Photocatalysis has emerged as a promising and environmentally sustainable solution to produce high-purity hydrogen through ethanol photoreforming. It is commonly accepted that adding co-catalysts, especially noble metals, significantly enhances the catalytic activity of semiconductors. However, the high cost of noble metals such as Pt may limit the real application of this emerging technology. Here we evaluate the possibility of reducing the noble metal loading by creating the appropriate interface between pre-formed semiconductor nanoparticles. Commercial titania (P25) was selected as the semiconductor due to its commercial availability, facilitating the straightforward validation and corroboration of our results. Pt was selected as co-catalyst because one of the most efficient photocatalysts for the ethanol photo-reforming is still based on the use of P25 in combination with Pt. We report that the creation of induced aggregates dramatically improves the total hydrogen produced when very low loadings (≤0.05 wt%) of Pt are used. We have developed a pioneering reactor designed for conducting photoluminescence studies under authentic operational conditions of nanoparticle suspensions in the liquid phase. This approach allows us to obtain the average photoluminescence emission from the P25 agglomerates what it would be impossible to obtain by using standard solid samples holders. Thanks to this equipment, we can conclude that this remarkable improvement of the activity is mainly due to creation of an interface that favors the charge transfer between the particles of the aggregates. According to this, the titania nanoparticles of the agglomerates act as an antenna to collect the photons of the sun-light and produce the photo-excited electrons that will be transferred to the platinum nanoparticles located in the same agglomeration. In contrast, raw P25 with low loadings of Pt would have a high number of titania nanoparticles without platinum, and therefore, inactive. This result would be especially relevant in the case of immobilized photocatalytic systems for real future photocatalytic reactors because the immobilization of the semiconductors would generate similar interactions to the one created by our method. Consequently, the initial semiconductor immobilization followed by the subsequent photo-deposition of the co-catalyst emerges as a promising approach for a substantial reduction of the co-catalyst content.