b

,

Yue-Hwa Yu

a

, Pen-Chi Chiang

a

, Je-Lueng Shie

c

, Chyow-San Chiou

c

a_{Graduate Institute of Environmental Engineering, National Taiwan University, Taipei 106, Taiwan} b_{Department of Environmental Engineering, Lan-Yang Institute of Technology, I-Lan 261, Taiwan}

c_{Department of Environmental Engineering, National I-Lan University, I-Lan 261, Taiwan} Received 8 August 2004; received in revised form 21 October 2004; accepted 27 October 2004

Available online 9 December 2004

Abstract

This study investigates the ozonation of 2-naphthalenesulfonate (2-NS) combined with UV radiation in the electroplating solution. 2-NS is commonly used as a brightening and stabilization agent in the electroplating solution. Semibatch ozonation experiments were conducted under various reaction conditions to study the effects of ozone dosage and UV radiation on the oxidation of 2-NS. The concentrations of 2-NS were analyzed at specified time intervals to elucidate the decomposition of 2-NS. Total organic carbon (TOC) is chosen as a mineralization index of the ozonation of 2-NS. In addition, values of pH and oxidation reduction potential were continuously measured in the course of experiments. As a result, the nearly complete mineralization of 2-NS via the ozonation treatment can be achieved. The mineralization of 2-NS is found accelerated by the introduction of UV radiation and has a distinct relationship with the consumption of applied ozone. These results can provide useful information for the proper removal of 2-NS in the electroplating solution by the ozonation with UV radiation.

Keywords: Ozone; Ozonation; UV radiation; 2-Naphthalenesulfonate; Electroplating solution

1. Introduction

Discarded aged electroplating solution is one of the major wastewater sources in the printed wiring board (PWB) indus-try. The substrates (the major chemical species) of solution recipe are inorganics such as sulfuric acid, copper sulfate, hydrochloric acid, etc., while the minor substances are or-ganics such as 2-naphthalenesulfonate (2-NS), which is com-monly used as a brightening and stabilization agent[1]. Con-sequently, the characteristics of wasted electroplating solu-tion exhibit high acidity (pH = 0.18–0.42) and ionic strength. All of the above features make the solution hard to be treated by the conventional treatment processes[1,2]. The current major method used to treat the waste electroplating solution of PWB is chemical coagulation, which produces hazardous

∗_{Corresponding author. Tel.: +886 2 2363 8994; fax: +886 2 2363 8994.}

E-mail address: cychang3@ntu.edu.tw (C.-Y. Chang).

chemical sludge because of its high heavy metal content such as copper.

In Taiwan, the yield of the waste electroplating solu-tion of PWB is approximately 1.23 m3/s, resulting in about 2.1× 107kg waste and hazardous sludge per year with 78 wt.% water content[3]. Furthermore, in view of the re-source recycling, the aged electroplating solution of PWB has high reclamation and recycling potentials with high copper concentration and electric conductivity. Note that the qual-ities of the organics in the electroplating solution become low and unreliable to the process after a certain period of time of operations of the electroplating and electrophoresis. For this reason, removal of the spent organic additives with the replacement by adding new additives is one of the key steps for the reutilization of process solution. Ozonation is an effective way to remove organics and reduce the total or-ganic carbons (TOCs). The compounds may be attacked via two different reaction mechanisms: (1) the direct ozonation

(2)

Nomenclature

CAGi gas concentration of ozone of inlet gas (mg/L

or M)

CAGe gas concentration of ozone of discharged gas

(mg/L or M)

CAL concentration of ozone in bulk liquid (mg/L or

M)

CALS dissolved ozone concentration of liquid at

gas–liquid interface (M)

CBL concentration of 2-NS in bulk liquid (mg/L or

M)

CBL0 initial concentration of 2-NS in bulk liquid

(mg/L or M)

Ce experimental data

¯

Ce average value of all experimental data

Cp predicted values

CTOC concentration of total organic carbon (TOC)

(mg/L or M)

CTOC0 initial concentration of TOC (mg/L or M)

HA dimensionless Henry’s constant of ozone

(M M−1)

He Henry’s law constant of ozone, pAi/CALS

(atm L/mol)

[IUV] photon flux of UV entering the reactor (W/m2) kB pseudo-first-order reaction rate constant of

2-NS (min−1)

mO3R ozone consumption (mol)

2-NS 2-naphthalenesulfonate ORP oxidation reduction potential

pAi partial pressure of ozone of gas (atm) QG gas flow rate (L/min)

R2 determination coefficient, 1−

[(Ce− Cp)2/

(Ce− ¯Ce)2]

TOC total organic carbon

VF volume of free space (L) VL volume of solution (L)

t ozonation time (min)

Greek symbol

ηTOC removal efficiency of TOC defined by Eq.(1)

(%)

by the ozone molecule and (2) the radical oxidation by the highly oxidative free radicals such as hydroxyl free radicals, which are formed by the decomposition of ozone in the aque-ous solution[4]. The radical oxidation is non-selective and vigorous.

Recently, Chen et al.[5]investigated the decomposition of 2-NS in the aqueous solution by ozonation with ultra-violet (UV) radiation. Their results indicated that the com-bined process of ozonation with UV radiation is an ef-fective way for the removal of 2-NS in the aqueous

solu-tion. However, the information about the ozonation of 2-NS in the electroplating solution, which has distinct prop-erties from the aqueous solution, is still scarce but de-sirable for evaluating the practicability of removal of 2-NS via ozonation treatment. Thus, the aim of the present study is to employ ozonation combined with UV radia-tion to eliminate 2-NS in the electroplating soluradia-tion. The purpose of introducing UV radiation in the ozonation pro-cesses is to enhance the ozone decomposition yielding more free radicals for achieving a higher ozonation extent

[6,7].

Semibatch ozonation experiments are proceeded under various reaction conditions. The decomposition of 2-NS ac-companies with the variations of TOC, gas and liquid ozone concentrations and oxidation reduction potential (ORP). The results obtained can provide useful information about the proper application of the process for the removal of 2-NS via ozonation with UV radiation in the electroplating solution.

2. Materials and methods 2.1. Chemicals

The substrate recipe of the electroplating solution is: [CuSO4·5H2O] = 200 g/L, [H2SO4] = 60 g/L, and

[Cl−] = 0.03 g/L. The concentration of target organic compound of 2-NS is 200 mg/L according to the pre-scription of electroplating solution [1]. The 2-NS with chemical formula as C10H7SO3Na, which was purchased

from Aldrich (Milwankee, WI, USA) and used without any further purification, has molecular weight of 230.22. The molecular structure of 2-NS is shown in Fig. 1. All experimental solutions were prepared with deionized water without other buffers. The initial values of pH and TOC of experimental solution are about 0.25 and 104 mg/L, respectively. The dimensionless Henry’s law constant (HA) of ozone in the electroplating solution was

deter-mined as 4.18 [8]. Accordingly, the dimensional Henry’s law constant of ozone (He) defined by pAi= HeCALS is

102.1 atm L/mol.

Fig. 1. The molecular structure and simplified scheme of the decomposi-tion pathways of the ozonadecomposi-tion of 2-naphthalenesulfonate (2-NS) with UV radiation.

(3)

2.2. Instrumentation

The airtight reactor of inside diameter of 17.2 cm was made of Pyrex glass with an effective volume of 5.5 L, and equipped with water jacket to maintain a constant solution temperature at 25◦C in all experiments. The design of re-actor was based on the criteria of the shape fre-actors of a standard six-blade turbine [9]. The gas diffuser in cylin-drical shape with pore size of 10␮m was located at the bottom of reactor. Two quartz tubes of outside diameter of 3.8 cm installed symmetrically inside the reactor were used to house the UV lamps. The low-pressure mercury lamps emitted principally at 254 nm supplied the UV radiation. The radiation intensity was measured by a digital radiome-ter (Ultra-Violet Products (UVP), Upland, CA, USA) with a model UVP-25 radiation sensor. About 3.705 L solution was used in each experiment, while the total sampling volume was within 5% of solution. The stirred speed was as high as 800 rpm to ensure the complete mixing of liquid and gas phases according to previous tests[10,11]. The generation of ozone gas was controlled by the power input of ozone generator (model SG-01A, Sumitomo, Tokyo, Japan) with constant gauge pressure (0.968 atm). The ozone generator used in this study employed two steel plate electrodes and ceramic dielectric. Ozone-containing gas generated by pure oxygen was introduced into the reactor with a gas flow rate of 1.94 L/min. A circulation pump was used to transport the liquid from the reactor to the sensors of monitors with a flow rate of 0.18 L/min, and to re-flow it back during the ozona-tion.

The concentrations of 2-NS (CBL) were analyzed

us-ing high-performance liquid chromatography (HPLC) sys-tem with column of 250 mm× 4.6 mm of model BDS C18 (5␮m) (Thermo Hypersil-keystone, Bellfonte, PA, USA),

and UV–vis detector (model 1706, Bio-Rad, Hercules, CA, USA) at 254 nm. The ratio of water:CH3CN of

composi-tion of effluent with flow rate of 1.0 mL/min was 83:17. The injection volume of analytic solution was 20␮L. The TOC concentrations (CTOC) of samples were analyzed by

a TOC analyzer (model 700, OI Corporation, Texas, USA). The instrument applies the UV–persulfate technique to con-vert the organic carbon for the subsequent analysis by an infrared carbon dioxide analyzer calibrated with the potas-sium hydrogen phthalate standard. The pH (model 300T, Suntex, Taipei, Taiwan) and ORP (model 900C, Apogee, Taipei, Taiwan) meters with sensors were used to measure the values of pH and ORP of solution. All fittings, tubings and bottles were made of stainless steel, teflon or glass. The experimental apparatus employed in this work is shown in

Fig. 2.

2.3. Experimental procedures

The semibatch experiments of ozonation of 2-NS were performed to examine the variations of CBL, CTOC, pH,

and ORP. Before starting the ozonation experiments, the ozone-containing gas was bypassed to the photometric analyzer (model SOZ-6004, Seki, Tokyo, Japan) to as-sure the stability and ozone concentration. A part of gas stream at the preset flow rate was directed into the reac-tor when the ozonation system was ready to start. In ad-dition, the concentrations of inlet (CAGi) and discharged

(CAGe) gas ozone were measured. Liquid samples were

drawn out from the reactor at desired time intervals in the course of ozonation of 2-NS. The photon flux of UV, [IUV] entering the reactor with value of 60.35 W/m2 was

employed to test the effect of UV radiation on the ozona-tion.

(4)

Fig. 3. Variation of CBL/CBL₀with time for the ozonation of 2-NS in semi-batch system. CBL=CBL₀at t = 0.

3. Results and discussion

3.1. Decomposition of 2-NS during ozonation with UV radiation

The time variations of the decomposition of 2-NS via the ozonations with and without UV radiation are shown inFig. 3. The effect of the ozone concentration of feed gas on the de-composition of 2-NS is remarkable. The time required for the complete decomposition of 2-NS with CAGi= 40 mg/L

is about one-third and half of those with CAGi= 10 and

20 mg/L, respectively. However, comparing the results of the cases with and without UV radiation indicates that the UV radiation affects the decomposition extent of 2-NS slightly.

In order to illustrate the possible roles of O3and UV

ra-diation on attacking the 2-NS and its intermediates, a sim-plified scheme of the decomposition pathways of 2-NS via the ozonation with UV radiation is proposed as shown in

Fig. 1. The major contribution of UV is to generate OH• via reaction II. This can enhance the radical reaction of hy-droxyl free radicals with 2-NS noted as reaction III, which is more vigorous than the direct reaction of O3with 2-NS via

reaction I. However, the decrease of dissolved ozone con-centration then decreases the extent of reaction I. As the result, the overall decomposition extent of 2-NS in ozona-tion is only slightly accelerated by the UV radiaozona-tion. Further, the pseudo-first-order kinetics for the elimination of 2-NS,

CBL/CBL0= exp−kBt_{, can be obtained from}_{Fig. 3}_{with k}_B

(min−1) = 0.0107± 0.0042CAGifor a 95% confidence

inter-val, where CAGiis in mg/L.

The decomposition of 2-NS accompanies a diminution of TOCs. The removal efficiencies of TOCs (ηTOC) de-fined by Eq. (1) at time about 8–25 min, for the complete elimination of 2-NS are between 1 and 4% as shown in

Fig. 4. The intermediates produced from the

decomposi-Fig. 4. Variation ofηTOCwith time for the ozonation of 2-NS in semibatch system.

tion of 2-NS contribute over 96% TOC relative to the initial value.

ηTOC=CTOC0 − CTOC CTOC0

(1) whereCTOC₀ and CTOC are values of CTOC at t = 0 and t,

respectively. The decrease of decomposition extent of 2-NS with ozonation time also counts for the generation of by-products, which are also competitors for oxidation. The initial attacks of ozone on 2-NS may be proceeded via (1) an ozone dipolar cycloaddition on the 1,2 bond or (2) an electrophilic substitution of ozone on carbon 1[5].

3.2. Variations of removal efﬁciency of TOC and ozone concentration

Fig. 4shows the variation of removal efficiency of TOC (ηTOC) under various experimental conditions. The value of

ηTOC increases with the ozonation time continuously. Ap-parently, the CAGi significantly contributes the

mineraliza-tion extent of 2-NS for both sole O3 and O3/UV systems.

Moreover, the UV radiation in the ozonation of 2-NS also improves the removal of TOC. The OH•induced by the pres-ence of UV radiation proceeds with higherηTOCvia the re-action noted as rere-action V[12,13]. Limiting values ofηTOC seem to appear for the sole ozonation of intermediates. The lower value of dηTOC/dt in the early stage of ozonation is due to the cause that ozone is first consumed for the opening of benzene rings and the oxidation of the sulfonic substituent, thus accompanying with the low diminution of TOCs. In ad-dition, the lower value of dηTOC/dt in the later ozonation stage (higherηTOC) is caused by the ozonation of resistant intermediates such as oxalic acid, phthalic acid, and formic acid, which have low reactivity towards ozone molecules

(5)

Note that the cause for the apparent contribution of UV ra-diation on the mineralization of 2-NS during ozonation is at-tributed to the vigorous attack of hydroxyl radicals, for which the generation is enhanced by the UV radiation, on the in-termediates that are more resistant to ozone than hydroxyl radicals. On the other hand, the insignificant effect of UV radiation on the decomposition of 2-NS is due to the high reactivity of 2-NS toward ozone. Similar phenomena were also observed during the ozonation of 2-NS in the aqueous solution[5]. Furthermore, the decomposition rates of 2-NS and TOC via sole ozonation in the electroplating solution seems to be faster in comparison with those in the aqueous solution. It may be caused by the catalytic effect of cop-per ions in the electroplating solution[16,17]. On the other hand, the enhancement of UV radiation on the mineraliza-tion of 2-NS during ozonamineraliza-tion in the electroplating solumineraliza-tion is weaker than that in the aqueous solution because of the lower transmission of UV radiation in the electroplating so-lution. In addition, the pH value of the electroplating solution is about 0.25. The variation of pH during the ozonation ex-periments was found to be very slight due to the high acid concentration.

Moreover, the variations of dimensionless liquid (CAL/(CAGi/HA)) and gas (CAGe/CAGi) ozone concentrations

withηTOCfor various cases are shown inFig. 5. In the first stage (withηTOC≤ 5%), the CALremains nearly undetectable

(Fig. 5a). In this regime, the rate of ozonation reaction is very fast so that the ozone transferred from the gas phase is imme-diately consumed in the solution. Then the dissolved ozone starts to appear with CAL/(CAGi/HA) having a value of about

0.11 in the regime with 34%≥ ηTOC> 5%. Later, the CAL

gradually increases with ηTOC when ηTOC> 34%. The ac-cumulation of dissolved ozone is attributed to the case that the consumption rate of dissolved ozone is lower than the gas–liquid mass transfer rate of ozone due to the lower reac-tivities of the refractory intermediates in the reacted solution. The variation of CALwithηTOCseems to be only slightly

de-pendent on CAGi. Obviously, the CALhas the smaller value

in the case of O3/UV than that of sole O3.Fig. 5b shows that

the CAGe/CAGi increases rapidly in the beginning and has a

value of 0.36–0.56 whenηTOC≤ 34%. The CAGeapparently

further increases withηTOCcorresponding to the increase of

CALasηTOC> 34%.

3.3. Removal of TOC associated with ozone consumption and ORP

In addition, the relation betweenηTOCand the ozone con-sumption per mol 2-NS,mO3R/(CBL0VL) (mol mol−1), is il-lustrated inFig. 6. The ozone consumption (mO3R) is calcu-lated by Eq.(2), where VLand VFare the volumes of solution

and free space in the reactor, respectively.

mO3R=

_t

0QG

(CAGi− CAGe) dt − CALVL− CAGeVF (2)

Fig. 5. CAL/(CAGi/HA) and CAGe/CAGivs.ηTOCfor the ozonation of 2-NS in semibatch system. Line: average and smooth curve of experimental data. (a) CAL/(CAGi/HA); (b) CAGe/CAGi.

As a result, ηTOC increases with mO3R/(CBL0VL) ap-parently and agreeably in all cases examined, indicating a high correlation between the mineralization of 2-NS and the ozone consumption. It reveals that the value of ratio of

ηTOC/(mO3R/(CBL0VL)) is about 4.3% during the ozonation treatments of 2-NS. The ηTOCapproaches to 100%, which stands for the nearly complete mineralization of 2-NS, as the value ofmO₃R/(CBL₀VL) is greater than 23. It is worth noting that the higher mineralization extent of 2-NS in the O3/UV

treatment is resulted from the greater consumption efficiency of applied ozone.

Fig. 7presents the variation of ORP of liquid withηTOC. The value of ORP can be used as a supplementary ozonation index of ozonation system [5,18]. The experimental ORP data of the present study were further used to plot the av-erage and smooth curve inFig. 7, which shows the distinct

(6)

Fig. 6.ηTOCvs.mO₃R/(CBL₀VL) for the ozonation of 2-NS in semibatch sys-tem. Line: average and smooth curve of experimental data with R2_{= 0.980.}

variation of ORP with ηTOC. The value of ORP reveals a slight variation in the initial period while a significant in-crease in the later period for 10% <ηTOC< 30%. The initial ORP value of the electroplating solution before ozonation is 426 mV. For ηTOC≤ 10%, the values of ORP are between 700 and 785 mV. In this stage, the organics of TOCs in the solution are predominant for controlling the ORP of system with small variation. As the concentration of TOCs (CTOC)

decreases below 93.6 mg/L (ηTOC> 10%), the ORP starts to increase withηTOCas shown inFig. 7. This regime may be called the transient stage. As theηTOCis close to 50%, the value of ORP of system approaches to that of organic-free solution of about 1360 mV. Thus, ORP is a useful index as-sociated with the oxidation state of the ozonation of 2-NS in terms ofηTOC.

Fig. 7. ORP vs.ηTOCfor the ozonation of 2-NS in semibatch system. Line: average and smooth curve of experimental data with R2_{= 0.947.}

In summary, the difference in performance between O3

and O3/UV treatments regarding the destruction of 2-NS is

not significant. However, the introduction of UV radiation combined with O3gives significant contribution for the

sub-sequent oxidation of intermediates after the disappearance of 2-NS. The combination of ozone with UV radiation is recom-mended for treating the 2-NS in the electroplating solution as far as the TOC reduction is concerned, although the process with O3alone may be sufficient for destructing the 2-NS.

4. Conclusions

Ozonation combined with UV radiation was employed as an effective way for the removal of 2-naphthalenesulfonate (2-NS) in the electroplating solution. The nearly complete mineralization of 2-NS can be achieved under the experi-mental conditions of this study. The decomposition of 2-NS accompanies the diminution of total organic carbons (TOCs) and the consumption of ozone. The following conclusions may be drawn:

1. The decomposition extent of 2-NS increases with the ozone concentration of feed gas. However, it is not signif-icantly enhanced by the presence of UV radiation. At the time when the decomposition of 2-NS is completed, the removal efficiency of TOC (ηTOC) is lower than 4%. 2. Both the ozone concentration of feed gas and UV

radia-tion can improve the mineralizaradia-tion extent of 2-NS. Lim-iting values ofηTOCof 2-NS in the electroplating solution appear for the sole ozonation of intermediates. The ex-planation for the phenomena can be addressed with the mechanism of ozonation of 2-NS.

3. Both the liquid- and discharged-gas ozone concentration values remain low whenηTOC≤ 34%, while they increase significantly with higherηTOCafterwards. Furthermore, the clear-cut relationship between ηTOC and the ozone consumption was obtained in this study.

4. The oxidation reduction potential (ORP) can be used as a supplementary index of oxidation state of 2-NS in the electroplating solution. The ORP value varies with the residual TOC concentration during the ozonation of 2-NS.

Acknowledgement

This study was supported by the National Science Council of Taiwan under Grant No. NSC 89-2211-E-002-107.

References

[1] C.L. Fang, General Concepts of Additives in the Electroplating So-lution, Finishing Science Pub. Co., Taipei, Taiwan, 1996.

[2] C.H. Liao, M.C. Lu, Y.H. Lin, Y.J. Chen, P.F. Hsieh, H2O2/UV oxidation of polyethylene glycol in copper sulfate electroplating

(7)

wastewater of printed circuit board manufacturing, J. Chin. Inst. En-viron. Eng. (Taiwan) 12 (1) (2002) 1–8.

[3] Taiwan Industrial Development Bureau (TIDB), Waste Recovery and Pollution Prevention of the Electroplating Business, Ministry of Eco-nomic Affairs of Taiwan, Taipei, Taiwan, 1997.

[4] J.L. Sotelo, F.J. Beltr´an, F.J. Ben´ıtez, J. Beltr´an-Heredia, Ozone de-composition in water: kinetic study, Ind. Eng. Chem. Res. 26 (1) (1987) 39–43.

[5] Y.H. Chen, C.Y. Chang, S.F. Huang, C.Y. Chiu, D. Ji, N.C. Shang, Y.H. Yu, P.C. Chiang, Y. Ku, J.N. Chen, Decomposition of 2-naphthalenesulfonate in aqueous solution by ozonation with UV ra-diation, Water Res. 36 (16) (2002) 4144–4154.

[6] H.W. Prengle, Experimental rate constants and reactor considerations for the destruction of micropollutants and trihalomethane precursors by ozone with ultraviolet radiation, Environ. Sci. Technol. 17 (12) (1983) 743–747.

[7] R. Andreozzi, V. Caprio, R. Marotta, A. Radovnikovic, Ozonation H2O2/UV treatment of clofibric acid in water: a kinetic investigation, J. Hazard. Mater. B 103 (2003) 233–246.

[8] Y.H. Chen, C.Y. Chang, C.C. Chen, C.Y. Chiu, Y.H. Yu, P.C. Chiang, C.F. Chang, J.L. Shie, Kinetics of ozonation of 2-mercaptothiazoline in electroplating solution, Ind. Eng. Chem. Res. 43 (22) (2004) 6935–6942.

[9] W.L. McCabe, J.C. Smith, P. Harriott, Unit Operations of Chemical Engineering, 5th ed., McGraw-Hill, NY, USA, 1993.

[10] H. Li, C.Y. Chang, C.Y. Chiu, Y.H. Yu, P.C. Chiang, Y.H. Chen, S.J. Lee, Y. Ku, J.N. Chen, Kinetics of ozonation of polyethylene

glycol in printed wiring board electroplating solution, J. Chin. Inst. Environ. Eng. (Taiwan) 10 (1) (2000) 69–75.

[11] C.Y. Chang, Y.H. Chen, H. Li, C.Y. Chiu, Y.H. Yu, P.C. Chiang, Y. Ku, J.N. Chen, Kinetics of decomposition of polyethylene glycol in electroplating solution by ozonation with UV radiation, J. Environ. Eng. 127 (10) (2001) 908–915.

[12] F.J. Benitez, J. Beltran-Heredia, J.L. Acero, T. Gonzalez, Degrada-tion of protocatechuic acid by two advanced oxidaDegrada-tion processes: ozone/UV radiation and H2O2/UV radiation, Water Res. 30 (7) (1996) 1597–1604.

[13] I.A. Balcio˘glu, N. Getoff, Advanced oxidation of 4-chlorobenzaldehyde in water by UV-light, ozonation and combination of both methods, Chemosphere 39 (9) (1998) 1993–2005.

[14] B. Legube, S. Guyon, H. Sugimitsu, M. Dore, Ozonation of naph-thalene in aqueous solution. 1. Ozone consumption and ozonation products, Water Res. 20 (2) (1986) 197–208.

[15] P. Berger, N.K. Vel Leitner, M. Dore, B. Legube, Ozone and hy-droxyl radicals induced oxidation of glycine, Water Res. 33 (2) (1999) 433–441.

[16] S.H. Lin, S.J. Ho, Catalytic wet-air oxidation of high strength in-dustrial wastewater, Appl. Catal. B: Environ. 9 (1996) 133–147. [17] Y.Z. Pi, M. Ernst, J.C. Schrotter, Effect of phosphate buffer upon

CuO/Al2O3 and Cu(II) catalyzed ozonation of oxalic acid solution, Ozone Sci. Eng. 25 (2003) 393–397.

[18] C.P. Yu, Y.H. Yu, Identifying useful real-time control parameters in ozonation process, Water Sci. Technol. 42 (3–4) (2000) 435–440.