Tembotrione (95% purity) was supplied from Institute for plants, Belgrade, Serbia. Silver electrode was supplied from Austin, USA. Methanol, potassium dihydrogen phosphate, sodium hydrogen phosphate, sodium hydrogen carbonate, sodium hydroxide, all of analytical grade purity, was obtained from Merck (Darmstadt, Germany).

As supporting electrolytes in voltammetric measurements different pH solutions were used: KH₂PO₄/NaOH buffer, NaHCO₃/NaOH buffer and Na₂HPO₄/NaOH buffer.

All voltammetric measurements were carried out using a computerized voltammetric analyzer CH Instruments (Austin, USA) driven by voltammetric software CHI (Version 4.03). The three electrode system consisted of working electrode, reference electrode and auxiliary electrode. Working electrode was mercury buttoned silver electrode (Ag/Hg). Reference electrode was Ag/AgCl (3M KCl) (CHI 221). Auxiliary electrode was platinum wire (CHI 111) (Figure 2).

For pH measurements, a pH meter, model 744, equipped with combined glass pH electrode and temperature sensor, all from Metrohm (Herisau, Switzerland) was used.

Tembotrione solution (concentration of 6 × 10^–3 mol/L) was prepared. First, 0.0053 g of tembotrione (95% purity) was measured on analytical balance and dissolved in 2 ml of absolute methanol. Then, 2 ml of supporting electrolyte was added, so the concentration of tembotrione was 3 × 10^–3 mol/L.

Silver/Amalgam electrode was prepared by a left to stand silver electrode in mercury for 24 hours. Silver is dissolved and silver amalgam (most probably crystalline Ag₂Hg₃) is formed. Thus obtained electrode has a lifetime of a month. Afterwards electrode needs to be regenerated. This is done by polishing it on powder of aluminum

(0.05 µm), then rinsing with water and placing it in an ultrasonic bath. When this is done, electrode is placed in mercury for 24 h.

Cyclic voltammograms were recorded using Ag/Hg electrode by gradually increasing scan rate. They were recorded at scan rates: 25 mV∙s⁻¹, 50 mV∙s⁻¹, 100 mV∙s⁻¹, 200 mV∙s⁻¹, 300 mV∙s⁻¹, 400 mV∙s⁻¹, 500 mV∙s⁻¹, 600 mV∙s⁻¹ and 700 mV∙s⁻¹ for pH values 6, 8, 10 and 12. Different scan rates were selected to see if there is a shift of peaks with increasing scan speed. That is the first condition to be seen whether occurring electrochemical processes is reversible and whether the tested compound behaves electrochemically correctly. Molecule of tembotrione is large organic molecule with many functional groups which have different forms at different pH, so it is expected that behavior would be different. Because of that it was necessary to examine his behavior at different pH values. Concentration of tembotrione in all experiments was 3 × 10^–3 mol/L.

Tembotrione contains several groups that may undergo electron transfer reactions and therefore the voltammetry is used for mechanistic and analytical studies [12,13]. For the investigation of electrochemical behavior of tembotrione cyclic voltammetry technique is used, with conventional three-electrode cell system and electrochemical workstation. On the basis of tembotrione structure, the reduction process probably involves reduction of two carbonyl groups on the cyclohexane ring, and third carbonyl group [11,14]. During the experiment pH values of aqueous solution variated while constant concentration of tembotrione was maintained. Cyclic voltammograms were recorded by gradually increasing scan rate.

Figure 3 shows the cyclovoltammogram of tembotrione at scan rates 25 mV∙s⁻¹, 50 mV∙s⁻¹, 100 mV∙s⁻¹, 200 mV∙s⁻¹, 300 mV∙s⁻¹, 400 mV∙s⁻¹, 500 mV∙s⁻¹, 600 mV∙s⁻¹ and 700 mV∙s⁻¹, recorded on the instrument CHI 760. Supporting electrolyte was 0.1 M phosphate buffer solution, pH = 6. Mercury buttoned silver electrode (Ag/Hg) was working electrode, Ag/AgCl (3 M KCl) was reference electrode, and platinum wire was auxiliary electrode.

On this cyclovoltammogram on potential (E) of −1.2068 V to the Ag/AgCl, well-defined spreading cathodic (reduction) peak was observed, which probably corresponds to the joint reduction of two carbonyl group on the cyclohexane ring. It is likely that the carbonyl groups on the cyclohexane have the same acidity at a given pH, which is leading to the spread of peaks. On the potential (E) of −1.4564 V another peak was observed. It probably corresponds to the reduction of the third carbonyl group [11,14]. The mechanism proposed is evaluated on the basis of the expected effects of structure [15,16].

No anodic peak is observed in the reverse scan. This suggests that the reduction of tembotrione is irreversible at Ag/Hg electrode at pH 6.

Joint reduction of carbonyl group on the cyclohexane ring:

Reduction of the third carbonyl group:

The dependence of the signal height with the square root of scan speed was shown at Figures 4 and 5 (R₁ = 0.99901 and R₂ = 0.99376). In electrochemistry processes in which the current dependence of the square root is linear behave electrochemically properly, only diffusion current occurs, i.e. reaction taking place on the electrode is diffusion controlled. Voltammetric peaks came from diffusion current [17].

Figure 6 shows the cyclovoltammogram of tembotrione at scan rates 25 mV∙s⁻¹, 50 mV∙s⁻¹, 100 mV∙s⁻¹, 200 mV∙s⁻¹, 300 mV∙s⁻¹, 400 mV∙s⁻¹, 500 mV∙s⁻¹, 600 mV∙s⁻¹ and 700 mV∙s⁻¹, recorded on the instrument CHI 760. Supporting electrolyte was 0.1 M phosphate buffer solution, pH = 8. Mercury buttoned silver electrode (Ag/Hg) was working electrode, Ag/AgCl (3 M KCl) was reference electrode, and platinium wire was auxiliary electrode.

On the potential of −1.2317 V cathode (reduction) peak was observed, which probably corresponds to the joint reduction of carbonyl group in positions 1 and 3 of cyclohexane ring. On the potential of −1.4342 V was observed peak (signal 2), which probably comes from the reduction of the third carbonyl group. This peaks (signal 1 and 2) have lower intensity than the peaks at pH = 6, which indicates possible participation of chemical reactions (protonation). No anodic peak is observed in the reverse scan. This suggests that the reduction of tembotrione is irreversible at Ag/Hg electrode at pH 8.

Figure 7 (peak 1) shows the dependence of the signal height with the square root of scan speed (for comparing the correction factors). It indicates the quasireverse process and possible changes at diffusion layer. Nonlinearity of curve (correction factor R = 0.92835) indicates the participation of chemical reaction (protonation).

Figure 8 (peak 2) also show the dependence of the signal height with the square root of scan speed (for comparing the correction factors). It indicates that the

electrochemical reactions are diffusion controlled, voltammetric peaks came from diffusion current [17].

Figure 9 shows the cyclovoltammogram of tembotrione at scan rates 25 mV∙s⁻¹, 50 mV∙s⁻¹, 100 mV∙s⁻¹, 200 mV∙s⁻¹, 300 mV∙s⁻¹, 400 mV∙s⁻¹, 500 mV∙s⁻¹, 600 mV∙s⁻¹ and 700 mV∙s⁻¹, recorded on the instrument CHI 760. Supporting electrolyte was 0.1 M carbonate buffer solution, pH = 10. Mercury buttoned silver electrode (Ag/Hg) was working electrode, Ag/AgCl (3 M KCl) was reference electrode, and platinum wire was auxiliary electrode.

On the potential of −1.4330 V only one, well-defined, cathode (reduction) peak was observed. At pH 10 there is only a reduction of third carbonyl group. Loss of signal 1 is maybe due to enolization of two symmetric carbonyl groups (have hydrogen of α type) in basic conditions. This is not happening at acidic pH values (6 and 8). Reverse peak wasn’t observed. This suggests that the reduction of tembotrione is irreversible at Ag/Hg electrode at

pH 10.

Figure 10 shows the dependence of the signal height with the square root of scan speed (for comparing the correction factors). It indicates that the electrochemical reactions are diffusion controlled, voltammetric peaks came from diffusion current [17].

Figure 11 shows the cyclovoltammogram of tembotrione at scan rates 25mV∙s⁻¹, 50 mV∙s⁻¹, 100 mV∙s⁻¹, 200 mV∙s⁻¹, 300 mV∙s⁻¹, 400 mV∙s⁻¹, 500 mV∙s⁻¹, 600 mV∙s⁻¹and 700 mV∙s⁻¹. Supporting electrolyte was 0.1 M phosphate buffer solution, pH = 12. Mercury buttoned silver electrode (Ag/Hg) was working electrode, Ag/ AgCl (3 M KCl) was reference electrode, and platinum wire was auxiliary electrode.

On the potential of −1.4198 V only one, well-defined, cathode (reduction) peak was observed. Reverse peak wasn’t observed. This suggests that the reduction of tembotrione is irreversible at Ag/Hg electrode at pH 12. Loss of signal 1 can be explained in the same way as for

pH 10.

Figure 12 shows the dependence of the signal height with the square root of scan speed (for comparing the correction factors). It indicates that the electrochemical reactions are diffusion controlled, voltammetric peaks came from diffusion current [17].

Figure 13 shows the dependence of potential (E) of the solution pH, for signal 2. It indicates that the peak position changes with change of pH value. In this case potential increased with increasing pH. Based on this, can be concluded that change of position for signal 2 probably comes from participation of both, chemical and electrochemical reaction. Signal 1 at pH 6 have potential of −1.2068 V, at pH 8 −1.2317 V.

Table 1 shows values of current 1) and pH, for peak 2. It indicates that the peak intensity change with pH. This change in signal intensity most likely is due to the participation of chemical and electrochemical reaction. From the table we can see optimum pH value at which the

Table 1. Current dependence of pH for peak 2.

current intensity is greatest. The optimum pH value is 12.