Degradation Of Atrazine By Manganesecatalysed Ozonationdinfl

JUN MA1M and NIGEL J. D. GRAHAM2*M
1School of Municipal and Environmental Engineering, Harbin University of Architecture and
Engineering, PO Box 627, Harbin, 150008, People‘s Republic of China and 2Environmental and Water
Resources Engineering, Civil Engineering Department, Imperial College, London SW7 2BU, UK
(First received 1 July 1999; accepted in revised form 1 January 2000)

Abstract DThe e?ect of radical scavengers such as bicarbonate and tert-butanol on the MnII catalysed ozonation of atrazine, an important herbicide and well-established radical probe substance, was studied using a conventional gas bubble-contacting column. It was found that the presence of a small amount of MnII (0.3亇1.2 mg/l) greatly increased the degradation rate of atrazine, with the formation of byproduct compounds of a lower molecular weight and a greater polarity (as indicated by shorter retention times in HPLC chromatography). However, the presence of either bicarbonate or tert-butanol had a negative e?ect on the degradation of atrazine. With an increase of bicarbonate concentration, the oxidation rate of atrazine by MnII-catalysed ozone was substantially reduced and a correspondingly higher residual ozone was observed. In the presence of tert-butanol, greater reductions of the degradation rate of atrazine were observed and the decomposition of ozone was greatly retarded,resulting in a higher residual ozone; this was the case for either ozone oxidation alone or MnIIcatalysed ozone oxidation. These results appear to conRrm that the degradation of atrazine by ozone in the presence of MnII follows a radical mechanism. It is believed that MnII catalyses the decomposition of ozone through the formation of intermediate manganese species (such as MnIV), leading to the generation of hydroxyl radicals. 7 2000 Elsevier Science Ltd. All rights reserved
Key words Datrazine, ozone, catalytic oxidation, manganese, advanced oxidation, hydroxyl radical scavengers
INTRODUCTION
Advanced oxidation is a process for generating hydroxyl radicals which have a much greater oxidation capability than ozone for the degradation of refractory organic micropollutants, such as pesticides, which are resistant to degradation by ozone alone. Various advanced oxidation processes have been investigated, such as UV/H2O2, O3/H2O2, UV/TiO2, UV/O3, and Fenton‘s reagent. Particular attention has been paid to metal-catalysed ozonation processes in recent years for the degradation of organic materials due to their potentially low cost.For example, Paillard et al. (1991) reported that TiO2-catalysed ozonation was more e.cient for the degradation of humic acid than ozone alone, but the e.ciency was relatively low for the degradation of atrazine. MnII has been shown to be e.ective for the catalytic degradation of oxalic acid (Andreozzi et al., 1992), where it was believed that MnII complexed with oxalic acid to form an intermediate product, that was more easily degraded by ozone. Ma and Graham (1997) reported that in the presence of a small amount of MnII (1 mg/l) the oxidation of atrazine by ozone was greatly enhanced, and at the same time the residual ozone in the aqueous solution clearly decreased with increasing MnII dosage,suggesting a possible radical mechanism as being responsible for the MnII-catalysed oxidation of atrazine. It was further observed that the colloidal form of hydrous manganese oxide prepared by the reduction of permanganate with divalent manganese had a similar e.ectiveness to MnII for the catalytic degradation of atrazine, and it was suggested that the catalytic oxidation of atrazine might be caused by the hydrous manganese oxide formed in situ in the process of oxidation between MnII and ozone,rather than by soluble MnII itself. In contrast, a commercial MnO2 (precipitated active ``Merck‘‘,BDH Ltd, UK) did not exhibit any catalytic e.ectiveness at all for the degradation of atrazine (Ma and Graham, 1997). Recent investigations (Ma and Graham, 1998) have shown that the presence of a small amount of humic substances (e.g., 1 mg/l as DOC) accelerates the degradation of atrazine by Mn-catalysed ozone oxidation, while higher concentrations of humic substances decreased the degradation rate of atrazine.
Further understanding of the fundamental reaction mechanism of manganese catalysed ozonation is considered very important in terms of its potential application in the treatment of refractory organics. A well-established radical probe, atrazine, has been used in this study, since it is an important micropollutant that can not be degraded easily by ozone alone (K=2.3 My1 sy1; Xiong and Graham,1992), but which can be degraded readily by hydroxyl radicals (K = 3  109 My1 sy1; HoigneA ,1997). The well-known radical scavengers, bicarbonate ion and tert-butanol, were used to evaluate their e?ect on the degradation of atrazine, in order to conRrm whether the MnII-catalysed ozone oxidation follows a radical type mechanism.
MATERIALS AND METHODS
Materials
A model water was prepared by spiking 3 mM atrazine (Promochem Ltd, St. Albans, 99.85%) in high purity(>18 O cm) Milli-Q water (Millipore Ltd, Watford). A phosphate bu?er (1 mM, NaHPO4, BDH Ltd) was added to the aqueous solution to keep a constant pH of 7.0.The manganese stock solution was prepared by dissolving MnSO4
4H2O (Laboratory Reagent, BDH Ltd) in to high purity Milli-Q water, which thus had an Mn concentration of 1000 mg/l. Bicarbonate solution (0.1 M) was prepared by dissolving analytical grade sodium bicarbonate (BDH Ltd) into Milli-Q water. The tert-butanol (BDHLtd) was diluted to speciRc concentrations before use.Borosilicate glassware was cleaned by soaking in a 3 w/v solution of Dri-Decon detergent (Decon Laboratories Ltd, Hove) and rinsing with tap water and then ultra-pure (Milli-Q) water. All the glassware except for the volumetric 丳asks was mu.ed overnight at 4008C. The volumetric 丳asks were washed by soaking them in chromic
acid and then rinsing with ultrapure water.
Ozonation procedure
The ozonation tests were conducted in a bench scale ozonation system, which consists of an ozone generator(Trailigaz Labo 76, Ozotech Ltd, Burgess Hill, Sussex,UK), an ozone contact column (1300 mm in height and 60 mm in diameter) of 3.65 l capacity, and an ozone online analyser (Model 26506 Indicating Instrument, Orbisphere Laboratories). A peristaltic pump (Watson-Marlow 502 S, Alexander Wright and Co.) was used for feeding the catalysts and a variable speed pump Rtted with Te丳on wetted contact parts and magnetically coupled gears
(Micopump Corp., Concord, CA) was used to recirculate the reacting solution in the ozonation column. A schematic diagram of the ozonation system is shown in Fig. 1.Ozone was produced from ordinary grade air at a generator power setting of 20 W, and was subsequently fed into the ozone contactor column through a porous glass frit at the base of the column. The column was shielded from daylight by an external aluminium foil covering to exclude possible photocatalytic e?ects. Before starting the test the column was preozonated for 5 min to satisfy any ozone demand in the column and then the column was washed several times with high purity Milli-Q water. In the ozonation tests, a model water (3.5 l) with an atrazine concentration of 3 mM was pumped into the column (3.65 l volume) by a magnetic pump and then circulated at a rate of approximately 85 l/h. The ozonation time was controlled at 0.5 min for all of the samples. A preliminary test showed that if the ozonation was carried out for 0.5 min and then the ozone generator was switched o?, the residual ozone in the water approximately reaches a maximum after 2 min. This indicated that the ozone gas left in the generator continued to pass into the column following the switching o? of the generator, up to the time of approximately 2 min. The experiments were conducted with and without the presence of radical scavengers (bicarbonate or tert-butanol).
The MnII aqueous solution was fed into the column continuously at the middle part of the column at a Rxed rate of 24.8 ml/min (49% of the pump speed) to react simultaneously with ozone in the water. The manganese solution was admitted to the column concomitant to the application of ozone to make the accumulative concentration of MnII equal to a speciRc amount (as mg/l Mn) within 2 min.

Water samples were taken from the contactor column at various reaction times to determine the residual concentration of atrazine. The oxidation reaction was stopped by the addition of a small amount of sodium thiosulphate solution (BDH Chemical Ltd, Analytical Grade Reagent), and the water samples were centrifuged before the atrazine analysis.
Analysis
The total applied ozone was calculated as the numerical sum of the exhausted ozone, leaving the ozonator column and the residual ozone in water when the column was Rlled only with phosphate-bu?ered Milli-Q water, and ozonated in the same way as water samples containing reactive species. The exhaust ozone from the column was analysed by the iodometric method of the International Ozone Association (Standardisation Committee, 1987).Residual ozone in the aqueous solution was analysed by an on-line ozone analyser whose sensor for ozone analyses was installed in the water circulation system. The total applied ozone in this experiment was controlled at around
2.5 mg/l.
Atrazine was analysed by a Waters modular high performance liquid chromatogram (HPLC) (Millipore Ltd, Watford), using a Jones Chromatography HPLC column(Spher ODS2) and acetonitrile (HPLC Grade) and water (65:35 in v/v) as the mobile phase at a rate of 1 ml/min by
the HPLC pump (Waters 501). The water sample was injected into the HPLC Chromatography column by an autosampler (Waters 712 WISP), and detected by a UV detector (Waters 484) at 220 nm. A standard calibration curve for atrazine was made by injecting atrazine standards into the HPLC column under the same conditions as with the analysis of the water samples.
RESULTS
E?ect of MnII dose
In previous studies by the authors (Ma and Graham, 1997, 1998), a clear and positive e?ect was observed for the degradation of atrazine in the presence of 1 mg/l of MnII when the applied ozone dose was about 2.5 mg/l. As MnII is a reducing agent, it can not only accelerate the degradation of atrazine but also may react with some oxidative intermediates, and therefore the concentration of MnII may have some in丳uence on the degradation of atrazine. It is possible that an optimal MnII concentration may exist for a certain dose of ozone. In natural waters the concentration of manganese is normally less than 1 mg/l, but groundwaters subject to reducing conditions can have much higher
manganese concentrations than well aerated surface waters (Twort et al., 1994). A preliminary test was conducted to evaluate the e?ect of MnII concentration on the degradation of atrazine at a speciRc ozone dose (accumulative applied dose within 2 min is about 2.5 mg/l). It is seen in Fig. 2 that only a very small amount of MnII (0.3 mg/l) is required to achieve a substantial reduction in atrazine concentration. This indicates, possibly, that only a small amount of manganese is needed to initiate the radical type reaction. As is seen from the variation of the residual ozone in the water (Fig. 3), the consumption of ozone is higher as the concentration of MnII is increased. The greatest removal of atrazine appeared to occur at the lowest MnII concentrations used in this test (0.3 and 0.6 mg/l). Further increase
of MnII concentration did not cause any increase in the degradation of atrazine, but rather it had a slightly negative e?ect. This phenomenon may suggest that an above-optimal concentration of MnII may scavenge the hydroxyl radicals generated, thereby causing a reduction in the degradation of atrazine. However, a more detailed investigation of the stoichiometry of this e?ect is required.
Figure 4 summarises a series of HPLC chromatographs for water samples taken at di?erent reaction times when the concentration of MnII was 0.5 mg/l. It can be seen that the peak corresponding to atrazine gradually disappears with reaction time, and


that some by-products with a lower molecular weight and higher polarity are formed. With increasing reaction time, these intermediate com-
pounds are further oxidised, as indicated by reductions in the peak areas. IdentiRcation of the byproduct compounds was not attempted in this study, but this phenomenon is consistent with the proposition that MnII-catalysed ozonation follows a radical type mechanism, given that atrazine can only be degraded signiRcantly by hydroxyl radicals, rather than by ozone alone (Legube et al., 1987).A comparison of hydrogen peroxide-catalysed ozone oxidation with MnII-catalysed ozone oxidation has been carried out (this will be the subject of a separate paper), and it was found that a relatively lower concentration of MnII (in molar terms)is required to achieve a similar e?ectiveness for the oxidation of atrazine than hydrogen peroxide. However, the pattern of compound peaks obtained by HPLC chromatography was very similar for both ozone systems, and this supports the proposition of a radical type oxidation mechanism in both cases.E?ect of bicarbonate Carbonate and bicarbonate are well-known radical scavengers (HoigneA and Bader, 1985; AWWA,1998), which have a strong in丳uence on radical type reactions. Carbonate is a stronger radical scavenger than bicarbonate due to its higher reaction rate constant with hydroxyl radicals .KCO2y 3 . 3:9  108 My1 sy1, KHCOy 3 . 8:5  106 My1 sy1: Acero and von Gunten, 1998). However, under neutral pH conditions, the inorganic carbon exists mainly in the form of bicarbonate, which is present in surface and ground waters at concentrations typically in the range of 50亇200 mg/l. Higher concentrations may


be encountered in high alkalinity waters. Bicarbonate ion is reported to take part in reactions with hydroxyl radicals in competition with refractory organic pollutants in which it has a relatively lower rate constant. For example, for bicarbonate ion the rate constant is 8.5  106 My1 sy1 (Acero and von Gunten, 1998), while for atrazine and nitrobenzene the rate constant is 3  109 (HoigneA , 1997). However, since the concentration of bicarbonate ions is substantially higher (mg/l level) than that of organic micropollutants (mg/l level or ng/l level) in the water, it is likely that bicarbonate is the principal consumer of the hydroxyl radicals, particularly when relatively high concentrations of bicarbonate are present in water. The scavenging e?ect of bicarbonate also lies in the fact that it reacts with hydroxyl radicals, to generate bicarbonate radicals .HCO.z:rad . 3 ). These act as a very selective additional oxidation species and which have a much lower reaction rate constant than hydroxyl radicals for the oxidation of organic micropollutants (HoigneA ,
1998). It has been reported that bicarbonate ions scavenge hydroxyl radicals to produce intermediates which do not release a radical-type chain carrier, thereby quenching the radical type chain reaction (HoigneA , 1998). In contrast, it has been reported that in certain cases (e.g., O3/H2O2 process) bicarbonate ion plays an important role as a promoter during ozone decomposition (Acero and von Gunten, 1998). Thus, an assessment of the e?ect of bi-

carbonate on the MnII-catalysed ozonation can assist in identifying the reaction mechanism of MnII-catalysed ozonation. Figure 5 shows the e?ect of bicarbonate ion concentration on the MnII-catalysed ozone degradation of atrazine. It can be seen that the presence of bicarbonate signiRcantly decreases the rate and extent of atrazine degradation, and that the e?ect is dependent on the bicarbonate concentration. Higher concentrations of bicarbonate caused greater reductions in the rate of atrazine degradation. However, even at the relatively high concentration of 200 mg/l, there is still a clear catalytic e?ect with MnII. This indicates that bicarbonate has a limited e?ect on the MnII catalysed ozonation and/or that some of the bicarbonate radicals generated may also contribute to the degradation of atrazine;
hence the overall e?ectiveness for atrazine degradation is still relatively high in the presence of bicarbonate.The corresponding results indicating the residual ozone at di?erent concentrations of bicarbonate (Fig. 6) showed that bicarbonate did not have a strong in丳uence on the residual ozone. In the case of ozone alone (absence of Mn), the presence of bicarbonate ion also retarded the degradation of atrazine (Fig. 5), indicating that the small, but measurable degradation of atrazine by ozone alone, may be caused principally by a small extent of radical formation generated by ozone decomposition at neutral pH, rather than oxidation by molecular ozone. It is well known that ozonation of blank

water also results in the production of hydroxyl radicals through ozone decomposition (HoigneA and Bader, 1976). Thus, the presence of bicarbonate inhibits the formation of hydroxyl radicals when only ozone is used, resulting in a lower degradation rate for atrazine.
E?ect of tert-butanol tert-Butanol is a stronger radical scavenger than bicarbonate as indicated by its higher reaction rate
constant with hydroxyl radicals .K .z:rad . OH is 5  108 My1 sy1: AWWA Research Foundation, 1991). tert-Butanol reacts with hydroxyl radicals, generating inert intermediates, thus causing termination of the radical chain reaction. Thus, tert-butanol is a
more suitable indicator for the radical type reaction because of its stronger scavenging e?ect on hydroxyl radicals.
In Fig. 7 it can be seen that the presence of tertbutanol has a very strong in丳uence on the degradation of atrazine, causing a major reduction in the degradation rate, even at a very low concentration(5 mg/l). At the higher dose of tert-butanol (50 mg/l) the scavenging e?ect on the degradation of atrazine is very strong, and there is only a minor degradation of atrazine. It is also seen that the
degradation rate for atrazine is decreased for ozonation alone when tert-butanol was present in water. Thus, the presence of tert-butanol e?ectively inhibits the generation of radicals in aqueous solution at neutral pH. It is seen from Fig. 8 that the concentration of residual ozone is correspondingly in丳uenced by the presence of tert-butanol, with the decomposition of ozone strongly retarded by the presence of tert-butanol. It is noted that the maximum concentration in the water appeared earlier than in either the case of ozonation alone or MnII-
catalysed ozonation.
CONCLUSIONS
The e?ect of radical scavengers, such as bicarbonate and tert-butanol on the MnII catalysed ozonation of atrazine, was studied using a conventional gas bubble-contacting column. It was found that a small amount of MnII (0.3亇1.2 mg/l) greatly increased the degradation rate of atrazine. Atrazine was shown to be degraded to by-products with a lower molecular weight and a greater polarity (as indicated by the shorter retention times in HPLC chromatography). The presence of either bicarbonate or tert-butanol had a negative e?ect on the MnII-catalysed oxidation of atrazine. An increase of the bicarbonate ion concentration caused a clear and systematic reduction in the degradation rate for atrazine by ozone catalysed with MnII. A much greater scavenging e?ect was observed for the MnII catalysed degradation of atrazine in the presence of tert-butanol, owing to its greater reaction rate with hydroxyl radicals than that of bicarbonate ion.These experimental results have conRrmed the proposition that the degradation of atrazine by MnII follows a radical-type mechanism. Thus, a relatively small amount of MnII is able to catalyse the decomposition of ozone via intermediate manganese species (such as MnO2), generating hydroxyl (and possibly other) radical species which have a very high oxidation potential for atrazine.
AcknowledgementsDThe support from the National Natural Science Foundation of China and the Royal Society of the UK are greatly appreciated and acknowledged.
REFERENCES
Acero J. L. and von Gunten (1998) In丳uence of carbonate on advanced oxidation process for drinking water treat ment. In Proceedings of Ozonation and AOPs in Water Treatment: Applications and Research, Poitiers, France, 23亇25 September, pp. 13.1亇13.9.
Andreozzi R., Insola A., Caprio V. and D‘Amore M. G. (1992) The kinetics of MnII catalysed ozonation of oxalic acid in aqueous solution. Wat. Res. 26(7), 917亇921.
AWWA Research Foundation and Compagnie GeA neA rale des Eaux (1991) Fundamental Aspects. In Ozone in Water Treatment Application and Engineering. Lewis Publishers, Michigan, USA, pp. 18亇19.
AWWA Research Foundation (1998) E?ect of Bicarbonate Alkalinity on Performance of Advanced Oxidation Processes, pp. 238.HoigneA J. and Bader H. (1976) The role of hydroxyl radical reaction in ozonation processes in aqueous solution.Wat. Res. 10, 377亇386.
HoigneA J. and Bader H. (1985) Rate constants of reactions of ozone with organic and inorganic compounds in water. III: Inorganic compounds and radicals. Wat. Res.19(8), 993.
HoigneA J. (1997) Inter-calibration of OH radical sources and water quality parameters. Wat. Sci. Tech. 35(4), 1亇8.
HoigneA J. (1998) Chemistry of aqueous ozone and transformation of pollutants by ozonation and advanced oxidation processes. Part C Quality and Treatment of Drinking Water II. In The Handbook of Environmental Chemistry, Vol. 5, ed. J. Hrubec, pp. 83亇141. SpringerVerlag, Berlin Heidelberg.
Legube B., Guyon S. and Dore M. (1987) Ozonation of aqueous solutions of nitrogen heterocyclic compounds:benzotriazoles, atrazine and amitrole. Ozone Sci. Engng 9(3), 233亇246.
Ma J. and Graham N. J. D. (1997) Preliminary investigation of manganese-catalysed ozonation for the destruction of atrazine. Ozone Sci. Engng 19(3), 227亇 240.
Ma J. and Graham N. J. D. (1998) Degradation of atrazine by manganese-catalysed ozonation-inuence of humic substances. Wat. Res. 33(3), 785亇793.
Paillard H., Dore M. and Bourbigot M. M. (1991) Prospects concerning applications of catalytic ozonation in drinking water treatment. Proceedings of the 10th Ozone World Congress, Monaco, March. Vol. 1, 313亇329 pp.
Standardisation CommitteeDEurope (1987) Iodometric method for the determination of ozone in a process gas,001/87(F), International Ozone Association.
Twort A. C., Law F. M., Crowley F. W. and Ratnayaka D. D. (1994) Water Supply, 4th ed. Edward Arnold,UK.
Xiong F. and Graham N. J. D. (1992) Rate constants for herbicide degradation by ozone. Ozone Sci. Eng. 14(4),283亇301.