Tantalum nitride films integrated with transparent conductive oxide substrates via atomic layer deposition for photoelectrochemical water splitting

The first example of tantalum nitride electrodes on transparent conductive oxide substrates, which enables solar water splitting, is presented.


Tantalum Nitride Films Integrated with Transparent Conductive Oxide Substrates via Atomic Layer Deposition for Photoelectrochemical Water Splitting
Ti: The Ti main peak before and after annealing in ammonia were located at the same position.
Upon annealing in ammonia, however, two new peaks associated to Ti-O-N (blue) and Ti-N (green peak) groups were appeared which confirms that beside Ta, N is also doped into TiO. [3][4][5] O: As expected, oxygen peak (metal oxygen bond) at 29-30 eV was not affected by annealing in ammonia, however, the atomic percentage of oxygen was changed by both annealing in ammonia and the atomic percentage of Ta. The other Oxygen peaks were assigned to C=O and C-OH groups.
N: As it can be seen, the as-deposited films did not have any nitrogen, however, after annealing in ammonia the nitrogen signal was emerged. At high concentration of Ta, there is a small shoulder (orange peak), therefore, it was considered as a different type metal-N bonding.
Ta: Before annealing in ammonia, there was only one type of Ta present which can be associated to Ta-O groups. Upon annealing in ammonia, however, Ta signals became broadened which is an indication of doping of Ta into the structure with a range of atomic interaction with its neighbors.
The Ta signal for the film with high concentration of Ta, i.e. %5 Ta, could not be fitted to one peak, and it was fitted to two peaks. Therefore, there are two types of Ta present. This observation is in line with the extra N peak for the film with high concentration of Ta. Then it was hypothesized that at high % Ta doping, TaNx would segregate off from TiO2 and form a separate phase.
The atomic percentages of Ta, Ti, O and N were calculated using the folowing equation:

% 100
The Sx is the normalized peak area associated to each element. For normalization, the raw peak area was divided to the sensitivity factor. The sensitivity factore of Ti, Ta The diffraction patterns were unambiguously matched to anatase TiO2. Since the ionic radii of Ta 5+ is slightly larger than Ti 4+ (0.64 vs 0.60), 6  This is consistent with the extra N and Ta peaks observed in XPS ( Figure S17). On the other hand, For the films with % Ta doping in between, however, the diffraction peaks are shifted to the lower angles. GSAS 7,8 was used to calculate the cell parameters. The calculated volumes vs % Ta doping level are shown in Figure S20b.       To eliminate the effect of flow rate, high throughput ammonia (~ 500 SCCM) was used and the annealing temperature and duration were optimized. Based on the XRD, the minimum temperature required to form pure phase Ta3N5 was found to be 750 °C.  The effect of the ammonia flow rate on crystallinity and phase purity of the films were studied by annealing at 750 °C for 2 hours with different flow rate of ammonia. From Figure S9, films annealed at low flow rate of ammonia, e.g. 50 and 100 ml min -1 , had poor crystallinity and were comprised of impure phases (possibly TaON). On the other hand, for the flow rate more than 200 ml min -1 , the observed peaks were sharper (more crystalline) and Ta3N5 was the only detectable phase. This observation is consistent with previous study, where thin films of TaOx and Ta3N5 were prepared on Ta foil. It was found that at lower flow rates the formation of TaON is more favorable where at higher flow rates only films are totally nitridized to Ta3N5. 2 Here we found that the optimum ammonolysis conditions to form pure and highly crystalline Ta3N5, is annealing at 750 °C for 2 hours with flow rate of ammonia of ≥ 200 ml.min -1 . The XRD of the ALD deposited TaOx and TaOxNy are compared in Figure S10. As it can be seen the ALD deposited TaOx were completely nitridized to Ta3N5 after ammonolysis at the same at 750 °C for 30 minutes.