INSTALACIÓN TUBERÍAS

Two experiments, to test the relationships of accumulated mass versus deployment time and diffusion layer thicknesses, were carried out to validate the principle of DGT for measuring pesticides. The masses of targeted chemicals accumulated by DGT increased linearly (for 7

chemicals sorbed by HLB and 5 chemicals with XAD18, r2 _{values were higher than 0.99) with}

55

chemicals (see Figure S3.5). For DGT devices with HLB resin gel, the results of ETH showed significant deviation from the theoretical line after the deployment of 36 hours. For devices with XAD resin gel, only three target chemicals, ATR, THI and CLO, followed the theoretical line. The other six chemicals showed different degrees of deviation at different deployment times. These results indicate that the performance of DGT with HLB is better than that with XAD18 gel for measuring pesticides. A further test of DGT principle for pesticides was carried out using HLB DGT devices with different thicknesses of diffusive gel in a well stirred solution. The measured mass of the target compounds that diffused through the diffusive gel layer was inversely proportional to the diffusion layer thickness (Figure S3.6). The experimental data agreed well with the theoretical line obtained from the equation (3.1). Both results of time dependence and diffusion layer thickness confirm the principle and mechanism of the DGT technique for pesticides in solution.

The results obtained from the different diffusion layer thicknesses also indicate the diffusive boundary layer (DBL) at the surface of the device is insignificant during the experiment under stirred condition and it can be neglected in calculations. In the real environment, the DBL usually cannot be negligible, the value of DBL can be obtained by deploying different thicknesses of DGT devices and calculated through Eq. 3.1 (Warnken et al., 2006).

3.3.6 Effect of pH, ionic strength and DOM

Pesticides are ionic organic chemicals, which possess at least one polar functional group. They can be neutral, cationic, anionic or zwitterionic, depending on the pH of the solution. Their physicochemical properties may change with the environmental conditions, which can also affect the performance of DGT.

To assess the pH effect on the DGT measurement, DGT devices were immersed in solutions with the pH ranged from 4.7 to 8.2. The ratio of the target compound concentrations measured

by DGT (CDGT) to their concentrations in the bulk solutions (Cb) were plotted against pH values

56

measurement by DGT with HLB binding gel as most of the ratios (CDGT/Cb) were between 0.9

and1.1. However, for DGT with XAD18 binding gel, the CDGT/Cb ratios were below 0.9 at the

pH 7 for all tested compounds and at pH 6 and 7.5 for most compounds. These results demonstrate that the DGT with HLB binding gel can accurately measure concentrations of pesticides in the aquatic environment with a wide range of pH, whereas DGT with XAD 18 binding gel has its limitation.

The effect of ionic strength on DGT measurements was investigated in solutions with ionic strength similar to freshwater, estuary water and seawater, ranging from 0.01 M to 0.5 M. For DGT with HLB binding gel, there was no significant effect observed in the range of 0.01 M to 0.25 M,

as shown in Figure S3.8. The ratios of CDGT to Cb were within 0.9 and 1.1 for all tested chemicals.

At the ionic strength of 0.5 M (close to seawater), the DGT measured concentrations were slightly

lower than expected. The ratio of CDGT to Cb was below 0.9 for ATR, THI and CLO, and close to

0.9 for other six chemicals. This is probably because the viscosity of the solution was higher on addition of a large amount of NaCl, which led to an impediment to the mass transfer process (Castells et al., 2003).

The effect of dissolved organic matter (DOM) on measurements of target chemicals by DGT

devices with HLB resin as binding phase was demonstrated in Figure S3.9. The ratios of CDGT/Cb

were between 0.9 and 1.1 for majority of the chemicals at various DOM concentrations up to 20

mg L-1_{. There was very little difference between the various DOM concentrations for all nine}

chemicals. The CDGT/Cb ratios of some chemicals, such as CHL, FLU, PIR and CLO were below

0.9, they are similar to the ratios for the control solution where the DOM concentration was zero. These findings suggest that the performance of DGT is independent of DOM concentration. Similar phenomena have been observed in the study of Li et al. (Li et al., 2011) on the impact of POCIS for PPCPs (pharmaceuticals and personal care products) and EDS (endocrine disrupting

substance), where Rs was not affected by DOM. Li et al.’s research (Li et al., 2016) on

57

In general, the performance of DGT devices with HLB resin gel was better than the DGT devices with XAD18 resin as binding gel. DGT with HLB resin gel was therefore selected as the suitable devices for the future experiments and measurements.