Protecting TiS3 Photoanodes for Water Splitting in Alkaline Media by TiO2 Coatings

Abstract

Titanium trisulfide (TiS₃) nanoribbons, when coated with titanium dioxide (TiO₂), can be used for water splitting in the KOH electrolyte. TiO₂ shells can be prepared through thermal annealing to regulate the response of TiS₃/TiO₂ heterostructures by controlling the oxidation time and growth atmosphere. The thickness and structure of the TiO₂ layers significantly influence the photoelectrocatalytic properties of the TiS₃/TiO₂ photoanodes, with amorphous layers showing better performance than crystalline ones. The oxide layers should be thin enough to transfer photogenerated charge through the electrode-electrolyte interface while protecting TiS₃ from KOH corrosion. Finally, the performance of TiS₃/TiO₂ heterostructures has been improved by coating them with various electrocatalysts, NiS_x being the most effective. This research presents new opportunities to create efficient semiconductor heterostructures to be used as photoanodes in corrosive alkaline aqueous solutions.

Keywords: KOH corrosion; TiO2 coatings; heterostructures; hydrogen production; photoelectrolysis; water-splitting.

Figure 1

(a) TGA curve of TiS₃ showing the time evolution of the relative mass (left axis) and temperature (right axis). (b) Mass spectrometric ionic currents (recorded at m/q = 18, 48, and 64) as a function of time, measured during the TGA experiment. (c) Relative mass vs time at different applied temperatures. (d) Relative mass of TiS₃ samples when heated at different temperatures between ambient temperature (25 °C) and 450 °C; inset: representation of the Arrhenius equation.

Figure 2

XRD diffractograms of three samples: TiS₃, ox-Ar-230, and ox-air-90. TiS₃ and TiO₂-anatase peaks correspond to PDF 00–015–0783 and PDF 01–071–1167 files, respectively. The most intense peaks of TiS₃ and TiO₂-anatase are indicated with a red asterisk and a green circle, respectively.

Figure 3

S 2p (left) and Ti 2p (right) XPS spectra for TiS₃, ox-Ar-20, ox-Ar-54, and ox-air-20 samples; the data are stacked along the vertical axis for clarity. Experimental data (dotted lines), single fitting curves (colored continuous lines), and total fitting (continuous black line). The single fitting curves and total fitting are shown only for the TiS₃ case.

Figure 4

(a) SEM image of sample ox-air-90. (b) EDX mapping analysis superposed with the image in (a). (c) EDX mapping results for oxygen (red), sulfur (yellow), and titanium (blue).

Figure 5

Optical microscopy images of (a) ox-Ar-140 and (b) ox-air-30 samples. (c) Raman spectra in the edge and the center of one ribbon of the ox-Ar-140 sample (marked with 1 and 2 in (a), respectively). (d) Raman spectra for the ox-air-30 sample in the edge and the center of one ribbon (marked with 1 and 2 in (b), respectively). TiS₃ peaks are indicated with *, and anatase TiO₂ peaks are indicated with “a”.

Figure 6

(a, b) Collection of high-resolution TEM images recorded in ox-Ar-140 and ox-air-30, respectively, and (c, d) the corresponding electron diffraction patterns. They are oriented with respect to the images and the spacing for the most prominent reflections is included.

Figure 7

Optical density spectra obtained from transmittance measurements of TiS₃ samples grown on fused silica and oxidized (a) in air (right axis in orange corresponds just to the OD of the sample oxidized for 300 min) and (b) under Ar flow for different times. Ratio between optical density values of the (2) and (4) bands (OD₂/OD₄) against the oxidation time for samples oxidized (c) in air and (d) under Ar flow whose OD curves are shown in (a, b). (e) Kubelka–Munk function obtained from diffuse reflectance measurements for samples TiS₃ (red), ox-air-12 (light blue), ox-air-20 (dark blue), ox-air-90 (yellow), and ox-Ar-240 (green) as a function of the incident photon energy. (f) Relationship between F(R)₂ over F(R)₄ as a function of the oxidation time. Corresponding samples to the data are labeled in the figure. Dashed lines in c, d, and f are used as visual guides.

Figure 8

(a) Chronoamperometry measurement at 1.95 V vs RHE for ox-air-12 in the dark (OFF) and light (ON). (b) Photocurrent as a function of the oxidation degree estimated by the ratio F(R)₂/F(R)₄, described in the optical characterization section. (c) Schematic diagram of the formation of the electron–hole pair and the pathway of the hole to the surface of the heterostructure. (d) Mott–Schottky plots for TiS₃, ox-air-12, and ox-air-20 at 500 Hz. TiO₂ data has been included from ref (24). (e) Zoom to the region closer to C^2– = 0 F^–2.

Figure 9

SEM images of (a) TiS₃ and (b) ox-air-12 at different magnifications after the PEC experiments. (c) Raman spectra of TiS₃ and ox-air-12 after the PEC measurements. For comparison, it is also shown the spectra of the sulfur powder collected from the ampules in which the TiS₃ samples are synthesized. TiS₃ Raman bands are indicated with a *.

Figure 10

Top: Photocurrents obtained at 1.95 V vs RHE for samples oxidized in air for 12 min, without photoelectrocatalyst (ø), with NiS_x, Ni thin film (4 nm thick), Ni nanoparticles (Ni-np), and BCN. Bottom: Schematic of the material.