A novel visible-light-response photocatalyst was prepared through the heat treatment of Ti-bearing blastfurnace slag with sodium nitrate and subsequently leaching processes in which most of the SiO2, Al2O3, and MgO in Ti-slag (TS) have been separated. The photocatalytic activity of the TTS was studied by observing the evolution of H2 under the UV–Vis and visible light. Compared with the TS and commercial perovskite CaTiO3, the sample prepared exhibited an exclusive visible-light-response activity and enhanced H2 evolution.

In the view of global environmental and energy issues, hydrogen (H2) production from photocatalytic water splitting is considered as a promising solar energy conversion method.[1,2] Over the past 40 years, many photocatalysts reported, such as La-doped NaTaO3,[3] La2Ti2O7,[4] Sr2M2O7 (M = Nb, Ta),[5] K2La2Ti3O10,[6] and b-Ge3N4,[7] have been reported and exhibited high photocatalytic activities for splitting the water in the ultraviolet (UV) light region. However, it is rather diffiffifficult to directly apply these photocatalysts with wide band gap (band gap energy, Eg>3.2 eV) for water splitting because it can only become effffective for the H2 production under the UV light irradiation (wavelength < 420 nm). In fact, the UV light accounts for only ca. 4 pct of the incoming solar energy, whereas the visible (vis) light occupies the most part of solar spectrum (46 pct). Therefore, in order to achieve effiffifficient water splitting under abundant visible light irradiation, two main approaches have been adopted thus far:[1] (1) narrowing band gap by doping metal and/or nonmetal ions;[8] (2) dye sensitization to induce visible-light response.[9]

As an important valuable resource, the titanium blastfurnace slag (Ti-slag) contains about 25 pct perovskite CaTiO3 (Eg = 3.5 eV), which has a high photocatalytic activity of decomposing water into H2 and O2 under the UV light.[10,11] Although many works have been done for the applications of Ti-slag (TS),[12–15] these reported methods are aimed at recovering only one of the component. Furthermore, silicates usually show the complicated phase behavior in the solution chemical process. Therefore, it is necessary to design a method to make use of the multicomponents of TS as much as possible in a simple step. In an earlier work, we had successfully prepared the multi-doped LiFePO4 from the steel slag and converter sludge.[16,17] And in order to explore the preparation of excellent silica sol from complicated silicic acid system, the condensation behavior of silicic acid in blast-furnace slag was also studied.[18] In the present work, a novel and simple approach for fabrication of the water splitting photocatalyst from the TS is reported.

The TS used was from Pangang Group Company Ltd, China. Sodium nitrate and hydrochloric acid acting as analytical reagents were provided by Sinopharm Chemical Reagent Co., Ltd, China. Commercial perovskite CaTiO3 (CPC, purity > 99 pct) and dihydrogen hexachloroplatinate (purity ‡ 99.95 pct) deployed were brought from Alfa Aesar, USA. Analytical reagent methanol was supplied by Xilong Chemical Co., Ltd, Shantou, Guangdong, China.

The chemical compositions were determined by X-ray fluorescence (ARL Advant’X Intellipower 3600, Thermo Fisher Scientific, Waltham, MA). Thermogravimetry–differential scanning calorimetry, TG-DSC (STA 449 F3, Netzsch, Germany) was employed to estimate the light elements such as ‘H’ in the physically and (or) chemically absorbed water, which cannot be detected by the XRF technology. Phase analysis of the photocatalyst was carried out using an X-ray diffractometer (D8 Advance, Bruker AXS, Germany) with Cu Ka radiation. The diffuse reflflectance spectra were recorded using a UV–Visible spectrophotometer (UV3600, Shimadzu, Japan). The Brunauer–Emmett–Teller (BET) surface areas were measured via nitrogen physisorption (SSA- 4300, Builder, China).

The chemical compositions were determined by X-ray fluorescence (ARL Advant’X Intellipower 3600, Thermo Fisher Scientific, Waltham, MA). Thermogravimetry–differential scanning calorimetry, TG-DSC (STA 449 F3, Netzsch, Germany) was employed to estimate the light elements such as ‘‘H’’ in the physically and (or) chemically absorbed water, which cannot be detected by the XRF technology. Phase analysis of the photocatalyst was carried out using an X-ray diffractometer (D8 Advance, Bruker AXS, Germany) with Cu Ka radiation. The diffuse reflflectance spectra were recorded using a UV–Visible spectrophotometer (UV3600, Shimadzu, Japan). The Brunauer–Emmett–Teller (BET) surface areas were measured via nitrogen physisorption (SSA- 4300, Builder, China). Photocatalytic reactions among the TS, the treated Ti-slag (TTS), and the CPC for H2 evolution were conducted in a water splitting system (CEL-SPH2N, AULTT, China) with gas chromatograph (GC7890II, TECHCOMP, China).

The TS (100.0 g) was first sieved using a 200 mesh sieve, then was mixed and milled with NaNO3 (30.0 g) and was roasted at 1523 K (1250  C) for 2 hours. And the 50.0 g sintered sample passed through a 200 mesh sieve was leaching by deionized water under a mechanical agitation at 323.0 K (50  C) for 1.5 hours. The filtrate was washed by deionized water till the pH of lixivium being about 7.0. Then, the suspension was filtered, and the fltrate was reacted with 500 ml 5 pct hydrochloric acid under mechanical agitation at 323.0 K (50  C) for 1.5 hours in order to remove the metallic oxides such as Al2O3, MgO, and Fe2O3 and the complex substances, which masked the perovskite CaTiO3. The sample treated by hydrochloric acid was washed by deionized water again until the pH of the lixivium being about 7.0. The suspension was filtered and the specimen was dried in a vacuum oven at 383 K (110  C) for 12 hours.