Physicochemical properties of agrochemicals - their effects on foliar penetration

Uptake and translocation of 26 chemicals (herbicides, fungicides, growth regulators, insecticides and model compounds), formulated with and without 'Ethylan TU' (NP8) surfactant, have been determined 24, 48 and 72 b after spraying onto leaves of maize, rape, strawberry and sugar beet. Eight of the chemicals may be classified as polar (log partition coefficient (P) < 0) and 18 as lipophilic (log P 0.8-4.6). Chemicals penetrated more readily into waxy leaves of rape and strawberry than into less-waxy sugar beet, whilst lowest rates were observed for maize. Rates of uptake varied from exceedingly low (< 2% 72 h after application) for uracil, maleic hydrazide, isoproturon and methyl phenylurea applied to maize, to rapid (> 98% within 24 h of application) for acephate and triadimefon applied to strawberry leaves. The effects of NP8 on rates of chemical uptake were classified into 4 types, viz:-(i) low rates, with limited enhancement by NP8; (ii) high rates unaffected by NP8; (iii) marked enhancement by NP8 within 24 h; and (iv) continuous enhancement by NP8 over 0-72 h. Translocation into the tissues immediately adjacent to the target site was greater for waxy than less-waxy plants, enhancement by NP8 was attributed to its effects on penetration rather than to surfactant interactions at the sub-cuticular level. Logarithms of physicochemical properties (partition coefficient, water solubility S, molecular weight MW), together with deposit area and melting point (m.p.) values were used to investigate relationships with uptake and translocation. In most cases the variability could not be described adequately in a simple linear form. Regressions using the entire group of 26 chemicals also showed poor correlations between uptake patterns and individual variables. Polynomials in log S, log P, log MW and m.p. were used to test for complex relationships. A quadratic polynomial in log P, log S and m.p. accounted for slightly more (77%) of the variability between the means for total uptake than a cubic polynomial in log P and log S (75%). Although both regressions were significant, the 'lack of fit' indicated that other factors need to be included to account fully for the variability in rates of uptake and translocation.