Recherche et identification des premiers sous-produits d'oxydation de l'isoproturon par le système ozone/peroxyde d'hydrogène Identification of the initial isoproturon oxidation by-products resulting from the ozone/hydrogen peroxide System

Résumé

Une solution aqueuse tamponnée par des phosphates (pH initial - 8) dopée en isoproturon (N- (isopropyl-4-phényl)-N-N'-diméthylurée) (~ 20 mg 1-1), a été oxydée par le système perozone, combinant l'ozone et le peroxyde d'hydrogène dans un rapport molaire de 0,5 à 0,6 moles de H2O2 par mole d'ozone. Les disparitions du composé parent, du carbone organique total (COT), du carbone total (CT) et de la consommation d'ozone, ont été suivies au cours de l'oxydation. Les premiers sous-produits d'oxydation, ceux susceptibles de conserver une formulation moléculaire proche de celle du composé initial, et par conséquent de posséder encore une activité toxique, ont été isolés et caractérisés par chromatographie gazeuse couplée à la spectrométrie de masse.

Il a été trouvé que l'isoproturon requiert un taux d'oxydation molaire de 10 moles d'ozone par mole d'isoproturon introduit, pour obtenir une élimination complète de cet herbicide. En revanche, le COT n'est pratiquement pas minéralisé, même avec de très forts taux d'ozone, ce qui indique la présence dans le milieu de sous-produits rémanents.

La plupart des premiers sous-produits d'oxydation détectés conservent le cycle aromatique dans leur structure, et au moins un atome d'azote, et sont présents à des concentrations significatives. Ces composés semblent aussi réactifs que l'isoproturon vis-à-vis de la perozonation puisqu'ils disparaissent lorsqu'on prolonge l'oxydation. De plus, l'identification de ces sous-produits laisse supposer que l'attaque des radicaux hydroxyles générés par le procédé perozone, entraîne la rupture d'une liaison C-N ou d'une liaison C-H, conduisant à la formation de composés oxygénés.

Mots-clés:

• Ozone,
• peroxyde d'hydrogène,
• isoproturon,
• premiers sous-produits d'oxydation,
• identification,
• CC/SM

Abstract

The goal of our study was to identify the initial oxidation by-products (IOBP) of isoproturon (N-(isopropyl 4 phenyl)-N-N'-dimethylurea) formed during a combined ozone/hydrogen peroxide (peroxone system) treatment. Solution of isoproturon (20 mg · l-¹ or 10[sup]4 M) were prepared in ultrapure water buffered with phosphate ions (45.9 mg · l-¹ KH2PO4 + 457.2 mg · l-¹ Na2HPO4) at an initial pH dose to 8 and an ionic strength of about 10-² mol · l-¹, and in the absence of radical scavengers (bicarbonate ions) and organic matter. Each experiment was conducted in a glass semi-continuous reactor (bubble column, capacity: 2.81, ID=40 mm, H=2 m) with recirculation of the aqueous phase (60-651 · h-¹) counter current to the gaseous phase. Ozonized air produced in the laboratory (TRAILIGAZ Labo 76 apparatus) was applied at the bottom of the column through a porous glass frit (porosity: 15 to 40 µm) at a flow rate of about 2.81 · h-¹ (ozone concentration in gas: 76 to 124 mg l-¹). The hydrogen peroxide solution (dilution from a 30 % solution FLUKA) was introduced at the level of the ozonized air entrance. The applied hydrogen peroxide/ozone molar ratio was equal to about 0.5 (or 0.4 g/g).

In the first phase of our work, primary experiments were conducted to determine the efficiency of peroxone oxidation (combined O3/H2O2) in removing isoproturon and carbon. For these experiments, the analysis of isoproturon was performed by HPLC on a SUPELCOSIL C8 column (15 cm x 4.6 mm) with UV detection at 236 nm (WATERS Model 500 pump with SPECTROMONITOR 3100 detector), using a methanol/water carrier phase (50/50 v/v, 1 ml · min-¹). Each five minute during the oxidation, total organic carbon (TOC) and total carbon (TC=TOC + mineral carbon) were controlled with DOHRMANN DC80 carbon analyser. The pH and ozone concentrations were also monitored (ozone introduced and in the off-gas by potassium iodure method, and dissolved ozone by indigotrisulfonate method). Calculation of consumed ozone was obtained by the following equation: consumed O3=introduced O3 - O3 in off-gas - dissolved O3.

The results are expressed as curves showing removals of isoproturon, TOC, TC versus the oxidation dosage (as moles of introduced ozone per mole of initial isoproturon). Their interpretation has shown that the complete disappearance of isoproturon is achieved in 12 minutes and requires about 10 moles of ozone per mole of pesticide. However, TOC was removed to only 50% for a three times higher ozone dose (27 moles per mole reached in 30 minutes). The presence of this remaining TOC (65 mg · l-¹) for such high ozone dose indicates that some by-products remain in the solution. These by-products visualised on the HPLC chromatograms for the isoproturon dosage (4 well-separated and significant peaks) seem to be as reactive as pesticide itself because of their disappearance during oxidation.

In a second phase of our work, a similar experiment was conducted over a period of 7 minutes for having up to 90% removal of isoproturon. A 1.5 litre of oxidized isoproturon solution was collected for liquid-liquid extraction with methylene dichloride ((50 ml (2 min), 25 ml (2 min), 25 ml (2 min)) after adding acid (HCI to pH 2) and salt (NaCl). After desiccation on anhydrous sodium thiosulphate (Na2SO4) and concentration under nitrogen flow, the methylene dichloride extract (extract A) was analysed by gas chromatography/mass spectrometry (VARIAN 3300 coupled with a FINNIGAN ITS 40, on-column injector: 280°C, carrier gas: helium) on DB5 capillary column (50°C to 250°C at 3°C · min), for structural identification of the oxidation by-products. Two other extracts were obtained by the same way and analysed as blanks: the initial isoproturon solution not oxidized (extract B), and the buffer solution without isoproturon oxidized under the same conditions as the pesticide solution (extract C). These two blanks have allowed to distinguish the peaks really appeared after oxidation of those either present before oxidation or produced by the oxidation/extraction of the buffer.

The GC/MS chromatogram of extract A has revealed 15 peaks really issued from the oxidation of isoproturon. The molecular weights given by the mass spectra have been correlated by chemical ionisation. The identified oxidation by-products (7 on the 15) are phenylated and/or nitrated compounds which are: 4-isopropylaniline, 4-amino-benzaldehyde, paraquinon, 4-isopropylnitrobenzene, 4-isopropylbenzene-N-for-mamide, N-(4-phenol)-N-N'-dimethylurea or « oxoisoproturon » and N-(isopropyl-4-phenyl)-N-N'- (methyl-formyl) urea. Mechanisms are suggested for the formation of these products from isoproturon. It seems that the hydroxyl radicals (OH) generated by the peroxone system attack either a C-N bond (as in the case of atrazine) or a C-H bond. The subsequent attacks (OH· or O3, O2) lead to the formation of oxygenated molecules (alcohol, carboxyl groups).

Keywords:

• Ozone,
• hydrogen peroxide,
• isoproturon,
• initial oxidation by-products (IOBP),
• identification,
• CC/MS