because solar energy is freely available and hydrogen fuel is clean with high gravimetric energy density. For practical applications, it is very important to develop a low-cost water-splitting photocatalyst with an ≈10% solar-to-hydrogen energy conversion. [4] Nearly half of the energy in the sunlight that reaches Earth's surface comes from visible light photons (400-700 nm); therefore, it is important to develop a visible light active photocatalyst for efficient solar-to-fuel conversion. [5] Most of the research on photocatalysts has focused on wide bandgap semiconductors, such as TiO 2 , [6][7][8] SrTiO 3 , [9,10] and carbon nitride (CN). [11][12][13] Some of these photocatalysts have achieved very good operation stabilities that surpass 1000 h, along with more than 50% maximum external quantum efficiency for watersplitting reactions. [14][15][16] However, the wide and difficult-to-tune bandgaps of such photocatalysts restrict light absorption to only the ultraviolet (UV) region and thus fundamentally limit the practical applications for solar light harvesting to fuel. [16,17] Therefore, novel photocatalyst semiconductor materials with narrower bandgaps that can absorb a greater proportion of solar spectrum are needed to achieve the maximum theoretical solar energy conversion efficiency for practical applications.The ligand-to-metal charge transfer (LMCT) facilitated activation of TiO 2 has noteworthy potential for solar energy harvesting. However, the fast back electron transfer from TiO 2 to an oxidized sensitizer is a key limiting factor causing low photocatalyst efficiency. Herein, a new catalyst design to both increase LMCT efficiency and minimize the back electron transfer is presented. A phase-selective modification of mixed-phase TiO 2 (anatase: rutile interface) with poly-salophen organic polymer is developed. The salophen and salen family organic monomers are selectively bound and polymerized on the anatase phase but not the rutile phase, which results in the formation of a three-phase system. Such a three-phase system converts an unfavorable polymer TiO 2 core-shell structure to an intimately mixed blend morphology, consisting of interfaced crystalline rutile TiO 2 and an amorphous polymer-covered anatase-phase TiO 2 . The developed mixed-blend morphology poly-S@P25 can produce H 2 of 37 410 µmol h -1 g -1 of polymer, which is ≈3.4 times higher than core-shell poly-S@anatase TiO 2 . This approach overcomes the drawback of the traditional core-shell structured system for efficient electron harvesting from the LMCT process.