Fig. 1. Schematic representation of the polymeric precursor solution preparation and thin films fabrication process for pristine and Zr-modified hematite photoelectrodes.

Fig. 2. STEM images of hematite single solvent (a) and mixing solvent (b) and Zr4+ addition layers. Bottom: Magnified regions represented by red squares highlighting the contrast of poor and rich contact between photocatalyst thin film and commercial FTO glass substrate.

Fig. 3. Schematic representation of the different molecular configurations leading to better contact interface between FTO and photocatalyst during thermal treatment. More suitable molecular packing along with the Zr4+ segregation at hematite surface is providing an appropriate environment for the growth of a back-contact-rich interface.

Fig. 4. TEM images of pristine hematite/FTO interface showing (a) low magnification image and HRTEM images at two different magnifications for (b) the tip of FTO/hematite interface (white box in image a) and (c) filtered HRTEM of lateral FTO (SnO2)/hematite interface (white box in image (d); (e,f) Fast Fourier Transform (FFT) images of hematite and FTO (SnO2) regions in image c indexed with the crystallographic planes in green and red dots along with the zone axis in the right side bottom corner of the images. TEM images of FTO/ZrH3% interface (hematite containing Zr) showing (f) low magnification image and HRTEM images with two magnifications for (g) the lateral interface and (h) filtered HRTEM of lateral FTO/ZrH3% interface (white box in image h), (i,j) FFT images of hematite (α-Fe2O3) and FTO (SnO2) regions in image h.

Fig. 5. (a) Linear sweep voltammograms of pristine hematite, ZrH1%, ZrH2%, ZrH3% and ZrH4% measured in 1.0 mol/L NaOH at scan rate of 10 mV s-1. The solid lines represent the response under AM 1.5G illumination (100 mW cm-2) and the dashed lines under dark conditions, and (b) Overall efficiency (ƞoverall) obtained for the ratio Jph/Jabs for all the prepared thin films, (c) Photocurrent densities extracted at 1.23 V (RHE) from literature using the same chemical procedure method and similar planar thin film morphology showing a Si4+-doped (black line, ) [17], Cu2+-doped (blue line, ) [31], pristine Hematite multiple layers (green line, ) [27], Sn4+-doped/FeNi (orange line, ) [28], and the ZrH3%/FeNi (present work, red line, ). The reported photocurrent densities were measured in 1.0 mol/L NaOH aqueous solution at pH = 13.6.

Fig. 6. (a) EIS spectra for pristine-H (black line, ) and ZrH3%/FeNi (red line, ) measured at 1.23 V vs. RHE under illuminated conditions; the dashed lines represent the fitting curves obtained from the equivalent circuit inserted in a). (b) Mott-Schottky plots recorded at 1 kHz in dark condition for Pristine-H and ZrH3%/FeNi. Solid lines represent the linear fit to the selected experimental data. The experiment was measured in 1.0 mol/L NaOH solution at pH = 13.6.

Fig. 7. Cross-section SEM images for FTO (a), pristine hematite (b), ZrH3% (c), ZrH3%/FeNi (d) and their corresponding AFM topographical images (e).

Fig. 8. STEM image and chemical mapping images obtained from EDS for the ZrH3%/FeNi thin film. The area marked with a yellow box was used to drift correction and the green box is the mapping region.

Fig. 9. STEM image and STEM/EELS energy shift mapping (ΔeV) for L3-edge of Fe in the rectangular green box. Variations of oxidation state and chemical environment of iron are displayed by the ΔeV color map where differences from the brown color (ΔeV equal 0) represents differences from the standard L3-edge peak position for iron in hematite. The square yellow box was used for drift correction in the STEM/EELS mapping.

Fig. 10. Linear sweep voltammograms of pristine-H, ZrH3% and ZrH3%/FeNi measured in 1.0 mol/L NaOH + 5% H2O2 for ZrH3%/FeNi at scan rate of 10 mV s-1. The solid lines represent the response under AM 1.5G illumination (100 mW cm-2) and the dashed lines under dark conditions.

Fig. 11. (a) Cyclic voltammetry of ZrH3%/FeNi photoanode up to 200 scans from 0.6 to 1.8 V vs. RHE under simulated light conditions (forward scan and 50 mV/s scan rate and light intensity calibrated at 100 mW; (b) Chronoamperometry curve of ZrH3%/FeNi at 1.23 V vs. RHE.