Oxidation of dichloracetic acid (DCA) by two titanium dioxide catalysts (P25 and a nanoparticulate rutile (NR)) was measured at pH 3 and the results compared with parallel measurements on the same TiO(2)s modified by deposition of platinum. Although the platinum loading corresponded to less than one platinum crystallite for each TiO2 crystal, platinization significantly increased photocatalytic activity - 30-fold in the case of NR, high area rutile. This has been attributed, by ourselves [7] and others, e.g. [9], to the increased efficiency of hole-generation as a consequence of electron trapping by the platinum. Each catalyst was then milled, and UV-vis transmission spectra were used to show that milling improved the catalyst dispersion. Milling unplatinized TiO2 caused either insignificant decreases (P25) or increased activity (NR). This contrasts with the previously reported [16] significant decrease in activity for propan-2-ol oxidation. The difference is attributed to a different dependence on UV intensity of DCA degradation and propan-2-ol oxidation. DCA oxidation rate increased linearly with UV intensity, I, whereas propan-2-ol oxidation varies as I-0.5. However, milling did decrease significantly the rate of DCA oxidation by both of the platinized catalysts, especially by the platinized rutile. It is suggested that most of the UV is absorbed by TiO2 crystals that have no platinum, but that in poorly dispersed TiO2, UV-generated electrons may be transported along a chain of TiO2 crystals to a platinized TiO2 at which they are trapped. This is analogous to the antenna model postulated by Bahnemann and co-workers [15]. It is proposed that milling breaks down the antennae. Consequently electron transport is disrupted and the effectiveness of electron trapping by platinum is reduced and the measured photoactivity decreases. (C) 2010 Elsevier B.V. All rights reserved.