In this paper, the influence of tetravalent dopants such as Si4+, Sn4+, Ti4+, and Zr4+ on the hematite (alpha-Fe2O3) nanostructure for enhanced photoelectrochemical (PEC) water splitting are reported. The tetravalent doping was performed on hydrothermally grown akaganeite (beta-FeOOH) nanorods on FTO (fluorine-doped tin-oxide) substrates via a simple dipping method for which the respective metal-precursor solution was used, followed by a high-temperature (800 degrees C) sintering in a box furnace. The photocurrent density for the pristine (hematite) photoanode is similar to 0.81 mA/cm(2) at 1.23 V-RHE, with an onset potential of 0.72 V-RHE; however, the tetravalent dopants on the hematite nanostructures alter the properties of the pristine photoanode. The Si4+-doped hematite photoanode showed a slight photocurrent increment without a changing of the onset potential of the pristine photoanode. The Sn4+- and Ti4+-doped hematite photoanodes, however, showed an anodic shift of the onset potential with the photocurrent increment at a higher applied potential. Interestingly, the Zr4+-doped hematite photoanode exhibited an onset potential that is similar to those of the pristine and Si4+-doped hematite, but a larger photocurrent density that is similar to those of the Sn4+- and Ti4+-doped photoanodes was recorded. The photoactivity of the doped photoanodes at 1.23 V-RHE follows the order Zr > Sn > Ti > Si. The onset-potential shifts of the doped photoanodes were investigated using the Ab initio calculations that are well correlated with the experimental data. X-ray diffraction (XRD) and scanning-electron microscopy (FESEM) revealed that both the crystalline phase of the hematite and the nanorod morphology were preserved after the doping procedure. X-ray photoelectron spectroscopy (XPS) confirmed the presence of the tetravalent dopants on the hematite nanostructure. The charge-transfer resistance at the various interfaces of the doped photoanodes was studied using impedance spectroscopy. The doping on the hematite photoanodes was confirmed using the Mott-Schottky (MS) analysis. (C) 2017 Elsevier B.V. All rights reserved.