Background

With more and more attention paid to the environmental issues nowadays, the study of water treatment materials emerged in a continuous stream [1,2,3,4]. Hundreds of strategies were proposed for the treatment of polluted water. However, there were many problems, such as low efficiency, low recycling rate, and secondary environment pollution, restricting their further applications [5,6,7]. The semiconductor materials were considered to be a promising candidate, and titanium oxide was recognized as one of the best photocatalyst materials due to its high photocatalytic activity and good chemical and mechanical stability [8,9,10,11,12]. Recently, TiO2 materials with nanotube (NT) array were widely studied, and the tubular morphology was proved to be a promising structure for photocatalysis. Compared with other microcosmic morphologies, TiO2 NT arrays owned several significant advantages [13,14,15,16,17]. Firstly, unique tubular structure could enhance electron-transporting efficiency and restrain the recombination of carriers, which will further produce more reactive oxygen species (ROS) [18, 19]. Secondly, TiO2 NT arrays are much easier to recycle than the TiO2 powder photocatalysts [20,21,22,23,24]. Thirdly, TiO2 NT arrays have large specific surface area and high surface energy. However, due to the relatively wide gaps (~ 3.2 eV). TiO2 NT photocatalyst is only active under UV irradiation [25,26,27,28]. In fact, a photocatalyst that is able to respond with visible light will surely take obvious advantages. At present, the focus of photocatalyst research is to adjust their light response band and improve their photocatalytic efficiency.

Building heterogeneous TiO2 photocatalysts with narrow band gaps is one of the hotspots as an attempt to overcome such impediments. Narrow band semiconductors, like Cu2O, CdS, CdTe, PbS, and Bi2O3, have been studied to build TiO2 heterojunction photocatalysts [29,30,31,32,33,34]. Among them, Cu2O (with the direct gap of ~ 2.2 eV) is regarded as one of the best candidates. For Cu2O, the response band is about 560 nm, and its band gap structure happens to well match with the energy level of TiO2 NTs. As schematically shown in Fig. 1, under the excitation of visible light, electron/hole pairs are generated and the photoinduced electrons are excited to the conduction band of Cu2O and then transfer to the conduction band of TiO2, which suppresses the recombination of electrons and holes. This heterojunction structure solves the problem that the TiO2 materials could not respond to visible light and the problem that electron/hole pairs generated on Cu2O get recombined easily. From this point of view, Cu2O-TiO2 NTA heterojunction structure materials guaranteed a natural advantage in visible light photocatalysis.

Fig. 1
figure 1

Schematic structural diagram of Cu2O-TiO2 NTA heterojunction. Under visible light illumination, the electrons were excited to the conduction band of Cu2O particles and then transferred to the conduction band of TiO2 for the matching band gap structure

Full size image

General approach to prepare Cu2O-TiO2 heterojunction photocatalysts includes chemical coprecipitation and electrodeposition, and the products have shown promising photocatalytic performances. But it is still a challenge to prepare Cu2O-TiO2 heterojunction photocatalysts with good quality by using a facile and low-cost method. Inspired by the concept of the precursor from Chemical vapor deposition (CVD), the idea, using acetate to carry copper ion to get into the inside of TiO2 NTs prepared by anodic oxidation, comes out. It is known that metal organic compounds are likely to get thermally decomposed. In this study, anodic TiO2 NTAs were functioned as “nano-container” to load copper acetate at first and then as “nano-reactors” to provide space for thermal decomposing the loaded copper acetate. After a thermal treatment, Cu2O-TiO2 TNA heterojunction films were successfully obtained. To the best our knowledge, this method has not been reported to prepare Cu2O-TiO2 TNA heterojunction. Furthermore, the phase composition, morphology, and photocatalytic activity were characterized by XRD, EDS, SEM, and spectrophotometer.

Experimental Section

The chemicals which were mentioned in the experiment process were purchased (Sinopharm Group Chemical Reagent Co. Ltd., China) and used without further purification, except the deionized water with a resistance of 18.3 MΩ cm.

Preparation of Pure TiO2 Nanotube Arrays

Anodic oxidation method was used to prepare uniform and stable TiO2 NTAs with vertical alignment [35, 36]. Metal titanium (Ti) sheets were cut into pieces of 1.5 × 5 cm2 and cleaned by a cleanser. After a sonication bath in ethanol, the Ti pieces were dried in oven. Electrolyte consisted of 535.45 g glycol, 10 g deionized water, and 1.6617 g NH4F, which were mixed and stirred for 2 h. Then, we took two pieces of Ti as anode and cathode, respectively. Immersing them into electrolyte, applying a constant potential of 50 V for 2 h, amorphous TiO2 nanotube arrays (TiO2 NTAs) were fabricated at room temperature.

Synthesis of Cu2O-TiO2 NTA Heterojunction

The amorphous TiO2 NTAs were crystallized into anatase by a thermal treatment at 450 °C. And then, they were used as substrate to prepare Cu2O-TiO2 NTA heterojunction film. First, cupric acetate (Cu(Ac)2) with different concentration was prepared, ranging from 0.05 to 0.3 mol/L. Then, annealed TiO2 NTAs were immersed into the solution transiently and dried in oven at 70 °C immediately. And the final products, Cu2O-TiO2 films, were marked as sample S1-S5 respectively by the different Cu(Ac)2 concentration of 0.05, 0.1, 0.2, 0.3, and 4 mol/L in this immersing process. After this process, the cupric acetate molecules had got into the TiO2 nanotubes. Next step was putting the samples into an atmosphere-sintering furnace of N2 with a sintering temperature of 400 °C for 150 min. The cupric acetate was thermal decomposed in a way described by Eq. (1). Finally, the Cu2O-TiO2 NTA heterojunction films were prepared. This process was schematically shown in Fig. 2.

$$ {\left({\mathrm{CH}}_3\mathrm{COO}\right)}_2\mathrm{Cu}\overset{\Delta}{\to }{\mathrm{Cu}}_2\mathrm{O}\downarrow +{\mathrm{CH}}_4\uparrow +{\mathrm{CO}}_2\uparrow +{\mathrm{H}}_2\mathrm{O}\uparrow +\mathrm{CO}\uparrow $$
(1)
Fig. 2
figure 2

Synthesis procedure of the Cu2O-TiO2 NTA heterojunction films. a Step 1, anodic TiO2 NTAs. b Step 2, fill the tubes with precursor solution. c Step 3, sintered the filled tubes at 400 °C to get the Cu2O-TiO2 NTA heterojunction. d Chemical reaction formula of the sintering process

Full size image

Just like holding a test tube containing cupric acetate, after heating, cupric acetate thermal decomposed into Cu2O which was left inside the TiO2 NTAs.

Characterization

A scanning electron microscopy (SEM, JSM-7000F, JEOL Inc., Japan) with energy dispersive spectrometer (EDS) was used for the observation of the morphology and structure. The samples were characterized by a D/max-2400 X-ray diffraction spectrometer (Rigaku, D/max-2400, Japan) and a UV-vis spectrometry (Ultrospec 2100 pro) was also used. To evaluate the photocatalytic activity of the as-synthesized Cu2O-TiO2 NTA heterojunction, we took methyl orange (MO), a typical organic indicator, as the degraded object. The Cu2O-TiO2 NTA films (3.0 × 1.5 cm2) were immersed in 5 × 10−5 mol/L of MO aqueous solution and irradiated with seven 4 W visible bulbs (Toshiba, Cool white, FL4W, Japan). Then, the solution was magnetically stirred in the dark for 30 min to ensure adsorption-desorption equilibrium prior to photocatalytic degradation. Photodegradation experiments lasted 180 min with 1.5 mL samples withdrawn periodically. The concentration of the residual MO was measured by a spectrophotometer at about 460 nm on the basis of the Beer-Lambert law. The degradation efficiency of the MO could be defined as follows:

$$ {C}_t/{C}_0=\left({A}_t/{A}_0\right)\times 100\% $$
(2)

And the varying of At/A0 referred to the changing in Ct, which represented the photocatalytic activity of the tested samples.

Results and Discussion

Figure 3 shows a typical SEM observation of the pure anodic TiO2 NTAs after annealing at 450 °C. Anodizing is an electrolytic process which converts the outer surface of metals into an oxide layer or pore structure. As shown in Fig. 3, the as-prepared TiO2 NTs have open-tube morphology with a uniform outer diameter distribution of ∼ 100 nm. The anodic TiO2 NTAs are highly ordered and oriented, and each single TiO2 NT owns very smooth tube walls with an average thickness of ∼ 10 nm. Our former studies have shown that tube length, the diameter, and morphology could be manipulated by adjusting the anodization protocols [37, 38]. The SEM results also indicate the thermal annealing at a high temperature of 450 °C does not destroy morphologies of the TiO2 NTAs. XRD is used to characterize the crystalline of the pure TiO2 NTAs (sample 1), see Fig. 4a. Results show that diffraction peaks locating at 25.3°, 36.9°, 37.8°, 48°, 53.9°, 55°, 62.7°, and 68.8° could be observed in sample 1, attributing to the (101), (103), (004), (200), (105), (211), (204), and (116) of anatase phase, respectively. As we know, there are three types of titanium dioxide phase, anatase, brookite, and rutile. Rutile could show relatively good photocatalytic ability with a granularity less than 10 nm. However, to get a rutile phase, the TiO2 sample needs to be heated up to a high sinter temperature of 800 °C, which could lead to the break of TiO2 tubes in this case. Brookite phase is hardly to be formed by using thermal annealing method for the bad thermodynamic phase stability, while anatase is the most common phase with good photocatalytic activity [39, 40]. The sharp diffraction peaks and the strong intensity of sample 1 (see Fig. 4a) indicated a highly crystallized anatase structure, which meant our TiO2 substrate was excellent not only in the morphology but also in the crystalline phase. The highly ordered TiO2 NTAs with open-tube mouth morphology were used as substrate to prepare Cu2O-TiO2 NTA heterojunction films in this study.

Fig. 3
figure 3

Typical SEM images of pure TiO2 nanotube arrays without modification. a Top view and b side view, indicating the highly ordered vertical alignment structure with open-tube mouth morphology. The tube diameter is about 100 nm, and the tube length is about 10 μm

Full size image
Fig. 4
figure 4

a XRD patterns of the Cu2O-TiO2 NTA heterojunction samples. Sample S0: pure anatase TiO2 NTA sample; samples S2, S4, and S5: thermally decomposed sample with immersing in 0.1, 0.3, and 4 mol/L Cu(Ac)2 solution, respectively. b EDS result of the Cu2O-TiO2 NTA heterojunction films, showing the existence of elements Ti, Cu, and O. The results confirm the successful loading of Cu2O on the TiO2 NTAs

Full size image

XRD patterns of the TiO2 NTAs loaded with Cu2O nanoparticles in concentration gradient ranging from 0.05 to 4.0 mol/L are also shown in Fig. 4a, and the 4.0 mol/L sample was prepared by a cycling immerse process described in Additional file 1, the “Experimental Details” part. The samples were named samples 2 to 4 with the increasing Cu(Ac)2 concentration. Except the TiO2 peaks, there was no peak of Cu2O showing up in sample 2 because of the tiny amount of the loading Cu2O particles. And particles might be decorated inside of the TiO2 “nano-container” which also raised the difficulty for characterization. In sample 3 and sample 4, obvious cuprite peaks could be observed at 29.6°, 36.4°, 42.3°, and 61.3°, attributing to the cuprite (110), (111), (200), and (220) of Cu2O, respectively. It should be noted here that sample 4 was only used to characterize the existence of Cu2O particles, and its synthetic details were described in Additional file 1. Moreover, the lattice parameters and the grain size were calculated based on the XRD data. After removing the background and Kα2 diffraction, and following the smoothing and fitting process, we got the average lattice parameters of our samples of a = b = c = 4.2646 Å, which matched with the standard PDF. The standard PDF showed that the lattice parameters of Cu2O are: a = b = c = 4.2696 Å, and Cu2O had a cubic structure [41]. The average grain size of Cu2O was calculated as ~ 47 nm, by using Debye-Scherrer formula:

$$ D=\frac{K\gamma}{B\cdot \cos \theta } $$
(3)

In Eq. (3), D is the grain size, K is the Scherrer constant, γ is the wavelength of X-ray, B is FWHM which needs to be in the radian, and θ is the diffraction angle. XRD results indicate that the Cu(Ac)2 were loaded into the TiO2 NTAs and successfully decomposed into Cu2O inside the same TiO2 NTAs, and then the Cu2O-TiO2 NTA heterojunction films got formed. To further investigate the Cu2O-TiO2 NTA heterojunction, an elemental analysis was carried out by using EDS. Figure 4b showed an EDS diagram of Cu2O-TiO2 NTA heterojunction film which was prepared with 0.2 mol/L Cu(Ac)2. The atomic percentages were 7.32, 28.96, 57.45, and 6.27% for elements Cu, Ti, O, and impurity C. This result showed that the Cu2O owned a relatively low content in the heterojunction sample, but it still brought about the visible-light activity, which would be discussed later in the MO degradation experiment. The EDS results well agreed with the XRD results in Fig. 4a that cuprite Cu2O was successfully loaded to anatase NTAs.

Figure 5 showed the top view SEM results of the modified TiO2 NTAs. Compared with the pure TiO2 NTA samples in Fig. 3, a few small particles could be seen near the upper and inside of the TiO2 tubes in Fig. 5a. Increasing the modification amount, a number of nanoparticles could be observed obviously in Fig. 5b. Figure 5c was sample 4 that we discussed before. Large parts of the tube surface were covered by the redundant Cu2O, indicating that sample 4 was over-decorated. Based on the SEM images, the size distribution of Cu2O particles was estimated ranging from ~ 30 to ~ 80 nm, which well agreed with XRD calculated grain size of ~ 47 nm. For the tubular structure of the three samples, they still retained the vertical alignment state, but some tubes got a little awry. It was considered as the influence of thermal decomposition process, which needed a heating process of 400 °C to get Cu(Ac)2 decomposed into Cu2O. High temperature in the decomposition step had a negative effect on the tubular structure, supported by the SEM images. However, if the heating temperature in the thermal process went too low to 240 °C, Cu(CH3COO)2·H2O would just get dehydrated instead of decomposed. So the temperature should be controlled in ~ 300 to 400 °C to keep the nano-scale tubular structure and ensure the fabrication of Cu2O-TiO2 NTA heterojunction. It can be concluded that the Cu2O-TiO2 heterojunction could be formed, and the morphology retained well, when the decomposition happens at 400 °C.

Fig. 5
figure 5

Typical SEM images of the Cu2O-TiO2 NTA heterojunction films. a Sample immersed in 0.2 mol/L Cu(Ac)2. b Sample immersed in 0.3 mol/L Cu(Ac)2. c Sample immersed in Cu(Ac)2 of over high concentration

Full size image

The Cu2O nanoparticles were loaded on TiO2 NTAs to fabricate the heterojunction, which was expected to enhance the photo-response ability in visible light range, so UV-vis characterization was adopted to investigate optical properties of the as-synthesized Cu2O-TiO2 NTAs. Figure 6a shows the UV-vis absorption spectra of the Cu2O-TiO2 NTA samples with Cu2O-loaded magnitude increasing from none to 4.0 mol/L. It could be seen in Fig. 6a that the pure TiO2 NTAs without loading Cu2O only showed out a high absorption in the ultraviolet region (< 380 nm), due to its intrinsic material properties. After loading the Cu2O particles, the absorption range was expanded to 600–700 nm. And when the intensity is increasing with the raising of the Cu2O modification magnitude, the absorption value of the Cu2O-TiO2 heterojunction films also got increased. Figure 6a indicated that TiO2 NTAs were given the visible light response ability by decorating Cu2O nanoparticles. UV-vis along with SEM, EDS, and XRD results proved that the Cu2O-TiO2 NTA heterojunction was fabricated successfully by the thermal decomposition method, and samples showed the enhanced visible light absorption.

Fig. 6
figure 6

a UV-vis spectra of the Cu2O-TiO2 NTAs and absorption get expanded to visible light range and raised with the loading amount of Cu2O. b Visible-light photocatalytic degradation kinetics of MO treated by the heterojunction films with different Cu2O content. When decoration magnitude of Cu2O on TiO2 NTAs increased, the MO decomposition efficiency under visible light irradiation raised. Sample S0 referred to pure TiO2 film, and sample S1–S5 were the samples immersed in Cu(Ac)2 solution with the concentrations of 0.05, 0.1, 0.2, 0.3, and 4 mol/L, respectively

Full size image

Photocatalytic activities, one of the most important properties of the Cu2O-TiO2 NTA films, were evaluated through degradation of MO aqueous solution. The visible-light photocatalytic degradation kinetics was shown in Fig. 6b. The MO degradation rate was proportion to the loading amount of Cu2O approximately. The more Cu2O particles were loaded on TiO2 NTAs, the faster MO got degraded. Sample S1 degraded MO to 91.0% in 3 h under visible light irradiation, while sample S4 degraded MO to 86.4% in 3 h under the same condition. MO degradation rate represented the photocatalytic activity of the samples. Comparing with the photocatalytic degradation rate to the ~ 2.73% of CdTe-TiO2 by a pulse electrodeposited method [29], ~ 45% of Bi2O3 by a ultrasonication-assisted successive ionic layer adsorption and reaction (SILAR) technique [32], and ~ 27.25% of Cu2O by a square wave voltammetry method [33], photoactivity of this as-synthesized Cu2O-TiO2 sample was improvable. However, as a facile new strategy, it still conduced to improve fabrication method. When the Cu2O loading amount went up, there was a trend that the photocatalytic activity of our as-synthesized Cu2O-TiO2 NTA heterojunction films increased. It indicated the Cu2O content had a positive influence of the visible-light photocatalytic activity. TiO2 itself only responded to ultraviolet, and the visible light range photocatalytic ability should come from the decoration of Cu2O. As shown in Fig. 7, the conduction band bottom of Cu2O was a little higher than that of TiO2, while the valence band top of Cu2O was higher than that of TiO2. So, the photoinduced electrons were excited to the conduction band of Cu2O and then transferred to the conduction band of TiO2. As a direct-gap semiconductor, wave vector of Cu2O was just the same at the bottom of the conduction band and the top of valence band. It meant that only the changes of energy were required, instead of the changes of momentum. This energy band structure led to the situation that carriers recombined easily. However, due to the help of heterojunction structure, the photogenerated electrons on Cu2O transferred to TiO2 NTAs which suppressed the recombination of electron/hole pairs. The longer the pairs existed, the more easily the ROS got produced which brought this photocatalytic activity. As more Cu2O loaded on TiO2 NTAs, the heterojunction fabricated better. And the photocatalytic ability got promoted. So, the Cu2O content showed a positive influence of the visible-light photocatalytic activity. However, further increase of Cu2O content as well as the photocatalytic ability is limited, due to the solubility of Cu(Ac)2 in aqueous solution which was 7.2 g (0.36 mol/L) at room temperature. And sample S5 with Cu(Ac)2 concentration of 4.0 mol/L is prepared by a cycling immerse process described in Additional file 1, the Experimental Details part. The photocatalytic degradation of the MO followed pseudo-first-order kinetics [42] and the kinetic reaction could be expressed as:

$$ {A}_t={A}_0{e}^{- kt} $$
(4)
Fig. 7
figure 7

Band gap structure of Cu2O and TiO2 before (left) and after (right) contact. When the Cu2O-TiO2 heterojunction is formed, the electron/hole pairs photogenerated on Cu2O could transfer to TiO2 NTAs

Full size image

While our degradation curve showed nearly a straight line, it is not an exponential function. So, there was still room for improvement. And the solubility limit could be broken by the repeated immersion method we mentioned before, with further investigation of the Cu(Ac)2 concentration and repetition times to avoid adverse effects. In this study, as this thermal decomposition method was what we concerned and tried to illustrate, we just took the 0.3 mol/L (close the solubility of 0.36 mol/L) as the maximum concentration of Cu(Ac)2 solution. And the photocatalytic activity in visible light range of our as-synthesized heterojunction was confirmed by the MO degradation results. Our previous study found that the Degussa P25 had similar ultraviolet photocatalytic activities with TiO2 NTAs, when the power P25 was placed on a glass substrate [28]. It can be concluded that we have successfully prepared Cu2O-TiO2 NTA heterojunction films with visible-light photocatalytic activities.

Conclusions

In summary, we have successfully prepared the Cu2O-TiO2 NTA heterojunction films by a simple thermal decomposition process. SEM, EDS, and XRD results show that TiO2 NTAs with a tube diameter of ~ 100 nm were loaded by Cu2O nanoparticles with an average size of ~ 50 nm. The anodic TiO2 NTAs functioned as both “nano-container” and “nano-reactors” to load and synthesize the narrow-band Cu2O nanoparticles, which has not reported before. UV-vis spectra indicate that the absorption range of the TiO2 NTAs was expanded from ultraviolent range to visible light range, due to the loading of Cu2O. Photocatalytic testing indicated that there was a visible-light photocatalytic activity of the as-synthesized Cu2O-TiO2 heterojunction. The photocatalytic abilities of the Cu2O-TiO2 NTA heterojunction films were found to be increased with the Cu2O content from 0.05 to 0.3 mol/L. Our current work has demonstrated a novel and facile method to prepare Cu2O-TiO2 NTA heterojunction films, which could also be promising for environmental and energy-related areas.