Photo-flow chemistry has become an important research area due to the ability of this technology to boost reaction rates and productivity. This field unites the mass transfer enhancement of flow chemistry with the high energy field density of microstructured geometries. However, even though the space-time yield increases dramatically with microphotoreactors, the overall productivity of a single microreactor module remains low for many applications. This study shows that for a photochemical rare earth element separation reactor, choosing a five-times thicker characteristic length compromises 40% of the space-time yield due to lower energy density. However, this can triple the product throughput and improve energy efficiency, which is important when numbering-up photoreactors. This work addresses the question of "How micro?" and aims to introduce a new methodology to seek an optimum point of energy efficiency without compromising the high productivity achieved by photo-flow chemistry. (C) 2016 Elsevier B.V. All rights reserved.