This project set the goals to identify the most suitable photocatalysts made of earth-abundant-elements for integration within a photoelectrochemical cell (also called artificial leaf) in perspectives of future fabrication of large-sized devices for large-scale solar H2 production.
Following are main objectives:
(i) Preparation of nanostructured BiVO4, Fe2O3, WO3 photocatalysts being suitable for the solar-driven water oxidation reaction (also called oxygen evolution reaction).
(ii) Preparation of nanostructured Cu2O, CZTS photocatalysts being suitable for the solar driven hydrogen evolution reaction.
(iii) Investigation on catalytic properties of these photocatalysts with aims to gain insights into the influences of nanostructures, chemical compositions, surface chemistry etc on their catalytic activities and stability.
(iv) Formulation of catalyst inks and their use for fabrication of photoelectrodes, devices having size of 1x1 cm2 and 10x10cm2. Investigation on the operation of these photoelectrodes and devices.

-    Theoretical results:
(i) We have successfully fabricated a series of nanostructed BiVO4, WO3 and -Fe2O3 photocatalysts using scalable solution methods, namely the electrodeposition, hydrothermal treatment, and solution precipitation.
The BiVO4 loaded with AgOx nanoparticles were found to show significant enhanced photocatalytic activity and stability in comparison to a bare BiVO4 counterpart. The best one, namely BiVO4/AgOx-11.32%, showed a photocurrent onset potential of 0.4V vs. RHE and a remarkable photocatalytic current density of 4.65 mA/cm2 at applied potential of 1.23V vs. RHE when being immersed in a pH 7 phosphate buffer electrolyte and illuminated by a simulated 1 sun light. The BiVO4/AgOx developed in this project is placed among the top-tier photocatalysts made of BiVO4 for the water oxidation reaction.
The WO3 photocatalyst developed in this project also displayed attractive photocatalytic properties: it showed a photocurrent onset potential of 0.4V vs. RHE and a photocatalytic current of 2 mA/cm2 at applied potential of 1.23V vs. RHE. It was found that its catalytic activity and stability were strongly dependent on the light illumination direction as well as the nature of the electrolyte employed (pH and nature of anion). Regardless of WO3 film thickness, the back-side illumination mode (light was sent from FTO substrate to WO3 layer) always showed higher photocatalytic current density but also a faster degradation in comparison to a front-side illumination (light was sent from WO3 layer). The dissolution of W into electrolyte was found as a reason of the degradation.
The -Fe2O3 photocatalyst was prepared by annealing FeOOH precursor in air, e.g. at 550 C. Different morphologies like nanodisk, nanodendrite were prepared. However, the materials obtained showed rather poor photocatalytic activity. In a pH7 and under 1 sun illumination, -Fe2O3 photocatalyst showed a photocurrent onset potential of 0.6V vs. RHE (which was not too bad in comparison with the current state-of-the-art, 0.6–0.8V vs. RHE) and a low photocurrent density of only 50 µA/cm2.
In another approach, we have prepared Fe@FeOx core/shell nanoparticles with average diameter of 10-15 nm and functionalized their surface with a [Ru(phenanthroline)3]2+ complex via a phosphonate linker. In this design, the [Ru(phenanthroline)3]2+ acts as a light harvester which absorbs the visible light and the Fe@FeOx acts as a water oxidation catalyst. We found that the electronic communication between the [Ru(phenanthroline)3]2+ and the Fe@FeOx was the limitation for the overall performant of the hybrid photocatalyst. The quenching of excited state of [Ru(phenanthroline)3]2+ by O2 molecule was suggested to be a poisonous event to the photocatalytic operation.
(ii) We have succesfully fabricated a series of Cu2O, CZTS photocatalysts using scalable solution methods, namely the electrodeposition and solution precipitation
The Cu2O photocatalyst was prepared by an electrodeposition process using a Cu2+ lactate alkaline deposition bath or by a reduction process using a Cu(OH)2 precursor and citrate reducing agent. Cu2O thin film, nanoparticles having cubic or hollow sphere mopholorgy were obtained. The surface of Cu2O was then modified by interfacing with TiO2 or CdS, that helped to create a p-n junction. In a pH 7 phosphate buffer electrolyte and under 1 sun illumination, the best Cu2O/TiO2 and Cu2O/CdS showed comparable photocurrent onsent potential of 0.5V vs. RHE. However, the Cu2O/CdS showed higher photocurrent density at applied potential of 0V vs. RHE in comparion to a Cu2O/TiO2: 1.9 versus 0.5 mA/cm2. The CdS was found to offer a better protection to Cu2O against its photocorrosion in comparison to a TiO2 protective layer.  
Efforts have been focused to prepare CZTS photocatalysts. However, the resultant CZTS, which showed crystalline structure and nice hexagonal nanodisks, showed NO photocatalytic activity for the solar H2 generation.
(iii) We have indentified appropriate formulation for preparing stable catalyst inks of BiVO4, WO3, -Fe2O3, Cu2O/TiO2, Cu2O/CdS and CZTS by using common solvents like H2O, MeOH, H2O/MeOH, H2O/EtOH, mixture together with polymer additives like Aristoflex velvet, Nafion. From these catalyst inks, electrodes were fabricated by drop-casting or doctor blade technics. A doctor blade setup was custumized during this project. By using this technics, electrodes of large sclae like 10x10 cm2 having catalyst layer of µm thickness were fabricated.
(iv) Two configuration were adopted to fabricate photoelectrochemical cell: the Honda-Fujishima design using a BiVO4/AgOx photoanode and a Pt wire cathode (design 1) and a Z-scheme two photoelectrodes using a BiVO4/AgOx photoanode and a Cu2O/TiO2 (or a Cu2O/CdS) photocathode (the design 2). The solar-driven water splitting performance of these two designs was assayed in a pH 7 phosphate buffer electrolyte and using a simulated 1 sun light illumination. The design 1 provided solar-to-hydrogen conversion yield of 1.6% which is stable for hours (e.g. 12 h experiment was tested). Whereas the design 2 obtained a conversion yield of 0.4% which tended to degrade quickly because of Cu2O based photocathode degradation.
-    Applied results: The fabrication and operation of photoelectrochemical cells should be further optimized before move to the step of prototype fabrication.

The work on BiVO4/AgOx demonstrated AgOx as an excellent hole collector that significantly enhanced the photocatalytic activity and stability of BiVO4 even AgOx itself was not a good electrocatalyst for the water oxidation reaction. This finding was important as it had been largely accepted that an excellent electrocatalyst for the water oxidation reaction was required to improve the performance of BiVO4. Apparently, an efficient electronic communication between the hole collector (or electrocatalyst) and the light harvester is more important to the overall performance than how fast the hole collector/ electrocatalyst uses the holes to drive the water oxidation reaction (the dark reaction).
The work on WO3 photoanode demonstrated significant influences of light illumination direction and the nature of electrolyte to the photocatalytic performance and stability of this material. We demonstrated (for the first time to the best of our knowledge) a close relationship between the photocatalytic current degradation and the dissolution of WO3 into electrolyte. We were able to propose a mechanism for the corrosion and degradation of WO3 wherein the dangerous impacts of H2O2 and OH. radical were emphasized.
The work on Fe@FeOx@Ru(phenanthroline) provided a new design at molecular level for fabrication of hybrid photoanodes. Even the photocatalytic performance of the new photoanode was not excellent but it provided two important fundamental insights (i) the covalent bond helps to promote significantly the electronic communication between the [Ru(phenanthroline)3]2+ light harvester and the Fe@FeOx water oxidation catalyst; and (ii) the quenching of the excited state of [Ru(phenanthroline)3]2+ by O2 was a factor that decreased the overall photocatalytic performance. Thus, a new challenge was identified: we should accelerated the O2 generation (the water oxidation reaction) but at the same time should evacuate efficiently the O2 molecule from the whole system.

- Scientific papers in referred journals (list):
1. Hoang V. Le, Minh D. Nguyen, Yen Thi Hai Pham, Duc N. Nguyen, Ly T. Le, Hyuksu Han, Phong D. Tran, Decoration of AgOx hole collector to boost photocatalytic water oxidation activity of BiVO4 photoanode, Materials Today Energy, 2021, Volume 21, 100762 (DOI: 10.1016/j.mtener.2021.100762)
2. Hoang V. Le, Phuong T. Pham, Ly T. Le, Anh D. Nguyen, Ngoc Quang Tran, Phong D. Tran, Fabrication of tungsten oxide photoanode by doctor blade technique and investigation on its photocatalytic operation mechanism, International Journal of Hydrogen Energy, 2021, Volume 46, 22852-22863 (DOI: 10.1016/j.ijhydene.2021.04.113)
3. Quyen T Nguyen, Elodie Rousset, Van TH Nguyen, Vincent Colliere, Pierre Lecante, Wantana Klysubun, Karine Philippot, Jérôme Esvan, Marc Respaud, Gilles Lemercier, Phong D Tran, Catherine Amiens, Covalent Grafting of Ruthenium Complexes on Iron Oxide Nanoparticles: Hybrid Materials for Photocatalytic Water Oxidation, ACS Applied Materials & Interfaces, 2021, DOI: 10.1021/acsami.1c15051
- Technological products (describe in details: technical characteristics, place):
+ 02 BiVO4/AgOx-11.32% photoanode sized of 1x1 cm2 and 10x10 cm2;
+ 02 Cu2O/CdS photocathode sized of 1x1 cm2 and 10x10 cm2;
+ A Photoelectrochemical cell consisted of a BiVO4/AgOx photoanode (10x10cm2) and a Pt plate;
+ An artificial leaf made of BiVO4/AgOx photoanode (10x10 cm2) and a Cu2O/CdS photocathode (10x10 cm2);

The research team expects to receive another funding support to further investigate on the degradation issues of BiVO4, WO3 photoanodes and on the identification of appropriate strategies to stabilize these materials. Such an investigation is critical to be able making sustainable photoelectrochemical cells for long-term operation, e.g. for few 100 to 1000 hours.