The effects of radio frequency atmospheric pressure plasma and thermal treatment on the hydrogenation of TiO₂ thin film

1.   Introduction

Titanium dioxide (TiO₂) is a widely used multifunctional n-type semiconductor material with several advantages, such as great accessibility, excellent stability, nontoxicity, high photocatalytic properties and moderate cost, making it an attractive candidate for a wide range of potential applications in H₂ production, photovoltaics, photocatalysts, lithium-ion batteries, fuel cells, gas sensors, detoxification and supercapacitors [1–11]. Nevertheless, the light absorption range of TiO₂ is limited to the ultraviolet (UV) region (< 5%) due to its wide band gap (anatase: 3.2 eV, rutile: 3.0 eV) [12]. In order to expand and improve the light utilization, many strategies have been developed to tailor the TiO₂ band gap, including doping, composite formation, sensitization, co-catalyst loading, etc [13]. Hydrogenation, as one of the common doping strategies, has also been investigated [14–17]. The black coloration and improved visible light response of hydrogenated TiO₂ have been considered as promising ways to enhance solar absorption and photocatalytic activity [18–22].

Table 1 summarizes the thermal and plasma hydrogenation strategies, preparation method, reaction conditions and possible hydrogenation mechanism. Chen et al [18] hydrogenated TiO₂ nanoparticles in a 20 bar H₂ atmosphere at about 200 ℃ for 5 days. Wang et al [19] treated TiO₂ in a tube furnace filled with ultrahigh purity H₂ atmosphere at 200 ℃‒550 ℃ for 30 min, and Lu et al [23] hydrogenated TiO₂ in a 35 bar H₂ atmosphere at about 15 ℃ for 3‒20 days. The hydrogenated (black) TiO₂ is obviously superior to pristine (white) TiO₂ in optical absorbance [18, 19, 23], photocatalytic H₂ generation [18, 19, 23] and methylene blue (MB) and phenol degradation [18].

Table  1.  Summary of the literature on the preparation of black TiO₂ by hydrogenation. Abbreviations: atmospheric pressure (AP), oxygen vacancy (OV) and inductively coupled plasma (ICP).

Preparation method	Reaction conditions	Hydrogenation mechanism	References
H2 thermal annealing	20.0 bar, pure H2, 200 ℃, 5 days	Surface disordered layer	[18]
	35 bar, pure H2, about 15 ℃, 3‒20 days	Ti3+, crystalline-disordered core–shell structure	[23]
	20‒40 atm, pure H2, 773 K, 2‒10 h	Ti3+ and OVs	[30]
	20 bar, ultrahigh purity H2, 500 ℃, 1 h	Ti3+ and OVs	[31]
	20 bar, ultrahigh purity H2, 500 ℃, 1 h‒3 days	OVs, Ti3+ states, Ti interstitials and surface reconstruction	[32]
	20 bar, pure H2, 500 ℃, 1 h	OVs	[33]
	AP, ultrahigh purity H2, 200 ℃‒550 ℃, 30 min	OVs	[19]
	AP, pure H2, 500 ℃, 1 h	Ti3+	[20]
	AP, Ar/H2, 600 ℃, 6 h	Surface amorphous layer and OVs	[34]
	AP, H2/N2, 450 ℃ and 650 ℃, 2 h	Ti3+ and OVs	[35]
	Ar/H2, 650 ℃ and 850 ℃, 5 min‒2 h	Oxygen loss and Ti3+	[36]
	AP, air, Ar, Ar/H2, H2, 300 ℃‒900 ℃, 15‒960 min	Ti3+ and OVs	[37]
	Hot wire annealing, 1 bar, pure H2, 245 ℃, 265 ℃ and 290 ℃, 20 min	Disordered shell (Ti3+ and OVs)	[38]
Low pressure H2 plasma treatment	200 mTorr, Ar/H2, 13.56 MHz, 300 W, 200 ℃, 1 min	OVs	[24]
	25.8‒27.1 mTorr, pure H2, ICP, 3000 W, 150 ℃, 0.5‒20 min	Trapped holes (O− centers) and Ti3+	[21]
	2 Torr, Ar/H2, ICP, 13.56 MHz, 200 W, 25 ℃, 5 min	Surface amorphous layer and OVs	[34]
	13.3 Pa, pure H2, 13.56 MHz, 200 W, 260 ℃, 8 min	Ti3+ and OVs	[39]
	2000 Pa, pure H2, HFCVD, 350 ℃‒500 ℃, 3 h	Surface disordered layer (H+, Cr3+, OVs)	[40]
	50 Pa, pure H2, 13.56 MHz, 40 W, 1.5 h	Ti3+ and –OH groups	[29]
	40 Torr, pure H2, microwave plasma, 2.45 GHz, 400 W, 30 min	Ti3+, surface disordered layer	[41]
	20 mTorr, pure H2, DC plasma, 25 ℃‒350 ℃	H-doping and/or OVs	[42]
AP H2 plasma treatment	He/H2, 13.56 MHz, 180 W, 425 ℃, 5 min	Ti3+ and/or OVs	[26]
	Jet, He/H2, 13.56 MHz, < 10 W, < 100 ℃	Ti3+	[25]
Thermal plasma treatment	Pure H2, 200 W, 500 ℃, 4‒8 h	Ti–H bonds	[20]

 | Show Table

DownLoad: CSV

However, the practical use of the above thermal methods is limited due to the harsh treatment conditions (high pressure and/or temperature H₂ atmosphere, long time). Comparatively, plasma presents an energy efficient and sustainable approach to preparing black TiO₂. Wang et al [20] hydrogenated TiO₂ in a thermal plasma furnace (200 W) using H₂ plasma for 4‒8 h at 500 ℃. Yan et al [21] treated TiO₂ using an inductively coupled plasma (3000 W) under a 25.8–27.1 mTorr H₂ atmosphere at 150 ℃ for different times (30 s, 1, 3, 5 and 20 min). Ren et al [24] reduced TiO₂ using radio frequency (RF) plasma (13.56 MHz, 300 W) under a 200 mTorr Ar/H₂ atmosphere at 200 ℃ for 1 min. These cases have shown that low pressure plasma has high hydrogenation efficiency, decreasing preparation time. However, the process needs a vacuum chamber and pump, with expensive costs and complicated operation. Atmospheric pressure (AP) plasma has also been used to prepare black TiO₂ and has shown high hydrogenation efficiency, reducing equipment and handling costs [25]. In our previous work [26], TiO₂ coated on Ni foam was successfully hydrogenated by RF AP He/H₂ plasma within a short time (5 min). The prepared black TiO₂ film presented good visible light absorption and photocatalytic activity in MB degradation. The temperature in the reaction chamber was about 425 ℃ at 180 W.

From these previous works, it can be seen that high pressure or AP H₂ thermal treatment usually takes a long time and generates both bulk and surface defects [27, 28], while plasma hydrogenation is mainly a surface activated process (high-energy species bombardment) and creates surface defects effectively without deterioration of the bulk [20, 21]. However, the temperature of RF AP plasma discharge usually reaches several hundred degrees Celsius without external heating [26]. Hence, the effects of high-energy species bombardment and high temperature on TiO₂ hydrogenation are coupled together for RF AP plasma hydrogenation. It is necessary to clarify what the different effects of the plasma and the discharge gas temperature on the hydrogenation process are. Furthermore, the presence of Ti³⁺ species, oxygen vacancies (OVs), Ti–OH groups [29] and Ti–H bonds [20] strongly depends on the hydrogenation process and parameters. The role of H₂ in creating this defect doping also deserves to be investigated.

In this paper, we report on and compare the TiO₂ film hydrogenation process using RF AP He/H₂ plasma and thermal annealing at the same discharge gas temperature in order to research the key effects of the processes. The color of the original TiO₂ film changes from white to black after being hydrogenated in He/H₂ plasma at 160 W (gas temperature ~381 ℃) within 5 min, while the color of the thermally treated TiO₂ film did not change obviously, even in pure H₂ or He/H₂ atmosphere with higher temperature (470 ℃) and longer time (30 min). The effects of He/H₂ plasma and thermal treatment on the optical absorbance, chemical composition, crystalline structure and surface morphology of TiO₂ film were investigated and discussed. A possible mechanism analysis of the hydrogenation process was also proposed. The investigation will provide a novel insight into the preparation of black TiO₂ thin film with high visible light response activity.

2.   Experiment

2.1   Preparation of TiO₂ thin films

TiO₂ thin films were prepared using AP dielectric barrier discharge-enhanced chemical vapor deposition. An AC power supply of 10 kHz (CTP-2000K, Nanjing Suman Electronic Co., Ltd, China) and a homemade quartz discharge chamber (discharge gap: 2.0 mm) were employed in this experiment. Titanium tetrachloride (TiCl₄, ≥98.0%, Sinopharm Chemical Reagent Co., Ltd, China) was selected as the titanium precursor. A bubbler containing TiCl₄ was immersed in a water bath which was heated and maintained at 40 ℃. The connecting tube from the bubbler to the discharge chamber was kept at 70 ℃ using a heating band to prevent the TiCl₄ vapor from condensing. Ar (99.99%, Shanghai Central Gases Co., Ltd, China) was used as the discharge and carrier gas. O₂ (99.99%, Shanghai Central Gases Co. Ltd, China) was used as the reactive gas. The gas flow rates were controlled by mass flow controllers (MFCs, Beijing Sevenstar Electronics Co., Ltd, China) and set as Ar: O₂: carrier gas (TiCl₄) =600:15:25 ml min⁻¹. A quartz slice (1.5 × 2.0 × 0.5 mm³) was used as the substrate and placed into the discharge chamber after cleaning with ethanol and deionized water. The quartz substrate was heated by an electric heater (at the bottom of the reactor) and the heating temperature was measured and monitored by a thermocouple. More details of the experiment can be found in our previous work [43].

When the substrate was heated to and maintained at 300 ℃, discharge gas (Ar) was put into the chamber for about 5 min to remove the air in the device. The precursor gas (TiCl₄) and O₂ were put in the reactor and the discharge was on. The input voltage and current were set at about 8.2 kV and 13.0 mA, respectively. The discharge frequency was adjusted to 14.68 kHz. After 30 min deposition, a white TiO₂ thin film was obtained, and this was named Original.

2.2   H₂ plasma treatment

TiO₂ thin films as-deposited were hydrogenated by RF AP He/H₂ plasma. Figure 1 presents a schematic diagram of the experimental setup for the hydrogenation process. The hydrogenation process of TiO₂ thin film was operated in the same quartz discharge chamber. An RF generator (RF-10S/PWT, Advanced Energy Co. Ltd, USA) was employed, and the discharge power was set at four different powers (40, 80, 120 and 160 W). He (99.99%, Shanghai Central Gases Co., Ltd, China) was used as the discharge gas and H₂ was used as the reduced agent. The H₂ was supplied by decomposing deionized water via a H₂ generator (QL-300, Shandong Saikesaisi Hydrogen Energy Co., Ltd). The gas flow rates were controlled by MFCs and set as He: H₂ = 600:15 ml min⁻¹. The treatment time was 5 min. The treated samples were obtained and were named P40, P80, P120 and P160, respectively.

Figure  1.  Schematic diagram of the experimental setup for the hydrogenation process.

DownLoad: Full-Size Img PowerPoint

There was no external heating during the plasma discharge. The gas temperatures (T_g) during different plasma power applications were obtained through simulation analysis based on the OH peak (309 nm) of the plasma optical emission spectroscopy (OES) images, measured five times. The results are displayed in figure 2 and show that T_g increases with plasma power. The four values (176 ℃, 213 ℃, 257 ℃ and 381 ℃) marked by blue dashed circles correspond to the plasma hydrogenation temperature when the plasma power was 40, 80, 120 and 160 W, respectively.

Figure  2.  The gas temperature (T_g) increases with the plasma power.

DownLoad: Full-Size Img PowerPoint

2.3   H₂ thermal treatment

H₂ thermal treatment of samples was conducted in the same quartz chamber. The four temperatures (176 ℃, 213 ℃, 257 ℃ and 381 ℃) corresponding to the four different powers of plasma hydrogenation and a higher value (470 ℃) were selected as the H₂ thermal annealing temperatures. The annealing time was 30 min. The atmosphere was He/H₂ (same as in the He/H₂ plasma treatment) and pure H₂. The flow rates were also controlled by MFCs and set as He: H₂ = 600:15 ml min⁻¹ and H₂ = 15 ml min⁻¹, respectively. The treated samples were named T176, T213, T257, T381, T470, H176, H213, H257, H381 and H470, and the treatment modes are shown in table 2.

Table  2.  The various treatment modes and conditions and the corresponding sample names.

Treatment mode	Treatment conditions	Sample name
Untreated	Ar: O2: carrier gas (TiCl4) = 600:15:25 ml min−1, 300 ℃, 18.00 kV, 1.00 mA, 30 min	Original
He/H2 plasma treatment	He: H2 = 600:15 ml min−1, 40 W, 5 min	P40
	He: H2 = 600:15 ml min−1, 80 W, 5 min	P80
	He: H2 = 600:15 ml min−1, 120 W, 5 min	P120
	He: H2 = 600:15 ml min−1, 160 W, 5 min	P160
He/H2 thermal treatment	He: H2 = 600:15 ml min−1, 176 ℃, 30 min	T176
	He: H2 = 600:15 ml min−1, 213 ℃, 30 min	T213
	He: H2 = 600:15 ml min−1, 257 ℃, 30 min	T257
	He: H2 = 600:15 ml min−1, 381 ℃, 30 min	T381
	He: H2 = 600:15 ml min−1, 470 ℃, 30 min	T470
Pure H2 thermal treatment	H2 = 15 ml min−1, 176 ℃, 30 min	H176
	H2 = 15 ml min−1, 213 ℃, 30 min	H213
	H2 = 15 ml min−1, 257 ℃, 30 min	H257
	H2 = 15 ml min−1, 381 ℃, 30 min	H381
	H2 = 15 ml min−1, 470 ℃, 30 min	H470

 | Show Table

DownLoad: CSV

2.4   Characterization

OES of the RF plasma was investigated using an Avantes AvaSpec-ULS4096CL-EVO spectrometer. The gas temperatures at different discharge powers were simulated with MTO software (a spectral simulation program written in python 3.7) using the OH radical peak (309 nm) in the plasma OES images [44, 45]. The morphologies of samples were examined using a field emission scanning electron microscope (FE-SEM, Hitachi S-4800, Japan) operated at 5.0 kV and transmission electron microscopy (TEM, JEM-2100, Japan). For the TEM measurements, a few samples of TiO₂ were scraped from the film surface with a clean stainless-steel knife and ultrasonically dispersed in alcohol. The crystalline phase analysis of samples was completed using x-ray diffraction (XRD, D/max-2550 PC, Japan) in a grazing incidence mode with Cu-K_α radiation (λ = 0.154 056 nm). The accelerating voltage and the applied current were 40 kV and 100 mA, respectively. XRD spectra were recorded from a 2θ angle of 10–80 degrees with a step size of 0.020 degrees at a rate of 5 degrees min^–1. The surface structural fingerprints were obtained using Raman spectroscopy (Renishaw inVia-Reflex, United Kingdom). The chemical states of the samples were investigated using x-ray photoelectron spectroscopy (XPS, ThermoFisher Scientific ESCALAB 250Xi, United States of America) equipped with Al-K_α x-ray radiation (1486.6 eV). For XPS analysis, Avantage software was employed to deconvolve and analyze the detailed binding energy and ratios of species. The binding energy was calibrated using the C 1s photoelectron peak at 284.8 eV as the reference. The optical properties (absorbance and reflectance) of samples were acquired using a UV–vis spectrophotometer (Shimadzu UV-2600, Japan) equipped with an integrating sphere accessory and using BaSO₄ as a reference (for reflectance). The data were recorded in the range of 200–800 nm.

3.   Results and discussion

3.1   Comparison of blackening effects on TiO₂ films after modification treatment

Figure 3 shows photographs of samples treated by various modes. In the first row, the colors of plasma treated samples (P40, P80, P120 and P160) exhibit a blackening trend with plasma power. For sample P160, the surface in the plasma zone presents as almost totally black with some accompanying dark spots sometimes. Generally, such a black TiO₂ sample has highly concentrated defects on the surface [46], which will be further discussed in the XPS, XRD and Raman analyses. While thermal annealing samples (in the second and third row) do not present a black color even under a pure H₂ atmosphere with higher temperature (470 ℃) and longer time (30 min), it is clearly shown that plasma can effectively blacken TiO₂ film under lower temperature, which suggests that some factors in the plasma can promote the TiO₂ hydrogenation process.

Figure  3.  Photographs of samples before and after treatment with various modes. The length of the plasma is 20 mm and is marked by the pink dotted lines.

DownLoad: Full-Size Img PowerPoint

It was found that the black spots present on the surface of P80, P120 and P160 strongly depended on the hydrogenation processes and plasma parameters. The heterogeneity of the plasma hydrogenation process is beyond the scope of this work and will be further studied later.

Figure 4 shows the optical absorbance of samples. In figure 4(a), it can be seen that plasma treated films present improved visible light response, and the absorbance intensity in the visible region increases with plasma power. This suggests that the electron energy level structure of black TiO₂ has been reconstructed and the optical band gap has been tailored [18, 20]. For thermally treated samples (figures 4(b) and (c)), the absorbance spectra exhibit an almost identical trend to the original ones. The optical absorbance results are consistent with those of their photographs (figure 3).

Figure  4.  The absorbance in the UV–vis region (200‒800 nm) of samples treated by (a) plasma and ((b), (c)) thermal annealing.

DownLoad: Full-Size Img PowerPoint

The color change of samples was measured based on the CIE L^*a^*b^* color space system, which was introduced in 1976 and is recommended by the International Commission on Illumination [47]. This system expresses color as a combination of three values, L^*, a^* and b^*. The overall change of the colorimetric value ΔE is calculated using the following equation:

Here, L^* is the lightness ranging from 0 (black) to 100 (white), and a^* and b^* represent the red–green and yellow–blue chromaticity indices which take both negative and positive values. The subscript o denotes the original sample, and the subscript m signifies the modified sample. The L^*, a^* and b^* values of various samples were calculated by transforming the optical reflectance spectra.

TiO₂ film color changes (ΔE) are presented in figure 5. The average overall color changes ΔE for plasma treated films are 103.79 (P40), 778.11 (P80), 1369.16 (P120) and 3430.19 (P160). The ΔE values for thermally annealed films are lower than 22.26 (H470). He/H₂ plasma treatment results in dramatic ΔE change, while thermal annealing shows low ΔE change.

Figure  5.  The color change (ΔE) of TiO₂ films treated by various modes.

DownLoad: Full-Size Img PowerPoint

3.2   The morphologies of TiO₂ films

The surface morphologies of samples have no significant changes after being modified by various modes, and FE-SEM images of the Original, P160, T470 and H470 samples are presented in figure 6. All TiO₂ film surfaces consist of many clusters and present a cauliflower-like structure. These clusters are composed of many smaller nanospheres. The FE-SEM results suggest that the surface morphologies of the TiO₂ films were not significantly changed through plasma or thermal hydrogenation.

Figure  6.  FE-SEM images of samples before and after being treated by various modes. (a) Original, (b) P160, (c) T470 and (d) H470.

DownLoad: Full-Size Img PowerPoint

3.3   The chemical composition of TiO₂ films

XPS was performed to detect the chemical composition and binding environment of the TiO₂ surface and the results are shown in figure 7. Curve fitting analyses of Ti 2p and O 1s were done based on Gaussian equations. The XPS survey scan spectra (figure 7(a)) collected from Original, P160, T470 and H470 are very similar. In addition to the Ti and O signals that are expected for TiO₂, C peaks were detected on the samples' surfaces. The C element content is low compared with the O and Ti elements and is believed to be included during sample preparation, storage and subsequent testing. Meanwhile, for Original, a very weak peak at 199.26 eV can be attributed to a Cl 2p peak, while for modified samples, this peak disappears. This is owing to the small amount of residual TiCl₄ or HCl on the Original surface during the deposition process, and these residues have been removed completely after modification processing. The results prove that the H₂ treatment is a clean process and, more importantly, the hydrogenated TiO₂ is not doped with other elements.

Figure  7.  (a) XPS survey scan spectra, and high-resolution XPS spectra of (b) Ti 2p and (c) O 1s. Open circles and lines represent the experimental and the fitted data, respectively.

DownLoad: Full-Size Img PowerPoint

The core level XPS spectra of Ti 2p are displayed in figure 7(b). For Original, two obvious peaks at 458.9 and 464.6 eV can be attributed to Ti⁴⁺ 2p_3/2 and Ti⁴⁺ 2p_1/2 with typical characteristic Ti⁴⁺ binding energies [48]. Usually, Ti³⁺ species will be created on the TiO₂ surface after He/H₂ plasma treatment [39]. For P160, two new peaks appear at positions of 457.4 and 460.8 eV, attributed to Ti³⁺ 2p_3/2 and Ti³⁺ 2p_1/2 [49]. Similarly, the Ti³⁺ peaks are also present in T470 and H470. The percentages of integrated areas of various Ti 2p species are listed in table 3. The content of Ti⁴⁺ species in the modified samples (P160, T470 and H470) is reduced compared with that of Original. The content of Ti³⁺ species for thermally annealed samples (T470 and H470) is very similar, while the content of Ti³⁺ species for P160 is about 1.5 times that for thermally annealed samples. The results indicate that the Ti atom has a different chemical bonding environment before and after modification processing, and He/H₂ plasma can create more Ti³⁺ species on the TiO₂ surface than thermal annealing.

Table  3.  The percentage of integrated area of the high-resolution Ti 2p and O 1s XPS spectra.

	Ti (%)					O (%)
Samples	Ti4+ 2p3/2 (458.9 eV)	Ti4+ 2p1/2 (464.6 eV)	Ti3+ 2p3/2 (457.4 eV)	Ti3+ 2p1/2 (460.8 eV)		Ti–O–Ti (530.0 eV)	Ti–OH (531.5 eV)	C–O (532.8 eV)
Original	34.55	17.05	—	—		42.46	3.75	2.19
P160	32.56	16.11	1.34	1.46		40.53	4.80	3.20
T470	34.04	16.90	0.42	1.28		41.27	3.60	2.50
H470	34.05	16.79	0.65	1.37		41.53	3.54	2.07

 | Show Table

DownLoad: CSV

The core level XPS spectra of O 1s were divided into three peaks and are displayed in figure 7(c). The first component at 530.0 eV can be ascribed to the Ti–O–Ti bond, which means the lattice O²⁻ bonded to the Ti⁴⁺ ion [50]. The second component centered at 531.5 eV can be assigned to the Ti–OH bond, which is related to hydroxyl groups [19] or OVs [22]. The third component at around 532.8 eV can be attributed to the C–O bonded species on the TiO₂ surface [51]. Additionally, the percentage of integrated area of O 1s is also shown in table 3. The C–O bond content is very low, which suggests a small amount of C contamination during storage or testing processing. It is consistent with the results of the XPS survey spectra (figure 7(a)). For Original, the peak at 531.5 eV may be relate to OH species adsorbed on the surface [39]. For P160, the Ti–OH bond content increases to 11.34%, which indicates the possible formation of H₂O from the interaction of H⁺ ions with the TiO₂ film and the creation of OVs [49, 52]. However, for T470 and H470, the content of Ti–OH is almost the same, suggesting that the two thermal treatments do not have different effects on the production of OVs or hydroxyl groups on the TiO₂ surface. Therefore, He/H₂ plasma treatment plays a critical role in the hydrogenation of TiO₂ as compared with thermal treatment.

3.4   Crystallographic structure of TiO₂ films

Figure 8 shows the XRD patterns of samples treated by various modes. The peaks identified at 2θ values of 25.4°, 37.8°, 48.1°, 53.8°, 55.0° and 62.7° correspond, respectively, to the crystal planes of (101), (004), (200), (105), (211) and (204) in the standard XRD pattern of anatase TiO₂ (JCPDS No. 21–1272). All samples show similar XRD patterns, indicating that no phase transformation occurred during the plasma or thermal processing. Comparing with Original, the (101) peak intensity of P40 decreases, which can be assumed to indicate that defects (such as OVs and surface hydroxyl groups) were introduced into the TiO₂ crystal by the bombardment of high-energy active species [24]. With increasing plasma power, the (101) peak intensity of samples P40, P80, P120 and P160 increases, indicating the improvement of samples' crystallinity. This can be attributed to the increase of substrate temperature. For thermally treated samples, the (101) peak intensities increase with the annealing temperature, also indicating the crystallinity improvement of the samples.

Figure  8.  XRD patterns of samples treated by (a) He/H₂ RF plasma, (b) He/H₂ and (c) pure H₂ thermal annealing.

DownLoad: Full-Size Img PowerPoint

The XRD spectrum intensity of plasma treated samples is weaker than that of thermally treated samples at the same temperature, which indicates that some factors attenuate the effects of high substrate temperature, such as high-energy species collision. Generally, the crystallization of TiO₂ will be improved by pure Ar or He plasma [53], while defects can be introduced into TiO₂ when treated by H₂ or N₂ plasma [20, 54]. Therefore, the XRD spectrum intensity of plasma treated samples is weaker than that of thermally treated samples, and this can be attributed to the defects introduced by He/H₂ plasma bombardment of the TiO₂ surface.

As the penetration depth of Cu-K_α x-rays is ~15 μm [39], XRD is known as a bulk sensitive method. Raman spectroscopy, as a very sensitive technique, contains more signals from the surface region than from the bulk region of the TiO₂ samples. Hence, to understand more deeply the effect of He/H₂ plasma treatment and thermal annealing on the phase transformation of TiO₂, Raman spectroscopy was also employed in the present study to assess the surface crystalline structure of the samples. The major Raman bands of the anatase phase appear at 144, 197, 399, 515, 519 (superimposed with the 515 cm⁻¹ band) and 639 cm⁻¹, which can be ascribed to the six Raman-active modes of the anatase phase with the symmetries of E_g, E_g, B_1g, A_1g, B_1g and E_g, respectively [39].

Figure 9 presents the Raman shift spectra of samples modified by various modes. All samples display anatase TiO₂ major Raman vibrational modes, consistent with the XRD results. In figure 9(a), the peak intensities, especially for the main peak (144 cm⁻¹), clearly decline with increasing plasma power, which exhibits a different result from the XRD. The reduction in Raman peak intensity indicates the increased disorder of the TiO₂ surface due to the bombardment of the He/H₂ plasma. It also suggests that, for the plasma treated samples, defects (disordered layers) are mainly concentrated on the TiO₂ film surface rather than inside the bulk. For the thermally treated samples, however, the intensity of the peak increases with the annealing temperature, indicating that the crystallization of TiO₂ film was promoted by increasing the annealing temperature [55]. The results for the thermally treated samples are in good agreement with the above XRD results, indicating that thermal annealing did not introduce the defects into the surface or bulk of the TiO₂ thin films. This is different from the results of Wang and Chou [34], and the reason may be that the temperature is not high enough. The above Raman results indicate that the hydrogenation occurs mainly on the surface of the films during He/H₂ plasma treatment processing.

Figure  9.  Raman shift spectra of samples treated by (a) He/H₂ RF plasma, (b) He/H₂ and (c) pure H₂ thermal treatment.

DownLoad: Full-Size Img PowerPoint

3.5   TEM analysis of TiO₂ films

Figure 10 shows the TEM results for Original, P160, T470 and H470, where the (101) lattice planes of anatase TiO₂ were confirmed on the basis of a lattice fringe of d = 0.35 nm. Compared with Original (figure 10(a)), the surface layer of P160 (figure 10(b)) seems to contain some degree of imperfection, while the inner layer (figure 10(c)) exhibits improved crystallinity. T470 (figure 10(d)) and H470 (figure 10(e)) show well-defined crystal structure with minimal lattice imperfection. The TEM results suggest that the bombardment of high-energy active species disrupted the crystal structure on the TiO₂ surface [56], while the high temperature increased the overall crystallinity of TiO₂. This validates the conjecture obtained from the XRD and Raman measurements.

Figure  10.  TEM images of samples. (a) Original, (b) surface and (c) inner of P160, and (d) T470 and (e) H470.

DownLoad: Full-Size Img PowerPoint

3.6   The active species analysis

Figure 11 presents OES of the plasma (160 W) and thermal (470 ℃) processes. For RF He/H₂ plasma (black line), the main emission peaks at 728.46, 706.78, 668.11, 578.91, 502.01 and 389.62 nm are detected and identified as He species [48]. The peaks at 656.62, 486.49 and 434.49 nm correspond to H_α, H_β and H_γ, respectively [57], which belong to the Balmer series. A weak peak at 309.23 nm is also detected and is attributed to the OH radical peak [58], which has been used to estimate the gas temperature of the RF He/H₂ plasma mentioned above. The presence of the OH peak is a common phenomenon observed typically in AP plasma. The OH peak originates from H₂O that was desorbed from the gas pipe wall or traces in the He or H₂ used. Obviously, the appearance of these emission lines demonstrates that RF plasma has a stronger ability to split the reduced agent (H₂) and generate H₂ active species. The process of H₂ agent splitting caused by high-energy electron collisions in the bulk plasma region can be summarized according to the chemical reactions in equations (2) and (3).

Figure  11.  OES of He/H₂ plasma at 160 W and thermal annealing at 470 ℃.

DownLoad: Full-Size Img PowerPoint

However, for thermal processing (red line at the bottom), no peak was detected in the spectra, suggesting almost no active species. OES results suggest that the reduced agent (H₂) and He were successfully split, and numerous active species were generated in the RF He/H₂ plasma due to the bombardment of high-energy electrons rather than high temperature. The high-energy species are beneficial to the surface activated process during hydrogenation processing [21].

Figure 12 shows the Gibbs free energy changes of ΔG⁰(T) related to oxidation reactions of TiO₂ involving one mole of gaseous O₂. Such a figure is called an Ellingham diagram [59]. The thermodynamic data used in figure 12 are cited from the compilation by Kubaschewski [60]. The relationship for oxidation of H and Ti approximates to a straight line over the temperature ranges in which no phase transformations occurred. Thermodynamically, the straight lines represent the Gibbs free energy changes ΔG⁰(T) at different temperatures, which indicates the relative stability of oxides. It is possible to directly visualize the affinities of elements for O₂ in their standard conditions by observing the relative positions of their lines in the diagram. A given element can reduce the oxides of all other elements whose lines lie above others in the diagram [59]. From figure 12, the lines of Ti₃O₅ and Ti₂O₃ unfortunately lie below the H₂ line, so elemental H cannot reduce TiO₂ at low temperature (< 1400 K). This explains why H₂ gas hardly reduced crystalline TiO₂ by thermal annealing below 470 ℃ in our experiment. However, for plasma processing, in addition to the thermal effect, He/H₂ plasma interacted with the TiO₂ surface and greatly promoted the hydrogenation process.

Figure  12.  Ellingham diagram of ΔG⁰ versus temperature.

DownLoad: Full-Size Img PowerPoint

The ionization degree of H₂ was calculated using the Saha equation:

Here, χ, P, T, e, V_i and k_B are the ionization degree, pressure (in Pa), gas temperature (in K), electronic charge, gas ionization energy (13.59 V for H₂), and Boltzmann's constant. The ionization degree of H₂ for 470 ℃ thermal annealing is about 1.958 × 10⁻⁴⁶. Therefore, no characteristic peak was detected in OES during 470 ℃ thermal annealing due to such a low decomposition rate.

Although there are a lot of articles on TiO₂ hydrogenation through H₂ thermal annealing [34], the black TiO₂ was prepared in H₂ atmosphere at 600 ℃ for 6 h. The ionization degree of H₂ at 600 ℃ estimated by the Saha equation is about 1.724 × 10⁻³⁹. It is seven orders of magnitude higher than that at 470 ℃ and 14 orders higher than that at 381 ℃ (9.031 × 10⁻⁵³). Therefore, according to the ionization degree of H₂ and the annealing time of Wang and Chou (6 h) [34], the reaction time of TiO₂ blackening in this experiment through thermal annealing is about 5.28 × 10⁷ h (more than 6030 years) at 470 ℃ and is about 1.15 × 10¹⁴ h (more than 13 billion years) at 381 ℃. High enough temperature and long enough treatment time are the key factors for successful thermal annealing in the TiO₂ hydrogenation process. From these analyses, the hydrogenation process using plasma in our experiment displays the obvious advantages of moderate temperature (below 381 ℃) and less time (less than 5 min) due to the bombardment of energetic species.

3.7   A possible hydrogenation mechanism

Based on the above experimental results and discussion, a possible hydrogenation mechanism was proposed as follows. Figure 13(a) is a photograph of the RF AP plasma discharge in our experiment. A sheath can be observed on the surface of the TiO₂ film and quartz dielectric. The schematic structure of our RF AP plasma discharge is shown in figure 13(b). The potential (φ) decreases gradually, and the direction of the electric field (E) is directed from the bulk plasma to the dielectric inside the sheath region. H active species are abundantly generated in the bulk region due to the collision of high-energy electrons with H₂ and diffuse toward the sheath region due to the concentration differences. Figure 13(c) shows how the charged species (H⁺) are accelerated by the strong electric field in the sheath and bombard the TiO₂ surface with sufficient energy. The interactions between the H₂ plasma and the TiO₂ film surface result in the dissociation of Ti–O bonds and desorption of O atoms [46], which can be summarized according to the chemical reaction in equation (5)

Figure  13.  Schematic diagram of a possible hydrogenation mechanism for RF AP He/H₂ plasma treatment. (a) The photograph (front view), (b) schematic diagram of RF AP plasma structure during the TiO₂ hydrogenation process, (c) the interaction of the He/H₂ plasma with the TiO₂ film during the blackening process, and the representative microstructures of (d) pristine and (e) black TiO₂ films. Size of the species is not to scale.

DownLoad: Full-Size Img PowerPoint

The formation of volatile molecules of H₂O is possibly responsible for the desorption of O atoms from the film surface after He/H₂ plasma treatment [39], creating Ti³⁺ species. This behavior may be associated with the high energy transferred to the sample during the plasma treatment, which favored the reduction of Ti⁴⁺ to Ti³⁺. Another point to be discussed is that some ions can be implanted into the TiO₂ film at depths of several nm, promoting the disordering by the formation of OVs [20, 39]. The representative microstructures of pristine and black TiO₂ films are exhibited in figures 13(d) and (e), respectively. After plasma hydrogenation, H species combine with surface O atoms to form hydroxyl groups (Ti–OH bonds) on the surface of the black TiO₂ film. Some O atoms are removed, forming OVs in the subsurface.

Therefore, it is the high dissociation rate of H₂ and the acceleration of its ion species through the sheath in the plasma that greatly modifies the surface structure and chemical bonding environment of TiO₂, creating OVs and Ti³⁺ and causing disorder in the film surface under relatively mild conditions. The existence of defects leads to the tailoring of the band gap of black TiO₂ and causes the light absorption to extend into the visible region.

4.   Conclusion

The effects of He/H₂ plasma and thermal treatment on the hydrogenation process of TiO₂ film are investigated in this work. The samples treated by RF AP He/H₂ plasma have been successfully blackened within a short time (5 min), while the color of the samples treated with He/H₂ or pure H₂ thermal treatment at higher temperature and with longer time (30 min) did not change obviously. The surface morphologies of the samples remained almost unchanged before and after plasma or thermal hydrogenation processing. XPS results indicate that OVs and Ti³⁺ species were introduced into the TiO₂ surface after He/H₂ plasma treatment. XRD and Raman results suggest that plasma hydrogenation improved the overall crystallinity of TiO₂ film, and at the same time, this process also created some defects in its surface crystalline structure. TEM results indicate that the surface layer of plasma treated TiO₂ film was amorphous, while the crystallinity of the inner layer was slightly improved compared with the original sample. TEM results validate the conjecture obtained from XRD and Raman spectroscopy (crystalline-disordered core–shell structure).

OES results suggest that numerous active H or He species were generated in RF He/H₂ plasma, while there were no peaks detected in the thermal processing. A photograph of the RF He/H₂ plasma was taken and a possible mechanism for the TiO₂ hydrogenation process by plasma has been proposed. Numerous active species were generated in the bulk plasma region, accelerated in the sheath region, and bumped toward the TiO₂ film, which will react with the TiO₂ surface to form OVs and disordered layers. This leads to the tailoring of the band gap of black TiO₂ and causes its light absorption to extend into the visible region, resulting in enhanced visible light absorption.

It is the high dissociation rate and the acceleration of the sheath that led to the better effects of plasma treatment than of heat treatment in the hydrogenation of TiO₂ films under a lower gas temperature and short time. It is worth noting that the hydrogenation on the surface of the TiO₂ film was uneven, which we will research in the future. Finally, RF AP He/H₂ plasma treatment, as a simple hydrogenation method, opens new opportunities for the design of advanced photocatalysts with enhanced solar energy utilization.