Abstract

Homogeneous p-type cobalt (II) oxide (CoO) nanoparticles were successfully deposited on n-type three-dimensional branched TiO2 nanorod arrays (3D-TiO2) through photochemical deposition and thermal decomposition to form a novel CoO/3D-TiO2 p-n heterojunction nanocomposite. Due to the narrow band gap of CoO nanoparticles (~2.4 eV), the as-synthesized CoO/3D-TiO2 exhibited an excellent visible light absorption. The amounts of deposited CoO nanoparticles obviously influence the hydrogen production rate in the photoelectrochemical (PEC) water splitting. The assynthesized CoO/3D-TiO2-5 obtains the highest PEC hydrogen production rate of 0.54 mL h−1 cm−2 after five-time CoO deposition cycles (at a potential of 0.0 V vs Ag/AgCl). The photocurrent density of CoO/3D-TiO2-5 is 1.68 mA cm−1, which is ca. 2.5 times greater than that of pure 3D-TiO2. The results showed that the formation of internal electrical-field between the CoO/3D-TiO2 heterojunction, which has a direction from n-type TiO2 to p-type CoO, facilitated the charge separation and transfer and resulted in a high efficiency and stable PEC activity.

Keywords

CoO nanoparticles . Three-dimensional TiO2 nanorod arrays . p-n heterojunction . Photoelectrochemical . Water splitting

Introduction

Photoelectrochemical (PEC) water splitting hydrogen production is a promising technology to transfer solar energy to hydrogen energy. During the past decade, one-dimensional (1D) TiO2 nanostructure (nanowire, nanorod, and nanotube) arrays have been extensively investigated to PEC water splitting hydrogen evolution due to its special advantageous properties, such as direct electrical pathways for photogenerated carriers and convenient recycling [1–4]. Furthermore, the 3D branched TiO2 nanorod array nanostructure with increased surface area has been also attracted the attention of scientists because of its distinctive physicochemical properties [5, 6]. It is well know that pure 1D nanorod TiO2 (rutile) arrays with the wide binding energy of 3.0 eV can hardly absorb visible light [7], which limits its efficiency of solar energy utilization. However, compared with 1D nanorod arrays, the 3D branched TiO2 nanorod arrays with a multiple scattering effect among the branches can significantly improve the light harvesting efficiency and therefore, possess a higher photon-to-current conversion efficiency [6]. At the same time, a great deal of researches has been done to increase the visible light absorption performance of TiO2 photocatalyst, including elementdoping [8, 9], self-doping [10, 11], semiconductorsensitizing [12, 13], and structure-varying [5, 14].

Sensitized TiO2 nanostructure, which was modified with other narrow bandgap semiconductors, such as CdS [15, 16], Bi2WO6 [17], dye [18, 19], etc., displays a promising activity under visible light irradiation. Gong et al. [15] sensitize 3D TiO2 nanorod arrays with CdS that achieve a nearly four times larger photocurrent than that of the pristine 3D TiO2 nanostructure. However, the antitoxic stability of the CdS seriously influences its practical application. Recent years, due to its nontoxicity, low cost, and highly electrochemically activity, p-type semiconductor cobalt (II) oxide (CoO) with direct band gap of 2.4 eV [20, 21] received great attentions in photocatalytic field [22]. It have been demonstrated that CoO plays an important role in photocatalytic hydrogen production [21] and photodegradation of organic pollutants under visible light irradiation [22]. Ngo et al. [20] fabricated CoO/TiO2/ SrTiO3/Si heterogeneous composite photocatalyst; the band offsets of CoO/TiO2 was measured by in-situ XPS test and farther calculated by density functional theory (DFT). It was suggested that the deposited CoO could greatly promote the water splitting performance of TiO2 under visible light irradiation.

In this paper, n-type 3D branched TiO2 nanorod arrays (3D-TiO2) modified with p-type CoO nanoparticles by a simple photochemical deposition and thermal decomposition strategy. It was indicated that benefiting from the unique p-n heterostructures and inside Schottky barriers, the asconstructed CoO/3D-TiO2 achieved a higher photocurrent density and more excellent PEC water splitting performance than pure TiO2 under the simulated solar light irradiation. Furthermore, to well understand the enhanced photocatalytic mechanism, a possible charge separation and transportation processes between CoO and TiO2 were also discussed.

Experimental

1D-TiO2 was prepared using a modified hydrothermal method according to our previous report []. Ten milliliters of deionized water was mixed with 10.0 mL of concentrated hydrochloric acid (36 wt% HCl) in a Teflon-lined stainless steel autoclave (30-mL capacity) and stirred for 5 min. Then, 0.24 mL of titanium butoxide (97 % Aldrich) was introduced and stirred for another 5 min. When the solution changed to transparent, two pieces of well-cleaned FTO substrates (3.5 × 2.0 cm2) were placed against the wall of the Teflon-liner with the conductive side facing down. The sealed autoclave was kept in an oven and maintained at 170 °C for 6 h. After which, the autoclave was removed out and allowed to cool to room temperature naturally. The FTO substrates with TiO2 nanorod was taken out and washed by deionized water.

Preparation of 3D branched TiO2 nanorod arrays

The 3D-TiO2 was prepared using a dip-coating method [6]. Typically, the FTO glass substrates with TiO2 nanorod arrays were placed upside down in a beaker containing 15.0 mL of deionized water, 15 μL of hydrochloric acid (37 wt%), and 150-μL TiCl3 solution. The beaker was kept at 80 °C for 1 h and cooled down to room temperature naturally. Then, the sample was removed and cleaned by deionized water and ethanol to remove the surface remnant. Finally, the FTO substrates containing 3D-TiO2 nanorod arrays were annealed in air at 350 °C for 1 h with a heating rate of 4 °C min−1.

Synthesis of CoO/3D-TiO2 heterojunction

The as-prepared 3D-TiO2 was dipped into a 0.1-M Co(NO3)2 and 0.1-M ethylene glycol aqueous solution for 30 s. Then, the 3D-TiO2 film was irradiated under UV-vis for 120 s followed by immersed in deionized water for 60 s to clean the film. The entire procedure was performed for one to several cycles to obtain a desired amount of deposition. The CoO/ 3D-TiO2-X heterojunction was achieved in argon (100 sccm) atmosphere at 300 °C in a tubular furnace for 1 h with a heating rate of 5 °C min−1 (X denotes the repeated deposition cycle times).

Characterization

Field-emission scanning electron microscopy (FESEM; Philips FEI Quanta 200 FEG) and transmission electron microscopy (TEM) (JEOL-2010 microscope operated at 200 kV) were used to observe the morphology and microstructure of the prepared samples. The X-ray diffraction patterns were carried out with an X-ray diffractometer (XRD; Rigaku, D/max 2500 v/pc) at a scan rate of 4 min−1 in the 2θ range from 20° to 70°. The chemical nature of Co was performed with X-ray photoelectron spectroscopy (XPS) in Krato Axis Ultra DLD spectrometer with Al Kα radiation. Optical diffuse reflectance spectra were recorded by a UV-vis diffuse reflectance spectrometry (DRS; V-560, Jasco). The photoluminescence (PL) spectra were carried out using a LS 50B (Perkin Elmer, Inc., USA) with an excitation wavelength of 260 nm.

PEC performance and PEC water splitting hydrogen evolution test

In the PEC test, the CoO/3D-TiO2-X was employed as a working electrode; a platinum wire and Ag/AgCl electrode were used as the counter electrode and reference electrode, respectively. A 0.1 M of NaNO3 aqueous solution was used as electrolyte. A 100 mW cm−2 of light irradiation was obtained from a simulated solar (CEL-HXF300E7, Beijing China Education Au-light Co., Ltd). An electrochemical workstation (AUTOLAB PGSTAT302) is used to measure current-voltage (I-V) characteristics of the working electrode with a scan rate of 20 mV s−1 . The PEC water splitting hydrogen evolution test was carried out in a CEL-SPH2N hydrogen evolution system (Beijing China Education Au-light Co., Ltd), the CoO/3D TiO2-X sample flake acts as work electrode, and the gas from the Pt wire is collected by a homemade device. The device is made of a heated Pt wire and impaled in the bottom of 5-mL graduated centrifuge tube.

Results and discussion 

Characterization of photocatalysts

X-ray diffraction

Figure 1 shows XRD patterns of 3D-TiO2 nanorod arrays and CoO/3D-TiO2-X heterojunction. The main diffraction peaks at 36.5° and 63.2° could be indexed to the (101) and (002) lattice planes of rutile TiO2 (JCPDS No. 88-1175) after subtracting the diffraction peaks of the FTO substrate. Besides, another distinctive peak with 2θ values of 42.6° is observed in the each CoO/3D-TiO2-X sample, matched well with (002) lattice plane of cubic structure of CoO (JCPDS No. 48-1719). Figure 1 also displays that the intensity of the CoO characteristic diffraction peaks is enhancing as the deposition times are increasing, implying the increased loading amounts of the CoO nanocrystal.

FESEM

Figure 2 displays the FESEM images of 3D-TiO2 and CoO/ 3D-TiO2-X heterojunctions fabricated with various deposition cycles, and EDX spectrum of CoO/3D-TiO2-5 sample. The pure 3D-TiO2 (see in Fig. 2a) is enclosed by short needle-shaped branches with a diameter of ca. 50 nm, which improves the specific surface area of TiO2 nanorod arrays [6]. Figure 2b–e reveals the amounts and the evolution of CoO nanoparticles deposited on 3D-TiO2 under various photochemical deposition cycles. It clearly indicates that as the deposition cycles were increasing, the amount of CoO nanoparticles is enhancing, the needleshaped branches are gradually covered totally by the CoO nanoparticles, and the shape changes to approximate shortrod; some of the branches even disappear completely. CoO nanoparticles congregate obviously as the deposition cycles were increasing to seven times. The EDX spectrum (see in Fig. 2e) of CoO/3D-TiO2-5 implies that the asprepared sample is composed of Ti, Co, and O elements (C comes from the based material of the carbon-conductive adhesives, Si and Sn come from the FTO glass substrate, and the Pt comes from Pt-sputtering of the samples before morphology observation.). No other impurity peaks are observed in both XRD and EDX patterns, revealing the fine purity of all the samples.